This page contains discussions that have been archived from Village pump (technical). Please do not edit the contents of this page. If you wish to revive any of these discussions, either start a new thread or use the talk page associated with that topic.
Latest comment: 9 years ago14 comments6 people in discussion
I created this category 10 days ago, and it's still not fully populated from the stub tag. The transclusion count states there are 70 articles with the stub, but so far it's showing 56 articles in the category (it had about 40 a few days ago). Why is it taking so long to update? Thanks. LugnutsDick Laurent is dead07:53, 29 April 2015 (UTC)
In the past couple of years, I have seen category population from templates take up to 100 days. It can be quite frustrating. I submitted a bug on Bugzilla somewhere, and there is a post in the VPT archives, I believe, but it's been a while. If you search for my user name, you might find it. – Jonesey95 (talk) 13:08, 1 May 2015 (UTC)
Its the bonus from one template fits all. Anybody know whats the setting for the job queue is? (fifo, random or timestamp) Christian75 (talk) 19:54, 2 May 2015 (UTC)
Right, I've made this edit. It is not a WP:NULLEDIT, since a diff exists (and it shows in the page history, my contribs, etc.) but it has no effect on the action of the template. However, it should get the transcluding pages into the job queue. --Redrose64 (talk) 19:01, 2 May 2015 (UTC)
I've just looked through all 70 transclusions of the template, and they all say that they are in the category. It occurs to me that this may not be a job-queue problem, but a category-counting bug (tracked as phab:23230). I'm just about to bot-null-edit all of the transclusions to see what happens. Watch this space... — Mr. Stradivarius♪ talk ♪14:57, 8 May 2015 (UTC)
@Lugnuts and Mr. Stradivarius: As Technical 13 noted elsewhere on this page, the job queue has been down for some days, so when a template is changed, the changes are only propagating through to articles as and when those articles are edited; and only at that point are the cats populated, one article at a time. I've been deliberately refraining from WP:NULLEDITing the transcluding pages, in order to see when the job queue picks up again. --Redrose64 (talk) 15:33, 8 May 2015 (UTC)
According to Legoktm, the job queue has been back up. There is something else going on because I've watched it go from 19.7 million yesterday morning to about 21.8 million right now and still climbing at a steady rate. I'm wondering at what threshold it will be time for a few more resources to be given to the queue to knock it back down to acceptable levels (or what an acceptable level even is). — {{U|Technical 13}} (e • t • c)15:50, 8 May 2015 (UTC)
Well the null edits worked, so it was definitely the job queue that was the problem. Sorry for ruining the job-queue test. (Although maybe something in userspace could be used to test that instead?) — Mr. Stradivarius♪ talk ♪15:56, 8 May 2015 (UTC)
Latest comment: 9 years ago9 comments3 people in discussion
Mobile web Chrome. I made this edit; now I have been experiencing a lot of issues. (1) Beta was enabled. (2) Picture loading was disabled. (3) I had this weird box full of technical writing at the top of pages (it is gone now; after I clicked on it, everything got worse). (4) I get a prompt nonstop that my edit was saved even though I'm not making edits (severe). (5) Pages are taking longer to load as if there is something new being loaded (this is the worst). (6) The page isn't loaded properly for a second then it is fine. (7) There is a big space on edit summary pages. (8) The page in which I thank editors is a mess. (9) Of course there is the notifications system bug above. I have tried logging out (while I was logged out testing things for a few minutes, all these things are still happening). I have tried making a null edit. I have tried clearing some "website settings" in Chrome. I tried setting my preferences to default. I tried turning off my phone. Of course everything is all fine on desktop view. It is not my phone because other websites are fine. This needs to be resolved. —DangerousJXD (talk) 23:25, 5 May 2015 (UTC)
It would be great if you could take a screenshot of what you see. Also, have you tried using another mobile browser? Is this on Android or iOS? Additionally, though you are experiencing a trial in your life, it's best not to demand results. Editing via mobile Wikipedia is new and full of bugs. Patience is required. Killiondude (talk) 03:45, 6 May 2015 (UTC)
Killiondude, thanks for responding. I have the high tech knowledge of a stereotypical 70 year old so uploading a screen shot is a no no. I just tried using the browser on my phone called "browser" but it's the same. It is Android. I am sorry if it sounded like I was "demanding results". —DangerousJXD (talk) 03:58, 6 May 2015 (UTC)
It would be too hard to time it right to take a picture of the 4th thing I listed for example. As I said, uploading pictures is not for me. Notifications appear to be fine now. 1 and 2 are obviously fine now (because I turned them off). 3 as I said is gone (that is what I think made the first 2 items happen. 4 and 5 are the most important thing here. 6, 7 and 8 are still a thing but they are minor compared to 4 and 5. —DangerousJXD (talk) 22:10, 6 May 2015 (UTC)
I have discovered that the loading issue is (most of the time) non-existent for the loading involved when you save an edit; otherwise, the loading issue is still happening. It is a pain to navigate. —DangerousJXD (talk) 08:56, 8 May 2015 (UTC)
Hi, ya'll – yes, I am aware of this situation and want to look into it. It seems to me to need more study, as I've found several items on the What links here pages of the rcats that don't appear to fall into the instructions found on the /Explanation subpage. Another seemingly mangled mess that I want to untangle and provide more clarification. I haven't yet taken the opportunity to give it the time it deserves. – Paine19:09, 7 May 2015 (UTC)
Thank you, FayenaticLondon! I updated the rcats and the explanation page, so they hopefully make more sense now. Feel free to help out with further edits because, as I'm sure you know, here on Wikipedia there is always room for improvement. Thank you! and Best of everything to you and yours! – Paine05:01, 11 May 2015 (UTC)
I found myself wondering the same thing, and that's one of several reasons that I had trouble understanding what was going on. It appears that the original namers thought ... heck, I don't know what they were thinking. Now that I have familiarized myself with what they were trying to accomplish, I'm pretty much okay with the cat and rcat page titles as long as things are clear enough in their content/documentation. – Paine17:21, 11 May 2015 (UTC)
Latest comment: 9 years ago5 comments4 people in discussion
In my Preferences, in the User Profile tab, it says, "How do you prefer to be described?" The choice is "She edits wiki pages" or "He edits wiki pages" or "I prefer not to say". It also says, "This information will be public." My question is, how do I see how another editor has set this preference? — Mudwater (Talk)13:52, 9 May 2015 (UTC)
mw:Help:Magic words#gender shows how to do it in wikitext. It's used by {{Gender}}. For example, {{Gender|Mudwater}} produces: he. It's currently nominated for merging and displays "See Tfm" before the pronoun. I don't know a simply way to see the setting without previewing such wikitext, but a script could be made if it doesn't already exist somewhere. PrimeHunter (talk) 14:17, 9 May 2015 (UTC)
At present, the "Mark all pages as visited" button on the Watchlist special page does so as at the time that the button is pressed. This has the effect that I need to take note of the last time on the watchlist, as pages that have not yet been displayed to me on my watchlist but have since been edited will be marked as visited. I then have to find that time in the updated list to see which entries I have missed. It would make sense to change the function so that only the updates up to the time at which the displayed page was generated get marked as visited. Would this be reasonable? —Quondum14:28, 9 May 2015 (UTC)
I am forced to, else I lose my place in reviewing changes. And if I forget to refresh just beforehand, I lose where I was. Which means, I have to review and edit everything on the page, then refresh, and review and edit everything new, then ... you get the idea. It forces me to get all the way up to date before I can clear everything that I have chosen to skip. With a large watchlist, this means I cannot take a break until I'm absolutely up to date, and even then I might miss the newest items that were updated in the seconds between finally scanning the list and clicking the button, so a still have to remember the last time on the list when I click. In contrast, I cannot see any problems if my suggestion is implemented. Go through the list, do all the needed edits, call it a day, hit the button, and take that much-needed break. Come back, and the watchlist tells me exactly where I left off. Isn't that what the watchlist is intended for? —Quondum21:26, 10 May 2015 (UTC)
Are you using NAVPOPS (or similar) to check pages on your watchlist? Because if you're going to the pages themselves, then there is never a need to click the 'Mark all pages' button. Whatamidoing (WMF) (talk) 02:19, 13 May 2015 (UTC)
I'm just using the standard Special:Watchlist page. I don't open every page, because several relate to things I'm not involved in at the moment, or I can see from the edit comment what it is. Why can't anyone understand what clean functioning is? I'm not interested in workarounds; this must impact everyone to some degree. —Quondum02:30, 13 May 2015 (UTC)
Here's my workflow, FWIW: 1. Refresh the watchlist page. 2. Cmd-click (Ctrl-click) on all diffs that I want to view, which opens them in new tabs in the background. 3. Refresh the watchlist again if I am worried that my clicking took too long or if I got distracted. 4. Mark all pages as watched. – Jonesey95 (talk) 03:48, 13 May 2015 (UTC)
I've been editing WP long enough to figure out a workflow that works (and I also use your approach at times). I'm not interested in workarounds, since I already have them. What I'm saying is that it can be changed so that it always works right, and people learning to use watchlists won't first have to stumble over this, then figure out a workaround. Tell me that it involves work that no-one is going to do, or put it on the worklist – or tell me to get lost. Just please stop behaving like my suggestion is not an improvement to consider. —Quondum04:10, 13 May 2015 (UTC)
This sounds like a perfectly reasonable feature request to me. To get it looked at and possibly implemented by the MediaWiki developers, you will need to file it in Phabricator, which is where bug tracking and feature tracking for the software are done. The devs will know better whether this is feasible and how much work it would take to implement. — Mr. Stradivarius♪ talk ♪05:59, 13 May 2015 (UTC)
Latest comment: 9 years ago13 comments7 people in discussion
I'm sure this has come up before but ... I blocked an IP for 5 years after it started vandalizing straight off a 2.5 year block. However, the block length has come up as 4 years, 365 days, 43 minutes and 12 seconds [1]. I'm sure there's a logical reason (and it doesn't really make a difference), but it just seemed odd. Black Kite (talk)11:13, 11 May 2015 (UTC)
There are 4 years, 364 days, 13 hours, 32 minutes and 14 seconds until 5 years have passed. (refresh) I know that Trappist the monk has been working with time on the Template talk:Countdown#Inappropriate precision template, and maybe they can shed some light on this subject for you. I'm guessing it's just a combination of how PHP parses dates and time and a difference in time zones and/or daylight savings time. — {{U|Technical 13}} (e • t • c)12:34, 11 May 2015 (UTC)
I've encountered this before: sometimes I've decided to block a user for one second less than whatever the long period of time will be (e.g. 729 days, 23 hours, 59 minutes, 59 seconds), so I'll pull out the calculator and figure out how many seconds two years is, and then I'll subtract one and block for that number. When this happens, I've sometimes noticed that it ends up several hours short; I've always attributed this to a calculator-operation mistake by me, not knowing that it would happen on a simpler thing like "5 years". Nyttend (talk) 15:02, 12 May 2015 (UTC)
Nyttend, how many seconds are in a year? 31,536,000; 31,557,600; 31,622,400, or some other number (365, 365.25, and 366 day years respectively)? — {{U|Technical 13}} (e • t • c)15:10, 12 May 2015 (UTC)
I guess I'd normally do 31,535,999 seconds, or 157,766,399 seconds for a five-years-minus-one-second block (four years of 365 days, one year of 366), but when that happens, it routinely ends up with random numbers like Black Kite mentions, rather than being short a day here or there. Nyttend (talk) 15:16, 12 May 2015 (UTC)
Ahh... I see. So, if you set a 5 year -1 second block anywhere from March 1, 2011/2015 to February 28, 2012/2016 you'd end up with the incorrect number of seconds (2012/2016 are leap years, so the seconds should have been for 366+365+365+365+366 days or 157,852,800 minus a second. If you wanted to be more precise, 157,788,000 minus a second for 5 365+1⁄4 day years should always be correct. — {{U|Technical 13}} (e • t • c)15:44, 12 May 2015 (UTC)
One might use the #time: parser function to figure out how many seconds there are between now (or some date) and a date 5 years hence:
{{#time:U|now + 5 years}} → 1889910658 seconds UTC
My guess is that this is because of the way strtotime extension is coded in MediaWiki. It causes blocks to be calculated by adding relative times to midnight of the current day (notice the '0' in mktime). For example, if you block at 23:59 for one day it will actually be for one minute. strtotime callMamyles (talk) 16:09, 12 May 2015 (UTC)
Of course, that would be assuming it's calling the extension rather than the PHP strtotime built-in function. Perhaps someone else wants to take a look at that code, which calculates blocks. Mamyles (talk) 17:04, 12 May 2015 (UTC)
For the actual block, a duration of "5 years" is calculated by PHP's strtotime function in the way you'd expect: the same date 5 years hence. But MediaWiki's Language::translateBlockExpiry function that turns an English expiry like "5 years" into something in the local language can't be so smart. When it's not something in MediaWiki:ipboptions (in the target language), it calculates the duration relative to the epoch to use as seconds-since-the-epoch, and then breaks that timespan into something more appropriate in the local language. For years it uses 31556952 seconds (365.2425 days), the mean length of a year in the Gregorian calendar, leading to more rounding errors. Anomie⚔01:47, 14 May 2015 (UTC)
{{Collapse}} with whole header cell as collapse toggle
Latest comment: 9 years ago4 comments3 people in discussion
Instead of only having a [show/hide] toggle, is there any way to make the whole header line the toggle with a collapsible table/section? Where it just adds an icon or something, and a user could click on anything in the header that isn't a link to another page in order to collapse/expand the content.
Basically sometimes a more descriptive line would be of use, like when there is 'Additional information'. This way there could be the basic info/links on the line too, without taking up an extra line to explain what the additional, collapsed information is. -— Isarra༆20:28, 11 May 2015 (UTC)
Technically, it could be done, but it runs the risk of clashing with other links in the header. I think no one is interested in investing time in patching up the old collapsible code however... we should switch to mw-collapsible, which has been available for years, but has never been used (due to lacking support for autocollapse). -- [[User:Edokter]] {{talk}}21:12, 11 May 2015 (UTC)
Alright, thanks, guys. Maybe we'll see some progress on that at the hackathon or something soon. *cracks knuckles* -— Isarra༆17:34, 14 May 2015 (UTC)
Undoing move from User to User:User
Latest comment: 9 years ago3 comments3 people in discussion
Assuming that this move was a mistake, will moving it back cause any weird namespace problems? Normally I would just fix it myself but this one makes me a little nervous. Regards, Orange Suede Sofa (talk) 22:46, 12 May 2015 (UTC)
Latest comment: 9 years ago5 comments3 people in discussion
For example Shashmaqam links to Feudal which is a redirect to Feudalism. Shashmaqam correctly shows in the WLH for Feudal, but is missing from the WLH for Feudalism, even though other second-level links for Feudal show up in the Feudalism WLH. In fact I've identified over 500 articles that are not showing up in the Feudalism WLH by using the MediaWiki API:Backlinks and finding the full list of backlinks that way. -- GreenC03:04, 13 May 2015 (UTC)
PrimeHunter. Ok the 500 limit is suggestive of a bug. The 500 limit is set by the API. After 500 you need to recursively call the API until all entries are received. However this process is slightly different for second-level backlinks, the coding is a little different, it changed with a previous version of MediaWiki. It looks like the code to display the backlinks on the main site was not updated to reflect the changes in the API and the end result is only 500 second-level backlinks are being displayed. It might be possible this was done intentionally, to limit second level backlinks to 500.. but I can't understand why since there is no mechanism for users to go beyond the 500 limit (short of writing their own code via the API) and there is no notice saying the backlinks are incomplete. -- GreenC16:27, 13 May 2015 (UTC)
After the new version of Mediawiki went live today, AWB users are unable to log in. See bug report. Ok, you don't have to cheer so loudly. I can't make heads or talks with the changelog for the new update. Anyone have suggestions? Bgwhite (talk) 03:43, 14 May 2015 (UTC)
Starting a paragraph with a space after a nowiki code block results in rendering that paragraph as code
Latest comment: 9 years ago5 comments4 people in discussion
Normally, one should be able to have some text rendered as code, like this,
and then the next paragraph after the proper end of that block should look normal, like this.
However, if I have some text rendered as code, and then
that next paragraph starts with a space, like this, it's rendered as code.
This does not appear to be consistent with Help:Dummy_edit or meta:Help:Newlines and spaces which seems to suggest that space shouldn't make a difference, and seems to violate ending tags. What's up with that?
So, if there was some reason you needed to start a sentence with a space (which you really shouldn't, for just about anything), you'll have to use one of the templates, an html entity, or a nowiki wrapped space. — {{U|Technical 13}} (e • t • c)21:54, 14 May 2015 (UTC)
The admin dashboard counts should all be fixed now. I'm double checking as we speak to make sure the counts all match, so don't yell at me if one is a little off, I'll fix it quickly... — {{U|Technical 13}} (e • t • c)20:34, 14 May 2015 (UTC)
The changes I just made to the template make it so that only pages are counted on most of those listing, there is no more manual modification of numbers using #expr: since there is no need for it. I also trimmed out a bunch of formatnumber: stuff because that is built right into the PAGESINCAT function itself using the |R for Raw data. Let me know if any of the counts that I trimmed subcats/files off of need to have those counted and I can adjust for that. — {{U|Technical 13}} (e • t • c)21:50, 14 May 2015 (UTC)
Latest comment: 9 years ago1 comment1 person in discussion
Just for documentation:
Our servers are currently experiencing a technical problem. This is probably temporary and should be fixed soon. Please try again in a few minutes.
If you report this error to the Wikimedia System Administrators, please include the details below.
Request: GET http://en.wiki.x.io/wiki/Special:Search/intitle:16, from 10.64.32.107 via cp1065 cp1065 ([10.64.0.102]:3128), Varnish XID 3994251179
Forwarded for: 166.82.105.36, 10.64.32.107
Error: 503, Service Unavailable at Thu, 14 May 2015 21:15:40 GMT
— Vchimpanzee • talk • contributions • 21:20, 14 May 2015 (UTC)
Does one not simply log out of mobile?
Latest comment: 9 years ago6 comments5 people in discussion
an untested workaround: in my experience, logging out from any machine and project in SUL will log you out from all the borwsers and projects. i did not verify that "mobile" will log out also, but i can't imagine it should behave differently, so the answer will be: logging out from any desktop browser will log out the mobile also. you can then log-in back on the desktop - the mobile will remain logged out. peace - קיפודנחש (aka kipod) (talk) 22:12, 19 May 2015 (UTC)
Interwiki searches and case-sensitivity
Latest comment: 9 years ago13 comments5 people in discussion
It's always irritated me that, though the English Wikipedia's search engine allows us to search other Wikimedia projects, such searches don't follow the same rules regarding case-sensitivity as searches that are confined to this wiki. For example, typing "kazuo ishiguro" into the search box takes me to Kazuo Ishiguro (not Kazuo ishiguro, which doesn't exist), because the search isn't case-sensitive. But typing "q:kazuo ishiguro" goes to the Wikiquote page q:Kazuo ishiguro (which doesn't exist), not q:Kazuo Ishiguro (which does), because in this case it is case-sensitive. (If I search for "kazuo ishiguro" using the search bar on a Wikiquote page, however, I'm taken to q:Kazuo Ishiguro.) As far as I know the same applies for the various other interwiki codes: n: for Wikinews, b: for Wikibooks, c: for Commons, and so on, and language codes too. Ideally such searches would work in the same way as searching the wiki you're looking at, rather than adding the extra level of complexity that case-sensitivity creates. Obviously it's not the biggest deal, but I wonder if a fix for this might be a possibility? – Arms & Hearts (talk) 05:45, 15 May 2015 (UTC)
Actually, an interwiki prefix in the search box doesn't make a search at all. It just jumps to a specific url determined by meta:Interwiki map. q:kazuo ishiguro jumps to https://en.wikiquote.org/wiki/kazuo_ishiguro which automatically capitalizes the first letter by redirecting to https://en.wikiquote.org/wiki/Kazuo_ishiguro. As you say, it doesn't go to the article at https://en.wikiquote.org/wiki/Kazuo_Ishiguro, but it also doesn't make a search like https://en.wikiquote.org/w/index.php?search=kazuo_ishiguro&title=Special%3ASearch&fulltext=1. You could technically make a search by including it in the pagename like q:special:search/kazuo ishiguro, but of course users will not do that. The interwiki map includes many non-MediaWiki sites which may have various url formats to perform a search if they even have a search function. I suppose it would be possible for the software to say that if you enter an interwiki prefix in the search box which is known to be a Wikimedia wiki, or possibly any MediaWiki wiki, then go to a suitable search url instead of the url for a pagename at the wiki. There could also be a feature where meta:Interwiki map can specify two target url's, one to use for wikilinks and one for the search box. Then searches could also work at some non-MediaWiki wikis. Pedantic note: The interwiki map includes some entries which are designed to always perform a search, also as wikilinks like google:foo. PrimeHunter (talk) 12:08, 15 May 2015 (UTC)
I just tried the google trick above on reference 19 of Plummer v. State and couldn't get it to work, so I didn't save the edit. Would someone like to take a shot at it so I can look at the syntax you used? --Guy Macon (talk) 00:21, 16 May 2015 (UTC)
Not exactly. Try putting quotation marks around the text in Google's search box and see how many hits you get. Then replace the underscores with spaces and see how many you get. --Guy Macon (talk) 00:51, 16 May 2015 (UTC)
A Google search performed by myself cannot be guaranteed to return the same result as one performed by yourself; hence, a Google search results page fails WP:V and so cannot be used as the linked source in a reference. --Redrose64 (talk) 04:52, 16 May 2015 (UTC)
The technical section of the village pump is used to discuss technical issues about Wikipedia, and thus your comment is off-topic. If you post the same objection on the reliable sources noticeboard (which I monitor) I will be glad to have a discussion there as to why I believe you are mistaken, but i will not discuss policy questions here, and I advise others here to do likewise.
Getting back to the topic at hand, to search Google for a phrase you should enclose the phrase in quotation marks and separate the words with spaces. I was unable to get that to work using the google.* method when I tried it on Plummer v. State. --Guy Macon (talk) 05:53, 16 May 2015 (UTC)
Solved. That solved the syntax question. A search with plus signs gives the same result as a search with spaces. Thanks! --Guy Macon (talk) 17:16, 16 May 2015 (UTC)
Problem with user contributions
Latest comment: 9 years ago10 comments4 people in discussion
Take a look at [2] and [3]. Jeannierebecca thought she was responding in the correct place but her response ended up two questions above. I put the response where she obviously intended it. This happens to me a lot on The Teahouse as a new question might have gotten asked since I started reading, but when I see something to respond to, I end up responding to the wrong question. But her contributions indicate she responded in the right section, which shouldn't have happened.— Vchimpanzee • talk • contributions • 20:24, 15 May 2015 (UTC)
I'm sure there are already have a dozen different tickets for this on Phab. It's caused by (a) section(s) above being added or removed while the person has an open edit window. This usually happens on the Teahouse because it is top posting, but it happens elsewhere due to the way things are archived. The Teahouse has a unique advantage to solving the issue in "most" cases for people that use the "Join this Discussion" JavaScript (part of the "Ask a Question default on gadget" in that the JavaScript "could" do a sanity check and ask the API which section has a header matching the one that it should be in (the only issue with this is in cases where there are multiple sections with the same header, in which case it will be difficult to discern which one to put it in). — {{U|Technical 13}} (e • t • c)21:26, 15 May 2015 (UTC)
I find that I end up trying to post in the wrong section if the number of sections changed after I started reading. However, the contributions in this case show the incorrect section, which sounds like a bug to me.— Vchimpanzee • talk • contributions • 21:38, 15 May 2015 (UTC)
That is the correct threshold whether you use the default gadget or not. The [edit] link for each section is coded in a way that specifies a specific section number (which causes another whole set of issues if there are transclusion on the page that add sections) and from when you start reading until you click "Save page" or add my comment, you are locked into that section. There is nothing that can be done for those not using the gadget (without a massive change in the way the core software works) but those using the script have an advantage that the script could check "just" before submitting and have a lower chance of hitting an (edit conflict) or posting in the wrong spot. — {{U|Technical 13}} (e • t • c)21:57, 15 May 2015 (UTC)
I haven't experienced this problem. I suspect the gadget causes it. The post is about [4]. Special:Preferences#mw-prefsection-gadgets has "Ask a question" feature for the Teahouse project. I don't use the gadget but maybe the user did. The gadget adds a "Join this discussion" link next to the section edit links, and makes its own little edit box and a button to save the edit. I don't know how the gadget works but maybe it registers the section number and heading when you click "Join this discussion", and later edits that section number when you click "Add my response", but uses the originally registered section heading in the edit summary field. In the example another user added three sections at the top [5] two minutes before the post which was inserted three section earlier than it should. This is consistent with my theory. I think the normal section edit link would either merge a new post in the correct place or give an edit conflict. PrimeHunter (talk) 21:43, 15 May 2015 (UTC)
It's been known to put posts in the wrong section as well as cause (edit conflict)s because I've had it happen on other pages. It's also part of the reason for "no diff" edit conflicts that people save through not knowing there is an issue and something changed. — {{U|Technical 13}} (e • t • c)21:57, 15 May 2015 (UTC)
I have only had the issue when using the gadget on the Teahouse, it places it in whatever section number it was when you clicked on it, not when you submit it. I don't know enough coding outside of C# to fix that (hell I don't know enough coding in C# either) EoRdE6(Come Talk to Me!)22:11, 15 May 2015 (UTC)
Also going to add that an easy way to tell if it was done by the gadget, the gadget uses a default edit summary of "response" lowercase, nothing else, and others normally use "re or reply" EoRdE6(Come Talk to Me!)22:13, 15 May 2015 (UTC)
I did all this because I had already created a fairly complex set of sandbox pages and sub-pages at User:Dodger67/Sandbox. Recently I have noticed that the previously suppressed default menu link to User:Dodger67/sandbox has reappeared in the top menu, the Gadget option has dissapeared and to cap it all the link is not following the redirect! I have a "grammar-nazi" dislike of page titles that start with a lower case letter so I'm rather reluctant to move my whole current Sandbox setup to "sandbox". How do I get rid of the redundant and wrong top menu item? Roger (Dodger67) (talk) 22:25, 15 May 2015 (UTC)
@Edokter: Sady untrue. A. Doesnt matter what line for redirect, more importantly B. the sandbox link adds &redirect=no if your sandbox is a redirect. Try it by putting a redirect at the very top of your sandbox, it won't follow it. EoRdE6(Come Talk to Me!)00:10, 16 May 2015 (UTC)
Then why don't I have '&redirect=no' in my sandbox link? (edit) Never mind, probably didn't update the page. -- [[User:Edokter]] {{talk}}06:40, 16 May 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
Is there an edit counter which tells me how many bytes on average I add in my 20,000 odd contributions? Thx in advance, AshLin (talk) 00:55, 16 May 2015 (UTC)
This raises a good question, why does X!'s tools display Ø change per page (bytes): extended on pretty much everyone's contributions... EoRdE6(Come Talk to Me!)03:20, 18 May 2015 (UTC)
Looking for answer and help!!!!
Latest comment: 9 years ago5 comments2 people in discussion
Please someone help me in this problm......
In this portal recently I add a template named 10 numbered subpages in Selected articl sectione. After adding this template, why that articles are not showing in the portal page.
Pls help me out.--Iqsrb722 (talk) 13:27, 16 May 2015 (UTC)
I've little experience with portals, but I've copied the way a working portal did things and at least its displaying articles and pictures. You, or someone, else will have to check what I've done and tweak the settings. - X201 (talk) 13:48, 16 May 2015 (UTC)
Latest comment: 9 years ago2 comments2 people in discussion
Would someone be kind and explain to me if external search engines such as google deal with redirects? Do they index them? Or is redirecting just a tool to facilitate intra-wiki searches?
If there is a site/article etc. where I could explore more about how google indexes articles on wikipedia, please point me there. Thank you in advance! — Preceding unsigned comment added by Wesalius (talk • contribs) 07:34, 17 May 2015 (UTC)
Google has a very general article about how their search works here, but it doesn't mention Wikipedia explicitly, and from this search it doesn't seem that an article specific to Wikipedia exists. Google's spidering algorithms are proprietary, and they have no obligation to reveal how they work to anybody. That said, a Google engineer might give you a better answer if you ask a question at Google's support site. If you hear back from them, please let us know. — Mr. Stradivarius♪ talk ♪12:09, 17 May 2015 (UTC)
File with empty name?
Latest comment: 9 years ago2 comments2 people in discussion
Check this. The query should list the oldest files on Wikipedia. It finds numerous files from 2002 which could easily be some of the oldest files. However, it also finds a file with empty name:
What is this? All files need a name, but this one doesn't have a name. Also check the timestamp: 2015-05-17T11:50:13Z was a few minutes ago when I looked at the query output in my browser. --Stefan2 (talk) 11:55, 17 May 2015 (UTC)
There is, somehow, an entry in the "image" table in the database with all empty fields. An empty timestamp is equivalent to "now" in MediaWiki's timestamp-reformatting function. Anomie⚔10:33, 18 May 2015 (UTC)
Clock missing
Latest comment: 9 years ago17 comments12 people in discussion
What happened to the clock that I had at the top right of my screen when on Wikipedia? It's just disappeared today, without explanation. Trying to recover it, I disabled the gadget (Special:Preferences/Gadgets, "Add a clock in the personal toolbar"), saved, enabled it, and saved, but it's not back yet. I wonder if it's somehow browser-related: I'm missing it in my default browser, IE11, but when I log in with Firefox 37, it's there. Nyttend (talk) 14:26, 17 May 2015 (UTC)
Sounds like you are encountering a JS error. With F12, you can bring up the developer console, and with Ctrl-2, you can go to the developer console, which should show errors in red. —TheDJ (talk • contribs) 11:53, 18 May 2015 (UTC)
@Nyttend and TheDJ: The clock is a gadget, specifically "(S) Add a clock to the personal toolbar that displays the current time in UTC and provides a link to purge the current page (documentation)" Over the last few days, several other gadgets - some being Javascript and some pure CSS - have failed to load properly (or at all) for me. These include:
(D)CharInsert: add a toolbar under the edit window for quickly inserting wiki markup and special characters (troubles?)
Display diffs with the old yellow-and-green colors and design
Move section [edit] links to the right side of the screen
there may have been others. A hard refresh of the page usually brings the gadget in properly, so I suspect slow servers. --Redrose64 (talk) 21:32, 18 May 2015 (UTC)
Another, when clicking the red "Notifications" number, I normally get it as a popup - just now, it went to Special:Notifications instead, which is the full-screen version. So I guess the popup is some Javascript that failed to be served. --Redrose64 (talk) 23:26, 18 May 2015 (UTC)
Adding my experience:
Sometimes my clock is there, sometimes it takes a refresh to see it.
Yesterday, my notification alert would say I had messages, but the left-hand link was blanked out and wouldn't let me look at them.
Several hours earlier today, while doing a Save on a short talk page, it took about 5 minutes to complete. For about half an hour after that, it took that long for a page to load when I pulled it up, and any Save took that long. OK now. But editing has been quirky these last few days. — Maile (talk) 21:42, 18 May 2015 (UTC)
Should've checked here earlier; I've been having similar problems with intermittent extremely slow load times for the last couple of days. I wasn't able to trace it to any specific user script, but turning javascript off entirely fixes the issue, as does disabling all my user scripts, gadgets, and twinkle but leaving the browser settings alone. I've noticed this only in Firefox, though - Chrome works just fine, but Firefox on two different computers (Mac and Linux) will frequently fail. Opabinia regalis (talk) 02:56, 19 May 2015 (UTC)
Today, it seems that the Twinkle menu is sporadically not loading, or at least is not visible. It's happening in Firefox and Chrome on two different computers. Is anyone else having this issue or does anyone know what the cause might be?- MrX13:03, 18 May 2015 (UTC)
I have or had an issue with gadgets not loading at random since today morning; though I just restarted my browser (Chrome, kept it open for way too long, had to restart to install update). Gadgets seem to be working now, I'll post again if they're not. — Jeraphine Gryphon(talk)13:11, 18 May 2015 (UTC)
Twinkle, the Clock gadget that when clicked purges a page, and the Syntax highlighter (maybe others) were affected here. Popups remained OK though. And a few minutes ago, on the "Action complete" page that appears after a revert, the user talk page link "(talk|block)" became capitalised "(Talk|block)" and this apparently broke Twinkle's automatic article-name entry into the form for issuing warnings - it's OK again now. Someone's fiddling about behind the scenes, I think. Please stop it! —SMALLJIM16:27, 18 May 2015 (UTC)
New problem, almost certainly a duff server as above: attempting to follow the URL of a diff such as this one, or use an "[edit]" link on this page, sometimes returns a blank page containing only the line
Exception encountered, of type "BadMethodCallException"
I just got that, also, after dropping a location map into an infobox and doing "Preview". Tried it several times and got that message and gave up.— Maile (talk) 23:29, 18 May 2015 (UTC)
The Wikimedia Foundation Board of Trustees is the ultimate governing authority of the Wikimedia Foundation, a 501(c)(3) non-profit organization registered in the United States. The Wikimedia Foundation manages many diverse projects such as Wikipedia and Commons.
The voting phase lasts from 00:00 UTC May 17 to 23:59 UTC May 31. Click here to vote. More information on the candidates and the elections can be found on the 2015 Board election page on Meta-Wiki.
Now Wikipedia has another glitch where the previous message I wrote in this section doesn't display the way it's supposed to according to the source code I typed? Blackbombchu (talk) 21:35, 11 May 2015 (UTC)
Actually, he's complaining about the section after the # which would need to be dot encoded (anchor encoded). — {{U|Technical 13}} (e • t • c)21:52, 11 May 2015 (UTC)
Oh yes. Different parts of a URL have different encoding mechanisms. But the main point is that spaces aren't allowed anywhere in a URL. --Redrose64 (talk) 22:34, 11 May 2015 (UTC)
Because it doesn't know when you want the space vs. when you don't want the space. Unlike in a browser url bar, where it can make assumptions, about what users would input there. —TheDJ (talk • contribs) 23:24, 11 May 2015 (UTC)
I copied this from the archive because I noticed it while looking at something else and although the OP didn't say anything, I'm not sure their question was actually answered. As I understand it, the OP is referring solely to the browser URL bar. Most if not all URL bars will convert spaces and other special characters as appropriate, in the URL proper, which are inputed in to the bar. However many, if not all won't general convert anything after the hash, i.e. that is part of the Fragment identifier. I'm not sure the reason, but it may be because I'm not sure there is actually a standard as to how to convert "white space" in the fragment identifier. Underscore may be the most common, but some webpages may use dash (-) or just remove the space. It perhaps helps if the OP understands the fragment identifier depends on the page. It doesn't have to actually have to be the same as what it points to. Nil Einne (talk) 19:35, 17 May 2015 (UTC)
Actually some quick testing shows Firefox, IE11 and Chrome all work fine with whitespace in the fragment identifier, i.e. they handle it as is. So they will go to the relevant id if it's what the id attribute is (with white spaces). Having white space in the id violates the HTML standards [7] but as I said earlier, I don't think there's actually any defined way to handle spaces other than "don't do it" so the behaviour of browsers is somewhat understandable. Nil Einne (talk) 19:49, 17 May 2015 (UTC)
To your second q: double square brackets are for internal links. For external links, use single square brackets. --Redrose64 (talk) 21:45, 11 May 2015 (UTC)
Constantly being logged out
Latest comment: 9 years ago6 comments4 people in discussion
I am constantly being logged out and given the message "You are centrally logged in as k6ka, please reload the page to complete your request." I am logged back in when I reload the page, but it's starting to become a pain when it does this, since my Javascript gadgets don't load during this time, and I'm not sure if saving an edit in this state will result in "editing whilst logged out." Anyone else having this? --I am k6kaTalk to me!See what I have done22:59, 17 May 2015 (UTC)
I got that three or four times yesterday. As mentioned, a reload fixes it straightaway. I suppose we should be grateful that it's obvious when you're logged-out these days, because of the default font! Black Kite (talk)13:46, 18 May 2015 (UTC)
Are you sure the email address you have set in your account preferences is the same as the one you are using for the test? If you have ever used other email addresses then it is worth checking those. QuiteUnusual (talk) 12:44, 18 May 2015 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
The new version of MediaWiki has been on test wikis and MediaWiki.org since May 13. It will be on non-Wikipedia wikis from May 19. It will be on all Wikipedias from May 20 (calendar).
References are now always in the right order. Also, the reference list now only shows references used on the page. [8]
You can no longer create an account with a colon ':' in it. If you already have one, it still works. [9]
The toolbar in VisualEditor now looks different. It is easier to see the icons. [10]
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on May 21 at 15:00 (UTC). See how to join.
@Anythingyouwant and Redrose64: Wikicode can handle the haversine formula: {{#expr:2*{{{r|6373}}}*asin(((sin((({{{lat2|0}}}*pi/180) - ({{{lat1|0}}}*pi/180))/2))^2 + (cos({{{lat1|0}}}*pi/180)*cos({{{lat2|0}}}*pi/180)*((sin((({{{long2|0}}}*pi/180) - ({{{long1|0}}}*pi/180))/2))^2)))^0.5)}} gives you the great-circle distance in km given {{{lat1}}}, {{{long1}}}, {{{lat2}}}, and {{{long2}}}. Of course, this doesn't take into account that the earth is elliptical. I created Template:Great circle distance as an example. --Ahecht (TALK PAGE) 23:09, 19 May 2015 (UTC)
Community tech project ideas
Latest comment: 9 years ago1 comment1 person in discussion
Latest comment: 9 years ago4 comments2 people in discussion
When I look at Magna Carta on the official Wikipedia app on my Android tablet, over the lead image display is shown the title and, underneath, "Angevin charter". Now, that is very erudite because the article doesn't seem to say that anywhere in such words. Where is this text coming from? Thincat (talk) 16:17, 19 May 2015 (UTC)
Ah ha! Thank you. I'll report this to the article's talk page because, while you were discovering that, I was discovering there was editwarring over this exact text at the beginning of last month! Thincat (talk) 17:16, 19 May 2015 (UTC)
I'll continue this for now as a technical matter but I've re-read the recent Policy and Proposal discussions concerning Wikidata and maybe things need discussing there too. In this case the prominent text had been by consensus deliberately removed from the article. Is there a way in which people on the main site can opt in to seeing such "imported" text like it can be done with Persondata? Or, generally, should such "imported" text appear in all versions of WP to improve the general usability and to help checking? In this case the text was not on Wikidata maliciously but the article's editors clearly did not think to check Wikidata. Even for people aware of Wikidata they may not realise how it is becoming more used. Thincat (talk) 18:09, 19 May 2015 (UTC)
Duplicate thanks
Latest comment: 9 years ago4 comments2 people in discussion
@Redrose64: I see, but what makes this unique is I made a single edit, and was thanked three times over a period of over 17 hours, much longer than it would be for just clicking it multiple times. @Rms125a@hotmail.com: Do you recall when and how many times you thanked me? EoRdE6(Come Talk to Me!)21:49, 19 May 2015 (UTC)
It's not unique. In November 2013, I was thanked on a daily basis by one person for one edit; after three thanks in three days I sent them a user talk: message asking them to stop. --Redrose64 (talk) 22:18, 19 May 2015 (UTC)
Can't edit with iCab Mobile
Latest comment: 9 years ago2 comments2 people in discussion
If I try to edit Wikipedia using ICab Mobile, I invariably receive this message after hitting "Save page": "Spam protection filter The text you wanted to save was blocked by the spam filter. This is probably caused by a link to a blacklisted external site." Since no links are involved I have absolutely no idea why this is happening. Consequently I have to resort to using one of the lesser mobile browsers – an aggravation in the extreme. [navigator.userAgent Mozilla/5.0 (iPad; CPU OS 8_3 like Mac OS X) AppleWebKit/600.1.4 (KHTML, like Gecko) Version/8.0 Mobile/12F69 Safari/600.1.4] Any help with this would be greatly appreciated. Modal Jig (talk) 19:42, 20 May 2015 (UTC)
Really difficult to help with this, since it's a very minor market share paid browser. I did see if i could find a problem with the iCab desktop browser. I had some initial problems with logging in, but after configuring it accept cookies from multiple domains, I've been able to make edits both as anonymous and logged in user. —TheDJ (talk • contribs) 21:00, 20 May 2015 (UTC)
Reminder about event logging test
Latest comment: 9 years ago1 comment1 person in discussion
Just a reminder that the first step in a long-planned A/B test for VisualEditor is scheduled for tomorrow. I think it's supposed to start around 15:00 UTC (full disclaimers apply) and run for 24 hours. The goal tomorrow is to find out whether the recent fixes to the logging software actually fixed stuff (or not, or broke some new stuff—you know how it goes).
This test shouldn't affect anyone who's reading this message. But if you see weird stuff, then feel free to {{ping}} me, leave a note on my talk page, drop by WP:VEF, find your favorite dev on IRC, or whatever's convenient for you. Just please let me know if you see anything odd. Whatamidoing (WMF) (talk) 21:34, 20 May 2015 (UTC)
Latest comment: 9 years ago5 comments3 people in discussion
There seems to be something wrong with the transclusion and the templates. I added "relist" and the signature appears as 4 tildes instead of my user name, too. Could somebody have a look? Kraxler (talk) 15:14, 19 May 2015 (UTC)
Hi. Just uploaded a few images to Commons and was surprised to find that they are not being recognised here, they just get redlinked, even though the file clearly exists under that name on Commons. Anyone know if there's a known issue at the moment? (Someexampleshere, and the full external links to their pages on Commons here: [16][17][18]. Ðiliff «» (Talk)21:40, 19 May 2015 (UTC)
Well, evidently that works but I guess it's not good practice to use interwiki links in articles. I was hoping that if there is a problem, it could be flagged and investigated rather than bypassed... Ðiliff «» (Talk)22:01, 19 May 2015 (UTC)
I don't think RedRose was proposing to only link to them, but to see if interwiki links work (maybe?). Anyhow, I tried purging the pages both on Commons and enwiki to no avail. I also created and subsequently deleted File:Saint-Séverin Sanctuary, Paris, France - Diliff.jpg, but it did nothing. What's curious is that popups shows the Commons page if I hover over the enwiki links. Killiondude (talk) 22:09, 19 May 2015 (UTC)
Came here to report this as well. Just uploaded some images to Commons that aren't showing up on Wikipedia. Examples: [19] vs [20]. --Golbez (talk) 22:13, 19 May 2015 (UTC)
Hmm. So it seems that the problem isn't specfically the English Wikipedia, it's perhaps the Commons side. I'll see if I can find if its been flagged over there. Ðiliff «» (Talk)22:23, 19 May 2015 (UTC)
All of the filenames have both e-acute (é) and a comma in them. Have you experienced similar problems with filenames that don't have that diacritic, or which don't have a comma? --Redrose64 (talk) 23:12, 19 May 2015 (UTC)
Well, all of those images of mine do... But User:Golbez's images don't have this and he's having the same problem. But I just uploaded a new file to Commons a short while ago and I'm not having the same problem with this one. So I'm confused. The problem hasn't been resolved for the previous files though, but yes I guess it could have something to do with the é and/or comma, although I did (in January this year) upload another file with the e-acute and a comma in it and I never experienced any problem at the time. Ðiliff «» (Talk)23:23, 19 May 2015 (UTC)
Update. Seems to be resolved, certainly for the images I uploaded, and all others that I was aware of that were experiencing problems. Ðiliff «» (Talk)12:45, 21 May 2015 (UTC)
You would need to upload each of them separately as a different image name (in the case of the French wiki example, appending the date that the map is meant to reflect). Fortunately as they are uploaded on commons, you can do this yourself - you take the revision you want, upload it as a new image with a different name and make sure to keep the same licensing and information as the original (plus I believe the "extracted from" template). You should also encourage the uploaders there to do not keep replacing the image but creating new ones as time progresses. --MASEM (t) 14:15, 20 May 2015 (UTC)
Latest comment: 9 years ago3 comments3 people in discussion
I used to use this report that listed people who are marked as living on enwiki but as dead on others. The report is no longer updated as it used Toolserver. Would it be possible to create a new version of this report ? Only the first two sections of the report are required. --Racklever (talk) 08:44, 21 May 2015 (UTC)
Latest comment: 9 years ago4 comments4 people in discussion
Recently an editor’s account User:AHLM13 was hacked, most probably by a banned user. The hacker caused much disruption using proxy/VPN servers, created many fake accounts, and pretended to be AHLM13 for long. The whole thing got very messy. The hacker has also targeted others including me, User:Mar4d, User:Lukeno94, User:CosmicEmperor as we have received threats. A discussion was going on here User_talk:Mar4d#My_account_is_hacked. So, password security has become a concern.
What are actually the ways through which a hacker can discover any password? Dose he randomly try to type a password, or anything more sophisticated? Apart from setting a strong password and using Wikipedia's secure connection, what others technical measures should we follow to protect the password from being hacked? I have a curiosity to know how strong Wikipedia's security system is? -AsceticRosé16:00, 21 May 2015 (UTC)
I do not know about security level, but this is what every user should do. This is very secure if used properly, because effectively you have two passwords one of which you never use (until you get hacked) and thus never type and thus whole bunch of hacking methods get useless (keyloggers, traffic sniffer, social engineergin etc.).
@AsceticRose:Sometimes insecure software is at fault, doing something like storing passwords in plaintext, I'm not sure how secure the MediaWiki software is but I assume it's fairly secure or hacks would be happening a lot more often. Most of the time when a hacker gains access to an account, the user is "at fault", that is, they could have easily done something to prevent the attack, not that they should necessarily be blamed for it. Bruteforce attacks are the ones commonly used by unskilled hackers, the kind that would go after low profile targets like random Wikipedia users, and they're basically just using a program to randomly type passwords, or making a guess based on information they know, like the target's username. MediaWiki probably does some things to make bruteforce attacks harder but there is no reason to rely on it, the attacks can be avoided easily by using good passwords. In general, passwords shorter than 11 characters are not secure and as technology advances that number increases quickly. A strong password should have at least one lowercase letter, uppercase letter, number and special character. To avoid being bruteforced, passwords should be random, not having a pattern recognizable to anyone, including yourself. Dogz$eemN1ce is a bad password, despite meeting the character requirements, 4^7%:_e}G!5%w:J is a good password, or at least it was until I posted it as an example. Man-in-the-middle attacks are another common attack, though they require the attacker to be somewhat close to you and don't always require the attacker to steal your password. They can be avoided, to some extent, by using a secure connection (Https instead of Http) and by using a different account when logging in on public connections (public being any connection where you don't completely trust everyone who can connect to it) or computers. There are other possibilities, but they are more complex and as they get more complex they get harder to avoid. As a final note, remember that your account can't be any more secure than the recovery options, that means that if your email password is password12, it doesn't matter what your Wikipedia account's password is. I recently proposed allowing multi-factor authentication here which would make attacks that rely on knowing your password harder against users who enable it. In reference to the discussion you linked, I only skimmed it but remember that "I was hacked" is a "great" excuse to try to get unblocked and that people tend to see things they're looking for, if you think someone's trying to hack you then you might notice things you otherwise wouldn't have and consider them an indication that you're being hacked when in fact they may be nothing that hasn't happened before. I'm not saying ignore anything that makes you think you're being hacked, change your password as often as you like, it's good for security, just be careful not to allow blocked users to lie about being hacked to get unblocked. Again, I haven't read enough to know if that's the case here, it's just a general thing you should keep in mind. PHANTOMTECH (talk) 01:06, 22 May 2015 (UTC)
In the early days, the first sockpuppets were easily exposed. Encrypted passwords were compared, and WP:DUCKS with the same encrypted passwords were blown away. Choor monster (talk) 10:38, 22 May 2015 (UTC)
Tool to scan category tree for article title matching regex?
Latest comment: 9 years ago5 comments4 people in discussion
Does anyone know if there is anything on the toolserver which will scan a category tree for an article title matching a regex?
I get this done the long way round by using Catscan2 to list all the pages. and then save the list to offline. But it would be handy if there was an online tool to do this simple job. --BrownHairedGirl(talk) • (contribs) 10:39, 21 May 2015 (UTC)
Latest comment: 9 years ago3 comments2 people in discussion
Has something gone wrong with Template:Refn? Previously, notes like the following ([a]) were working perfectly fine. Today, while editing i noticed the following was displaying post-editing: "Template loop detected: Template:Refn". The template article also shows said error, without pointing out how to correct it. Any thoughts?EnigmaMcmxc (talk) 23:42, 21 May 2015 (UTC)
^The large intake of men into the army had considerably increased the infantry arm. The reforms intended to address this, with many of the newly raised battalions being "converted to other arms, particularly artillery and armour".[1] In addition to this, historian F.W. Perry comments, there was considerable pressure "to increase the armoured component [of the army] and build up raiding and special forces". These pressures, and the re-balancing of the military, resulted in seven of the nine County Divisions being disbanded and only two being reformed as infantry divisions[2]
^Perry 1988, pp. 53–54. sfn error: no target: CITEREFPerry1988 (help)
^Perry 1988, p. 65. sfn error: no target: CITEREFPerry1988 (help)
Tool to scan category tree for article title matching regex?
Latest comment: 9 years ago5 comments4 people in discussion
Does anyone know if there is anything on the toolserver which will scan a category tree for an article title matching a regex?
I get this done the long way round by using Catscan2 to list all the pages. and then save the list to offline. But it would be handy if there was an online tool to do this simple job. --BrownHairedGirl(talk) • (contribs) 10:39, 21 May 2015 (UTC)
Latest comment: 9 years ago3 comments2 people in discussion
Has something gone wrong with Template:Refn? Previously, notes like the following ([a]) were working perfectly fine. Today, while editing i noticed the following was displaying post-editing: "Template loop detected: Template:Refn". The template article also shows said error, without pointing out how to correct it. Any thoughts?EnigmaMcmxc (talk) 23:42, 21 May 2015 (UTC)
^The large intake of men into the army had considerably increased the infantry arm. The reforms intended to address this, with many of the newly raised battalions being "converted to other arms, particularly artillery and armour".[1] In addition to this, historian F.W. Perry comments, there was considerable pressure "to increase the armoured component [of the army] and build up raiding and special forces". These pressures, and the re-balancing of the military, resulted in seven of the nine County Divisions being disbanded and only two being reformed as infantry divisions[2]
^Perry 1988, pp. 53–54. sfn error: no target: CITEREFPerry1988 (help)
^Perry 1988, p. 65. sfn error: no target: CITEREFPerry1988 (help)
No, this is intentional, and has been discussed before. It means that they have registered but have made no edits. --Redrose64 (talk) 19:43, 24 May 2015 (UTC)
Rendering an article into a PDF- skip the 2 column, or leave it as an option, OR expose the raw .tex file
Latest comment: 9 years ago2 comments2 people in discussion
The ability to create a PDF from a wikipedia entry is very powerful- but
the execution is
flawed when there are large math mode tex or latex formula sections. A
lot better lob is
done in the option to prepare the page for printing. This is because the
pdf option makes
two columns of output, and large equations don't fit and are cut off.
Makes the whole thing
not usable. Turn off that *2 columns* pdf option and you have a lot
better shot at making
this powerful feature really work. Or give the user the option maybe.
Even making the option
to render the entry suitable for printing could be used as a template
for production of a
simplified pdf.
If you really want to make something neat and useful, how about an
option to produce latex
output as a straight text file. Now that would be useful!!!!
Thanks for a great resource.
Here is an example of the problem I am talking about.
http://en.wiki.x.io/wiki/Bicubic_interpolation
is an example. Make the pdf, and look at the horrible mess that is
produced simply because the thing is made for 2 columns. Just give me
the .tex or .latex and most windows people that want this stuff have
miktex and Linux people have pdflatex — Preceding unsigned comment added by 71.99.54.129 (talk) 22:12, 23 May 2015 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Problems
VisualEditor was broken for 30 minutes on Tuesday. The problem was due to a tool it uses. [23]
The new version of MediaWiki has been on test wikis and MediaWiki.org since May 20. It will be on non-Wikipedia wikis from May 26. It will be on all Wikipedias from May 27 (calendar).
Meetings
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on May 28 at 18:00 (UTC). See how to join.
Latest comment: 9 years ago12 comments5 people in discussion
The articles on Eurovision Song Contest, such as Eurovision Song Contest 2015, appear to regularly use some kind of DISPLAYTITLE, so that the page titles appear to be italicized while this is not required by WP:MOSTITLE. I suspect it could be somewhere in the Infobox Song Contest, but can't find where exactly. Adding | Italic title=no doesn't solve the issue. Could someone have a look? Brandmeistertalk15:55, 24 May 2015 (UTC)
The original query was about the article title, which was incorrectly being italicised, and |Italic title=no fixed that. There is no problem with the album title which should be italicised and is, per WP:ITALIC.--JohnBlackburnewordsdeeds22:56, 25 May 2015 (UTC)
(edit conflict) The original post explicitly said "the page titles appear to be italicized". Album titles are always italicized in an infobox about the album. {{Infobox album}} has no feature to avoid this. Italic title parameters in infoboxes are about whether to omit the normal automatic italics of the page title. PrimeHunter (talk) 23:00, 25 May 2015 (UTC)
@PrimeHunter:, yes I am aware that the OP is about article title name. Per diffs like this, Brandmeister added |Italic title=no to the {{Infobox album}}, which is located towards the bottom of an article. So how could adding that parameter bear an impact on the article title itself? That is what has confused me. Wes Mouse | T@lk23:09, 25 May 2015 (UTC)
Oh! Thanks for clearing that confusion. I thought when BM was adding that italic no thing, that he was trying to remove the italics of the album title, when in fact he was removing the italics from the article title. Makes sense now. Thank you! Wes Mouse | T@lk23:22, 25 May 2015 (UTC)
Odd item at the edit counter page
Latest comment: 9 years ago5 comments3 people in discussion
I have just noticed (though it may have been this way for some time) that the word describing edits to article space at the "wmflabs.org/xtools" is now in a pictograph language. I am not sure if it is Chinese, Korean or Thai. When I checked the "wmflabs.org/supercount" it still reads "Main" as it always has. Now I know this isn't a big deal in the grand scheme of things, I was just curious if anyone knows how or why this happened. Also if there is a better place to ask this just let me know and I will move this thread there. MarnetteD|Talk04:07, 26 May 2015 (UTC)
I'm in Denmark and see Danish by default at https://tools.wmflabs.org/xtools/pages/. There is a long list of clickable language codes which includes en for English, but if I select en and make a search then the interface language sometimes changes to French. I can make it English with &uselang=en in the url. PrimeHunter (talk) 13:05, 26 May 2015 (UTC)
Thanks for your replies. I chose "en" and "en.gb" in the language codes but neither had any effect on the item in question. It makes me wonder if it is a leftover from an April Fools Day :-) MarnetteD|Talk13:40, 26 May 2015 (UTC)
{{Infobox rail accident}} is adding pages to Category:Pages with script errors but I cannot see why. The one at right is one of those with problems, copied from Thirsk rail crash (1892). But it is very ordinary, containing no other templates. It's none of the fields as the same error occurs if I remove all of them. It seems to be map related as if I add a map and coords the error goes away. But {{Infobox rail accident}} should work without a map, looking at its source, and has been working for a long time in those articles, while none of the templates (this one, {{Location Map}} or its module) have changed recently.--JohnBlackburnewordsdeeds02:24, 26 May 2015 (UTC)
Jackmcbarn recently updated Module:Infobox and my guess is that something there is responsible, although an invisible Lua error is new to me. As JohnBlackburne said, previewing "{{Infobox rail accident}}" in a sandbox is enough to show the sandbox as being in the error category. Possibly something in the new module version attempts an operation which may fail, and which somehow is suppressed as an expected error, but which activates the error category. Johnuniq (talk) 02:51, 26 May 2015 (UTC)
Something similar is going on with {{Infobox historic site}}. As they are more geographic and touristy I just went ahead and added maps to them, but oddly two of them, Fort Whoop-Up and Mother Armenia, adding maps fixed the error, i.e. removed them from the category, but the map did not display. Even though I could see the relevant Location map template in the list of templates it was using, so I had the name right and it was using them. That infobox was recently merged by Pigsonthewing though I can't see anything the merge that would cause this.--JohnBlackburnewordsdeeds02:55, 26 May 2015 (UTC)
It looks like this instance of the problem has been tracked down and fixed. If it becomes more widespread, I'm not opposed to reverting my changes until a better solution can be found. Jackmcbarn (talk) 03:21, 26 May 2015 (UTC)
That seems to have fixed it in the rail accident infobox, and hopefully those are the only cases as others would have popped up in the category at the same time I think. Despite the Luafication of the underlying template there's still a lot of this fairly complex parser function code in all these infoboxes, doing the same thing with slight variations, waiting to cause problems when something else changes in a module, or now also in the Wikidata. I don't know if there's any way it could be replaced with more Lua or at least merged so there weren't so many templates that can break.--JohnBlackburnewordsdeeds03:43, 26 May 2015 (UTC)
Latest comment: 9 years ago1 comment1 person in discussion
Quite often, I will be in a discussion on Facebook or elsewhere, and want to link to a Wikipedia article (or section). But having gone to the Wikipedia Android app to look up what I want, I cannot find a way to get a copy of the URL for the page or section, that I can paste there. There's an option "Share", which looks as if it will help; but "Share to Facebook" creates a new posting, somehow without a link that I can even cut; while "Share to Clipboard" puts the text on the clipboard, not the URL. Surely there must be a way to get the Wikipedia URL onto my clipboard?
(I don't even know away to run the browser version on my Android phone, because when I go to Wikipedia in the browser, it automatically redirects to the App. This is useful, but I thought there might be a workround). --ColinFine (talk) 16:53, 26 May 2015 (UTC)
Weird talk page notification
Latest comment: 9 years ago23 comments9 people in discussion
You didn't use "add section" - I can tell because this sub-section has a level 3 heading. New sections created using the "new section" tab always have level 2 headings. --Redrose64 (talk) 21:37, 21 May 2015 (UTC)
I did create this with "add section" which uses §ion=new, I just put a custom header in the edit box instead of using the header box (which caused me to have to click save twice since there was no edit summary). Redrose64, I hope this clears that up. — {{U|Technical 13}} (e • t • c)01:43, 22 May 2015 (UTC)
And I can confirm that the ===section heading=== was reported in the notification (on wiki; I don't use e-mail norifications for en.wp). Whatamidoing (WMF) (talk) 02:34, 22 May 2015 (UTC)
Maybe "I wonder if..." was a polite version of: If you want to know why an edit caused a message then why on Earth are you concealing which edit it was and making us waste time speculating? After a bunch of experiments at User talk:PrimeHunter2 I was able to generate a similar "footer" mail with [28] which is unclear about which section is being edited, so perhaps the secret edit was [29]. #footer is an anchor automatically inserted at the bottom of all pages. "footer" is not linked in the notification mail so it doesn't make much sense there, but the web notification said "footer" which makes a little more sense without being ideal. PrimeHunter (talk) 22:09, 21 May 2015 (UTC)
@Timtrent: Oh my God, what will it take to make you reveal theedit? I'm not wasting more time on speculation and experiments while you continue to conceal the edit you want help about. There is no privacy reason to say "(username redacted)". We can all see who has edited your talk page. PrimeHunter (talk) 13:03, 22 May 2015 (UTC)
I am very confused, PrimeHunter about what I interpret as your strange attitude towards me that you display here. The username is unimportant. The actual edit is unimportant. I am reporting something that has happened multiple times with multiple editors. I have only now determined that your prior weird barb was directed at me. I redacted the editor's name because I had no interest in their being involved in this. Why? Because they are irrelevant to it. Go look at the edit history and "reveal" it, if reveal is the word that you choose. But please tell us why this is so important to you? Your attitude as displayed makes me wish I had not taken the trouble to report this. Your words make me think that you consider that I have done you a great and personal wrong. Would you like a puppy? Perhaps a kitten? Are you having a bad day? Did I kick your cat by accident? Do you need a cuddle? FiddleFaddle13:10, 22 May 2015 (UTC)
I'm annoyed by wasting a lot of time trying to reproduce the issue because you refuse to reveal what caused it. Some edits cause it. Some edits don't. I examined every recent edit to your talk page before starting with guesses and experiments. I find it odd that you insist it's "unimportant" which edits cause an issue you want help with. It's important to the people trying to help you. Count me out of them now. PrimeHunter (talk) 13:28, 22 May 2015 (UTC)
I had already. Put plain, it has been logged by Whatamidoing (WMF). I have no idea about, nor interest in, Echo, or whatever the internal notification is. I thought I might be doing the edifice a service by reporting a strangeness. Others seem to have seen it as well. Since it started in the recent past, it is caused by a change made in the recent past. I do not know when it changed, nor do I have a clue what changed, nor do I have any idea when it started. The earliest one I have seen is in this diff. If you had asked me to find that out, and done so plainly and simply then I would have been very sweet and told you. I simply scrolled down the special notifications page. I had to track that down.
Reading your mind about what you felt you wanted, something I think that turned out to be not really what you wanted at all, was quite hard. If you would like more information, first refine what it is you want, then tell people what it is and how to find it for you. Do not take out your frustrations on other people, PrimeHunter, It does you no service. FiddleFaddle13:55, 22 May 2015 (UTC)
I don't think it should require mind reading to understand that if somebody says "why on Earth are you concealing which edit it was and making us waste time speculating?", then they want to know which edit it was. And I think it should be common sense to give the edit in the first place, and especially to give it when you later write that a guess about it was wrong. PrimeHunter (talk) 14:25, 22 May 2015 (UTC)
@Timtrent: The notifications system is notoriously delicate, the discussion page at WT:Echo is filled with threads beginning "I should have been notified, but wasn't" or similar. When a problem is perceived, those with the ability to do something about it will usually find their task easier if they know as much as possible about the circumstances giving rise to the problem. If they know who made the edit, and how the edit was made, they may be able to replicate the conditions. It may turn out to be undesirable behaviour (a bug), or it may be a known consequence of those events, etc. --Redrose64 (talk) 14:48, 22 May 2015 (UTC)
Getting back on track to the underlying problem: This diff resulted in me getting a similar notification ("Phil Copperman left a message on your talk page in "footer"."). No other talk page notification I've received in the last few days has done that; they've all correctly specified what section they were in. This edit resulted in a correct notification, even though it had nothing but the automatically-generated section in the edit summary, so it looks like that isn't it. The only pattern I see is that this is the only diff from the last couple of days that both (a) had no edit summary, and (b) had no signature. --Floquenbeam (talk) 16:31, 22 May 2015 (UTC)
Here's another one recently that did the same: [31]. I note it didn't have an edit summary (except for the default section title), but it was signed. I'm going to stop listing them here when they happen, though, unless someone tells me it's useful to get more diffs. --Floquenbeam (talk) 18:40, 26 May 2015 (UTC)
+1 - I too have got "MediaWiki message delivery left a message on your talk page in "footer"".- Looking through my notifications I seem to have got this 6 times from the 4th May and it's only took this section for me to even notice . –Davey2010Talk14:30, 25 May 2015 (UTC)
Collapsible groups
Latest comment: 9 years ago2 comments2 people in discussion
The groups "Interaction", "Tools", "Print/export", and "Languages" on the left side of a page are no longer collapsible. GeoffreyT2000 (talk) 00:23, 27 May 2015 (UTC)
Latest comment: 9 years ago4 comments4 people in discussion
I produce the WP:5000 report weekly. The analysis is done on a personal machine and the report is updated/edited using a single API call that passes the entirety of the page content (750kb, or so). This week my report is unable to post because the server keeps kicking back HTTP 503 errors and rejecting the edit. When I copy-paste the raw wikitext and try to update the report manually in-browser I get the "our servers our experiencing a technical problem" error. It takes a long time to get that error screen, so I suspect it might be a server/parser timeout? I've tried multiple attempts across the last 24 hours. I've also split the wikitext into 2/4/8 pieces and tried to update the page piece-wise without luck. This is the first time a couple of years I've had this problem, so there hasn't been issue with ingesting ~750kb of content in the past. Thanks, West.andrew.g (talk) 15:13, 25 May 2015 (UTC)
Can anyone in WMF address this problem? I've come to rely on the WP:5000 and Top25 and was wondering what had happened this week. LizRead!Talk!21:21, 26 May 2015 (UTC)
As an interim solution, break it up into sections with a link to each section at the original page? Samsara07:27, 27 May 2015 (UTC)
I (TB) to have had problems uploading longer reports. Some tricks to try are:
Blanking the page in a separate edit before uploading the new version. A significant portion of the processing window that is 'timing out' is spent removing entries from the pagelinks table. Blanking the article prior to upload reduces the work being done in single operation.
Pre-expanding templates; use subst to expand any expensive templates out before the 'big upload'. Normally I cut and paste the offending section of a report into a temporary wiki page, save, edit and cut and paste the results back into my local copy.
If you are working from your own network, reduce your local MTU by 80 bytes. I kid you not.
And of course, schedule your upload for off-peak editing times.
Latest comment: 9 years ago1 comment1 person in discussion
The WMF has a two programmers that are going to be supporting development of community lead ideas (listed here).
We the community likely need to put together processes to identify which ideas have sufficient support from us before tech time is spent developing them. I am wondering if this notice board would the appropriate place for EnWP discussions of potential ideas to occur? Doc James (talk · contribs · email) 07:07, 27 May 2015 (UTC)
Watch page section?
Latest comment: 9 years ago2 comments2 people in discussion
Is there any way to have edits from only a specific section of a page show up in my Watchlist?
This would be especially useful, for example, in tracking particular discussions on this page without getting loads of Watchlist entries for every update to every discussion.
Latest comment: 9 years ago4 comments3 people in discussion
Hiya, anyone know if there's a way, perhaps through some Javascript add-on or something, to sort one's watchlist? My watchlist is jam-packed every day and I'm trying to figure out a more efficient way of tackling it. Being able to sort by title would allow me to see which articles have been edited the most, at which point I could click "Hist", compare all the new diffs to see what the net improvement was to the article, and then clear out the if everything's okay. In the current system, as I go up my watchlist, I might not notice right away that Article X had 10 edits, and I might wind up opening five windows each with a unique diff, rather than seeing up front that Article X had 10 edits since I was here last, and accessing them all at once.
Making watchlist use more efficient would save tons of volunteer time. For example:
If Wikipedia knows which diffs I've seen and not seen, couldn't a query be run to compare all unseen diffs for Article X, and if the result of those diffs is "no change", my watchlist is updated, and these edits are de-bolded? The current system requires me to pick an article in my watchlist, access the history, pick start and ending diffs, and compare—for each article in my watchlist. Very time consuming. If someone's been on a vandalism spree, but an admin finally locks up the article, I don't need to see that entire exchange.
You might want to check out the "Group changes by page in recent changes and watchlist" option in the user preferences (I think it's under Recen5 Changes). Should help with the specific problem you cited. Writ Keeper⚇♔16:40, 26 May 2015 (UTC)
I'm guessing/assuming that the problem is the equal sign in the name? If anyone familiar enough with the complicated templates to undertake the challenge to fix this? --B (talk) 15:03, 27 May 2015 (UTC)
@B: To be clear, by "a problem with the template grammar", Cryptic means that the problem is in the way that the template is being used, not in the template itself. All templates behave like this when a value containing an equals sign is passed through a positional parameter. --Redrose64 (talk) 15:25, 27 May 2015 (UTC)
I need help finding this page (probably in the MediaWiki namespace)
Latest comment: 9 years ago5 comments3 people in discussion
I'm trying to locate the page of box that displays when an editor tries to view a page that does not exist and/or is created protected. The box starts with the phrase "Wikipedia does not have an article with this exact name". Could I please be directed to this page? (I'm posting it here since it's most likely a page in the "MediaWiki:" namespace, and thus, technical.) Steel1943 (talk) 18:45, 27 May 2015 (UTC)
I don't know PHP or the implementation details but edits in the MediaWiki namespace have immediate effect on the interface. Many messages in the MediaWiki namespace can contain wiki markup including parser functions, templates and so on. There are also messages which can only contain plain text or have to use html. PrimeHunter (talk) 21:31, 27 May 2015 (UTC)
Video on the main page
Latest comment: 9 years ago3 comments2 people in discussion
I've heard that posting videos on the main page may cause trouble for mobile viewers. Is this true? The reason I ask is because we have a video scheduled for the June 8th TFA, and I don't want a repeat of the GIF incident earlier this month. — Chris Woodrich (talk) 07:30, 28 May 2015 (UTC)
This will not be a problem. They just get what you see here. So there will be no player embedded, just a link to the movie, and depending on what kind of device/browser they have, that link may be playable or not. The problem with GIF is that a GIF is automatically downloaded to a user's device, without user intervention. So if you embed a 20 MB animated GIF, then everyone visiting that page will download 20MB (we have that problem everywhere throughout the encyclopedia btw, it's just more annoying when it's on the main page). With videos, people get to choose wether or not they click that play link. —TheDJ (talk • contribs) 08:18, 28 May 2015 (UTC)
Click "View logs for this page" on the page history to see [32]. It was deleted and restored after your protection. I don't think a restore remembers the protection level from when the page was deleted. PrimeHunter (talk) 02:15, 28 May 2015 (UTC)
Latest comment: 9 years ago12 comments5 people in discussion
Is there someway that yes/no toogle can be added to Rollbacks, similar to that used for the Thank a User function? I edit mostly from a tablet, and accidental rollbacks occur far too often for me. It seems like a yes/no option would be very useful for this feature, especially if it could be enabled or disabled from preferences. Thanks. - BilCat (talk) 21:47, 26 May 2015 (UTC)
Yes, I've done that already, but I still have issues on the diff screen, ironically, usuallywhen I'm trying to toggle the Yes on the Thanks option! - BilCat (talk) 23:17, 26 May 2015 (UTC)
The most robust way I've found to deal with it is to hide the rollback links in css (which gets cached, so it always works instantly), and add a button in javascript to turn them back on. function morelinks() in my user javascript implements the latter, though it'll need some work to extract from all the other detritus I've accumulated. —Cryptic00:17, 27 May 2015 (UTC)
Could you extract all the other detritus, and list the relevant part somewhere? I haven't a clue which is relevant. Thanks. BilCat (talk) 08:33, 27 May 2015 (UTC)
This goes in your common.js - warning, completely untested
I implemented MusikAnimal's script yesterday, as it was the simplest script available. It works well, but as noted above, one does have to wait for the page to fully load. first. At least this will reduce accidental clicks of Rollbacks to a rare occurrence, and that's a great help for me. Thanks to all who contributed here. I still hope this can be fully implemented into the Wiki software at some point I the near future so we don't have to rely on hacks. - BilCat (talk) 14:03, 28 May 2015 (UTC)
I've just finished writing User:Mr. Stradivarius/gadgets/ConfirmRollback, a highly configurable script to hide rollback links or to ask for confirmation when clicking them. If you like, why not try it out? I also welcome feedback on its features. For example, the confirmation dialog is currently implemented as a pop-up window, but after reading the latest comments here I'm wondering if it wouldn't be better done like the Thanks confirmation is. — Mr. Stradivarius♪ talk ♪16:39, 28 May 2015 (UTC)
I'd like to try your script out eventually, as I like the options available. MusikAnimal's script also uses a pop-up confirmation dialog. On a touchscreen, it's probably easier to use that than the the Thanks confirmation style. As noted above, I would sometimes hit rollback while trying to select Yes for Thanks! - BilCat (talk) 17:01, 28 May 2015 (UTC)
Hovering over links doesn't show "exact" target page
Latest comment: 9 years ago2 comments2 people in discussion
It is very good that you can hover over (and I do not want it disabled until it is the few flaws are fixed).
I've come across a number of issues (some might be fixed since I first saw them): Extra space before comma, unexpected pictures (see e.g. Troy McClure, is second not first the rule?), and propably more I've forgotten. The latest one:
Latest comment: 9 years ago3 comments2 people in discussion
Hi there, I've recently enabled the "Group changes by page in recent changes and watchlist" feature in Preferences --> Recent changes. This is really cool and it has sped up my gnoming significantly. I'm curious though if there is a way to reposition the links, and if there is, would someone with the talent be so kind as to help set that up for me? I don't know if it's a simple matter of tweaking my personal CSS preferences, or what, but I have no CSS skills. Here's what I'm interested in changing:
While this may not seem like a significant difference, it is for me, because it's more efficient to have all the diff links in the same spot the way they appear in the regular watchlist. This would help me burn through more articles quicker, instead of having to hunt down the links in the Before example. Any assistance would be appreciated. Thanks! Cyphoidbomb (talk) 18:05, 29 May 2015 (UTC)
This will require Javascript and I think it's probably more complex than that since the information being dumped to the watchlist page is not marked up with convenient classes to target items in the rows in question. --Izno (talk) 18:27, 29 May 2015 (UTC)
Sounds like the same bug as I'm not even seeing the show and hide buttons. Clearing my cache didn't fix it. (Using ios8.3 with latest ios version of safari) Beeblebrox (talk) 20:04, 29 May 2015 (UTC)
It's not; it's happening to me in Chrome. The problem goes away when I log out, along with the following error. The code is minified, so I can't debug it without a source map - given my limited knowledge of JavaScript. Alakzi (talk) 20:39, 29 May 2015 (UTC)
Exception in module-execute in module mediawiki.toc:
TypeError: Cannot read property 'get' of undefined TypeError: Cannot read property 'get' of undefined
at eval (eval at <anonymous> (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:4:681), <anonymous>:1:721)
at fire (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:45:124)
at Object.self.add (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:45:664)
at eval (eval at <anonymous> (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:4:681), <anonymous>:1:114)
at mw.loader.implement.hidetoc (eval at <anonymous> (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:4:681), <anonymous>:2:81)
at runScript (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:162:744)
at checkCssHandles (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:163:281)
at execute (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:164:60)
at Object.mw.loader.implement (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:170:400)
at eval (eval at <anonymous> (https://en.wiki.x.io/w/load.php?debug=false&lang=en&modules=jquery%2Cmediawiki&only=scripts&skin=vector&version=XOBi6sgc:4:681), <anonymous>:1:11)
Fwiw... the 'Hide/Show' function works for me (Win 8.1 / IE 11) but my console has 2 or 3 warnings about missing end tags upon editing this (or any other) section. They don't/can't point to anything specifically found on this page however but all point to that god damn wiki-nsa tracking scheme starting at: ' https://meta.wikimedia.org/w/index.php?title=Special:RecordImpression '.
Note: I can't say if these warnings are related to the issue at hand or not but figure best to mention it just in case they somehow are related. -- George Orwell III (talk) 23:35, 29 May 2015 (UTC)
Getting the same in Firefox 38, the error message is ""TypeError: mw.cookie is undefined" TypeError: mw.cookie is undefined". MER-C10:03, 30 May 2015 (UTC)
Then it is a generic JavaScript problem, and collapsible not working is just collateral. Last week, I saw some code in MediaWiki being switched from jQuery.cookie to mw.cookie, and most likely, some of those changes do not have their dependencies in order. -- [[User:Edokter]] {{talk}}10:50, 30 May 2015 (UTC)
A very small change in the code managed to trigger an apocalypse at the caching level at the server side. It took some 5 hours before people figured out what was going on and how to undo it. This means that a whole lot of people who visited the website within that timeframe have a whole lot of junk in their browser caches that conflicts with each other. WP:BYPASS can help. —TheDJ (talk • contribs) 13:51, 30 May 2015 (UTC)
More specifically: Clear your cookies from en.wikipeida.org. That removes the offending cookie that causes this issue. (The small change TheDJ refers to was a changed cookie name.) -- [[User:Edokter]] {{talk}}14:53, 30 May 2015 (UTC)
Eventually, it should resolve itself. Either the bad cookies will expire by themselves, or (hopefully faster) our cache will start to ignore them one way or another. -- [[User:Edokter]] {{talk}}16:56, 30 May 2015 (UTC)
My watchlist includes block log entries of newly blocked users if I have previously posted to that users talk page. The watchlist continues to show the entry as an unread item after I have visited the block log and the user and talk pages of both the performer and target of the block. How are these ever meant to be marked as read short of clicking "Mark all pages as visited"? Using the nuclear option is not always appropriate. SpinningSpark15:01, 30 May 2015 (UTC)
Latest comment: 9 years ago4 comments2 people in discussion
Resolved
Is there a styling option to show a paragraph that has only the initial line indented, and then returns to regular left-side justification (after regular unforced line wrappings)?
A constructed demo:
Lorem ipsum doloR
Sit amet, consectetur adipiscing elit, sed do eiusmod tempor
(that is: after 'doloR' follows an unforced line wrapping, and then 'Sit' alignes left-justified). -DePiep (talk) 21:46, 30 May 2015 (UTC)
Latest comment: 9 years ago10 comments7 people in discussion
At a time when the long - term viability of Wikipedia is in doubt, editors have been discussing invigorating the community by enabling the creation of articles by IPs subject to safeguards. There have been no opposes. WMF have been asked to comment and have raised no objections. The mood of the community appears to be that the change should be implemented. Presumably it would involve reversing the post - Siegenthaler dev changes and moving onto a "level 3" pending changes regime in which pending changes would be turned off immediately an autoconfirmed editor edited the article. What is the procedure for turning an agreed change into reality? 156.61.250.250 (talk) 10:22, 21 May 2015 (UTC)
That change is far from agreed upon, I see more people rallying for complete removal of IP editing than this. We have WP:AfC that let's IP editors create an article, and that is all that has been agreed so far. NPP is badly backlogged, and it is well documented that most articles by new editors fail, no references, pure advertising etc. I can't see consensus on this anytime soon. EoRdE6(Come Talk to Me!)13:09, 21 May 2015 (UTC)
That's the whole point of allowing new users to do things, so they learn enough to become valuable contributors down the line. Checking the new pages feed I saw one article at the top of the list awaiting review. A few seconds later it had been patrolled. If PC is turned off, all the PC reviewers can turn their attention elsewhere.
If you think people are agitating for the removal of IP editing you're looking in the wrong place. The proposition in the March proposals RfC was to ban IP editing or never unprotect a semiprotected page. It was defeated nemo contradicente. Some quotes:
Wikipedia is the encyclopedia that anyone can edit.
There are plenty of people that come on Wikipedia reading, notice a problem and, as an IP, fix it.
If we block IP editing, we'll move one step closer to being like Citizendum. With all due respect to Sanger's efforts, visit Citizendum's website and see how spotty their articles are. There are some very well - known topics that don't have articles there. That's what happens when you make it hard to edit a wiki. Do you want that to happen to us?
Strong agreement with 156.61.250.250. If you want to count edits that occurred as a result of our IP editing policy, count all of mine, because I started as an IP editor and would have never started if this had been one more website that requires registration. --Guy Macon (talk) 20:56, 21 May 2015 (UTC)
Discussion has now slowed. There is a clear consensus that this useful change should be implemented. Can someone please close accordingly. 156.61.250.250 (talk) 08:12, 26 May 2015 (UTC)
I can count one person (Guy Macon) in both this discussion and the one at VPP who agrees with you. How is that clear consensus? --Ahecht (TALK PAGE) 16:16, 26 May 2015 (UTC)
I am tempted to apply a close as you requested. My close would be NO CONSENSUS based on a majority of one, and NO CONSENSUS based on a participation of 3 being grossly inadequate to propose a major sitewide change. If you want to seek a consensus for change I suggest you slap an RFC notice on here. That will get you participation and a close. Alsee (talk) 07:18, 27 May 2015 (UTC)
I could use tool, which allows me to enter all the required variants, and the target, and creates all those which do not already exist, adding a {{R from alternative name}} or similar template. If it can take a list like that above and automatically include title-case variants (e.g. System Integration Lab, etc.), so much the better.
I think the whole of Wikipedia has got gremlins somewhere. As when I delete or archive content from my own talk page, it flags up that I have a new talk page message; when I haven't. Wes Mouse | T@lk21:01, 29 May 2015 (UTC)
Prevent (or at least warn against) broken reference names
Latest comment: 9 years ago4 comments3 people in discussion
I wonder why it's impossible to make edits that introduce blacklisted URLs, yet it is possible to make edits that break reference names. Not only it is possible, but not even a warning is issued. No wonder Category:Pages with broken reference names keeps growing: it is fairly easy to break the articles this way, yet it's sometimes very difficult to fix them.
Issuing a warning would be fine, and even blocking the edit completely - like URL blacklist does - would probably be OK, as I don't think there is ever a good reason to break ref names. Is there maybe a plan to enhance the MediaWiki software in this direction? GregorB (talk) 20:03, 30 May 2015 (UTC)
Blocking the edit would not be OK. Non-techies are often confused by HTML and template syntax. They come to grok it after a while; we don't wanna drive them away at their first attempt. Alakzi (talk) 21:53, 30 May 2015 (UTC)
That's a good point - yet what happens when then try to insert a blacklisted URL is not any better. Anyway, a warning would do the trick, and it shouldn't be that hard to implement. I'm not sure where to go with this idea - certainly others have thought of this before, but everything depends on the MediaWiki developers. GregorB (talk) 09:06, 31 May 2015 (UTC)
I wonder if an edit filter or edit tag could be created to mark the diffs that do this. In my experience, it's sometimes easier to fix these things if you can see the exact edit in which it happens. Whatamidoing (WMF) (talk) 01:30, 1 June 2015 (UTC)
Section expanding
Latest comment: 9 years ago4 comments3 people in discussion
I tried to replicate this error the day this was posted and didn't see anything odd. Geof, go to Settings>More>"About device" to list your Android version here. It might help in narrowing down potential issues. Killiondude (talk) 04:30, 1 June 2015 (UTC)
Section expanding
Latest comment: 9 years ago4 comments3 people in discussion
I tried to replicate this error the day this was posted and didn't see anything odd. Geof, go to Settings>More>"About device" to list your Android version here. It might help in narrowing down potential issues. Killiondude (talk) 04:30, 1 June 2015 (UTC)
Search Field on Wikipedia's Mobile Homepage
Latest comment: 9 years ago3 comments3 people in discussion
Phone is a Blackberry 9930 with OS7. (Long live Blackberry; I hear OS10 works so much better.)
On the phone, the search field on en.m.wiki.x.io/wiki/Main_Page appears as a dark gray bar across the top of the webpage. When I click the gray bar, a "new menu" appears on the left side of the display: Home, Random, Nearby, etc. The new menu leaves the right quarter of the display uncovered, but faded out. If I click on the right-side area that's uncovered but faded, the menu goes away. But I never did get access to the Search field that I wanted to type into.
I've tried clicking on the gray bar, to the left of the gray bar, and to the right of the gray bar. I also tried reducing my display font size to minimum, just in case a larger font was covering something. Either way, I just get that new menu, but I can't access the Search function.
When I hit Random (in the new menu), I get a random page. Once, it did not have the dark gray bar at top and once it did, but when I clicked on the Search field at the top of the page whether the gray bar was there or not, I only got the new menu, again, anyway, with no chance to search.
Since searching on the mobile homepage stopped working a few weeks ago, my best work-around has been to use Wikipedia's desktop homepage on my phone. But, the mobile homepage was a little easier to use when it worked.
I would like to acknowledge that the search field on the mobile homepage was also out of commission about a year ago. I could click on the search field, but was unable to effect a cursor from which to type. The field was visible but not accessible. I appreciated when that problem was fixed.
"Print page" is malfunctioning, it is including the navigation menu and other non-article content, even my username! This behaviour is new, though it has been quite a long time since I last printed an article. I'm using Chrome on Windows 7, both fully updated. Roger (Dodger67) (talk) 09:07, 1 June 2015 (UTC)
Latest comment: 9 years ago2 comments2 people in discussion
When you move a page over a redirect the associated talk page can't always be moved, as it is not always a redirect to the page or it may have content. This leads to an awkward mess of redirects between articles and talk pages and eventually speedy deletions and moves that nobody wants to deal with, so can anything be done about this? Making some way for the talk pages to be replace, that is? Kharkiv07 (T)20:41, 30 May 2015 (UTC)
The limitations are there for a reason (because it is a dangerous action). Convenience is traded in for security here. —TheDJ (talk • contribs) 13:49, 1 June 2015 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Problems
Many wikis were slow for a few hours on Wednesday due to a code error. Sometimes the pages did not load at all and showed an error. [33]
Some tools in Labs were broken on Wednesday and Thursday. [34]
Edit tags added by the software were broken on all wikis from May 23 to May 28. [35][36]
Changes this week
The new version of MediaWiki has been on test wikis and MediaWiki.org since May 27. It will be on non-Wikipedia wikis from June 2. It will be on all Wikipedias from June 3 (calendar).
You won't be able to use e-mail lists for a few hours on Tuesday. [37]
Meetings
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on June 3 at 18:00 (UTC). See how to join.
Latest comment: 9 years ago10 comments7 people in discussion
Is there any way (as an admin) I can turn rollback off? I've never used the feature because on the rare occasion I need to revert multiple edits I prefer Twinkle's ability to annotate good-faith or vandalism. The rollback button has no "are you sure" prompt, so it's possible to hit the button by accident (easily done when scrolling on an iPhone and accidentally registering a click event in the wrong place) and rolling back something I didn't want to. Ouch!
Are there any settings to turn it off, or at least bring up an "are you sure" prompt? It doesn't seem like a very well designed tool. I'll try some of the upthread JavaScript hacks but they seem like a bodge job. Ritchie333(talk)(cont)06:48, 27 May 2015 (UTC)
If you just want to turn it off, just put:
.mw-rollback-link{display:none;}
in your common.css page, which will get rid of it without all the loading nonsense that the Javascript things entail. Writ Keeper⚇♔06:57, 27 May 2015 (UTC)
I wish I had known about that three days ago, before I gave up the rollback right because of too many accidental clicks. Perhaps someone could add something about this to WP:ROLLBACK. Better yet, someone could make rollback require a confirmation click, per well-established UI design principles. ―Mandruss☎08:23, 27 May 2015 (UTC)
I asked the same question above. If they can make an "Are you sure - Yes/No" toggle for the "Send Thanks" feature,someone should be able to do the same for Rollbacks. Since you asked for Rollbacks to be removed, you should be able to have it restored if you chose to implement on of these solutions. I'm still looking at them to see if one will work for me, barring the built-in solution. z BilCat (talk) 08:29, 27 May 2015 (UTC)
Yeah, I have now read the earlier thread and I'm not interested in local hacks. If the powers-that-be choose not to make the software act like software is supposed to act, I'll live without rollback. ―Mandruss☎08:39, 27 May 2015 (UTC)
Hopefully some from of this can be implemented, even as an option. It would certainly make editing from touchscreens easier. - BilCat (talk) 13:58, 28 May 2015 (UTC)
Reflinks still exists and works, but you need to add some code to your common.js page to get Reflinks it on your menu. IMHO "refill" is not (yet) as flexible and comprehensive as Reflinks.
// Add WP:Reflinks launcher in the toolbox on left
$(function () {
mw.util.addPortletLink(
"p-tb", // toolbox portlet
"http://dispenser.homenet.org/~dispenser/cgi-bin/webreflinks.py/" + mw.config.get('wgPageName')
+ "?client=script&citeweb=on&overwrite=&limit=20&lang=" + mw.config.get('wgContentLanguage'),
"Reflinks" // link label
)});
So far as I can see, we've lost the effectiveness of CSD U5. Can the grandfathering process be blocked for blocked users? Bazj (talk) 13:52, 1 June 2015 (UTC)
No, I don't think so, and in the presumably rare case where this happens, it's easy to fix. I simply created a blank user page to override the global user page brought over from Meta. —DoRD (talk) 14:00, 1 June 2015 (UTC)
Yes! Creating a blank page IS the blocking process. Looks like this will need to be standard procedure for the deletion of a blocked user's user page from now on. Bazj (talk) 14:06, 1 June 2015 (UTC)
Well, only if they already have a page at meta. This explains a few things as i happened just yesterday to be looking at an article on another project that I had never edited and they had a copy of my meta user page, and I was wondering why on earth they would have done that. Beeblebrox (talk) 16:51, 1 June 2015 (UTC)
The functionality is useful, but some of these consequences were perhaps not considered. I wonder if we should seek a local configuration change to automatically prevent global userpage display of blocked users, and whether prevention of global userpage display can be done through protection (or perhaps a switch during deletion), which would be more elegant than making a blank page. –xenotalk19:00, 1 June 2015 (UTC)
Well, I took the liberty of blanking the meta page. It seems that Beeblebrox tried raising the issue over there but ran into one of the local cranks, so who knows how far that'll go. Tarc (talk) 19:09, 1 June 2015 (UTC)
I didn't run into him, he stalks my edits over there. Luckily some more thoughtful users not motivated by personal grudges have showed up in the conversation now. I like Xeno's ideas, that way we can just deal with it locally and won't have to depend on meta for help that may or may not be forthcoming. Beeblebrox (talk) 22:50, 1 June 2015 (UTC)
Updating the tz database to fix articles
Latest comment: 9 years ago1 comment1 person in discussion
Articles such as CST6CDT are poorly broken because it can't refer to the tz database at List_of_tz_database_time_zones. I have no idea about how to correct the database (or whether I should) to ensure articles such as CST6CDT are not broken. Does anyone have any understanding of this? ~NottNott ( ✉ -☻) 23:24, 1 June 2015 (UTC)
Wikipedia, the free encyclopedia
Latest comment: 9 years ago4 comments4 people in discussion
The following discussion is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
For some pages, the title displayed on the tab and the title bar ends with "Wikipedia" rather than "Wikipedia, the free encyclopedia". Also, "From Wikipedia, the free encyclopedia" is not shown below the page title and other things look different on the page. GeoffreyT2000 (talk) 01:34, 29 May 2015 (UTC)
The discussion above is closed. Please do not modify it. Subsequent comments should be made on the appropriate discussion page. No further edits should be made to this discussion.
Can we get a bot to check the Internet Archive for dead link solutions?
Latest comment: 9 years ago1 comment1 person in discussion
Possibilities for thread/content closures: by proposal, by future bot visit or by setting of closure time
Latest comment: 9 years ago1 comment1 person in discussion
How may times have editors worked on a thoughtful thread input and hit the save page button only to be faced by an edit conflict screen and then to find that the thread / section of content concerned had been closed.
I was wondering about possible delay mechanisms that might be introduced to permit a closure of a thread to be set into motion in advance of actual closure. This would mean that an editor visiting a page would know that s/he would have the options of either contributing within a set time limit or, perhaps, protesting the closure. It seems to me that it would be more civil to oppose a proposed closure than to overturn an enacted closure decision.
An option of closure following proposal could simply involve an initial placement of a template perhaps based on:
I think that an option for personalisation would with individualised comment would work here.
An option of closure by future bot visit might work in a similar way as in regard to RM templates etc. The move could simply occur at something like the next or the next but one, two, three, etc. bot visits to the page.
An option of closure by setting of closure time might work by waiting for a bot to visit so that the bot could record the time set for the visit in a process that would then book in the time for a bot to return so as to close the thread/content. On final visit the bot could also make a check that the proposed time for closure had not been changed.
I am unsure how positive it might be for editors to start threads with a timed closure already set but i presume that this might also be a possibility.
Prior to proposal I was wondering how practical any of these ideas might be and how much effort would be involved in each. Thanks. GregKaye08:10, 2 June 2015 (UTC)
Latest comment: 9 years ago2 comments2 people in discussion
Hi! I am a SPI Clerk and I'm using the Mark-blocked script (User:NuclearWarfare/Mark-blocked script.js, it is imported from ru.wikipedia). That script marks all linked blocked users with a strikethrough. That is very useful for the SPI. But, sometimes, user is not blocked but is globally locked. This script does not mark such users. I am not familiar with those scripts, so I'm asking somebody more experienced: Is it possible to change this script so that it also marks globally locked users somehow? Thanks a lot! Vanjagenije(talk)15:49, 1 June 2015 (UTC)
I sure won't do it if it can be sorted out. I haven't yet. This not my area of expertise. It's just it keeps blanking it and the bot's operator, while online, hasn't responded... Shawn in Montreal (talk) 22:51, 1 June 2015 (UTC)
Oh, well, I can't hit the button anyway as I'm not an administrator. Shows what I know. (Couldn't come at a worse time as I finally started to get some help with a walled garden RPP request. Dang.) Shawn in Montreal (talk) 22:56, 1 June 2015 (UTC)
Monty845 has discovered that pushing the button didn't stop the problem. There are a couple of threads at Wikipedia talk:Requests for page protection about the change from posting new entries at the bottom instead of the top of the page. Automated tools like twinkle are still putting new items at the top. I wonder if that is causing this bot to blank the whole page. MarnetteD|Talk00:47, 2 June 2015 (UTC)
As mentioned on the user talk page, I'm going to see if manual archiving removes whatever is breaking the bot. The bot does a lot of other useful stuff, so I really hope we don't need to block it until its fixed. Monty84501:00, 2 June 2015 (UTC)
I shut it down manually. Now I know what to do when I have nightmares about robots. :p I will be investigating first thing tomorrow morning. I'm in Germany.—cyberpowerChat:Limited Access02:41, 2 June 2015 (UTC)
Changed Since Last Visit indicators on enhanced watchlist (and recent changes)
Latest comment: 9 years ago5 comments2 people in discussion
By default, the standard watchlist shows green bullets for changed (since last visit) pages, but the enhanced watchlist does not. I would like to enable this for enhanced watchlist as well, but like to gauge the community first. To that end, there is now a test gadget available to enable the indicators for the enhanced watchlist. It will be especially visible when both "Group changes..." (under Recent changes tab) and "Expand watchlist..." options are enabled.
If feedback is positive, it will be enabled as a default gadget, and the current CSS for the standard watchlist will be moved to this gadget as well. -- [[User:Edokter]] {{talk}}17:02, 28 May 2015 (UTC)
If global.css works, you can use @import"//en.wiki.x.io/w/index.php?title=MediaWiki:Gadget-WatchlistGreenIndicators.css&action=raw&ctype=text/css";. Otherwise, in global.js: mw.loader.load('//en.wiki.x.io/w/index.php?title=MediaWiki:Gadget-WatchlistGreenIndicators.css&action=raw&ctype=text/css','text/css');-- [[User:Edokter]] {{talk}}08:07, 1 June 2015 (UTC)
Latest comment: 9 years ago1 comment1 person in discussion
While I was checking the active watchers of my talk page via Equazcion's Active Watchers tool, it instead redirected me to the page in which it said, "no redirect found", assuming that it was one of the technical difficulties after the tool server shutdown, so a new link for that tool should be re-created on Labs, please. The Snowager-is awake04:57, 1 June 2015 (UTC)
Polluted discussion pages by wikiprojects
Latest comment: 9 years ago5 comments5 people in discussion
Is there any Gadget or user script already for making "Talk" link red if the discussion page contains only templates like WikiProject-xx?
I'm not aware of any such script, although I agree that it would be useful. WP:Flow (warning: early development stages) uses the concept of separating headers from actual discussion, which might make it possible eventually. If this were a very popular idea, then it might be possible to put the templates/header stuff into a subpage, and automagically transclude that onto the talk pages (sort of the way global user pages work, in which you set up a user page at Meta and it appears at any project where you haven't created a user page). But that would be a major change involving multiple dependencies, plus an annoying bot run to move everything and delete all the otherwise blank talk pages, and I'm not sure that it would be worth it. Whatamidoing (WMF) (talk) 01:15, 1 June 2015 (UTC)
Latest comment: 9 years ago10 comments7 people in discussion
Across the top of a lot of pages, I see a big banner that reads:
"Voting Has Begun in Wikimedia Foundation Board of Trustees Elections! Voting Ends at 23:59, 31 May 2015 (UTC). Verify your eligibility and vote now."
Yet, if I click on it, I get an error: "We apologize, but you do not appear to be on the eligible voter list."
I can see how I might not be eligible, because my edit count isn't high enough between 15 October 2014 and 15 April 2015.
However, it seems like the computer can easily count and realize this (it does, when I click on the banner) and should be able to just stop showing me this banner that I've deliberately closed many times.
Is there any way to manually stop it from showing up? It especially gets in the way of comparing pages by changing between tabs back and forth. Also, it's just training new users to ignore the banners, when you're so often showing banners that are irrelevant for that user.
Really, I don't think this banner should be shown at all, *especially* after a user has already clicked through and the system has already determined that member's ineligibility.
The service that checks those criteria wouldn't be able to handle the load, since the central notice needs to run on the foundation scope and only runs in JS. Not very easily doable. —TheDJ (talk • contribs) 12:38, 24 May 2015 (UTC)
Is there any way for the service to set some bit in the user profile, or have an election-specific table of users and their eligibility status, and only that gets checked prior to banner display (instead of recomputing all eligibility criteria each time?) There seem to be lots of settings that get checked per-user in the process of generating various page elements; checking election eligibility before generating an election banner just makes sense. --ProtectorServant (talk) 19:00, 24 May 2015 (UTC)
I don't care about being shown this banner the first time, but I am getting bloody sick of seeing it repeatedly. Why is there no "Dismiss" option as there usually is to prevent re-display. I can click on the X to stop it on a given page, but the next page I go to it just pops right back. It has been going on for over a week now and I've hadit with this BS. This is very sloppy coding and is just not good enough. - Nick Thornetalk04:42, 29 May 2015 (UTC)
I'm observing the same as Nick. My fairly vanilla web browser doesn't refuse cookies (I can log in just fine and stay logged in, and Wikipedia knows I'm a logged-in user and it could know I'd previously clicked the X on this banner several times) and scripts aren't disabled (lots of other JS on Wikipedia also work just fine). --ProtectorServant (talk) 14:17, 3 June 2015 (UTC)
Pageview data missing
Latest comment: 9 years ago12 comments9 people in discussion
And I emailed Henrik directly. No results so far. Do we have an alternative for stats, other than trying to sift through raw data? — Maile (talk) 18:47, 26 May 2015 (UTC)
Curious question. Henrik usually kick starts the stats if he gets an email. Not seeing anything happen, I'm wondering if (1) Henrik might be on vacation; or (2) May have just dropped off the project forever. Does anyone have any knowledge of Henrik's whereabouts? — Maile (talk) 21:41, 27 May 2015 (UTC)
It's so bad, because last year very often we have troubles with stats... Maybe someone could do "new project", that it work well and without troubles... Stats are very usefull for create new pages. — Preceding unsigned comment added by Swd (talk • contribs) 10:05, 29 May 2015 (UTC)
With all due respect to individuals like Henrik who have put their time and effort into creating tools that provide us with these statistics but it is not a good situation to be dependent for this on the presence or absence of a single individual. We need to move on from that. Statistics (page, user) are a core Wikipedia functionality and should be provided by the WMF with sufficient guarantees regarding back-up and up-time.--Wolbo (talk) 12:09, 31 May 2015 (UTC)
I work on the WMF's analytics team and this is our top priority now. See my update here and please chime in with any help or opinions: https://phabricator.wikimedia.org/T44259#1301784 . We are more than thankful to Henrik for doing this amazing work for all this time. But this needs to be provided by WMF with guarantees, as Wolbo says. And it will be, we just have to find a compromise between what people are asking and the hardware resources we have. — Preceding unsigned comment added by Milimetric (talk • contribs) 14:06, 2 June 2015 (UTC)
Thanks for this note. I asked the Ops folks about this, and they said that it happens when one of the tokens they use during the sign-on process disappears unexpectedly. It's not very worrisome if it happens rarely (as you said, you managed to login anyway), but if this happens a lot, then they definitely want to know about it, because it could be a sign that something's gone wrong with WP:SUL software. Whatamidoing (WMF) (talk) 11:06, 3 June 2015 (UTC)
Identical citations.
Latest comment: 9 years ago6 comments4 people in discussion
Is there a way to hide identical citations so there isn't 3 or 4 of the same reference in the reflist? I'm working on Tales of Frankenstein, and I have 7 or 8 in my reflist with only 3 sources. Inkwell765 (talk) 19:59, 31 May 2015 (UTC)
BTW, wouldn't it be possible to track down such articles (where are references with the same content)? Or – couldn't the system automatically group such references as it does with {{sfn}} – when you're using for example {{sfn|Example|2006|p=6}} multiple times, in the reference list it is shown as one reference, not multiple. I understand, that it is a little bit different, but why not? --Edgars2007 (talk/contribs) 20:22, 31 May 2015 (UTC)
@Edgars2007 Thank you so much. That was fast service, you're a Computer Guru²!
I will try to figure it out, I'm learning, incrimentaly. Inkwell765 (talk) 20:37, 31 May 2015 (UTC)
@Edgars2007: In essence, the {{sfn}} template merges duplicate refs because of the way that it composes a value for the name= attribute of a <ref>...</ref> tag. It relies on the fact that when two <ref>...</ref> in the same group have identical name= attributes, only the first one is displayed by the <references /> tag. {{sfn}} builds a name= from the full list of parameters; for example, if you have
First sentence.{{sfn|Smith|2015|p=12}} Second sentence.{{sfn|Smith|2015|p=34}} Third sentence.{{sfn|Smith|2015|p=12}}
this is the same as
First sentence.<ref name="FOOTNOTESmith201512">[[#CITEREFSmith2015|Smith 2015]], p. 12.</ref> Second sentence.<ref name="FOOTNOTESmith201534">[[#CITEREFSmith2015|Smith 2015]], p. 34.</ref> Third sentence.<ref name="FOOTNOTESmith201512">[[#CITEREFSmith2015|Smith 2015]], p. 12.</ref>
notice how the first and third ref names ("FOOTNOTESmith201512") are identical, whereas the second ("FOOTNOTESmith201534") differs only in its last two characters, because a different page number was used. So for the system to automatically group references that don't use {{sfn}}, there would need to be some way of looking between the <ref> and </ref> tags in order to make a name= attribute for that <ref>. --Redrose64 (talk) 21:04, 3 June 2015 (UTC)
"wikitable sortable" tables are no longer sortable V2.0
Latest comment: 9 years ago30 comments6 people in discussion
I just asked a user at IRC: Latest version (07:27, 1 June 2015) is sortable in Chrome 43.0.2357.81 (64-bit), Firefox (38.0.1 on Mac), and Safari. Anna Frodesiak (talk) 09:52, 1 June 2015 (UTC)
I do not know if the table was sortable for others before I modified it. If the table as it is now is not sortable on Firefox Win7 38.0.1, then I think all tables are not sortable across Wikipedia using Firefox Win7 38.0.1 Win7. This is because I simplified it to be like any other sortable table. Try List of commercially important fish species and tell me if that sorts. Anna Frodesiak (talk) 10:06, 1 June 2015 (UTC)
Anna Frodesiak & Brudder Andrusha, that table was sortable all along. Only, the sort buttons were invisible. It's a well know issue with using the wrong code to provide a background-color in a header cell containing a sort button, as explained here. Anyway, I fixed it now. Tvx1 10:30, 1 June 2015 (UTC)
Clearly the sort buttons are visible when you see a sortable table in preview mode while editing. However they are NOT visible and do not function as buttons when you are reading the article in normal mode. Brudder Andrusha (talk) 10:39, 1 June 2015 (UTC)
They work just fine for me on Firefox WinXP 38.0.1. I don't think the problem that is being experienced lies within the coding of the tables.Tvx1 11:03, 1 June 2015 (UTC)
I have made snapshots of the screen so its clear what is visible and what is functional. Take a look at the snapshots is here. As you will see the function arrows for sorting are not visible and cannot be activated when viewing the article.Brudder Andrusha (talk) 11:27, 1 June 2015 (UTC)
You can type as loud as you like but my eyes are telling me that there is a difference between with what is on the screen when I see the article and when I'm in Show Preview mode & the sort button wherever it is DOES NOT WORK. Brudder Andrusha (talk) 12:34, 1 June 2015 (UTC)
I'm not sure why personal frustrations are even raising up here. Andy Dingley, Brudder Andrusha is saying that it does work in preview, but does not do so in normal mode.Tvx1 12:50, 1 June 2015 (UTC)
Friends, I only raised this issue because this has only been broke a couple of days ago. I had not touched the wikitable sortable header setting ever on that example page and had been using it for at least 4 years.Brudder Andrusha (talk) 12:55, 1 June 2015 (UTC)
Sorry, but I disagree. There is something definitely different with the behaviour of wikipedia sortable on Firefox 38.0.1. Again have a look at the images that I have saved here. Clearly the article in Firefox does not show the sortable arrows and the button is non functioning for sorting and it is available in Show Preview mode. Brudder Andrusha (talk) 17:37, 1 June 2015 (UTC)
Do you have a non-Windows XP platform on which you can give it a try? If I was going to suspect anything being broken in any of these scenarios, it'd be an unsupported operating system. Tarc (talk) 17:49, 1 June 2015 (UTC)
If you think the problem is with Firefox 38.0.1, then how do you explain these sortable tables to work perfectly with that same Firefox 38.0.1 on Mac OS 10.10.3 and on my ten year old WinXP? And if it is a problem with the browser there is nothing we can do about it. We can only fix the wik and thus far we have not found the problem to be with the wiki. It's not that we don't want to help you. Tvx1 19:36, 1 June 2015 (UTC)
The table at List of Australian Group races sorts just fine for me, in Firefox 38. I have two suggestions: First, would anyone who knows something about that subject please edit the lead to say what kind of race it is? Second, Brudder Andrusha, would you please open a "New Private Window" in Firefox and then go to that list and see whether it works in "private" or "incognito" mode? That will tell us something about whether the problem is in your Wikipedia account or with your personal computer. WhatamIdoing (talk) 23:12, 1 June 2015 (UTC)
WhatamIdoing, thanks for your suggestions. The lead to the example has been changed, but I want to add a bit more when I get a free moment. Private mode worked! Thanks to all who spent the time to be involved here. Cheers! Brudder Andrusha (talk) 06:19, 2 June 2015 (UTC)
I believe the general rule of thumb is that when something's busted while logged in, but works in private mode, then the problem could be a cached page, but could also be something in your personal js or css files (like User:Brudder Andrusha/common.css) or a gadget that you've enabled. WhatamIdoing (talk) 08:21, 3 June 2015 (UTC)
Stats grok no longer maintained???
Latest comment: 9 years ago4 comments4 people in discussion
If you look at any stats.grok page, it used to say it was maintained by Henrik. His name has been removed. Does this mean no one maintains this? That would explain why it broke and nobody is doing anything. Maybe Henrik's user page should have some kind of Inactive banner at the top, so people don't post useless threads there.— Maile (talk) 18:20, 1 June 2015 (UTC)
Latest comment: 9 years ago6 comments4 people in discussion
The article name is displayed in Italics, but this is a person. Could somebody fix this? (Could somebody tell me how to do this, or does one need special tools?) Kraxler (talk) 14:42, 2 June 2015 (UTC)
Fixed. It was happening because {{Infobox album}} is used in the article and this automatically chnages the article title to italics unless the parameter |italictitle=no is set within the infobox. Nthep (talk) 14:49, 2 June 2015 (UTC)
@Nthep: That shouldn't have worked. As advised at the top of {{infobox album}}, the parameter is |Italic title=no and looking at its code, no variants on that are recognised - whether decapitalisation or removal of the space. --Redrose64 (talk) 20:44, 3 June 2015 (UTC)
I've tried that to no avail. It does work perfectly in those other Wikis though, I guess I'll try and cut and paste into them to preview in future :( Gareth E Kegg (talk) 23:56, 2 June 2015 (UTC)
I don't know what the problem is but you could try another browser or try to disable any Firefox extensions you may have. However, if pt: works then it doesn't sound like an issue with a browser extension. You can also try Special:ExpandTemplates but it would be inconvenient to copy back and forth. PrimeHunter (talk) 00:07, 3 June 2015 (UTC)
This may be the same intermittent fault I am having on en. Using the latest Firefox on Linux Mint- save appears not to work! On doing the edit I press button save page- something happens, a new page is rendered. It is the page I had before the edit. Click refresh page and I see my newly edited page. It is if the process to render the page had pulled something out of the cache- instead of waiting for the new page to be generated. When I tried to replicate this and take screen dumps, the fault was not there.-- Clem Rutter (talk)15:13, 3 June 2015 (UTC)
Formatting section headers subsequent to an image
Latest comment: 9 years ago5 comments4 people in discussion
Let's suppose that we have two sections at the same level. The first one contains an image on the left (or binding for any direction, but consider text that flows left-to-right).
Example (part 1)
Text blah blah blah.
Example (part 2)
And, more text. (←) So, here we see that the image causes an indent which is carried forward to some lines proceeding. If an editor preferred to start a new paragraph that was not indented, they would need some way to present that to the MediaWiki parser or whatever.
Preferrably, this would be something that applied only to section headers at the same level as the parent of the image. I.e., a subheader defined with 3 '=' would not need to be flushed to the left — the binding for the text — because it could be ancillary to the image.
The help is much appreciated. I didn't notice until i had already submitted it that a <code> tag didn't envelope wiki markup, but that template works. — JamesEG (talk) 21:27, 3 June 2015 (UTC)
Looking for in depth Wikipedia statistics
Latest comment: 9 years ago4 comments2 people in discussion
I was looking for combine article stats including deleted articles created on Wikipedia and articles which have survived one or more AfDs which are currently deleted. Is there anyway to get this? Valoemtalkcontrib13:34, 3 June 2015 (UTC)
I believe it is useful if not important to known the overall retention rate of this encyclopedia. I assume most deleted article are hoaxes or blatant promo, but still the total number of articles does interest me particularly the actual depth of hidden material. A few numbers should be accessible to everyone such as:
Total number of pages created including gibberish
Total number of article with discernible content (At least a paragraph of 3 comprehensible sentences or more)
Total number of article with at least one source before deletion
Total number of articles which have survived at least one AfD as no consensus or keep, but is currently deleted.
The latter of which is particularly important in determine whether there is a large among of content which may pass GNG and NPOV guidelines if expanded. I am interested in comparing these numbers with my predictions, I've estimated based on our guideline we should have had a lot more articles by this point. Valoemtalkcontrib14:18, 3 June 2015 (UTC)
The last one does seem interesting, and is the one that is easiest to determine by the heuristic "is there more than one AFD for an article that does not exist?". That would miss ones that survived AFD but then were deleted for other reasons, but there's not much policy that would allow it. The false-positive would be articles that are recreated after AFD-deletion and then go through a second AFD (and then deleted again from it) rather than simply WP:CSD#G4. It's easy to find most subsequent-to-first AFDs in Wikipedia:Archived deletion discussions by looking for various parenthetical notes in the AFD names (possible false-negative would be specific pages that are part of bulk nom). Look for redlinks, or a quick script. DMacks (talk) 16:45, 3 June 2015 (UTC)
Can someone more experienced with scripting help a new user out?
Latest comment: 9 years ago2 comments2 people in discussion
Catherinesjy is trying to do something or other with their js, creating the mainspace page common.js before also creating their own common.js page in their userspace. The page they are trying to import code from User:Catherinesjy/swl_editor.js does not exist. Can someone with scripting experience go over to their talk page and see what they're trying to do and help them? KonveyorBelt19:49, 3 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
Greetings, Of all the many pages I am doing assessments for, this page Talk:Catholic bishops in Nazi Germany is showing WP Germany of class=C when the Wikicode actually is class=B. The categorization at bottom is also incorrect, showing class C. I have tried many things: moving Germany line above and below; filling in the importance value. For last attempt, I removed WP Germany all parameters, Saved, did the Edit, then Purge, added the Class=B and save, and then class C re-appears. Looks like some kind of a bug? For some unusual reason this is the only page I have seen with this specific issue. Help please . Regards, JoeHebda (talk) 11:37, 4 June 2015 (UTC)
Click "[show]" at "WikiProject Germany" and then "[show]" at "This article has not yet been checked against the criteria for B-Class status". You need |b1=yes| b2=yes| b3=yes| b4=yes| b5=yes before B-class is accepted. PrimeHunter (talk) 12:02, 4 June 2015 (UTC)
Thanks PrimeHunter (talk), so now I understand the assessment process for WP Germany is different than for WP Catholicism, and that all WP assessments are not exactly the same, i.e., may be some variations for each WikiProject. Learned something new today. Thanks again. Cheers, JoeHebda (talk) 12:19, 4 June 2015 (UTC)
@JoeHebda: It's essentially the same: in order to be granted B-class, an article needs to satisfy all of the criteria at WP:BCLASS, whatever the WikiProject. Some articles may meet some of these criteria, but not all; and so some WikiProject banners provide a five- or six-point checklist. These are usually described on the documentation for the WikiProject banner; in the case of {{WikiProject Germany}} it says
class – valid values are fa, a, ga, b, c, start, stub, fl, list, na (case-insensitive). See the project's quality scale for details. Setting an explicit value of NA is rarely necessary; for this, either leave |class= blank or omit the parameter. B-class checklist (available only when |class=stub|class=start|class=c or |class=b):
B-Class-1 – Referencing and citation (see the project's B-class criteria for details).
B-Class-2 – Coverage and accuracy (see the project's B-class criteria for details).
B-Class-3 – Structure (see the project's B-class criteria for details).
B-Class-4 – Grammar and style (see the project's B-class criteria for details).
B-Class-5 – Supporting materials (see the project's B-class criteria for details).
Set these to "yes" or "y" if the article meets the criterion; "no" or "n" if it does not. If the criterion has not been checked, omit the parameter or leave it with a blank value. When |class=b is set, all five criteria need to be satisfied, otherwise the article is treated as if |class=c were set.
Whether there are five (as here) or six (as with {{WikiProject Catholicism}}), all of them need to be set to yes in order for |class=b to be recognised, otherwise it will show C class. If the template does not provide a checklist (as with {{WikiProject Judaism}}), |class=b will always be recognised. --Redrose64 (talk) 18:01, 4 June 2015 (UTC)
502 Bad Gateway error
Latest comment: 9 years ago1 comment1 person in discussion
I wasn't able to access any Wikipedia page for about one minute. It just said "Bad Gateway", and underneath that was "nginx/1.6.2". What was that? Was it on my side? Dustin(talk)19:41, 4 June 2015 (UTC)
needs-persondata
Latest comment: 9 years ago5 comments4 people in discussion
Several Wikiproject banner templates, such as {{WikiProject Rugby union}}, have a |needs-persondata= parameter. I can't see where (in a parent template?) this is, and I'm about to board a plane for several hours.
I use Firefox with NoScript, so I use most web sites, including Wikipedia, with JavaScript disabled. (More specifically, I am using the latest version of TenFourFox (31.7.0) which is equivalent to the same version of Firefox.) I noticed that recently the behavior of footnotes on Wikipedia has changed. Before the change, clicking on a footnote marker jumped the page down to the footnote and highlighted the citation with a light blue background (as described at Help:Footnotes § Overview). After the change, clicking on a footnote marker jumps the page down to the footnote but no longer highlights the citation (when JavaScript is disabled; it works as expected when JavaScript is enabled). I imagine that there was some change to the Wikipedia CSS and JavaScript so that the footnote highlighting functionality that was previously implemented purely in CSS now requires JavaScript.
I mention this in the hope that someone who implements these technical changes might consider modifying the behavior of footnotes so that clicking on the footnote marker jumps the page down to the footnote and correctly highlights the citation when JavaScript is not enabled, as it did before the change. I consider it to be a software bug that footnote highlighting no longer works when JavaScript is disabled. It is a reduction of functionality.
I expect that some people would respond: "Well, just turn JavaScript on!" I realize that I could whitelist Wikipedia in NoScript so that JavaScript is always enabled when I am using Wikipedia, but I prefer to leave JavaScript disabled with NoScript because pages load faster on my 10-year-old computer when JavaScript is disabled. I imagine that there are other people who also use Wikipedia without JavaScript. Thanks, Biogeographist (talk) 18:33, 5 June 2015 (UTC)
@Biogeographist: I think that the change was made in March this year, and judging by this edit, I would say that if you add
/* Highlight clicked reference in blue to help navigation */ol.referencesli:target,sup.reference:target{background-color:#DEF;}
I appreciate your help. That works (only when I am logged in, of course). But I pity those people who have JavaScript disabled and who do not have Wikipedia accounts, and who therefore cannot benefit from your advice like I can. Thanks, Biogeographist (talk) 19:13, 5 June 2015 (UTC)
@TheDJ: It's not the only one. If you have the gadget "Display diffs with the old yellow-and-green colors and design" enabled, which is a pure CSS gadget with no JavaScript, there is a noticeable delay when you follow a diff link before MediaWiki:Gadget-OldDiff.css kicks in. I eliminated the CSS as a problem by disabling the gadget and making this edit, since when there has been no delay. --Redrose64 (talk) 20:35, 5 June 2015 (UTC)
Reported and patch submitted. Thanks for pointing it out Biogeographist. @Redrose64:, I think Gadgets are always JS modules, even if they don't include any JS at all. But I will have to check to figure out how it works exactly. —TheDJ (talk • contribs) 21:03, 5 June 2015 (UTC)
Ah i know what is happening. It's a bottom loaded module. All css files used to be top loaded by default, but the new default is bottom loaded (performance reasons. —TheDJ (talk • contribs) 14:23, 6 June 2015 (UTC)
Post not showing up immediately
Latest comment: 9 years ago24 comments16 people in discussion
Has anyone else noticed in the last couple weeks that they occasionally have to refresh a web page to get the post they just made to show up? --NeilNtalk to me15:18, 3 June 2015 (UTC)
I saw that for the first time today, and I've been on every day the past couple of weeks. It was confusing and disorienting at first! ―Mandruss☎15:21, 3 June 2015 (UTC)
Also happened to me today for the first time I remember. Something must have changed when so many say it. I use the current Firefox version 38.0.5 if it matters. PrimeHunter (talk) 15:32, 3 June 2015 (UTC)
It just happened again when I saved the above post. It's only the second or third time today out of several more posts. PrimeHunter (talk) 15:34, 3 June 2015 (UTC)
I've had that today - the post shows when you preview but when you hit save it isn't there until page is refreshed. I'm using FireFox too. SagaciousPhil - Chat15:35, 3 June 2015 (UTC)
I have noticed this since at least yesterday. Its intermittent though. I sort of suspect there may be some lag somewhere in the system. (Firefox as well) Monty84515:56, 3 June 2015 (UTC)
Several times, over at least the last week, poss ten days. Firefox 38.0.5 and whichever it was upgraded from a day or two back. F5 reloads the proper version. --Redrose64 (talk) 20:39, 3 June 2015 (UTC)
I have recently found that my edits sometimes just don't go through when I press the "Save page" button. This problem only seems to have started affecting me recently. It will reload the Wikipedia page I was editing, but then I'll see that nothing has changed. Luckily, I can just use the "go to previous page" arrow on my web browser to recover the change and I can just press "Save page" again to get my edit to go through, but this problem concerns me. Dustin(talk)17:14, 4 June 2015 (UTC)
I guess making an edit that is exactly the same as the previous revision just refreshes the page? If so, then that is likely the same problem. Dustin(talk)19:17, 4 June 2015 (UTC)
It's working ok for me on three random articles, so it's either a transient problem that has passed or a local problem. Check again, and if it persists perhaps you could be more specific. ―Mandruss☎11:33, 4 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
Hi all, I've made a few mistakes recently by using this usernames when I mean to be editing with new WMF account, which I'm using to edit their blog posts and similar things (I will only rarely have to use it here). Then today I moved a page with the WMF account, which is potentially much worse. Is there a way with javascript or CSS that I can use to tell the two apart, say with a background color or something similar? Thanks! Ed[talk][majestic titan]02:21, 5 June 2015 (UTC)
If you put that on your global css page at Meta, it will turn up everywhere.
I reduce this risk by using different browsers. Firefox = personal account (where I've edited for years), Safari = staff account. It mostly works for me. Whatamidoing (WMF) (talk) 05:53, 7 June 2015 (UTC)
Confirmation for rollback on touch devices
Latest comment: 9 years ago4 comments3 people in discussion
The reason for making this default-enabled is because not requiring confirmation isn't a preference, it's a problem. Accidental rollback happens quite a lot, often with the user unaware they even did it. If users still feel they do not want confirmation (maybe they're using a mouse on their tablet, for instance, or they prefer to use another similar script), the gadget can be turned off in Special:Preferences#mw-prefsection-gadgets. I hope and am confident this is a good move, but if not we can easily remove the gadget or make it opt-in. Thanks! — MusikAnimaltalk14:31, 6 June 2015 (UTC)
@Mr. Stradivarius: I am guessing the library is not optimized for mobile? I tried this on my Nexus 5 and the dialog appeared very tiny. Zooming in I can get to the dialog but we may end up running into the same issue where we accidentally click on the wrong button. — MusikAnimaltalk23:47, 6 June 2015 (UTC)
I guess it isn't. I thought OOjs UI was optimised to work across all browsers, but it looks the devs haven't been concentrating on making it work with mobile. in that case, I'd say that visible should be preferred over beautiful. :) — Mr. Stradivarius on tour♪ talk ♪00:12, 7 June 2015 (UTC)
VisualEditor News #3—2015
Latest comment: 9 years ago2 comments1 person in discussion
Did you know?
When you click on a link to an article, you now see more information:
The link tool has been re-designed:
There are separate tabs for linking to internal and external pages.
The user guide has more information about how to use VisualEditor.
Since the last newsletter, the Editing Team has created new interfaces for the link and citation tools, as well as fixing many bugs and changing some elements of the design. Some of these bugs affected users of VisualEditor on mobile devices. Status reports are posted on Mediawiki.org. The worklist for April through June is available in Phabricator.
A test of VisualEditor's effect on new editors at the English Wikipedia has just completed the first phase. During this test, half of newly registered editors had VisualEditor automatically enabled, and half did not. The main goal of the study is to learn which group was more likely to save an edit and to make productive, unreverted edits. Initial results will be posted at Meta later this month.
Recent improvements
Auto-fill featuresfor citations are available at a few Wikipedias through the citoid service. Citoid takes a URL or DOI for a reliable source, and returns a pre-filled, pre-formatted bibliographic citation. If Citoid is enabled on your wiki, then the design of the citation workflow changed during May. All citations are now created inside a single tool. Inside that tool, choose the tab you want (⧼citoid-citeFromIDDialog-mode-auto⧽, ⧼citoid-citeFromIDDialog-mode-manual⧽, or ⧼citoid-citeFromIDDialog-mode-reuse⧽). The cite button is now labeled with the word "⧼visualeditor-toolbar-cite-label⧽" rather than a book icon, and the autofill citation dialog now has a more meaningful label, "⧼Citoid-citeFromIDDialog-lookup-button⧽", for the submit button.
The link tool has been redesigned based on feedback from Wikipedia editors and user testing. It now has two separate sections: one for links to articles and one for external links. When you select a link, its pop-up context menu shows the name of the linked page, a thumbnail image from the linked page, Wikidata's description, and/or appropriate icons for disambiguation pages, redirect pages and empty pages. Search results have been reduced to the first five pages. Several bugs were fixed, including a dark highlight that appeared over the first match in the link inspector (T98085).
The special character inserter in VisualEditor now uses the same special character list as the wikitext editor. Admins at each wiki can also create a custom section for frequently used characters at the top of the list. Please read the instructions for customizing the list at mediawiki.org. Also, there is now a tooltip to describing each character in the special character inserter (T70425).
Several improvements have been made to templates. When you search for a template to insert, the list of results now contains descriptions of the templates. The parameter list inside the template dialog now remains open after inserting a parameter from the list, so that users don’t need to click on "⧼visualeditor-dialog-transclusion-add-param⧽" each time they want to add another parameter (T95696). The team added a new property for TemplateData, "Example", for template parameters. This optional, translatable property will show up when there is text describing how to use that parameter (T53049).
The design of the main toolbar and several other elements have changed slightly, to be consistent with the MediaWiki theme. In the Vector skin, individual items in the menu are separated visually by pale gray bars. Buttons and menus on the toolbar can now contain both an icon and a text label, rather than just one or the other. This new design feature is being used for the cite button on wikis where the Citoid service is enabled.
The team has released a long-desired improvement to the handling of non-existent images. If a non-existent image is linked in an article, then it is now visible in VisualEditor and can be selected, edited, replaced, or removed.
The weekly task triage meetings continue to be open to volunteers, each Wednesday at 12:00 (noon) PDT (19:00 UTC). Learn how to join the meetings and how to nominate bugs at mw:Talk:VisualEditor/Portal. You do not need to attend the meeting to nominate a bug for consideration as a Q4 blocker. Instead, go to Phabricator and "associate" the Editing team's Q4 blocker project with the bug.
If your Wikivoyage, Wikibooks, Wikiversity, or other community wants to have VisualEditor made available by default to contributors, then please contact James Forrester.
If you would like to request the Citoid automatic reference feature for your wiki, please post a request in the Citoid project on Phabricator. Include links to the TemplateData for the most important citation templates on your wiki.
Subscribe, unsubscribe or change the page where this newsletter is delivered at Meta. If you aren't reading this in your favorite language, then please help us with translations! Subscribe to the Translators mailing list or contact us directly, so that we can notify you when the next issue is ready. Thank you! Whatamidoing (WMF) (talk) 17:31, 6 June 2015 (UTC)
Quick update: The weekly bug triage meeting has moved to Tuesday, 9 June 2015 at 12:00 (noon) PDT (19:00 UTC). I don't know if that's permanent or just for this next week. Whatamidoing (WMF) (talk) 06:01, 7 June 2015 (UTC)
Pywikibot compat will no longer be supported - Please migrate to pywikibot core
Sorry for English, I hope someone translates this. Pywikibot (then "Pywikipediabot") was started back in 2002. In 2007 a new branch (formerly known as "rewrite", now called "core") was started from scratch using the MediaWiki API. The developers of Pywikibot have decided to stop supporting the compat version of Pywikibot due to bad performance and architectural errors that make it hard to update, compared to core. If you are using pywikibot compat it is likely your code will break due to upcoming MediaWiki API changes (e.g. T101524). It is highly recommended you migrate to the core framework. There is a migration guide, and please contact us if you have any problem.
There is an upcoming MediaWiki API breaking change that compat will not be updated for. If your bot's name is in this list, your bot will most likely break.
Thank you,
The Pywikibot development team, 19:30, 5 June 2015 (UTC)
When is ContentTranslation coming to enwiki?
Latest comment: 9 years ago3 comments3 people in discussion
During my ventures on the French Wikipedia, I have noticed a new tool for translation that has been introduced. It is more or less in a state of uselessness, as it can't even translate French yet, but I am interested, what plans are there to introduce this tool onto enwiki? My nameisnotdave (talk/contribs) 20:31, 6 June 2015 (UTC)
My name is not dave, thanks for your interest. As a matter of fact, ContentTranslation (a.k.a. CX) was used to create more than 500 articles in the French Wikipedia, so while my opinion about ContentTranslation is certainly biased, I'm pretty sure it's not useless :)
It currently works for translation from English. There's a plan to enable it as a beta feature in the English Wikipedia in June. There will be a proper announcement about this very soon.
In the meantime, you are welcome to peek at the FAQ.
Recently (not sure when it started, maybe about 3-6 days ago), whenever somebody posts on my talk page, the notifications says "...posted on your talk page in "footer"...". I don't have a section called "footer" on my talk page. Weird. —DangerousJXD (talk) 02:35, 7 June 2015 (UTC)
phab:T99989 reports a patch in the works. They're shifting the deployment schedule right now, so I'm not sure when to expect it. Assuming that it isn't considered urgent enough for a special deployment, then this might reach us as early as Thursday of this week, or as late as next Wednesday. Whatamidoing (WMF) (talk) 06:27, 8 June 2015 (UTC)
Code tag broken?
Latest comment: 9 years ago7 comments2 people in discussion
Never mind, I read the documentation for the quote template and figured out that I needed nowiki(?) tags and also that quote isn't an inline template so I needed to use Tq instead. 3gg5amp1e (talk) 13:15, 8 June 2015 (UTC)
@3gg5amp1e: It wasn't a problem with either the <code>...</code> tags or the {{quote}} template, it was the bare pipes - you can't use these in the parameter of any template, since it terminates the parameter and starts another. You could have fixed it by using either of these methods:
You could use {{=}} but you'd need to do it for each instance. By "explicitly number" I mean to put 1= right at the start - and you only need that once per {{quote}}. Without that, everything up to the first bare equals is taken to be a parameter name, and everything after that as the value for that strangely-named parameter. --Redrose64 (talk) 14:54, 8 June 2015 (UTC)
Latest comment: 9 years ago6 comments3 people in discussion
I think the wikipedia app is not working properly: the capitalism, history, business and economics portals are not updating at all. I checked someone's Iphone, it seems to be working ok there. Lbertolotti (talk) 19:12, 3 June 2015 (UTC)
Toolserver complaint #68973: sorting by article class & importance no longer works
Latest comment: 9 years ago8 comments6 people in discussion
First encountered this last night, in the middle of sifting thru articles needing to be rated. One minute, the links to tools.wmflabs.org worked as expected, the next I encountered an error page. I thought it was some manifestation of transient weirdness, so I let it slide until this morning only to verify it's no longer working.
Steps to reproduce
Go to an article assessment overview table, where statistics for article class & importance are displayed.
Click on any of the numbers in the grid (say Importance "???", & Quality "Unaccessed")
Expected result: be directed to a page with a list of articles matching those features
screenshot of Tool server error page
Actual results: Tool server error page with the heading "No web service". See image file to right.
(Adding a screen shot for a bug report is an effing pain. It would be much easier if there was a licensing rationale "bug report" instead of having to go thru several tedious steps explain why I want to upload a file at en.wikipedia & not to commons, & IDGAF what license it appears under. As if anyone else would want to use the file.)
User environment: Firefox 24.1.0, KDE, Slackware Linux 14.1 (Although I doubt this has any effect on the server end.)
You don't have to choose a license for your file on phabricator. Adding the direct link to the failing tool, would have been simpler than a treasure hunt through wikipedia to find the link. And that error page tells you that the people owning this tool are Theopolism and Hedonil.. Have you considered contacting them ? —TheDJ (talk • contribs) 06:53, 6 June 2015 (UTC)
@Llywrch: For uploading Wikipedia/etc. screenshots for use in a Wikipedia discussion, see WP:WPSHOT - there are only two license templates that you need to consider - they differ only in whether or not the puzzleball logo (or equivalent) is included in the screenshot. --Redrose64 (talk) 13:38, 6 June 2015 (UTC)
Much thanks for the tips on uploading screenshots, but anyone else seen this malfunction, or have a clue what's going on? Having these links work helps to speed up clearing up the backlog of unassessed & ungraded articles. -- llywrch (talk) 15:35, 6 June 2015 (UTC)
Thanks for this note. I've e-mailed User:Coren, who usually knows what's going on with Tool Labs, but I don't know how soon you might expect a response/if he's around this weekend. Whatamidoing (WMF) (talk) 05:57, 7 June 2015 (UTC)
As far as I can tell, at this time, the enwp10 tool is functionning properly. If the issue is still live, could someone provide me with a link to the broken page? — Coren(talk)15:40, 8 June 2015 (UTC)
The problem seems to have fixed itself; no sign of the issue when I followed the links earlier today. Apparently someone was doing something in the Wikimedia labs (not the toolserver, my confusion), decided the change wasn't what they wanted, & stopped doing that something -- thus the enwp10 tool is now working properly. (This is one of the drawbacks of abstraction in software: change something on one level, & it can cause failures in an interface on a higher level you didn't think was related.) -- llywrch (talk) 07:16, 9 June 2015 (UTC)
Phantom citations
Latest comment: 9 years ago5 comments2 people in discussion
Currently, we have a phantom reference appearing twice in the reference list, but not visible in the article. Can we have the backend software fix this problem? Frietjes (talk) 14:50, 7 June 2015 (UTC)
I think the parser adds it once as the caption (which won't display, but it doesn't know that at this step yet, and once as the alt attributes (from which it will get sanitized, but it doesn't know about that either at this stage. I'm not sure if that is fixable... You should probably file a bug report in phabricator however. —TheDJ (talk • contribs) 15:08, 7 June 2015 (UTC)
It's another symptom of phab:T101390 I suspect. This seems to be NEW behavior. (grah, i see in the discussion that Jack was already on this.. (and the cause of it)) —TheDJ (talk • contribs) 17:24, 7 June 2015 (UTC)
Latest comment: 9 years ago2 comments2 people in discussion
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Changes this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from June 9. It will be on non-Wikipedia wikis from June 10. It will be on all Wikipedias from June 11 (calendar). [44][45]
If you use the Monobook skin, the buttons and other controls now look more the same in VisualEditor and other tools. [46]
When you edit links and other items in VisualEditor, you now need to apply your change before closing the tool. [47]
The title of dialogs is now easier to see when it is near long buttons. [48]
Meetings
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on June 9 at 19:00 (UTC). See how to join.
Minor clarification for another item: "The title of dialogs is now easier to see when it is near long buttons" refers specifically to VisualEditor. (Also, that problem doesn't seem to appear if your account is set to display in English.) Whatamidoing (WMF) (talk) 01:21, 9 June 2015 (UTC)
Old versions (moved from Policy Pump)
Latest comment: 9 years ago2 comments2 people in discussion
It should be possible to transclude or substitute an old version of a template, as well as to redirect to an old version of a page and to display old versions of images. GeoffreyT2000 (talk) 04:31, 8 June 2015 (UTC)
@GeoffreyT2000: It might be, but I would ask what are you trying to achieve by doing this? I think you might have a specific example in mind. --User:Ceyockey (talk to me) 00:10, 9 June 2015 (UTC)
Reporting error: Navbox template, title not updating
Latest comment: 9 years ago7 comments2 people in discussion
Greetings, Sorry if I'm doing something wrong but even after doing the purge option, the updated title refuses to change. Help please... The template is at Template:Vatican City topics. Note: I had to go back to the template's creation to find the Title line as it disappeared at the next update. Regards, JoeHebda (talk) 20:30, 9 June 2015 (UTC)
Thanks, I missed that. Just wondering how or where is the linkage that connects the word topics with the correct article? I looked thru the docs & can't find anything that shows. JoeHebda (talk) 20:51, 9 June 2015 (UTC)
Yikes, sort of like stuffing a variable into a string of characters (as in good-old BASIC programming). Thanks for the answer. Unfortunately there is now a Redirect from the List of xxxx -related topics to another article name instead. Is there any way of overriding this? Or maybe a different template? JoeHebda (talk) 22:58, 9 June 2015 (UTC)
I wouldn't worry about it; redirects are not a big deal. So long as the link takes you to the right place... Alakzi (talk) 23:17, 9 June 2015 (UTC)
Latest comment: 9 years ago3 comments3 people in discussion
I don't know if it's just my cellphone (an iPhone 5), but I cannot edit on the Mobile setting. I can press "Edit" and get the markup language screen to appear, but whenever I try to edit text, the cursor zips back up to the top of the paragraph. Is anyone else having trouble with Mobile editing?OnBeyondZebrax • TALK02:07, 9 June 2015 (UTC)
The mobile version has always been crappy for me, iPhone 5 as well. I've been using the desktop version on my phone for years now.—cyberpowerChat:Limited Access20:23, 9 June 2015 (UTC)
Latest comment: 9 years ago6 comments3 people in discussion
I am trying to edit a list of references, but I can not figure it out. I was trying to remove an unneeded reference, and I was using VisualEditor. I tried editing the source, but I saw a reflist template. Help? ThatKongregateGuy (talk) 20:46, 9 June 2015 (UTC) — Preceding unsigned comment added by ThatKongregateGuy (talk • contribs) 20:39, 9 June 2015 (UTC)
@ThatKongregateGuy: Unless list-defined references are used in the article (which is rare), the References section does not actually contain the references. They are in the body, and the software collects them and displays them where the {{reflist}} template is located. So, to remove a reference from the reflist, you locate it in the body and remove it there. ―Mandruss☎20:44, 9 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
I've come across an empty category, Category:LGBT politicians from the Maldives, which was formerly not empty (among other proofs, it's existed since 2011 and was created by a user who knows very well not to create empty categories) — so what I need to do is to locate who used to be in the category, so that I can properly determine whether its emptying was appropriate or out-of-process.
I know there's a way to run this check, as I previously encountered a similar situation earlier this year (see this discussion) which was successfully resolved. The solution involved scanning for the category in an older database dump; as was true the last time, I've already tried Wayback Machine with no luck. So I wanted to ask if anybody could help out with this matter as well.
At the same time, I also need to run a check on Category:LGBT people from the Maldives, which currently contains only the first empty subcategory — that might have always been the case even before the first category was emptied, but I can't let it go without verifying that.
If I wonder about the context of an edit then I often examine other edits at the time by the same editor. In this case [51] shows the categories were created after [52]. The article history shows what happened to the LGBT claim and category which was sourced to [53]. I haven't investigated beyond that. PrimeHunter (talk) 22:49, 9 June 2015 (UTC)
@Bearcat: Another data point: I scanned the 20141208 dump for articles containing "LGBT politicians from the Maldives" and found only Abdulla Hameed, and for articles containing "LGBT people from the Maldives" and found none. -- John of Reading (talk) 07:40, 10 June 2015 (UTC)
Okay, thanks. Going back to look at the history on that article, it looks like the category was created and added on the basis of an allegation about Hameed's sexuality by a person who may have had a POV agenda — and not, as required by WP:BLP, any indication that he had ever actually come out in his own words. So I'm not going to repopulate it, as it looks like the removal was actually proper and correct. Bearcat (talk) 16:29, 10 June 2015 (UTC)
This might be a massive problem
Latest comment: 9 years ago11 comments6 people in discussion
No. I don't want to watch Wikidata. My suggestion is since the term "any can edit" doesn't apply to Wikidata, that only registered users should be allowed to touch Wikidata, and only then when they are autoconfirmed. Before that they can only submit changes through talk page requests. The potential for damage is extraordinary in this case.—cyberpowerChat:Limited Access21:19, 9 June 2015 (UTC)
Even that's not good enough. There are hundreds of thousands of BLPs out there. The App needs to stop picking up Wikidata information. Black Kite (talk)21:24, 9 June 2015 (UTC)
If you watch Wikidata - which I did for about two days - your watchlist fills up with crud like some bot adding or removing a property that is only of relevance to the French/German/Hindi/Italian/Japanese-language Wikipedias. I mean, how is this of relevance to me here on En.wp? That's what happens if Template:S-start is watchlisted and you go for Show Wikidata. --Redrose64 (talk) 22:20, 9 June 2015 (UTC)
Which is why I suggested that Wikidata should only be touchable by auto-confirmed users. Wikidata is not Wikipedia, and shouldn't be treated as such. It's a data platform for managing interwiki connections, and vandalizing it can have global consequences, as evidenced here.—cyberpowerChat:Limited Access05:56, 10 June 2015 (UTC)
Latest comment: 9 years ago6 comments4 people in discussion
We have an alias called [[tools:]] (tools:) and it goes to a dead site. toolserver.org has moved to tools.wmflabs.org .. is it possible to update the alias? It won't break anything since the alias is currently broken. -- GreenC00:39, 10 June 2015 (UTC)
There's a new prefix, toollabs:. Presumably, the reason it's not just "tools" is that they were both used during the migration from Toolserver to Labs; you might want to suggest that the "tools" prefix is retargeted on meta. See meta:Interwiki map for the full list of prefixes. Alakzi (talk) 00:43, 10 June 2015 (UTC)
@Green Cardamom: First, commenting on your post to Wikipedia talk:Namespace#Tools alias (tools:), "tools:" isn't a namespace, it's an Interwikimedia link (or if you prefer, an interwiki prefix, although this term is also used for interlanguage links). Second, toolserver.org didn't move to tools.wmflabs.org - they are two totally separate systems operated and funded by different organisations. If toolserver.org had been moved somewhere (anywhere) it would have made a lot of things so much easier, but it wasn't - it simply had the funding cut and the plug was finally pulled at the start of July 2014. Tool operators were given something like two years notice that this would happen.
To return to the problem. Interwikimedia links are not maintained locally, there's a central table, it's done that way to ensure that each prefix has exactly the same meaning regardless of which wiki it's being used on. I doubt if they will be willing to repurpose the tools: prefix, for several reasons: it's less than a year since toolserver.org went down; the paths (and hence the URLs) on tools.wmflabs.org are all different from toolserver.org (there isn't a one-to-one mapping either); not all of the toolserver.org tools were migrated to tools.wmflabs.org (and some will never be). --Redrose64 (talk) 09:41, 10 June 2015 (UTC)
Changing tools: wouldn't fix links but break them. If a tool on the old toolserver has a replacement somewhere then there is often a redirect so old links still work. The url structures are different so if tools: pointed to https://tools.wmflabs.org/ then those links wouldn't work. PrimeHunter (talk) 10:50, 10 June 2015 (UTC)
Yeah, this is called the "Bonnie and Clyde problem" on Wikidata. Since, at the moment, the German Wikipedia doesn't have articles (or even redirects) that correspond directly to either of the English articles, manual interwikis are the way to go for now. --Yair rand (talk) 12:19, 11 June 2015 (UTC)
Thanks. That is the way it was done the last time I was active here, but I was trying to use this new-fangled way. I don't know why the German Wikipedia has one article on two separate units— I will have to contact one of my German Air Force friends. --21lima (talk) 12:24, 11 June 2015 (UTC)
Edit conflict bug?
Latest comment: 9 years ago2 comments2 people in discussion
This morning, I was edit conflicted while making an edit to this section. I'd edited just that section by clicking on the 'edit' link to the right of the section title. When I attempted to save, I was edit conflicted. Yet, in the new edit window given to me, there was no additional comment in that section. So, I re-wrote what I said and saved it. This inadvertently removed a comment from the person with whom I edit conflicted...IN A DIFFERENT SECTION [54]. Thankfully someone else picked up on it and restored it [55]. I never noticed. Anyone want to opine as to what is happening here? --Hammersoft (talk) 14:20, 11 June 2015 (UTC)
I have a new user script to share...: Actual Live Preview.
With this script, if your window is wider then 1200px, it will split it in half and show you your edit surface on the left and a 'update while you type' content preview on the right. The only requirement is that you enable "Live preview" in your preferences. Enjoy. —TheDJ (talk • contribs) 07:55, 10 June 2015 (UTC)
Found another problem. At least in Latvian Wikipedia, I got such problem: screenshot. The same is at Vector and Monobook, tested in some few articles. Firefox+XP and 1440×900 big resolution. --Edgars2007 (talk/contribs) 15:34, 14 June 2015 (UTC)
Wikidata arrogance
Latest comment: 9 years ago5 comments2 people in discussion
It appears that wikidata people actually write: We need ..., We have to ..., You can do that manually [for 7500 facts?]. While really the (institutional) editor offers to add facts to wikipedia. Sic transit whatever gloria wikidata. -DePiep (talk) 01:20, 11 June 2015 (UTC)
By Jimbo, a "wiki" is a website with open editing, where anyone can edit. Four years on, and wikidata still is not a wiki. -DePiep (talk) 11:17, 12 June 2015 (UTC)
Wikidata, after hosting the iw's nicely, is a broken wiki. It does not serve. Nothing perspective, it does not function. -DePiep (talk) 21:14, 12 June 2015 (UTC)
number of contributions by a user
Latest comment: 9 years ago4 comments3 people in discussion
The simplest way is via your own Special:Preferences, it will show your number of edits as well as other account information on the User Profile tab. There are also links at the bottom of everyone's user contribution page (yours is at Special:Contributions/Bevo) linking to various tools to see Edit Count and Global contributions and associated statistics. Nanonic (talk) 20:44, 11 June 2015 (UTC)
For about a month now, I've been unable to perform searches on the mobile site's search box from my Blackberry Bold 9900, due to a big, black bar that covers the entire thing. What gives? lavender|(formerly HMSSolent)|lambast23:54, 11 June 2015 (UTC)
BTW. This looked like a very small thing and I'm sure all reporters were like: "Why don't you 'just fix it' already"... Just to bring some perspective: This issue seems to have required multiple debug sessions over multiple weeks by multiple people, was not always reproducible and in the end changed changed 8 files... Simple things often aren't simple :) —TheDJ (talk • contribs) 10:02, 12 June 2015 (UTC)
Latest comment: 9 years ago15 comments9 people in discussion
At some point between 07:38 and 10:00, 12 June 2015 (UTC), the user preference "Always use a secure connection when logged in" lost its effect, and regardless of its setting, Wikipedia became HTTPS only. Why was this done, and how can it be undone? Some users are unable to use HTTPS connections. --Redrose64 (talk) 10:33, 12 June 2015 (UTC)
Having the same problem, which for me means I will be making very terse edit summaries until this is fixed. Windows uploaded and installed updates yesterday, so it may have something to do with that. I get forced into the secure server even before logging on, and no amount of "s" deletion (https^h) takes me off the secure server. I checked the Favorites properties and it still just has "http", so something's rotten in IE methinks. – Paine11:12, 12 June 2015 (UTC)
@Paine Ellsworth: Are you referring to phab:T55636? Which version of IE are you using? As for the Firefox issue (EchetusXe), I note that autocompletion of edit summaries works very well for me in Firefox (latest version) with HTTPS. Perhaps you need to check browser or add-on settings. — This, that and the other (talk)13:10, 12 June 2015 (UTC)
To editor This, that and the other: – Yes re. phab:T55636. I've known for a long time that my IE-10 in Win8 would not work with WP as it concerns the edit summary autofil challenge. My workaround was to simply not check the HTTPS box in prefs. This, however, is a new problem that doesn't even allow me access to the Wikipedia non-secure server, HTTP, when I'm not logged in. I open the browser, I click the Fave link (HTTP only) and then am whisked away to the HTTPS Wikipedia server. Even if I delete the S in HTTPS up in the URL field, the moment I hit ENTER or GO, the S reappears. Apparently, the WMF has found a way to make it so even non-registered users must use the HTTPS, which is a total drag all around! – Paine14:07, 12 June 2015 (UTC)
I feel your pain, paine, this also means users from China can NO LONGER access the English wikipedia (this virus is currently limited to enwiki) since https is BLOCKED there, and also, the developers have enforced this without discussion and as per our community first policy and thus we must now vote on removing this..this is the 2nd time this has happened, btw, there is no longer a "forcedhttps" cookie as i have been told by the developer on IRC that the "Always use a secure connection when logged in" option is now well 'utterly useless'....--Stemoc14:28, 12 June 2015 (UTC)
Just to be clear, nobody is geographically exempt. The geographic exemptions only applied to the previous use SSL for logged in users. The use SSL for everyone applies to everyone visiting one of the 8 languages that it has been turned on for (Which includes ZH, but not the Iranian languages) regardless of geographic location. By what I have heard, users from China have not been able to access zh.wiki.x.io (both https and normal http) since several weeks prior to this switch, and they are still able to access en.wiki.x.io over HTTPS, so this switch hasn't had any immediate affect on Chinese users. Bawolff (talk) 06:07, 14 June 2015 (UTC)
Wikimedia devs, you know those people we pay to mess up the site decided our input doesn't matter even though Jimmy explicitly said that we will NOT force IP users into HTTPS during the last Wikimania..--Stemoc13:53, 12 June 2015 (UTC)
Eh nope. It is unsupported. wikilook. It needed a user "patch" to work with later firefox versions, but AFAIK there is no equivalent out there. This dictator driven change is un-representative... ˥ Ǝ Ʉ H Ɔ I Ɯ (talk) 11:43, 13 June 2015 (UTC)
You are complaining to you family because the guy who has been driving you around for 4 years was arrested for driving without a license. You should be thankful for a place where we collectively concluded that driving people around without a driving license is a bad idea, and you should stop hitching rides with that person, no matter how useful he might be to you. —TheDJ (talk • contribs) 12:08, 13 June 2015 (UTC)
GreaseMonkey scripts
It seems that for some people Greasemonkey scripts don't work anymore on https. I'm not entirely sure how and why, but I have figured out that if you add explicit @grant's to the script that they DO work. I'm a little dumbfounded that we didn't get more reports about such failures over the past 2 years... I guess it proves that greasemonkey users don't really understand what they are using. :) —TheDJ (talk • contribs) 19:20, 12 June 2015 (UTC)
Template works on some pages but not others (possible Lua/Wikidata issue?)
Latest comment: 9 years ago3 comments2 people in discussion
Resolved
Template:Tracks Wikidata is a new template I've recently created. It generates a box with a list of Wikidata properties, with Module:Uses Wikidata allowing for unlimited parameters, and pulling the labels for the properties from Wikidata – or the text "NO LABEL" if it can't find a label (usually because the property has been deleted, or has not yet been created). While the template works in the /testcases page, in my sandbox, and Template:Commons category/doc, it is not working for Template:Coord/doc (shows NO LABEL for P625 instead of coordinate location). The Parser profiling data at the bottom of the edit window for Template:Coord/doc doesn't show anything out of range:
Latest comment: 9 years ago12 comments5 people in discussion
There seem to be two problems with Graphs. The new template {{GraphChart}} has a number of examples on its documentation page, Template:GraphChart/doc. But none of them appear on the main template page despite being transcluded. Setting up a test at User:JohnBlackburne/GraphTest to try and reproduce this I noticed another problem. That transcludes fine but the animation where e.g. the colours change as you mouse over the chart or part of it only work in preview, not after the page is saved. This is the case on both pages, including where one is transcluded. I tried disabling navigation popups in case that was causing it but it made no difference. I'm using the latest version of Safari in Mac.--JohnBlackburnewordsdeeds19:02, 12 June 2015 (UTC)
The first, I'm not sure why that is. But the latter is intentional. In preview it uses Javascript (also with future live editing in mind for instance), but in content, you only get the image representation (because the JS is slow and adds a lot of bytes to the pages). —TheDJ (talk • contribs) 19:10, 12 June 2015 (UTC)
Renders the JS rather pointless then, if it's only seen when editing. I can see why it would be a performance problem, but could it be made an option? I won't be the last to notice this and wonder why it’s not working.--JohnBlackburnewordsdeeds19:16, 12 June 2015 (UTC)
Ah, purging didn't fix the transclusion but a null edit did; very odd. And the second issue seems to be a will not fix one so this is now resolved.--JohnBlackburnewordsdeeds19:24, 12 June 2015 (UTC)
Render to js happens during preview because there is a limitation in how its implemented, making the static image only work when the page is saved (Not only saved, also the most recent version). Similarly, if your viewing the graph of an old revision of the page, the graph won't work unless the image happens to be cached. Bawolff (talk) 22:29, 12 June 2015 (UTC)
I see. The problem is it’s not just rendering to JS but doing mouseover based highlighting/animation. It’s a problem as I’m sure I won’t be the first person to see this and wonder why it’s working in preview and not once saved. If there is no way to make it an option then it would probably be best to disable the animation code altogether, so preview returns to being WYSIWYG at least for this feature.--JohnBlackburnewordsdeeds23:39, 12 June 2015 (UTC)
JohnBlackburne: A bit of context, this new extension was imported from the German Wikipedia. The Germans also have a map template that I can't seem to import the JSON file for, so I left it alone for now. Yurik pointed out on my talk page that, in the absence of the should-have-happened-ages-ago interwiki transclusion feature, the best solution for the quickly oncoming onslaught of template sprawl is the creation of a bot that copies all of the templates' content every once in a while from a master template somewhere: either on MediaWiki or on Meta. I am afraid that it is unlikely that someone will be found to do it fast enough, before the organic growth Wikipedia is so famous for smothers us all in the heat death of yet another thousand modules and templates on the various wikis all maintained independently but doing the same damn thing. ResMar03:50, 13 June 2015 (UTC)
Don't hold your breath for interwiki transclusions. People have been talking about it since forever (There was even a gsoc project in 2010), and so far we are nowhere closer to having it then we ever were before. Bawolff (talk) 05:20, 13 June 2015 (UTC)
A framework bot is something that can be written, however. It just takes someone with a few days of free time and some experience with Labs to do the work. ResMar13:40, 13 June 2015 (UTC)
The behavior seems to be coming from the graph extension's caching mechanism. If you look at the generated image URLs you will notice that they follow the name scheme "<name of transcluding page>/<ID>.png" for whatever reason as <ID>.png should just suffice. Therefore if a page A transcludes a template B with a graph image, it expects the image at A/<ID>.png but there was only an image generated for B/<ID>-png. If make a null edit at A the graph extension is then forced to generate the A/<ID>.png. --Mps (talk) 16:28, 13 June 2015 (UTC)
This should not be too much of a problem. It probably happened with the template as it was created first: seems I’m not the only one who puts the {{documentation}} link in a template then follows it to complete documenting it. Normally for transcluded templates the template will be made first. It will mean a lot of redundancy though with one copy of the image for every time the template is transcluded. Plus it will still break with changes to templates, especially adding a graph.--JohnBlackburnewordsdeeds16:47, 13 June 2015 (UTC)
Where do you find a browser without SSL support these days? The only program I could find which didn't support SSL was Wget on my phone, although Wget supports SSL on my computer. --Stefan2 (talk) 22:39, 13 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
I don't know if this the correct place (if not please point me to the proper one), but I want do update the VisualEditor special character list with typographic quotation marks so that users can more easily use them.
Latest comment: 9 years ago5 comments3 people in discussion
Hi all, am trying to retrieve a page for my nearest photographer tool (source). It worked like a charm some days ago, but the (apparantly with that switch to https only) it doesn't anymore.
I'm tying to retrieve page source code by this http request
GET https://en.wiki.x.io/w/index.php?action=raw&title=Wikipedia%3AWikipedians%2FPhotographers HTTP/1.0
Host: en.wiki.x.io
User-Agent: script by de_user_Flominator
> GET /w/index.php?action=raw&title=Wikipedia%3AWikipedians%2FPhotographers HTTP/1.1
> User-Agent: curl/7.37.0
> Host: en.wiki.x.io
> Accept: */*
>
< HTTP/1.1 200 OK
[...]
Notice the difference in the GET line. It should start with the path, you might be confusing the proxy by also including the protocol and host there. --cesarb (talk) 17:57, 13 June 2015 (UTC)
Works for me, using the full GET link. (Are you sure you are actually connecting on port 443 and encrypting on TLS, instead of just making an HTTP request where the GET line is a full uri with https. For the connection to work, you need a real https connection)
bawolff@Bawolff-L:~$ openssl s_client -connect en.wiki.x.io:443
CONNECTED(00000003)
---
Certificate chain
0 s:/C=US/ST=California/L=San Francisco/O=Wikimedia Foundation, Inc./CN=*.wiki.x.io
i:/C=BE/O=GlobalSign nv-sa/CN=GlobalSign Organization Validation CA - SHA256 - G2
1 s:/C=BE/O=GlobalSign nv-sa/CN=GlobalSign Organization Validation CA - SHA256 - G2
i:/C=BE/O=GlobalSign nv-sa/OU=Root CA/CN=GlobalSign Root CA
---
Server certificate
-----BEGIN CERTIFICATE-----
[.. certificate omitted..]
-----END CERTIFICATE-----
subject=/C=US/ST=California/L=San Francisco/O=Wikimedia Foundation, Inc./CN=*.wiki.x.io
issuer=/C=BE/O=GlobalSign nv-sa/CN=GlobalSign Organization Validation CA - SHA256 - G2
---
No client certificate CA names sent
---
SSL handshake has read 3658 bytes and written 434 bytes
---
New, TLSv1/SSLv3, Cipher is ECDHE-RSA-AES128-GCM-SHA256
Server public key is 2048 bit
Secure Renegotiation IS supported
Compression: NONE
Expansion: NONE
SSL-Session:
Protocol : TLSv1.2
Cipher : ECDHE-RSA-AES128-GCM-SHA256
Session-ID: A12E6886D254C01CC094FA3254BD7C774CFAA1C6DBFD0CA2DC7DA4DBEAF4D17B
Session-ID-ctx:
Master-Key: 7BA289973BB568325ECAF13803B607441D32C9A56ECB0D7D8B7CB33B315D42B1976582F28228A8972BB278900AA3BA30
Key-Arg : None
PSK identity: None
PSK identity hint: None
SRP username: None
Start Time: 1434215498
Timeout : 300 (sec)
Verify return code: 20 (unable to get local issuer certificate)
---
GET https://en.wiki.x.io/w/index.php?action=raw&title=Wikipedia%3AWikipedians%2FPhotographers HTTP/1.0
host: en.wiki.x.io
user-agent: bawolff-test
HTTP/1.1 200 OK^M
Server: nginx/1.6.2^M
Date: Sat, 13 Jun 2015 18:06:31 GMT^M
Content-Type: text/x-wiki; charset=UTF-8^M
Content-Length: 46877^M
Connection: close^M
X-Powered-By: HHVM/3.6.1^M
X-Content-Type-Options: nosniff^M
Vary: Accept-Encoding^M
Last-Modified: Sun, 07 Jun 2015 10:06:03 GMT^M
X-Varnish: 744213003, 3153064536^M
Via: 1.1 varnish, 1.1 varnish^M
Accept-Ranges: bytes^M
Age: 0^M
X-Cache: cp1066 miss (0), cp1066 frontend miss (0)^M
Strict-Transport-Security: max-age=86400^M
Cache-Control: private, s-maxage=0, max-age=0, must-revalidate^M
X-Analytics: https=1^M
Set-Cookie: WMF-Last-Access=13-Jun-2015;Path=/;HttpOnly;Expires=Wed, 15 Jul 2015 12:00:00 GMT^M
^M
A great many people have contributed photographs to Wikipedia. Many of those photographs were found on the Net; others were created by the Wikipedian. To appear on this list you must be a [[Wikipedia:Wikipedians|Wikipedian]] who has either contributed '''your own work''' to the project or is willing to do so.
[.. rest of response body omitted..]
p.s. It should be noted, that while it technically works (If you use right port and tunnel over tls), per RFC 1945, you are only supposed to use absolute URI's in the GET line if you are talking to a proxy. Bawolff (talk) 18:25, 13 June 2015 (UTC)
How to question: How to get a list of all .svg files, i.e. their really correct locations, on commons.wikipedia.
Latest comment: 9 years ago2 comments2 people in discussion
Hi! I asked at https://en.wiki.x.io/wiki/Wikipedia:Help_desk first (#commonswiki-20150417-all-titles_contains_names_of_deleted_files), and it was suggested that I best try to ask here. I try here second so.
So I'm searching for all svg-files on wikipedia. While doing this I found out that commonswiki-20150417-all-titles seems to contain deleted files.
1. Is there another list-file similiar to commonswiki-20150417-all-titles that contains all deleted files, so I could do a cross check?
2. Is there an all-titles list-file that doesn't contain the deleted files?
3. Bonus question: Some of those files in commonswiki-20150417-all-titles have been renamed. In line with the two first questions: How to find out the current name?
I only want to download the svg-files, and not to download like 250% overload. Out of the 20 files I tried around half were renamed.
And some were deleted. Only around 5 were actually to find under the name given in the all-titles file.
By now I would have been faster if I just grabbed a whole wikipedia-dump. :/ But then that I really don't want to do .
Latest comment: 9 years ago4 comments2 people in discussion
A change was recently made to the {{Phanerozoic 220px}} template so that it uses the Ꞓ symbol (Unicode 792, I believe). This does not display in my browser (Firefox), and it's a template that's very widely used on the pages that I read, so I rather wish that it did. I've tried following the instructions here, but, insofar as I can even understand that page (it's quite poorly written, in my personal opinion) I appear to have done everything requested, and it's not made the slightest bit of difference. So, the question is, is this a widespread problem, or just me, and if the latter, is it possible for me to fix it? There was some brief discussion of the change to "Ꞓ" from the previously used "Є" n the relevant template talk page, which may indicate that the issue isn't entirely unique to me... but how common it is for users in general, I couldn't say. Anaxial (talk) 07:41, 14 June 2015 (UTC)
The template appears at the top of the infobox in thousands of articles. For example, see Amphibian where "PreꞒ Ꞓ" denote periods under "Temporal range". Prior to the 20 April 2015 edit, that would have displayed as "PreЄ Є". It's fine to use correct Unicode symbols in articles that need to discuss relevant characters, but presenting a box to general readers in these circumstances is not helpful. The correct approach would be to raise the matter at a suitable wikiproject and get consensus about what is needed—try WT:BIOL. Johnuniq (talk) 08:01, 14 June 2015 (UTC)
I just noticed that Cambrian has an endorsement of "barred capital C" ⟨Є⟩, so reverting the change seems desirable, but a note at WT:BIOL may as well happen first. Johnuniq (talk) 08:04, 14 June 2015 (UTC)
Yes, if it isn't just me (or some small minority of readers) then that would be the obvious solution. Thanks. Anaxial (talk) 08:12, 14 June 2015 (UTC)
Most of them probably do. It was removed in [57] for no apparent reason. We have 4.9 million articles written by tens of thousands of people. It's not possible to do everything identically everywhere. If you want coordinates in an article then just add them. No discussion is required here. PrimeHunter (talk) 16:31, 14 June 2015 (UTC)
Thanks. Did that. Thought there was a consensus about not having coordinates in country articles or something. --Flominator (talk) ~ — Preceding undated comment added 18:08, 14 June 2015 (UTC)
"Loss of session data" error on Save page
Latest comment: 9 years ago24 comments12 people in discussion
Sorry! We could not process your edit due to a loss of session data. Please try again. If it still does not work, try logging out and logging back in.
For months, I guess, I've been getting this on about one in ten saves. The second attempt always works, so it's not a serious hindrance, but it's definitely annoying, distracting, and a bit nerve-rattling, especially in addition to the other weird editing glitches we've seen lately.
A look at this page's archives shows other users having the problem, going back years, but I don't see a clear resolution there.
I have Firefox 38.0.1 on Windows 7. I'm using the NoScript 2.6.9.26 Firefox extension, but these WP-related sites are whitelisted: wikipedia.org, wikimedia.org, wmflabs.org, wmfusercontent.org. I also have the Adblock Plus 2.6.9.1 extension.
(edit conflict) A few times today already, I'll get the loss of session data error after taking a mere five seconds to make a correction to an article when I used to have to wait at least a few minutes with the edit window open before saving for that to happen. Luckily, I just need to press save again to fix the problem. Dustin(talk)06:15, 8 June 2015 (UTC)
If someone could say that they have the problem without one or both of the Firefox extensions (or the Chrome version), that would rule them out as the culprit. ―Mandruss☎06:23, 8 June 2015 (UTC)
This happens to me, and I have both of these extensions installed (but disabled for these domains). However, it doesn't happen to me nearly as often. Usually, I have to have an editing window open for ≥15 minutes to get this error. WhatamIdoing (talk) 06:28, 8 June 2015 (UTC)
It's possible it's far less than one in ten for me. I of course notice when the error happens, but I'm not aware of how many saves have worked since the previous error; I'm not paying attention to that. I could start keeping a record, but I don't know that a more accurate number would help identify the problem. ―Mandruss☎08:55, 8 June 2015 (UTC)
I think I see it more for edits that have been open for awhile, but I'm certain it happens sometimes after a ~30-second edit session. ―Mandruss☎06:59, 8 June 2015 (UTC)
"Please try again" suggests that the developers anticipated the possibility of "loss of session data", and were unable to come up with a way to prevent that possibility. The career software developers among us can attest that such situations are possible, although rare. Some aspects of the software environment are beyond our control, and the rare nut is uncrackable. But it would be nice to have the developers look at the problem and say whether that is the case here. If it is, that would be a good thing to know going forward. ―Mandruss☎07:52, 8 June 2015 (UTC)
Firefox 38.0.5 and some earlier versions. I don't have eother NoScript or AdBlock. I've not counted the rate at which the "Loss of session data" error occurs, but it's more likely for a longer interval between "edit" and "save page". But it has happened for a simple sub-10-second typo fix on more than one occasion. --Redrose64 (talk) 08:14, 8 June 2015 (UTC)
It suddenly became far more frequent about 10 days ago - and most of my edits are just typos (Windows 7, IE11, Vector & no fancy add-ons) - Arjayay (talk) 09:04, 8 June 2015 (UTC)
This edit took no more than ten seconds from clicking "[edit]" to Save page, during which time I did not navigate off the page, switch windows or do anything other than type - yet it still threw "Sorry! We could not process your edit due to a loss of session data." Ridiculous. --Redrose64 (talk) 12:48, 10 June 2015 (UTC)
I have strong suspicions that this is related to #Post_not_showing_up_immediately. I suspect the same change that is causing that problem (still not tracked down btw), is also causing this problem. Possibly because outdated tokens or timestamps are included in edit pages or something. —TheDJ (talk • contribs) 12:58, 10 June 2015 (UTC)
For me, the "session data" issue pre-dated the other by at least months — maybe many months, I can't recall. ―Mandruss☎16:59, 11 June 2015 (UTC)
Is this related? When these errors start happening, I typically also get "token" errors when I use the Upload Wizard. (If my memory is correct, the message is "token not recognized".) In this case, all my work is completely lost, and I have to restart the upload from scratch. Browser back buttons work just fine with loss of session data, but not with these token failures. I presume this is simply because Uploading doesn't come with a Preview button, so my browser when I upload never has a record of what I just did. I also presume I could avoid problems if I got in the habit of right-clicking the Upload button. I also presume that the simplest temporary fix would be that when the Upload button is activated, the page as it is also added to the browser history. Choor monster (talk) 13:08, 10 June 2015 (UTC)
About once a day, I try to make an edit, and get the error message that my edit could not be made due to a loss of session data. It tells me to try again, and says that if that does not work, I should log out and log back in. I press the Save button again, and it works. (I don't have to log out and log in.) Should I just ignore it, or can I do something to minimize the frequency of these glitches? Robert McClenon (talk) 15:23, 10 June 2015 (UTC)
I'm occasionally getting these too. It seems that this message is most likely to occur when I start editing an article and keep it open in the edit mode for a long time (over an hour). For me, too, the message goes away after pressing Save. I'd also be interested in knowing the cause of this message and whether it can be prevented. (I'm using Firefox 22.0 on Win7).—Ëzhiki (Igels Hérissonovich Ïzhakoff-Amursky) • (yo?); June 10, 2015; 15:29 (UTC)
It happens for me when I haven't been editing very long. Since we are using different web browsers, they might be different issues. Robert McClenon (talk) 15:53, 11 June 2015 (UTC)
I was coming here to complain myself. I'm getting this 70% of the time now. It's becoming more than just an annoyance. Also some pages when edited load out of date and need to be refreshed. There are some weird glitches going on right now.—cyberpowerChat:Online18:25, 11 June 2015 (UTC)
Noting that I too have had this happening with increasing frequency, about every third edit today, and it has also occurred when accessing restricted pages (in my case, the checkuser interface). Using Win7 with FF38.0.5. I have noted this on the current phab report. Risker (talk) 23:08, 14 June 2015 (UTC)
false positive reports not being handled
Latest comment: 9 years ago3 comments3 people in discussion
There's some attacks on AES 256 that suggest AES 256 bit is not really all that much better than 128 bit AES (From what I understand. Not a crypto expert). Bawolff (talk) — Preceding undated comment added 20:06, 13 June 2015 (UTC)
@Rich Farmbrough: I am not an expert either, but it seems like currently the best cipher suite recommended in the industry is ECDHE-RSA-AES128-GCM-SHA2561, which is what Wikipedia servers are actively preferring right now. After all, we are using 2048-bit RSA asymmetric keys for key exchange, which should be of similar strength to a 112-bit symmetric key. 2 (Less than 128-bit) If someone is able to bruteforce a 128-bit key for AES (which is almost impossible right now), they might as well be able to factor a 2048-bit RSA key and perform a successful Man-in-the-middle attack. Tony Tan·talk20:58, 13 June 2015 (UTC)
Interestingly that document suggests 15360 bit RSA keys to be equivalent to 256 bits of security. This much data might create difficulties in the TCP/IP window, causing more overhead, but again it might not. It also suggests that 256 128+ bits of security is needed for data that should be secure to 2030. It's certainly an interesting perspective on the security of the sites. All the best: RichFarmbrough, 00:41, 14 June 2015 (UTC).
The certificate that signs Wikimedia's certificate uses 2048 bit RSA. Using an RSA modulus higher than that is pointless, since then you'd just forge the CA's signature instead. Bawolff (talk) 04:53, 14 June 2015 (UTC)
Doesn't matter. You would have to forge the certificate now for a MITM attack. Compromising the key exchange can be done in (let us say) 2030, when <128 bits of security is considered insufficient. (Perfect forward secrecy key exchanges would defeat this, if we don't have it it might be a better enhancement than forced HTTPS.)
We do indeed use forward secrecy in most cases (It requires the client to support it as well as the server). Bawolff (talk) 21:01, 15 June 2015 (UTC)
@Bawolff:@Rich Farmbrough: It seems like we are using 256-bit ECDH keys for perfect forward secrecy exchanges, which is roughly equivalent to 3072 bits RSA, which provides 128-bits of security. (Which I think is adequate for now.) 1 Also, an attacker would only need to factor a 1024-bit RSA public key to pull off a successful MitM right now, because of the "Equifax Secure Certificate Authority 1024-bit root". 2. But HPKP (public key pinning) would mitigate such an attack, when the WMF implements it. Tony Tan·talk05:39, 16 June 2015 (UTC)
I think its worth repeating the note on the mozilla page on why they recommend AES 128 over AES 256 - "AES 128 is preferred to AES 256. There has been discussions on whether AES256 extra security was worth the cost, and the result is far from obvious. At the moment, AES128 is preferred, because it provides good security, is really fast, and seems to be more resistant to timing attacks. ". It should be noted that our servers will use 256 bit AES if the user's browser indicates it does not support 128 bit AES. Bawolff (talk) 21:16, 13 June 2015 (UTC)
AIUI a 32 bit phone might not like 256 bit AES, whereas a 64 bit phone will gobble it up. (Some blog about "Password1".) All the best: RichFarmbrough, 00:44, 14 June 2015 (UTC).
The attacks on AES 256 (that I am aware of) are differential key attacks, provided the RNG generating the session keys is robust, AES 256 should be fine. The concerns over AES 128 (again that I am aware of) relate primarily to quantum computing, which (for reasons I am not aware of) are expected to square-root the difficulty of cracking these types of cyphers, making AES 128 the difficulty of AES 64 without quantum computing.
This is way out of my expertise, but... If someone had a quantum computer, that could break 128bit AES using Grover's algorithm, it'd need time proportional to 264. To break 2048 bit RSA (which we also rely on), they only need time proportional to 750 20482*11*log2(11) = 159608986 ≈ 227 using Shor's algorithm (If I read the wikipedia article right which I don't actually understand), albeit possibly more qubits. I really don't think we're at the point where we have to worry about practical quantum computers yet. Bawolff (talk) 04:38, 14 June 2015 (UTC)
A way to determine the number of files in the history
Latest comment: 9 years ago6 comments3 people in discussion
Is there a non-Javascript method to determine the number of files in a page history? I have a Javascript method for admins that looks like this:
Personally I think the count of the number of versions a file has, would be the sort of thing that would make sense to add to the ?action=info of a file page. Bawolff (talk) 02:11, 15 June 2015 (UTC)
I'm having trouble when I type in a missing entry. Rather than the usual "you can start it" it comes up with a red error message "An error has occurred while searching: Search is currently too busy. Please try again later." and doesn't allow me to create it.♦ Dr. Blofeld09:20, 15 June 2015 (UTC)
And here. I get a red message: "An error has occurred while searching: Search is currently too busy. Please try again later." This has been going on for more than an hour. Searches for existing pages usually work, but searching for a deleted page (in order to find why it was deleted) consistently gets the error. JohnCD (talk) 09:23, 15 June 2015 (UTC)
It's affecting all Wikimedia wikis. The Wikimedia operations team is currently trying to solve the problem - that's all I've heard from them so far. Yunshui雲水09:47, 15 June 2015 (UTC)
Search being down is of course a big problem but if you just want to create a page or see a deletion log then you can for example preview a red link and click it. PrimeHunter (talk) 10:11, 15 June 2015 (UTC)
If you want to actually search the English Wikipedia then you can say site:en.wiki.x.io in Google searches and some other external search engines. Special:Preferences#mw-prefsection-gadgets has the option "Add a selector to the Wikipedia search page allowing the use of external search engines." The selector is only on Special:Search (reached by making any search) and not the search box on every page. PrimeHunter (talk) 10:18, 15 June 2015 (UTC)
Note that Nik Everett, the WMF search guru, has just restarted the entire search server cluster, which is a pretty serious measure. Operations staff are working hard to resolve the problem. (Nik says he has "been awake for >>24 hours", so he is certainly one very dedicated employee.) — This, that and the other (talk)10:55, 15 June 2015 (UTC)
Yeah, but why go through the hassle of searching for it when you can just change the URL? Much easier to remove everything after /wiki/ and replace it with bfouesbgusbfouesbgusbfouesbgus than to go through Special:Search and hope that it doesn't exist, or hope that it exists, or something else. Nyttend (talk) 12:56, 15 June 2015 (UTC)
It appears from translatewiki:MediaWiki:Search-error/qqq that the pagename is not available. But we could give a better message like "The search function is currently down. We are working to get it back as soon as possible." The message is currently called with $1 = "Search is currently too busy. Please try again later", so it displays: "An error has occurred while searching: Search is currently too busy. Please try again later." PrimeHunter (talk) 11:16, 15 June 2015 (UTC)
I have deleted MediaWiki:Search-error since search works again, at least for me. When it was down, "(search-error: (cirrussearch-too-busy-error))" for uselang=qqx in the url meant that MediaWiki:Search-error was called with $1 = the content of MediaWiki:cirrussearch-too-busy-error. It wouldn't have helped to make a red link to the searched term, but it could have helped to only display "The search function is currently down" when the software claims the Search-error is cirrussearch-too-busy-error. PrimeHunter (talk) 13:18, 15 June 2015 (UTC)
Category-list comparison
Latest comment: 9 years ago2 comments2 people in discussion
You can do this using the "List comparer" pane in AWB. In either list's dropdown, pick "Links on page". If you've not got AWB, drop me a note on my talk page, and I'll run the tool for you. Alakzi (talk) 15:15, 15 June 2015 (UTC)
Latest comment: 9 years ago1 comment1 person in discussion
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Users have created more than 5000 articles with the new translation tool. [60]
Editing a page is now faster. This is because some statistics about edit filters were removed. [61]
Changes this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from 16 June. It will be on non-Wikipedia wikis from 17 June. It will be on all Wikipedias from 18 June (calendar).
Meetings
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on June 16 at 19:00 (UTC). See how to join.
Editing a page is now faster. This is because some statistics about edit filters were removed
...at the cost of making it harder for sysops to debug filters and detect which ones can be optimized to consume fewer conditions. Helder15:11, 15 June 2015 (UTC)
Pending Changes Weirdness
Latest comment: 9 years ago1 comment1 person in discussion
If you go to special:preferences, in the Internationalisation section, in the Language drop-down box, there are three English locales: en, en-GB, and en-CA. A UK user, in this remark in VPProposals, not unreasonably supposed that "en" was being used for American English.
That does not appear to be true — witness the fact that the section is called "Internationalisation" and not "Internationalization". My guess is that the "en" locale was coded first, and spellings simply reflect the preferences of whoever did it; can anyone confirm or deny?
But it is still very strange. Why is there no en-US? Has just no one gotten around to implementing it? Isn't there some prefabricated corpus somewhere that could just be plugged in? We could at a stroke provide "Internationalization" for American users, and remove a perception of American bias on the part of non-American users. --Trovatore (talk) 20:15, 13 June 2015 (UTC)
If anything, they should be getting rid of the options for en-GB and en-CA, not adding more choices. They generate extra work, for no obvious benefit. – iridescent20:18, 13 June 2015 (UTC)
That's a separate discussion, I think. Having multiple English locales may not add a lot of value, but I think it adds more than having, say Alemannisch or Nordfriisk or Interlingua, which are all current choices. But in any case the original concern was that British and Canadian English are so specified, whereas just plain "English" is apparently American English (even if somewhat imperfectly American, given "Internationalisation"). I agree that that's a problem, and it sort of looks like it might be dealt with at low cost, just by presenting the en locale as en-US in the user interface, without really changing much else. But I know that these apparently simple fixes sometimes turn into huge headaches in implementation, so I'm hoping that someone with MW developer experience can comment. --Trovatore (talk) 21:57, 13 June 2015 (UTC)
I did not know that. So then maybe it's as simple as specifying it as "en-US English (United States)" in the UI, and nothing else would actually change? That would address the original complaint. (Still doesn't explain "internationalisation", though.) --Trovatore (talk) 20:29, 13 June 2015 (UTC)
I mean, in mediawiki it officially corresponds to US english (I get yelled at when I accidentally use Canadian english when adding new messages). In terms of official standards, EN means english generically. Edit: Trovatore is correct that all new system messages get added to the en locale by developers, and then translators make the other languages. Occasionally a dev will accidentally use the wrong variant, and occasionally you see people fixing that sort of mistake (e.g. gerrit:51994. Although looking at the code comments, it appears that that is somewhat disputed. I was always told it was supposed to be American English.) Bawolff (talk) 20:30, 13 June 2015 (UTC)
Speaking as a simple, non-technical writer of article content the present situation is confusing. The existence of a simple "English" option is very misleading. I imagine that many people click on "English" and then find that certain words are either spelt oddly or that some of their text spellings when editing are highlighted as being wrong. I cannot be the only person this has happened to. Given that options such as "British English" and "Canadian English" are available on these menus, that there is a simple "English" option renders the menu unsystematic at best. Perhaps the least confusing way to set out such menus would be in the form "English: American", "English: Canadian" etc. That way anyone looking for 'English' will find all available varieties together. Urselius (talk) 20:38, 13 June 2015 (UTC)
Just a side note here — the spellchecking is done by your browser, not by MediaWiki. There's nothing we can do about that; you'll have to adjust your browser settings yourself. --Trovatore (talk) 20:42, 13 June 2015 (UTC)
User:Bawolff wrote "I mean, in mediawiki it officially corresponds to US english". Does MediaWiki actually have a document that officially defines this sort of thing? Where?
Even if MediaWiki does have such a definition, conventions adopted by developers who make MediaWiki work are not presented to, or accepted by, end users. So "en - English", when presented to an end user on a menu, means whatever it means in ordinary everyday English. Jc3s5h (talk) 20:39, 13 June 2015 (UTC)
I thought it did, but then I looked and couldn't find any sort of official thing, so maybe I'm wrong. At the very least people are encouraged to use American english in the en locale. Bawolff (talk) 20:41, 13 June 2015 (UTC)
@Iridescent: I asked about that a few years ago, and (iirc) was told that these English variants were (at least partially) to assist with testing internationalization code, including aspects such as "fallback language".
Going back to what Iridescent said: As a general rule, only the 'main variant' gets customized. That means that if you don't use plain "en", then you will miss locally customized messages here, many watchlist and sitenotices, and other information that you may have wanted. Any system that encourages people to move away from the default (e.g., by labeling "en" as being American) will have the unexpected side effect of cutting them off from messages that the community expects them to receive. I know a couple of people do this intentionally, but most don't learn about this side effect until they miss something that was important to them. Whatamidoing (WMF) (talk) 06:04, 14 June 2015 (UTC)
Ah, thanks. I think maybe I had heard something like that before. That certainly does add a wrinkle to the discussion.
It's not really a very satisfying answer, though. As long as we have en-GB and en-CA, some people are going to choose them, and I don't see any warning notice about the "side effect". Should there be a warning when you pick them? Or should we just get rid of them, at least for ordinary users, but let developers keep them for testing purposes? If we got rid of them, then probably we should prevail on MW to drop the "consistency" requirement, and let en be a mix of US and Commonwealth spellings (which it already is; I mentioned the example of "Internationalisation" right on the page we're discussing.) --Trovatore (talk) 06:12, 14 June 2015 (UTC)
It is very unlikely we will get rid of them on the MediaWiki side (Disabling for Wikipedia is a possibility, but MediaWiki is used by lots of groups outside of Wikipedia. Some of them want to use these locales). The real problem here, is that some messages which get modified by Wikipedia should be treated as content language messages on wikipedia, but in the general case are interface messages. We should fix that underlying issue instead of the band-aid of removing some languages (For those not familiar with MW terminology - content language messages are in the locale of the site, interface messages are in the locale of the user's preferences). Bawolff (talk) 06:37, 14 June 2015 (UTC)
Let me rephrase and see if I've got it. If we were able to get all these modified messages (can I have an example of one?) be "content-language messages" rather than "interface messages", then that would mean that a user who picked en-GB would still see them, albeit in MW's version of en, which is American English, or at least American-except-when-it-isn't English?
And then we would be able to relabel en as en-US for Wikipedia?
Without the last part, I don't think it solves the problem as it arose, namely that Commonwealth users see en, en-GB, and en-CA, and think isn't that strange, apparently AmE is now the default English. That would still be true under the hood for the modified messages, but hopefully that's a relatively small subset of messages, and at least it wouldn't be so in-your-face. --Trovatore (talk) 07:01, 14 June 2015 (UTC)
I think that the translations cause notices to go away issue, and EN being really en-US issue, are entirely separate issues. Adding more content langauge messages that people can use to write notices is very easy. Getting EN renamed en-US, or creating en-us as a dummy language in mediawiki is easy technically but probably would involve some politics and lobbying of different people.
For an example of a content message, see MediaWiki:Recentchangestext. Note how the message shows up the same on Special:Recentchanges, no matter if you're viewing it in en, en-gb, or even something obscure like sco. It would be very easy to add a new message named something like watchlist-info, which would be output immediately above where mediawiki:watchlist-details, but instead is a content message so would be unaffected by translations. Bawolff (talk) 02:31, 15 June 2015 (UTC)
Any further thoughts on this? If the answer is, "we know it's suboptimal, but we've thought about all possible solutions and they're all worse than the problem, so WONTFIX", then so be it, but I'm not convinced yet that that's the case. --Trovatore (talk) 02:16, 15 June 2015 (UTC)
I remember seeing discussion of a warning (along the lines you mention above) earlier this year; I believe you'll find it in the archives for this page. Whatamidoing (WMF) (talk) 05:02, 15 June 2015 (UTC)
Sorry, I wasn't clear. I suppose the warning thing is a little bit of a problem but I don't consider it very important. The problem I was talking about is that what is just called "English" in the user interface is actually American English (except when it isn't). Can we find a solution to that problem? (Obviously, not by making the same mistake in the other direction.) --Trovatore (talk) 05:08, 15 June 2015 (UTC)
Trovatore, I don't know if it's possible to rename this locally (and I doubt that the WMF has a formal policy on the question, although I haven't asked), but I know that in previous discussions, the community here has said that they do not want to do that. If you label it "en-us", then more people will choose something else. The community here wants fewer people to choose something else. Therefore, the community does not want to rename it.
The reason the community wants fewer people to choose a different language is that choosing something else removes local messages. The community does not want to encourage people to remove local messages, even through the indirect method of signalling that 'this English (the one with all the useful local messages) is not the right English for people like you' in the prefs menu. Whatamidoing (WMF) (talk) 15:54, 15 June 2015 (UTC)
So I understood that bit. That's a good reason not to just rename en as en-US without doing anything else. So, what else can we do? Can we fix the issue that prevents people who choose (say) en-GB from receiving local messages?
How about setting it up so that, if there's a local message, even if it's an interface message, then if there's no translation into the individual user's interface language, the system falls back to displaying the local message in the content language? (And then rename en as en-US.) --Trovatore (talk) 17:32, 15 June 2015 (UTC)
That would cause problems with trivial translations. People often set other languages because they have difficulty reading English. They may be coming here and pasting everything into machine translation, or they may be looking for templates to port to another wiki, or whatever. If a non-default language setting gets overriden whenever there is a local translation, then even the most trivial translation change (e.g., from "Discussion" to "Talk" to save horizontal space on the screen) will overwrite the non-English language – all non-English languages, not just the English variants, where the language differences make almost no difference in comprehension at all. Whatamidoing (WMF) (talk) 19:35, 16 June 2015 (UTC)
I've unSALTed them, since the situation has clearly changed since the SALTing (the articles in question were both seleted at the time of the SALTRing of the talk pages, and now they exist). You should have no problem creating these. עוד מישהוOd Mishehu11:46, 16 June 2015 (UTC)
Now the color of every article title changes. How can I reset this to default or reset everything to the default colors. — Preceding unsigned comment added by LT391 (talk • contribs) 01:03, 17 June 2015 (UTC)LT391 (talk) 02:17, 17 June 2015 (UTC)
Latest comment: 9 years ago7 comments4 people in discussion
My global contributions page has me registered on 16 March 2009 but my first edit seems to have been 10:50, 23 April 2006, which is about right. Any idea why it's wrong? Thanks. Doug Weller (talk) 16:05, 16 June 2015 (UTC)
Thanks User:Floquenbeam. But my bad, I shouldn't have said global contributions, but "Global account information".[67]. That doesn't give a clue that I was editing before 2009. Or passed my RfA 18 September 2008. I know it says "attached", but it also says "Registered: 08:27, 16 March 2009 (6 years ago)" which the average reader will think means that was when I created my account. Doug Weller (talk) 12:11, 17 June 2015 (UTC)
Doug Weller, Floquenbeam: I can't remember a link but I remember that this is something that the SUL people are aware of as a problem. If I remember correctly it requires a one-time sweep through account information to correct the problem. No idea if that's something that's in the works atm, @Keegan (WMF):? ResMar15:23, 17 June 2015 (UTC)
Yeah, there's a suggestion to back-fill CentralAuth information with the information on the oldest registered account. I'm not sure if it's going to happen, however, or if so, anytime soon. SUL work is down to one engineer, part-time, so there's only so much that can be done. Keegan (WMF) (talk) 17:36, 17 June 2015 (UTC)
Incorrect category: Wikipedia pages with incorrect protection templates
Latest comment: 9 years ago5 comments3 people in discussion
WP:TEMPLATE LOOP says: "A template can call itself, but will stop after one iteration to prevent an infinite loop." This for example means a template can transclude examples of use on its own documentation page. The bottom of edit windows have a list "Pages transcluded onto the current version of this page". For Template:R fully protected that list currently includes the page itself. The list is also at "Page information". PrimeHunter (talk) 00:10, 17 June 2015 (UTC)
@Redrose64: Replacing <includeonly>...</includeonly> with {{#ifeq:{{FULLPAGENAME}}|Template:R fully protected|<!--nothing-->|...}} should fix the problem - Evad37[talk]04:57, 18 June 2015 (UTC)
Edit toolbar - Delay of display
Latest comment: 9 years ago4 comments3 people in discussion
Greetings, Wondering why the time delay? When the Edit window first opens, the text displays instantly, then there is a 2 to 3 (or more) second time lag waiting for the edit toolbar to appear. Both with the default editor and wikEd editor. Before posting this, I searched the archives (here and at Teahouse) and I attempted a few Preferences changes with no improvement. Is there some trick preferences WP setting or on my laptop brower settings (Chromodo, a version of GC) to make this time lag go away? Not urgent, but it sure would help up my volume of article updates. Regards, JoeHebda (talk) 16:19, 17 June 2015 (UTC)
Because it's Javascript, and it needs time to execute. For me it's almost instant, but it's very dependent on the speed of your computer. —TheDJ (talk • contribs) 00:44, 18 June 2015 (UTC)
Thanks TheDJ for this info. My laptop is several yrs old-Lenovo E535, 8Gb memory, Win7-64bit. I did the following changes that seem to have helped somewhat: Newest ver. Java 8.45 (was at v7.67) and changed Java buffer from max of 32, reduced to 16Mb. Cheers, JoeHebda (talk) 03:01, 18 June 2015 (UTC)
Java and JavaScript have nothing at all to do with each other. Funny thing, if you read our JavaScript article you will see that Netscape came up with that name to intentionally confuse people into thinking JavaScript had something to do with Java, because Java was at the time the trendy new thing in programming and Netscape wanted to ride its buzz. Anyway, you are either seeing a placebo effect, or, if you're reloading the edit page to test the speed, possibly the effects of caching and the just-in-time compilation that modern browsers' JavaScript interpreters use. Or, it could just be random variations in network speed (your computer can't display things any faster than it can download them from the Wikimedia servers). I doubt it's your hardware; a computer with 8 GB RAM running Windows 7 should have few problems with JavaScript, unless the CPU is being heavily throttled (which could happen in laptops due to temperature). If you really wanted to troubleshoot, you might want to start by trying a few different browsers and seeing if there's a difference. You might want to start with Mozilla Firefox as it's very popular and has a completely different browser engine from the Chromium-based browser you're using. --108.38.204.15 (talk) 04:37, 18 June 2015 (UTC)
Gradually enabling VisualEditor for new accounts
Latest comment: 9 years ago1 comment1 person in discussion
TL;DR: The recent test results show that there's no negative impact from offering VisualEditor to newly registered accounts. Here's a plan for how we might offer it more widely in a gradual manner.
Latest comment: 9 years ago2 comments2 people in discussion
Until now, we had the crap {{Look from}} template which just produced a link to Special:PrefixIndex with the variable in place. Now, I have created a new template which actually inserts the list of pages beginning with a particular prefix onto the page. The following parameters are supported:
First unnamed parameter for the prefix
|columns= to split the list into two or more columns
|redr=(any value) to hide redirects from the list
|stripprefix=(any value) to strip the given prefix from the page names.
This was created using the {{Special:PrefixIndex}} markup and is an improvement over it since {{Special:PrefixIndex}} produces a three-columned list and has no option for changing that. Also, this template produces a clean, bulleted list. 103.6.158.193 (talk) 15:20, 18 June 2015 (UTC)
Latest comment: 9 years ago17 comments11 people in discussion
Resolved
Starting a few hours ago, about 98% of the time, nav popups do not pop up. And wikEd does not show up. I logged off WP and back on, updated preferences (gadget section) to turn off these options and then turn them back on. Is something broken in WP? Am I the only editor who has these problems? Christhe spelleryack04:44, 18 June 2015 (UTC)
I am using Firefox 36.0.1 (have been running it for months), and also noticed that its tabs have a spinning progress symbol indicating that WP pages are not finished loading, even though all the content looks like it is there. Christhe spelleryack04:50, 18 June 2015 (UTC)
On this page I'm seeing something like "Waiting for tools.wmflabs...". I was able to fix this by turning off the preference "Enable tracking bugs on Phabricator using the {{tracked}} template". -- John of Reading (talk) 06:00, 18 June 2015 (UTC)
Hello, Wikimedia Labs, the infrastructure used for the Phabricator bug-tracking bot, is currently down. We are working on it. Mobrovac-WMF (talk) 08:01, 18 June 2015 (UTC)
I'm also having a problem with pages that never finish loading. I'm running Firefox 38.0.5 and using the Vector skin if that helps. Jenks24 (talk) 14:25, 18 June 2015 (UTC)
If I'm logged out it works properly again, regardless of whether I'm in private browsing or using HTTPS. Just checked on Google Chrome and it doesn't seem to be a problem with that browser whether I'm logged in or not. Jenks24 (talk) 14:41, 18 June 2015 (UTC)
Also down are DeltaQuadBot, CorenSearchBot, Duplication Detector, and Earwig's copy vio detector. @Mobrovac-WMF: Could we please get an update when you have a minute? WMF Labs has been down for 15 hours already, and we need these critical functions. Thanks, -- Diannaa (talk) 15:10, 18 June 2015 (UTC)
The outage is still ongoing, I'm afraid. Without getting too deep into technicalities, the file system and disks used to power all of your favourite bots are being replaced right now, after which the state will be restored from back-ups. Hold on for a couple of ours more. We understand your concerns, and there are people working hard on solving this as quickly as possible (most of which haven't slept at all during this period). You can follow updates on the incident report pageMobrovac-WMF (talk) 15:53, 18 June 2015 (UTC)
Everything appears to be back to normal, now that the tools have been started again on the afflicted server. Thanks to all who worked on this. Christhe spelleryack13:56, 19 June 2015 (UTC)
+1 for all who worked on it and all who patiently answered questions. Hope WMF realises that there is some better backup needed - as they permanently claim themselves to operate the #5 world's website... --Trofobi (talk) 17:54, 19 June 2015 (UTC)
Wikipedia on a mobile browser is not showing the mobile version of the page
Latest comment: 9 years ago2 comments2 people in discussion
Since about a week ago, when doing a Google search and clicking on Wikipedia in the results, the page is showing the desktop version of the site rather than the mobile version. I notice that the URL doesn't have the usual /wiki/Example format, but rather is using /?title=Example, which is the cause of it. You can test this by searching Google for "Microsoft acquisitions" and clicking on the Wikipedia article.
But is this problem on Google's end or Wikipedia's? I tested on multiple iPhones in Safari so the problem isn't just me. Gary(talk · scripts)17:35, 19 June 2015 (UTC)
The problem is mostly on our site, though it's third parties linking to weird URLs that are populating the Google search index. An immediate problem, mobile users not being redirected, should be fixed with this, the rest is being worked on at phab:T67402. Max Semenik (talk) 22:08, 19 June 2015 (UTC)
Table cell contents spilling over.
Latest comment: 9 years ago17 comments3 people in discussion
When viewing the following table in the mobile version of the site, some text from the first column spills over into the adjacent cell of the second column.
Thanks for mentioning that coding error. That didn't cause the problem however. Actually it did not do any harm at all. If you click on the link to the mobile view and then reduce the width of your browser screen to the minimum you will see the text from the first column spilling over. Using the identity template doesn't solve it. Tvx1 05:38, 1 March 2015 (UTC)
I already tried the minimum width on mobile and it works for me. I get a horizontal scroll bar and no overlap. I will stop guessing. It's too hard when I don't have the problem. PrimeHunter (talk) 05:51, 1 March 2015 (UTC)
Since I'm not able to explain the issue with text, I have made a screenshot:
It doesn't spill over for me. Based on the amount of spillover in your screenshot, maybe your browser doesn't reserve space for the flag icon when the column width is calculated. PrimeHunter (talk) 00:05, 20 March 2015 (UTC)
Multiple user have reported this to me though, regardless of which browser they use. Tvx1 06:54, 21 March 2015 (UTC)
I think you're spot on there. Either the flag icon isn't taken into account, either it's just counted as a 1px character. Is there a way to solve this? Tvx1 19:07, 26 March 2015 (UTC)
Well, cells without flagicons don't spill over into other, no matter how wide they are. But we want to have them in that table. Tvx1 10:13, 28 March 2015 (UTC)
And does making the icons larger make the problem worse? What I'm trying to determine is how the browser is interpreting what it's given and why it isn't making the cells large enough. If the browser isn't making sufficient space for all of the elements the cell contains, should we worry about that, since it's then the fault of the browser, or find a workaround and try to fix the problem on our end, even if the mobile browser isn't working as expected? I didn't seem to catch what your mobile browser even is. If I could replicate the problem on my end, I could attempt to fix it, but the desktop version of Chrome I'm running handles element widths as expected. Let me add a table with the changes I'm thinking of for illustration:
Of course, this is an extreme size change, but it's to troubleshoot things here. Does this make the problem worse, fix it, or make no difference whatsoever? Googol30 (talk) 23:42, 28 March 2015 (UTC)
Googol30, that makes the problem much worse. Here is a screenshot:
In case I hadn't made it clear yet, this is an issue that only occurs on the mobile version of the site. The only mobile browser I have thus far identified not to be affected by this issue is Firefox. All other mobile browser have this problem. I mostly use mobile Safari, but as said other mobile browsers are affected as well. Tvx1 00:37, 29 March 2015 (UTC)
@Tvx1: yes, and I've done some research into the problem, finding that this is actually intended behavior of the nowrap attribute. Sadly, I do not have enough technical knowledge of its use within MediaWiki or on Wikipedia to properly fix this issue, so I regret to inform you that although (I think) I know what is causing the problem, I cannot properly fix it without potentially causing more problems elsewhere. Googol30 (talk) 11:04, 4 May 2015 (UTC)
And do you know anyone who could be able to fix it, or do you think it would be better to file a bug report over on Phabricator? Tvx1 15:55, 4 May 2015 (UTC)
I'd say, if you haven't already, to file a report on Phabricator, possibly linking to there from here to give anyone looking for the problem here a place to continue searching.Googol30 (talk) 20:09, 10 May 2015 (UTC)
^"@McLarenF1: 13 December 2014". Twitter. 13 December 2014. Retrieved 14 December 2014. Good news: the McLaren-Honda MP4-30 has passed all its @fia chassis crash tests.
^"Christian Horner Q&A: Red Bull's 2014 a big achievement". Formula1.com. Formula One Administration.ltd. 14 November 2014. Retrieved 12 December 2014. Adrian will still be around. Yes, he will take a bit of a step back, but we will still be able to draw on Adrian's expertise and he's still very much involved in the design of the RB11.
^"@McLarenF1: 13 December 2014". Twitter. 13 December 2014. Retrieved 14 December 2014. Good news: the McLaren-Honda MP4-30 has passed all its @fia chassis crash tests.
^"Christian Horner Q&A: Red Bull's 2014 a big achievement". Formula1.com. Formula One Administration.ltd. 14 November 2014. Retrieved 12 December 2014. Adrian will still be around. Yes, he will take a bit of a step back, but we will still be able to draw on Adrian's expertise and he's still very much involved in the design of the RB11.
^"@McLarenF1: 13 December 2014". Twitter. 13 December 2014. Retrieved 14 December 2014. Good news: the McLaren-Honda MP4-30 has passed all its @fia chassis crash tests.
^"Christian Horner Q&A: Red Bull's 2014 a big achievement". Formula1.com. Formula One Administration.ltd. 14 November 2014. Retrieved 12 December 2014. Adrian will still be around. Yes, he will take a bit of a step back, but we will still be able to draw on Adrian's expertise and he's still very much involved in the design of the RB11.
^"@McLarenF1: 13 December 2014". Twitter. 13 December 2014. Retrieved 14 December 2014. Good news: the McLaren-Honda MP4-30 has passed all its @fia chassis crash tests.
^"Christian Horner Q&A: Red Bull's 2014 a big achievement". Formula1.com. Formula One Administration.ltd. 14 November 2014. Retrieved 12 December 2014. Adrian will still be around. Yes, he will take a bit of a step back, but we will still be able to draw on Adrian's expertise and he's still very much involved in the design of the RB11.
The last couple of days I have experienced something unfamiliar and unwelcome when copying words on Wikipedia. To be exact, when I, for example, copy a category or article title in order to paste it into an article, the result is like this: " Category:British military personnel killed in World War II ". In other words, all of a sudden I get a space on either side of the copied text. It occurs when I click the titles repeatedly to mark them, and then right-click to copy. When I mark the words by drawing the mouse pointer over the word, I only get an extra space at the end, like so: "Pointer (graphical user interfaces) ". This is something I've never before come across in more than 10 years on the project, and it's pretty annoying. Any reason why this has started happening now? Surely, this is some sort of technical malfunction? Cheers. Manxruler (talk) 18:30, 17 June 2015 (UTC)
Latest comment: 9 years ago6 comments4 people in discussion
Hello, in patrolling new pages, I often like to add wikiprojects to the pages so that they can be found by interested editors. The AfC helper script has an easy HotCat-like tool for doing this. I was wondering if such a thing existed separately. I asked at the Help Desk, but no one new. Thanks! Happy Squirrel (talk) 01:30, 20 June 2015 (UTC)
I don't know of one, but I suspect that if one existed, it could lead to over-bannering of talk pages. It's better to add just one which is totally relevant than to add eight or nine of which six or seven only have a tenuous connection. One of the things about WikiProjects is that participation rises and falls, so it's not always a good idea to add the project banner for an inactive wikiproject, and unless the topic is particularly significant, adding more than two or three can be a bit bloaty. It's an established principle that each WikiProject reserves the right to decide what falls within its scope, so if you add a WikiProject banner template that somebody else later alters or removes, don't get upset (recent example). In some cases it's an easy decision: biographies must be given {{WikiProject Biography}}, even if the subject is dead. --Redrose64 (talk) 08:35, 20 June 2015 (UTC)
Thanks for the quick replies! I understand that WikiProjects get to define their own scope. I really don't mind when people let me know more details, and I keep them in mind for the next time. I also try not to over-tag. I will definately look into the AWB option. Thanks! Happy Squirrel (talk) 14:16, 20 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
On my userpage, User:Joseph2302, and my talkpage, User talk:Joseph2302, the "New section" heading has disappeared. I first noticed it after I created a Javascript yesterday, but that Javascript is now deleted (as it didn't work). Any ideas? If you click the "Edit" or "View History" headings, then on those pages it's visible. Joseph2302 (talk) 10:52, 20 June 2015 (UTC)
Probably, one of the template on your talk page is emitting a "__NONEWSECTIONLINK__ " directive, which hides that link. You can remove that template, or add "__NEWSECTIONLINK__" to make it reappear. See also Help:Magic words for beginners. -- [[User:Edokter]] {{talk}}11:41, 20 June 2015 (UTC)
Both the user edit count and articles history functions of xTools do not seem to be working for me. Does anyone know what the issue is? Buffaboytalk22:36, 16 June 2015 (UTC)
I think it may be the newly enforced HTTPS, but I can't be sure. I'm trying to look into and get help. For the record, if anyone is a strong PHP developer, or has good experience settings up linux servers, we are actively looking for new maintainers! — MusikAnimaltalk22:42, 16 June 2015 (UTC)
MusikAnimal, I talked to YuviPanda this morning and he said it appeared to be because it was moved from trusty to precise and needed to be moved back because it currently isn't coded to work on precise. I'm not sure how to move it back, so if you can do it, that would be great (I'm the PHP guy, not the configuring Ubuntu guy). If not, I'll have Yuvi walk me through it tomorrow. — {{U|Technical 13}} (e • t • c)23:14, 16 June 2015 (UTC)
Technical 13, it's the other way around. xTools was built for precise, and the move to trusty broke it. I'm trying to find the failing code source so I can get it to work on Trusty, but the code does not allow for easy debugging and the problem seems to be deeply nested somewhere. Somebody shoot me. :-(—cyberpowerChat:Online12:51, 17 June 2015 (UTC)
On the other hand, our custom environment being developed is using precise. If we can recruit some more active users familiar with Linux servers and setting them up, we can get the move started. That should provide xTools with the permanent stability it needs. If you are interested in joining xTools, please shoot me an email.—cyberpowerChat:Online12:54, 17 June 2015 (UTC)
I believe there is a labs-wide issue right now (not sure if anything is running), should be fixed today from what I can tell. Hold tight! — MusikAnimaltalk14:41, 18 June 2015 (UTC)
I can't view Edit Count. For last two days it shows:
Internal error
The URI you have requested, /xtools-ec/?user=CosmicEmperor&project=en.wiki.x.io, appears to be non-functional at this time.
If you have reached this page from somewhere else...
This URI is part of the xtools-ec tool, maintained by Cyberpower678 and Tools.xtools.
Perhaps its magical script elves are temporarily ill, or the link you've followed doesn't actually lead somewhere useful?
If you're pretty sure this shouldn't be an error, you may wish to notify the tool's maintainers (above) about the error and how you ended up here.
If you maintain this tool
The error might be caused by incorrect permission, or by an error in the script or CGI that was meant to execute here. You may wish to check your logs or common causes for errors in the help documentation.
Don't forget that there is more than one edit counter out there. When the one at "edit count" is down the one at "tools.wmflabs.org/supercount/" is usually working. My apologies for not being able to leave a link to the start page for this edit counter. Hopefully, someone else can do that for everyone. MarnetteD|Talk23:07, 17 June 2015 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
Recent changes
Some wikis can now be used only with HTTPS. This includes the English, Russian and Chinese Wikipedias, among others. Soon all wikis will use only HTTPS for all users. [68][69]
You can't use HTTPS wikis with Internet Explorer 6 on Windows XP. You need to use another browser. [70]
Problems
On June 15, search was broken on all wikis for several hours. [71][72][73]
Many Labs tools were broken for several days. Almost all of them are back now. They may be missing some data for the last 10 days. [77][78][79]
Changes this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from June 23. It will be on non-Wikipedia wikis from June 24. It will be on all Wikipedias from June 25 (calendar).
Meetings
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on June 23 at 19:00 (UTC). See how to join.
Latest comment: 9 years ago33 comments10 people in discussion
Hidden in the link at the end of today's Tech News is a bit of a bombshell. For people who didn't notice, most of our bots are going to stop working on 2 July if their code isn't updated.
According to the announcement, the following bots need to be fixed. (I've restricted the list to bots active on this wiki.)
We are reliant on quite a few of those bots for the smooth functioning of our wiki, so it's very important that they are fixed. Also, as TheDJ says above, there are also a number of user scripts that need fixing as well.
This was also announced at WP:BOTN, with the complete list of bots that are known to be at risk. If you have know bot owners who don't frequent BOTN here (especially people at other projects), then please reach out to them. (Also, someday we need to create a template for major issues like this—maybe something like {{warning|All your bots are going to break}}. This only seems to happen once or twice a year, but I worry that people won't notice these messages in time.) Whatamidoing (WMF) (talk) 01:29, 9 June 2015 (UTC)
@Whatamidoing (WMF): You know we have a feature called Wikipedia:Mass message senders, and I do have a talk page that yells at me everytime someone edits it. I only check the bot operators' pages once in a blue moon, and I'm not subscribed to any email lists. I learned of this after my bots broke. – Wbm1058 (talk) 18:17, 13 June 2015 (UTC)
Is there a full list of affected bots? The announcement only names those with over 10,000 deprecation warnings over the course of a week. Some bots operate at a lower frequency but are equally as important — MusikAnimaltalk01:49, 9 June 2015 (UTC)
I'm guessing that the non-bot users will mostly be people using Huggle, Twinkle, AWB, STiki, and other similar tools. I'm not sure which of the tools is affected, though. — Mr. Stradivarius♪ talk ♪22:08, 9 June 2015 (UTC)
Debugging related to mandatory HTTPS implementation. Confusion results when an unannounced breaking change is implemented right after the announcement of a scheduled breaking change.
Impending?? Both of my bots are already down, and failing in a library API call. I guess I need to scramble to fix it as soon as I can. – Wbm1058 (talk) 17:17, 13 June 2015 (UTC)
/** * Sends a query to the api. * @param $query The query string. * @param $post POST data if its a post request (optional). * @return The api result. **/functionquery($query,$post=null){if($post==null)$ret=$this->http->get($this->url.$query);else$ret=$this->http->post($this->url.$query,$post);returnjson_decode($ret,true);}/** * Returns an array with all the members of $category * @param $category The category to use. * @param $subcat (bool) Go into sub categories? * @return array **/functioncategorymembers($category,$subcat=false){$continue='';$pages=array();while(true){$res=$this->query('?action=query&list=categorymembers&cmtitle='.urlencode($category).'&format=json&cmlimit=500'.$continue);if(isset($x['error'])){returnfalse;}foreach($res['query']['categorymembers']as$x){// this is line 263$pages[]=$x['title'];}if(empty($res['query-continue']['categorymembers']['cmcontinue'])){if($subcat){foreach($pagesas$p){if(substr($p,0,9)=='Category:'){$pages2=$this->categorymembers($p,true);$pages=array_merge($pages,$pages2);}}}return$pages;}else{$continue='&cmcontinue='.urlencode($res['query-continue']['categorymembers']['cmcontinue']);}}}/** * Returns all the pages $page is transcluded on. * @param $page The page to get the transclusions from. * @param $sleep The time to sleep between requets (set to null to disable). * @return array. **/functiongetTransclusions($page,$sleep=null,$extra=null){$continue='';$pages=array();while(true){$ret=$this->query('?action=query&list=embeddedin&eititle='.urlencode($page).$continue.$extra.'&eilimit=500&format=json');if($sleep!=null){sleep($sleep);}foreach($ret['query']['embeddedin']as$x){// this is line 496$pages[]=$x['title'];}if(isset($ret['query-continue']['embeddedin']['eicontinue'])){$continue='&eicontinue='.$ret['query-continue']['embeddedin']['eicontinue'];}else{return$pages;}}}
I don't find that the email announcement linked above adequately explains this. It may be a slow slog through the documentation if I don't get some help here. Can anyone confirm that there are breaking changes in the latest MediaWiki release?
I'm guessing that this is due to the switch to https-only, which happened a couple of days ago. As Whatamidoing said, the change in the continuation format will not happen until 2 July on this wiki. — Mr. Stradivarius♪ talk ♪08:15, 14 June 2015 (UTC)
@Whatamidoing (WMF): My bot's last successful update was at 06:18, 12 June 2015. Per § Forced HTTPS, "At some point between 07:38 and 10:00, 12 June 2015 (UTC), the user preference "Always use a secure connection when logged in" lost its effect, and regardless of its setting, Wikipedia became HTTPS only." I've inquired about solutions over at m:Talk:HTTPS#Bots. – Wbm1058 (talk) 21:20, 14 June 2015 (UTC)
OK, my HTTPS issues have been solved, but just a reminder that I still have only about two weeks to figure out what this is about. – Wbm1058 (talk) 18:41, 16 June 2015 (UTC)
I've just been looking at the number of user scripts that need fixing, and according to this search the number is 111. That includes the Popups and Contribsrange gadgets, along with some popular user scripts like Dabfinder. (And a lot of clones of the old version of closeAFD.) We should probably start going through and just fixing these. — Mr. Stradivarius♪ talk ♪05:39, 20 June 2015 (UTC)
I've just fixed a few dozen more scripts, bringing the number of search results down to 79 (and there are a few false positives in there as well). If anyone wants to help, be my guest. :) — Mr. Stradivarius♪ talk ♪03:42, 25 June 2015 (UTC)
Done Hurrah! Now all the scripts on enwiki are fixed. (Or at least, all the ones that the search found.) The seven remaining results are all false positives. I went and fixed quite a few people's common.js, monobook.js and vector.js as well. — Mr. Stradivarius♪ talk ♪16:07, 26 June 2015 (UTC)
Spurious warnings
Sorry, maybe that word is too strong here, but the bulk of warnings that RMCD bot is receiving are coming from simple requests to read the contents of the current talk page.
These are prop=revisions requests, where rvlimit=1 ... the application doesn't need to pull up historical revisions of talk pages.
So, I have two options to make the warning go away:
It doesn't matter which I use here, so this is much to do about nothing. I assume that even the largest possible talk page given the software limits may be retrieved in full with a single API request, so there is never any need to continue here. Couldn't you limit the warning messages to when applications actually make the continuation requests? Was this distinction made in determining who to send the mass-message notice to? I'm still looking to see whether my applications actually make any continue requests, but I see some in my library, so that needs to be updated whether my applications use those functions that continue or not. Wbm1058 (talk) 19:54, 22 June 2015 (UTC)
The warnings will presumably go away when the default changes on 2 July, so you can just wait it out and then your problem will be solved. I suppose displaying the warning by default could be seen as a waste of bandwidth, but that's probably better than not having the warning on by default and no-one knowing about the API change before it happens. — Mr. Stradivarius♪ talk ♪23:31, 22 June 2015 (UTC)
There will be a different warning (see the code here) for some time, so that when people who somehow missed all the notices try to debug their queries they have a helpful message to see.
As for the spurious warnings, you're free to ignore them if you're not actually continuing the query. Anomie⚔14:41, 28 June 2015 (UTC)
Horizontal numbered lists
Latest comment: 9 years ago6 comments4 people in discussion
Thanks. I was mainly looking at template hlist because of its name which alludes to "horizontal list". The only downside with this implementation of horizontal numbered/ordered lists is that one wikitext line is needed for each item of the list.SoSivr (talk) 09:33, 23 June 2015 (UTC)
T:Hlist should probably implement numbered lists... Under the hood, both of them use the hlist HTML class for styling.
You could also provide the Html 5 representation of the list, but there are bots that clean that up. --Izno (talk) 13:40, 23 June 2015 (UTC)
I meant simply the use of HTML markup rather than wiki markup. Though that's interesting to hear. Surprised it isn't in html/css yet since the markup of horizontal lists has been of interest for at least the past decade (just googling around)... I suppose you can do it with classing regardless? --Izno (talk) 17:08, 23 June 2015 (UTC)
Session data loss message
Latest comment: 9 years ago9 comments7 people in discussion
Does anyone know the name of the MediaWiki message that pops up if you are trying to save an edit, but session data has been lost? Happens if you disconnect from the internet for a while, and then reconnect and submit changes. Conifer (talk) 10:09, 20 June 2015 (UTC)
I see former threads about this, but I can't make heads or tails of the discussion on Phab (which doesn't display the year in the posts, so I don't really know if this is recent or not). Can someone offer an update? This is happening all the time to me. Maury Markowitz (talk) 16:15, 25 June 2015 (UTC)
It's happening to me, too, over on Wiktionary, FWIW (=confirming that it's not a WP-specific issue). -sche (talk) 16:24, 25 June 2015 (UTC)
@Maury Markowitz: There are no years shown in phab:T102199, this is true: it's because phabricator doesn't display the year for posts or actions that are newer than a month or so. It uses the space saved to show the day of the week instead, for some reason. --Redrose64 (talk) 16:32, 25 June 2015 (UTC)
HHVM
Latest comment: 9 years ago5 comments5 people in discussion
Latest comment: 9 years ago6 comments4 people in discussion
I do not know if this already exists, but I would suggest creating a bot that finds articles lacking a talkpage and creates one by using info from the page's categories.--Catlemur (talk) 12:50, 21 June 2015 (UTC)
It sounds like the editor wants a bot to tag non-existing talk pages for WikiProjects (based on the reference to the "page's categories"). --Izno (talk) 16:21, 21 June 2015 (UTC)
This happens from time to time but on a demand basis. A Wikiproject requests a bot to tag all articles in a category tree, or page list generated from one, with headers, creating talk pages as it goes. The problem is without that curation it potentially can do far more harm than good. The category tree has many inclusions of categories within categories it which make sense individually but lead to category trees based on a topic being much larger than would be expected. As an example there was a recent run to add WP:NZ related articles to the project, which required a lot of manual pruning of the categories. See Wikipedia:NZWNB#Bot request.--JohnBlackburnewordsdeeds18:08, 21 June 2015 (UTC)
Latest comment: 9 years ago6 comments3 people in discussion
As requested on Help talk:Score#Version?, it would be useful to know the version of Lilypond that Wikipedia uses in its implementation of the Score Extension. Different versions of Lilypond have different syntax; for example, it seems that Wikipedia does not allow for the \tuplet function, which wasn't present in earlier versions. If the version is outdated, I think it would be a good idea to update to 2.18, the latest stable version. Fern 24 (talk) 14:04, 21 June 2015 (UTC)
@Beeswaxcandle: Thanks for the information; I've added it to Help:Score. Do you know, does the extension needs to be updated to update Lilypond? I was under the impression that each MediaWiki implementation could install a new version of Lilypond and simply point the extension to the new version. There's no great rush to update, but 2.14 is more than four years old now and the latest versions offer some useful features that are currently missing. Fern 24 (talk) 13:24, 22 June 2015 (UTC)
Sorry, that's not something I can help with. I'm just the prime user of the extension over on Wikisource. I wonder if the best approach would be to wait for 2.20 (the next stable release expected in a few months' time) and then log a phabricator request for an update—in whatever form that needs to happen. With respect to tuplets, these can be done with the \times command. Beeswaxcandle (talk) 20:50, 22 June 2015 (UTC)
Thanks for your help. I think I'll do as you suggest about the update, and I'll submit a request at Phabricator when the new version comes out. About the tuplets, I ended up using \times, but with \tuplet you no longer have to type the command for each set of triplets - perhaps it would make more sense if you see what I mean, here. Fern 24 (talk) 20:10, 23 June 2015 (UTC)
Failure to download as pdf
Latest comment: 9 years ago4 comments3 people in discussion
The problem you are reporting sounds like a potential issue in the code of the MediaWiki software or the server configuration. It would be nice if you could send the software bug to the Phabricator bug tracker by following the instructions How to report a bug. This is to make developers of the software aware of the issue. If you have done so, please paste the number of the bug report (or the link) here, so others can also inform themselves about the bug's status. Thanks in advance! --AKlapper (WMF) (talk) 07:47, 22 June 2015 (UTC)
Ironic that "dimension too large" is coming up on an article about very long linear, entangled macromolecules :-) Nyttend (talk) 21:35, 22 June 2015 (UTC)
c-uploaded didn't get deleted
Latest comment: 9 years ago4 comments3 people in discussion
I just discovered File:High Line 20th Street looking downtown.jpg by accident. Can anyone imagine why it didn't get deleted? I uploaded it back last November with {{c-uploaded}}, and it should have been deleted as soon as it was off the Main Page, but it's still there. I can't remember what bot is (or was then) responsible for doing c-uploaded deletions. Nyttend (talk) 18:58, 22 June 2015 (UTC)
I just happened to see this by accident. I'm not sure that I was aware that one of my images was on the Main Page - but then, my memory ain't what it used to be. BMK (talk) 22:51, 22 June 2015 (UTC)
I found a few more with a "what transcludes here" search and took care of them. There were eight, with dates ranging from November 2014 to June 2015. I don't know the answer to the bot question though. If there's a bot that is supposed to delete these, it would have to be an admin-bot. -- Diannaa (talk) 18:56, 23 June 2015 (UTC)
None of the collapsing templates on AN/I (collapse, hat etc.) are working for me. I tried changing a cllapse to a aht, and no difference, I tried creating a new collapse elsewhere (i.e. not on AN/I), and it worked. I closed and opened Firefix, no joy. I'll see what it looks like with another browser. BMK (talk) 23:10, 22 June 2015 (UTC)
It's an intermittent problem affecting various pages, I first noticed it when protecting a page and saw that the "Instructions and special-case notes" box wasn't collapsed. The problem is that some collapsible boxes are displaying uncollapsed and without the [hide] link; this normally suggests that the JavaScript file that handles collapsing hasn't got through. --Redrose64 (talk) 23:15, 22 June 2015 (UTC)
Judging by the error I'm seeing in the JavaScript console I think it may have to do with this class having recently been removed. This is also affecting the Warn module of Twinkle, where you are unable to issue anything other than a level one warning. I've reported it and I was told it is being fixed. — MusikAnimaltalk23:30, 22 June 2015 (UTC)
Glad to hear that someone has a handle on the problem. I just checked and the page is collapsing fine under Chrome. I also purged the page on Firefox and got no change, still no collapsing. I assume that the bug will be fixed at some point and not worry about it. BMK (talk) 23:44, 22 June 2015 (UTC)
This is because the code that uses this module is not actually declaring the usage. It seems this module is no longer guaranteed to be provided by default, so any code making that assumption was already broken, but now it it's visibly broken. —TheDJ (talk • contribs) 09:53, 23 June 2015 (UTC)
Latest comment: 9 years ago8 comments4 people in discussion
When I go to "All public logs" for Marian Dawkins I can only see one action logged - semi-protection on 8th January 2014. However, the page appears to be PC1 protected. Am I missing something? Is it just displaying differently for me? Is PC protection listed somewhere else? Wouldn't it make more sense to include it in "All public logs"? Also, the log implies that the article is still semi-protected, which it can't be since some of the edits weren't automatically accepted.Striking through this coment as the log gives the expiry date. 12:34, 23 June 2015 (UTC) Yaris678 (talk) 08:42, 23 June 2015 (UTC)
@Yaris678: Logs for moved pages can be confusing; there is a summary at WP:MOVE#How to move a page, item 4. It's certainly an anomaly that if a page is protected at the time of the move, you get a log entry like 'moved protection settings from "Foo" to "Bar" (Foo moved to Bar)'; but if it is under PC at the time of the move, you don't get that log entry even though the PC setting is copied. --Redrose64 (talk) 10:26, 23 June 2015 (UTC)
Thank you for the explanation. That is needlessly confusing. Page logs should follow the moved page, as with page history. Has anyone tried logging that as a bug? Yaris678 (talk) 12:34, 23 June 2015 (UTC)
Page lots should not follow the moved page, because logs frequently are related to the page title itself. We need the deletion log to stay with the deleted title, for example (what would happen to the deleted edits otherwise?), and what would you do with protection logs for salted pages? Nyttend (talk) 18:34, 23 June 2015 (UTC)
@Nyttend: It would be a start if PC behaved like prot, i.e. if you moved a page under PC, there would be a log entry on the new page name like 'moved pending changes settings from "Foo" to "Bar" (Foo moved to Bar)'.
Better still would be if this log entry showed what that setting was, and when it expires. Something like 'moved protection settings [edit=autoconfirmed] (expires 23:59, 23 June 2015 (UTC)) from "Foo" to "Bar" (Foo moved to Bar)'. --Redrose64 (talk) 18:46, 23 June 2015 (UTC)
So you're suggesting that the page's log get a note mentioning the protection? Now that sounds unambiguously helpful, and it hadn't occurred to me at all. Note that PC and protection do behave similarly in that both of them create page history entries; if you go to [80] and look at 31 December, you'll see the PC entry. Nyttend (talk) 21:08, 23 June 2015 (UTC)
If I want to know when the current prot or PC expires, the easiest thing for me (as an admin) to do is to click the "protect" tab and press End; second easiest is to click the "history" tab and then "View logs for this page". This is fine when there have been no moves since the last protection; if the page has been moved, I see the enigmatic 'moved protection settings from "Foo" to "Bar" (Foo moved to Bar)' for a protected page, nothing at all relevant for a page with PC. Yes it's recorded in the page history, but for frequently-edited pages (as pages with prot or PC often are) it can mean going back through several screens to find the entry.
This brings me to the next problem. Even when I do find the history entry for the prot, the information that I want isn't always there, because of the 255-byte limit on history entries: a typical entry might be "Changed protection level of Foo: Violations of the biographies of living persons policy ([Edit=Allow only autoconfirmed users] (expires 23:59, 23 June 2015 (UTC)) [Move=Allow only administrators] (indefinite))". Now imagine that the page name is quite long, also that the reason for protection selected from the dropdown has been supplemented by a fairly lengthy custom reason; this can mean the loss of some information, which may include the expiry date and time, even the protection level. I see an obvious way to economise on space here: don't include the page name. If you're viewing a page history, it's at the top; if you're looking at the watchlist or somebody's contribs, it's shown earlier on the line, so for all three lists, including it in the log is totally redundant. Space can also be gained by shortening "Allow only autoconfirmed users" and "Allow only administrators" to "autoconfirmed" and "sysop". We do this in the page logs - why not in the history? --Redrose64 (talk) 22:37, 23 June 2015 (UTC)
Template:Left
Latest comment: 9 years ago6 comments4 people in discussion
Something has changed recently in relation to the {{left}} template. The last changes to the template were in March this year, made by Frietjes (talk·contribs) and Plastikspork (talk·contribs), but it's not clear whether that was the direct cause. The problem I'm seeing is at Template:Lea Valley Lines where the red vertical lines should be continuous, but instead there are gaps: and these gaps are occurring at the points where {{left}} is used. The intent, I think, is so that on a row like the one for Cheshunt, the distance (in this case 14m 01ch) is positioned to the left of the station name - instead, it's appearing below, and causing vertical separation. Has some CSS changed that may have affected this template? --Redrose64 (talk) 09:01, 23 June 2015 (UTC)
I can't see the spacing problem that you are describing. However I do see some mismatched double braces, which are likely to be causing an error somewhere. — Martin (MSGJ · talk) 10:23, 23 June 2015 (UTC)
Time for me to try all browsers then. I see the problem in Firefox 38.0.5 (under Windows XP); more to follow. --Redrose64 (talk) 10:28, 23 June 2015 (UTC)
OK, the accompanying screenshot is in three panes: the top one is from IE8 (Chrome Version 43.0.2357.130 m and Opera Version 12.17 are similar), and is the correct display; the middle one is what I see in Firefox, notice that two rows are broken (there were many more); the bottom one is Safari 5.1.7, where only one row shows breakage. --Redrose64 (talk) 11:15, 23 June 2015 (UTC)
I'm thinking that this might be the cause. Looks like quite a mess, and it might take a while to get the tools functioning completely as before. --Alan W (talk) 03:11, 24 June 2015 (UTC)
Thanks, Andre. I work with things like this all the time (mostly reporting them, not fixing them). Not claiming to be an expert in any of the technologies involved, but I understand just enough to say Ouch! I wouldn't want to be responsible for fixing this problem. The average Wikipedian probably has no idea of all the work that must go on behind the scenes to get these tools to function correctly. We tend to take such things for granted. --Alan W (talk) 04:57, 26 June 2015 (UTC)
Letterhead class?
Latest comment: 9 years ago13 comments6 people in discussion
I like to make use of a certain snippet of code, div class="letterhead", to make visually attractive quote delineations—it makes text like it's been written into a yellow notepad. However, this appears not to work anymore: for example it fails to appear here or, as I used it, here. I like this formatting option...where has it gone? ResMar17:46, 23 June 2015 (UTC)
I know that Philippe and I talked about it before removing and it was only removed because of the question and because we couldn't think of anywhere it was used anymore (it was originally used for a strategy announcement about 5 years ago). If the community is using it there is certainly no reason not to add it back :) it's sadly tough to really gauge how much it's used which is why we didn't realize. Jalexander--WMF19:08, 23 June 2015 (UTC)
I doubt it's used much, but I used it in several Signpost stories last year, and it's perfect for them when quoting long-form announcements. I'll add it back now. Thanks, James! Ed[talk][majestic titan]19:14, 23 June 2015 (UTC)
And this is why ppl shouldn't put 1 off crap like that in MediaWiki:Common.css.... People start using it. We now have a global CSS rule for a few signpost articles ???? That's crazy. —TheDJ (talk • contribs) 19:25, 23 June 2015 (UTC)
After four years of it laying around did you seriously not expect it to be taken advantage of? I'm used to hearing that kind of logic from the WMF—less and less now, thankfully—not from a fellow member of the community. ResMar00:00, 24 June 2015 (UTC)
First of all, I'm critiquing WMF for putting it in, in the first place and then NOT remembering to remove it. And second, I translate the above as "we don't care about site performance for our users, if you do, that's your problem". I and a few others spent years trying to keep the size of MediaWiki:Common.css down to somewhat reasonable proportions. This is one more example of the total insular view of the community towards technical matters. And the reason why we have so few users able to look after it properly. Whatever, I shouldn't let my own frustration cloud this. This was bad judgement by WMF and now we (community) can't easily fix it anymore. That doesn't mean we should keep it in, it will just be very hard to get it out and fix all current uses to inline the style instead of depending on a class. PS. with global i mean site scoped, instead of scoped to those users who need it because their page has such an element. —TheDJ (talk • contribs) 06:37, 24 June 2015 (UTC)
TheDJ: I know enough code to know that it's ultimately a problem, but also enough about matters in the movement to see it used (by the election committee no less) here and want to have access to it too. You have to break things to make them better, sure, but merely turning off a feature a lot of editors rely on is the wrong way to approach the problem. I suggest that this is one more example of the total insular view techies towards community matters. Make a list of the class's uses, provide an alternative, and create a deprecation schedule. I am reminded of a certain "routine" removal that the WMF made at one point in the past of some parsing code at the end of the text of a page load that broke a large number of bots relying on raw dumps. That content is back there again now. ResMar14:00, 25 June 2015 (UTC)
Text is not centered
Latest comment: 9 years ago11 comments7 people in discussion
Examples? If you use IE9 and saw centered text (headers) in infoboxes, you saw an IE bug which has now been fixed. -- [[User:Edokter]] {{talk}}18:58, 23 June 2015 (UTC)
@Edokter: I can see the same problem, demonstrated in the same (now) left-aligned information in Template:MedalTableTop. Conversion to a normal wikitable style allows normal text centering for columns, but use of "text-align:center" doesn't seem to work. I'm using latest Firefox. SFB19:18, 23 June 2015 (UTC)
I'm not sure if enough users are technically-inclined enough to understand MediaWiki markup (I'm sure not), but I just would like to add that entries are also no longer centered in chronology-related fields located inside music infoboxes ({{infobox single}}, {{infobox album}}, etc). Erpertblah, blah, blah...03:16, 25 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
When trying to edit Wikipedia:Village_pump_(policy)#MOS:IDENTITY_clarification, a rather long section with 15 subsubsections and counting, in one of the subsubsections, I hit the "edit" link and was taken to the source of the previous subsubsection. After a bit of futility, I found that I could go to the next subsubsection, edit it, and get the subsubsection I wanted. It worked. Meanwhile, in doublechecking to see how reproducible the error is, I am now finding the error is not happening.
I have default skins, use Firefox Portable on Windows. I'd tell you version numbers, but I have no idea what they are, and both seem to keep the information Top Secret. Choor monster (talk) 18:55, 23 June 2015 (UTC)
The error message is: "TypeError: $.isDomElement is not a function".
Two users have reported the error (which manifests itself in the entire script failing to load) on the wizard's talkpage since 22 June, and I can replicate on my machine under Firefox 37.0.2. It doesn't seem to apply to all users though, as occasional uploads using the wizard still show up in the logs (the latest one I can see at 18:44, 23 June 2015).
I don't use the upload wizards - they never worked properly when they were new, so after about five attempts with no success, I went back to the old ways. If you're interested, the links are: Wikipedia:Upload/old (English Wikipedia); c:Commons:Upload (Commons). --Redrose64 (talk) 10:04, 24 June 2015 (UTC)
The maintainers are still active and we are working on recruiting more, but Hedonil left xTools' code in such a convoluted state, it's hard to follow and virtually impossible to debug. Our primary goal is to get a new environment set up to restore stability to xTools, then we work on a rewrite that the maintainers can more easily maintain. Following that, the gadgets will be moved on to the xtools environment and Hedonil's JS will be redirected directly to xTools where everyone can maintain the script. We then plan to revive Wikiviewstats, which will likely need to be rewritten too, as I spent 3 hours looking for the bug and couldn't find what is causing Wikiviewstats to be broken. But first we need more maintainers. As xTools gets more maintainers, things will get done faster. We are currently discussing methods to recruiting new users.—cyberpowerChat:Limited Access02:36, 24 June 2015 (UTC)
I didn't get a thing. Did you send it to the right user? I'm Cyberpower678. Do you perhaps have Yahoo!?—cyberpowerChat:Limited Access18:40, 24 June 2015 (UTC)
Page revision History statistics has changed
Latest comment: 9 years ago2 comments2 people in discussion
Latest comment: 9 years ago3 comments3 people in discussion
jQuery.mwExtension was apparently deprecated a few days ago.
Also apparently, the function has already been removed from the deployed site? At least we were getting issues at WT:TW.
I've replaced the calls in Twinkle, but I'm noticing javascript errors in site script as well. In debug mode, mediawiki.util.js is throwing errors since it's using $.escapeRE.
It's no longer depended on by any of the default modules which are enabled on all pages. Since it was an undeclared dependency for Twinkle, it was causing errors. The issue with mediawiki.util is only in debug mode and caused by the way the caching is different for debug mode. —TheDJ (talk • contribs) 19:12, 24 June 2015 (UTC)
Latest comment: 9 years ago3 comments2 people in discussion
After two or three days of this, I'm pretty sure it's not just me. And it's sporadic. Edits I do, that may show up in a preview, vanish when I do a save. Some that do show up, the next day aren't there, not in my Contributions or anything. And some of them will be there if I open the edit window, and will be there in my Contributions, but not immediately on the Watch list, and not showing up on the article for several minutes. And then some of my edits just work like they're supposed to. — Maile (talk) 21:33, 24 June 2015 (UTC)
Latest comment: 9 years ago6 comments4 people in discussion
Hello all; since User:Citation bot is down, does anyone know of any other good tools for expanding bare citations consisting only of a doi or bibcode (NOT a sole URL, so no ReFill), or any other good replacements for the bot? StringTheory11 (t • c) 23:13, 24 June 2015 (UTC)
mw:Citoid (in VisualEditor; opt in via Beta Features in your prefs to try it out) can handle most, but not all, dois. The docs don't say anything about Bibcode, but it does mention "bibtex" as a supported format, so it might be planned. I've added phab:T103900 to its list in case they hadn't thought about it yet. Whatamidoing (WMF) (talk) 19:54, 25 June 2015 (UTC)
Thanks, will try Citoid. However, can the WMF please actually make it work with the source editor as well, which I prefer for everything other than editing tables? It seems like the WMF is trying to force us to use VE with this move, which doesn't come across well. StringTheory11 (t • c) 01:23, 26 June 2015 (UTC)
The citoid service will be added to the wikitext editor when it covers more sources. If you want to try it out in the wikitext editor, then perhaps User:Salix alba could tell us whether his user script to do that is still working. Whatamidoing (WMF) (talk) 21:23, 26 June 2015 (UTC)
Major scripting breakage
Latest comment: 9 years ago6 comments3 people in discussion
Seems most of the scripting based functions suddenly failed in the past hour or so - major failure on things like Hotcat or other normally working scripts. Dl2000 (talk) 03:02, 25 June 2015 (UTC)
Cease fire... looks like things are functioning normally again. Did some resetting of the Preferences, plenty of cache clearing reloads, cleared some browsing history, ensuring the browser scripting is allowed. Not sure what bombed, not sure which side(s) it came from, but probably was worth doing plenty of browser-side and preference resets anyway. Dl2000 (talk) 03:14, 25 June 2015 (UTC)
Please file a ticket in phab about that codeEditorMonitorFragment thing if you have time, so that I can remember to fix it. —TheDJ (talk • contribs) 07:44, 25 June 2015 (UTC)
Latest comment: 9 years ago7 comments2 people in discussion
When I do searches for multiple accounts, etc., I often get hits for userpages that have a transcluded Special:RecentChanges template. I can't find this template to see how many users do this. I ask because it drives me rangy getting all those false positives. Thoughts? Anna Frodesiak (talk) 06:30, 25 June 2015 (UTC)
@Anna Frodesiak: Do you mean pages that transclude the actual list of recent changes? You can do that with the code {{Special:RecentChanges}}, but it's not a template; you're transcluding the actual special page itself. You can't use Special:WhatLinksHere with recent changes (unless, apparently, you are using Flow), but if the users in question are transcluding recent changes indirectly through another template, you could find the links from that template instead. What do you mean by doing searches for multiple accounts, by the way? There might be a better way of doing whatever it is you're trying to do. — Mr. Stradivarius♪ talk ♪06:49, 25 June 2015 (UTC)
Hi, Mr. Stradivarius. Yes, pages like this. It contained {{Special:RecentChanges|limit=1000}}. This is a problem in this example: I find a new account. He's posted some inappropriate content at his userpage, such as "Everyone at school hates ███████ ███████ because he is a poo-poo-head...". Five minutes earlier he registered another account and did the same. So, sometimes I will search for such a string and what comes are recent changes transclusions at userpages containing (←Created page with Everyone at school hates ███████ ███████ because he is a poo-poo-head...) Anna Frodesiak (talk) 08:36, 25 June 2015 (UTC)
Hi, Mr. Stradivarius. You are very sweet. That page could be upside down and make as much sense to me. :) I must confess, I'm still trying to figure out how to get this to appear. My browser doesn't show that at all. I didn't want to bug you for help yet because you'd think me daft. Anyway, I'll live with the false positives. Maybe when I encounter them, I'll drop the user a line and ask if he could remove the transclusion if they're not using it. Anna Frodesiak (talk) 09:16, 25 June 2015 (UTC)
@Anna Frodesiak: I probably just wasn't very good at explaining it - sorry for talking in technobabble. :) If you make a normal search for "Everyone at school hates ███████ ███████ because he is a poo-poo-head..." then MediaWiki will look for that text found in the displayed page. So it will find all of the false positives from people who put {{Special:RecentChanges}} on their user pages. However, if you search for insource:"Everyone at school hates ███████ ███████ because he is a poo-poo-head...", then MediaWiki will look for that text in the source wikitext, not the displayed page. The source wikitext of the false positives won't contain the actual text "Everyone at school hates ███████ ███████ because he is a poo-poo-head..." - it will only contain something like {{Special:RecentChanges}}, and so MediaWiki won't include those pages in the search results. — Mr. Stradivarius♪ talk ♪10:15, 25 June 2015 (UTC)
Ahhhhh I see, Mr. Stradivarius.I get it now. No need to be sorry. I am very duncy when it comes to this sort of thing. Okay, I just add "insource:" and then the text and it will give me only the text where it is actually typed. Splendid. Thank you for this. It will be a huge help. Anna Frodesiak (talk) 10:34, 25 June 2015 (UTC)
Pages without Wikidata equivalent
Latest comment: 9 years ago4 comments3 people in discussion
Hi. Does somebody know whether Wikipedia has a special page for articles without a Wikidata item? Not to be confused with Special:WithoutInterwiki, which addresses pages without equivalents in other languages, regardless of having / not having an item on Wikidata. --Gikü (talk) 08:55, 25 June 2015 (UTC)
Latest comment: 9 years ago3 comments2 people in discussion
It's taking 60+ seconds to load all of my JS. Half of it loads immediately but the other half takes 60+ seconds. This just started recently.—cyberpowerChat:Online13:42, 25 June 2015 (UTC)
Did you check which specific scripts from which servers have problems to load, e.g. via the "Network" tab of your web browser's developer tools? Wondering if there's any kind of pattern. Also, which browser and version is this about? --AKlapper (WMF) (talk) 08:53, 26 June 2015 (UTC)
Latest comment: 9 years ago3 comments2 people in discussion
Okay, this might just be me being a careless idiot, but several times over the last couple of days, I've clicked to edit a section, typed a comment, then either previewed or saved and found that my comment was inserted in the section above the one I intended to edit. For example, just now I was looking at this revision of an Rfd thread, clicked beside "what is the best story" to edit that section, typed a comment and saved (no preview, lazy I know) and found my comment inserted in the section above: [82]. Clearly, according to the edit summary, I edited the section above ("sampernandu"), but I'm sure that I clicked beside the "best story" section. I think a clue might come from the fact that another user inserted a new section at the top of the page ([83]) and it could be that the system just counts the number of sections down from the top when you click to edit (I don't know how this works, just guessing) so then my second-from-the-top edit link was really (according to the database) for the section that I ended up editing. But I've never had this happen before this week. Has there been a recent change that would cause this behaviour, and is it something that can/should be fixed? Ivanvector (talk) 15:33, 25 June 2015 (UTC)
However that template, as used in this example has four rows of data input for each row of data output. Because I will have the stats in an Excel spreadsheet, it will be natural and easy for me to have one row of data for each year. I think it should be possible to use this template with one long row of input data for each year. However, if I simply concatenate the data, it does not work.
I would be grateful if someone could tell me what I am missing. Is it necessary to break up the Rose RMI missing some character or some other way to combine the rows?
@Sphilbrick: The "|-" lines need to be their own line, but you can concatenate everything else, provided you replace newlines followed by pipes with double pipes. Example:
That’s an intriguing possibility. I’ve used visual editor for simple text but not for tables. On the one hand the templates automatically create links to seasons and team years. On the other hand, I want to use it for a college team so the AutoLink to NBA seasons is an appropriate and I’ll have to create a new template for that. Second in my specific case I’m talking about Pepperdine and there are no team years for them, so perhaps I’ll try generating a table and copy and pasting using visual editor.--S Philbrick(Talk)20:50, 25 June 2015 (UTC)
Good luck. Don't forget that you (usually) still have to set table class separately in the wikitext editor for copy-pasted tables, and background colors still have to be done manually. If all the lines between cells are invisible after the save, then that's the likely cause. Whatamidoing (WMF) (talk) 22:11, 25 June 2015 (UTC)
I'm not quite getting it. You mentioned that I needed to set the table class separately which was a hint. Plus you mentioned that the lines between cells might be invisible and they are.
I tried two different things. First, (which is also documented at User:Sphilbrick/basketball_stats) I clicked on the visual editor edit button and simply copied and pasted a table. That table doesn't have shading for the header rows and doesn't have lines between cells. Is it possible to enter a table this way and then set the table class?
As an important aside, because I get a bit annoyed when people come and ask questions they can easily look up, I checked out the visual editor help; specifically:
That largely is has a placeholder, if there is other documented help please point iy out and I'll be happy to read it. The second thing I tried was to go into visual editor and tell it I wanted to insert a table. However that brings up a blank 4 x 4 table and if I copy and paste the whole table it seemed to copy it into the upper left most cell.
You also mentioned CSV format. The data I'm working with is in a file I've saved in CSV format but I'm simply copying and pasting it so I don't know that the format means anything. Should I be importing the CSV file? If so, I don't see how.--S Philbrick(Talk) 15:11, 26 June 2015 (UTC)
Whatamidoing Ping.--S Philbrick(Talk)18:34, 26 June 2015 (UTC)
If you save the spreadsheet as a CSV file, then you can actually drag-and-drop the file into VisualEditor. Steps: Open VisualEditor, pick up mouse, drag the file into the middle of the Wikipedia article, and voilà, it auto-imports immediately. You can set the header cells (select them, and go to the menu in the main toolbar, where you would normally set paragraph vs section headings, and choose "Header cell" rather than the default "Content cell"). Copying and pasting also works for normal tables.
But when you save the page, there's no wikitable class set (that's phab:T85577), so the lines between the cells are invisible. And you can't set or change the class (yet) in VisualEditor. So you have to save (or switch) and go back to it in the wikitext editor, and add the standard class="wikitable" to the first line, after the {| that starts the table to make it display properly to readers.
I apologize for the incomplete documentation. It's on my list. I'm hoping to get something solid written before Wikimania, when a lot of translation work usually happens. Whatamidoing (WMF) (talk) 22:33, 26 June 2015 (UTC)
Thanks that worked. No problem regarding documentation I fully understand the challenge of getting everything done at the same time. I mentioned it mainly for two reasons, first to let you know I wasn't simply begging for help I was trying to figure it out myself, and second there was a possibility that there was some documentation and I was simply looking in the wrong place. I know I recently had some issues with with references and after questioning I learned I was looking in the wrong page for the documentation so I wanted to check in case it existed and I just didn't know where to look.--S Philbrick(Talk)23:17, 26 June 2015 (UTC)
Request to add collapsible sections to a nav template
Latest comment: 9 years ago3 comments2 people in discussion
The template:Oral pathology is getting out of hand, and there are still many links to be added.
Another user has suggested collapsible sections, however after a few hours of messing around, apparently this is beyond my ability.
Please would someone be able to add collapsible sections to this nav template? Please note also that it would be desirable for the overall feel of the template to be retained (colors, font, layout), as these are standardized across medical pages. Many thanks if you can help. Kind regards, Matthew Ferguson (talk) 21:00, 25 June 2015 (UTC)
Latest comment: 9 years ago7 comments2 people in discussion
Sample text for the Centralnotice/Sitenotice:
Should Wikipedia run a site-wide banner protesting the proposed amendment for freedom of panorama in EU? Discuss
I'd like to request assistance in putting the existence of the discussion (Wikipedia talk:Freedom of Panorama 2015) to the centralnotice/sitenotice. Community is largely unaware of the discussion. This was done during the SOPA debate as well. -- A Certain White Catchi? 22:10, 25 June 2015 (UTC)
No, that's for watchlist notices, which only appear on the watchlist - for this you need either SiteNotice or CentralNotice. SiteNotice only affects the English Wikipedia, and can be updated by admins at MediaWiki:Sitenotice. CentralNotice can add a banner on multiple wikis simultaneously, and requests are handled on Meta. — Mr. Stradivarius♪ talk ♪23:29, 25 June 2015 (UTC)
I am not familiar which one would be more relevant. I do not believe I have ever requested something to be added in mass notification before. Needless to say this notice should be visible by editors only since it is a discussion to weather or not put a banner to mass notification. -- A Certain White Catchi? 07:12, 26 June 2015 (UTC)
Ah, sorry, you were talking about a notice about the discussion, not the proposed notice itself. I misunderstood. A watchlist notice does indeed sound like a good idea for advertising the discussion itself, so I've added it. — Mr. Stradivarius♪ talk ♪08:05, 26 June 2015 (UTC)
(edit conflict) Just for the sake of completeness, I should also mention that we have the option of restricting a SiteNotice to logged-in editors only. That's achieved by adding the SiteNotice as normal, and then adding <p></p> or similar to MediaWiki:Anonnotice. That notice is for anonymous editors only, and SiteNotice functions differently depending on whether it is enabled or not. — Mr. Stradivarius♪ talk ♪12:27, 26 June 2015 (UTC)
When would the discussion be sufficient/adequate to put the banner on site notice for guests? There will be no point to the banner if it is too late. -- A Certain White Catchi? 20:57, 26 June 2015 (UTC)
Deprecated cite parameters
Latest comment: 9 years ago2 comments2 people in discussion
I can guess about which parameters to fix, but I know of no way I can quickly verify that I have corrected all the problems. What can I do to verify that my citations are correct?
As a guess, |pages= contains a dash and a second number does not come after the dash. In any of the citations. The rest of each of the citations look fine. --Izno (talk) 13:28, 26 June 2015 (UTC)
Image previews
Latest comment: 9 years ago10 comments3 people in discussion
Can anyone explain why, in the infobox at Eden Land, hovering over the A Fool Who'll article wikilink shows the image on that article .... but hovering over the other album in that chronology (Our Swan Song) fails to show the image? Similarly, at A Fool Who'll, the wikilink to the previous album in the chronology (Eden Land) also fails to show the image as a preview. Just hovering over the wikilinks in the thread you are reading also shows the difference between articles in which the image appears and those in which it doesn't appear. I create and edit many album articles, but I don't recall seeing any where the primary image in wikilinked articles fails to appear as a preview. BlackCab (TALK) 11:53, 26 June 2015 (UTC)
Baffling. I ticked the Popups box (it hadn't been ticked) and it is still quite selective about what it shows as a preview. Those two album articles are the first I've noticed that don't reveal the image on preview. BlackCab (TALK) 12:26, 26 June 2015 (UTC)
Alright. You're talking about Hovercards, which I did not have enabled. You're right - not displaying. My mistake. - NQ(talk)12:34, 26 June 2015 (UTC)
I added the A Fool Who'll image an hour or two ago, but the other two articles, with images, have been in existence for some years. I wondered whether it was the way the images had been uploaded, so I replaced the FUR template on one, but with no change. BlackCab (TALK) 12:54, 26 June 2015 (UTC)
(edit conflict) Maybe the 200×200 image in Our Swan Song is too small for Hovercards. I don't know the details but the description to the right of mw:Talk:Beta Features/Hovercards says: "First image is not always used due to image proportions, size, or quality". In my limited tests, Hovercards scaled larger images down to 300px and didn't display smaller images. PrimeHunter (talk) 13:46, 26 June 2015 (UTC)
Latest comment: 9 years ago5 comments2 people in discussion
Is there an issue with pop-ups? Mine aren't working at the moment. Check my preferences to make sure it wasn't accidentally turned off.--S Philbrick(Talk)14:29, 26 June 2015 (UTC)
When Hovercards first came out, I tried it, however I got a conflict because I was getting information from that gadget and pop-ups and it was quite confusing, so I turned it off. I just try turning it on, and when on, I see Our Swan Song but not much else. I just turned it off so now the swansong image doesn't appear I'm not getting much of anything. If I hover over your name I get a small box that says "user: NQ" but none of the links to contributions etc.--S Philbrick(Talk)14:53, 26 June 2015 (UTC)
Sometimes popups freeze when used alongside hovercards. Since you seem to have it disabled, perhaps another script in your common/vector/monobook.js is causing a conflict? - NQ(talk)15:07, 26 June 2015 (UTC)
Popups now working. I've done nothing special, unless enabling howvercards and disabling it counts. FTR, popups was not working immediately after that. I'm going to mark this as resolved, although that doesn't mean the problem was identified.
Latest comment: 9 years ago5 comments2 people in discussion
In this edit, I tried to include two simple-format tables similar to the example at Help:Table#Multiplication table. However, when I previewed the edit, the paragraph immediately following the first table appeared to the right of the table (without a gutter in between) and not below.
I was able to work around this by inserting a {{-}} after the table, but why did I need to? Was my table markup wrong in some way, or what?
Your first table uses the attribute align="left" which floats the table. All you need do is omit that attribute. --Redrose64 (talk) 16:06, 26 June 2015 (UTC)
The align= attribute (which is obsolete in HTML 5, and was deprecated in HTML 4) has different meanings depending on whether it's used on the table as a whole, on the table caption, on a row, or on an individual cell. It only affects cell content when used on the cell itself, or on the row that the cell is part of. --Redrose64 (talk) 11:26, 27 June 2015 (UTC)
Latest comment: 9 years ago2 comments2 people in discussion
I have over 8000 pages on my watchlist. If I try to cut it down using the raw watchlist editor in either Chrome or Firefox, the text field never quite finishes loading so I can't actively select which pages I want deleted. My only other options are to nuke my watchlist, remove them one at a time, or let it grow until either somebody fixes it or the the heat death of the universe occurs. Mark Schierbecker (talk) 05:57, 27 June 2015 (UTC)
Latest comment: 9 years ago8 comments3 people in discussion
There seems to be a problem with the Special:DeadendPages cache or something as many of the pages listed are not dead end pages. Any advice and or suggestions? User:Mr._Stradivarius suggested making a post here for advice first and making a request in phabricator to raise awareness of the problem. Is anyone else already active in resolving this fault?. Bulgarios (talk) 15:41, 27 June 2015 (UTC)
You didn't link to that talk page (whether the page in general or the specific thread), and whereas you did link the phab ticket, its number wasn't visible. --Redrose64 (talk) 20:28, 27 June 2015 (UTC)
Special:DeadendPages currently says (UTC): "The following information is cached, and was last updated 04:11, 23 June 2015."
For all of them except Nieuw-Milligen, the last edit before the dead-end caching was User:ClueBot NG reverting a vandalism edit which had briefly turned the article into a dead-end. The ClueBot NG edit was often weeks before 23 June. PrimeHunter (talk) 20:34, 27 June 2015 (UTC)
User:Redrose64, in my previous comment above, I was referring to " User:Mr._Stradivarius suggested making a post here for advice first and making a request in phabricator to raise awareness of the problem" being advice taken from that talk page. I apologize for my precision in use of language which sometimes causes confusion for people not familiar with my verbal reasoning style. Best regards. Bulgarios (talk) 08:13, 28 June 2015 (UTC)
User Contribution Search
Latest comment: 9 years ago3 comments2 people in discussion
It seems that syntax highlighting has changed. Yesterday (26 June 2015) I noticed that the syntax highlighting colours and the boldfacing on .css and .js pages had changed - no big deal really; but just now I've noticed that the <source>...</source> extension (or <syntaxhighlight>...</syntaxhighlight> if you prefer) no longer respects the enclose=div attribute. This means that you can no longer indent blocks of code wrapped in those tags, because this happens:
/* start of CSS rule */.exampleclass{color:red;}/* end of CSS rule */
This should be a single box, containing three lines of code, all monospaced. It still works if you put it hard left:
/* start of CSS rule */.exampleclass{color:red;}/* end of CSS rule */
Well, the purpose of was to wrap long lines, which is now done by default; it won't be reintroduced. That it permitted linebreaks within list items was accidental. Alakzi (talk) 17:38, 29 June 2015 (UTC)
Does this problem also effect tab spacing and syntax highlighting for large Lua module pages? See Module:Convert and Module:Citation/CS1 for examples where the code is not highlighted (at least for me using current version of Chrome – the smaller subsidiary pages of Module:Citation/CS1 are properly highlighted).
The old classes used by GeSHi are gone, since Pygments is now responsible for syntax highlighting (despite the extension name still being SyntaxHighlight_GeSHi). -- [[User:Edokter]] {{talk}}19:57, 29 June 2015 (UTC)
Ok, but how do I tell this Pygments thing that I want tab size to be four spaces instead of 8 spaces? How do tell this Pygments thing to render Module:Citation/CS1 with syntax highlighting?
There is a resource limit for highlighting of 100KB right now. The exact limit is still under discussion on phabricator somewhere. —TheDJ (talk • contribs) 06:45, 30 June 2015 (UTC)
Latest comment: 9 years ago8 comments5 people in discussion
I have been using the Cat-a-lot script tool over the past few weeks to do some large-scale editing. Unfortunately, it stopped working for me close to a week ago, and nothing I've tried has gotten it to jump-start. I can begin the process, but it won't actually let me transfer articles between categories any longer.
I just tried to copy a category with it, and got the following error: "Uncaught TypeError: $.escapeRE is not a function: index.php?action=raw&ctype=text/javascript&title=MediaWiki:Gadget-Cat-a-lot.js:385." It looks like this is another victim of #jQuery.escapeRE vs. mw.RegEx.escape above. — Mr. Stradivarius♪ talk ♪06:33, 29 June 2015 (UTC)
Latest comment: 9 years ago5 comments5 people in discussion
For the past several days, whenever I try to use the "Edits by user" function on any article or talk page, I get the following message:
ERROR: Unhandled exception.[Errno socket error] [Errno -2] Name or service not known
Please try again.
Is this happening to other people, and is it being looked into and fixed? I hope so. By the way, if a bug report needs to be filed, could someone else do it? I'm lousy with that and I find it confusing and time-consuming. Thanks. Softlavender (talk) 06:55, 28 June 2015 (UTC)
Hi User:Softlavender, it is not easy to file a bug-report without direct personal experiential knowledge of the bug, but I am happy to report a problem here for research if you don't want to, and if you can explain the problem more for me? I am quite new here and can't find the function you mentioned. Bulgarios (talk) 08:19, 29 June 2015 (UTC)
Filing bugs for something hosted on Labs is generally a waste of time and if they involve the WMF they are invariably put on a back burner just like non life-threating medical attention by the NHS or ignored completely, leaving the tool owner/designers without any support for their efforts. The person to go to is probably cyberpower678 who by no fault entirely of his own appears to be trying to maintain such tools single handedly. If he is not directly responsible for this particular tool, he will probably know who is. Kudpung กุดผึ้ง (talk) 09:02, 29 June 2015 (UTC)
Not really sure what you are rambling about here, but as you can see in the ticket linked above, the issue has been resolved by one of the tool's maintainers within minutes of him being made aware of it.--Anders Feder (talk) 11:57, 29 June 2015 (UTC)
Template trouble
Latest comment: 9 years ago5 comments2 people in discussion
Recently I had a problem that is related to the template {{quotation}}, although I can't say that this template is the ultimate source of the problem. It may or may not be. By commenting out the template and using instead quotation marks for the quoted material, this workaround seemed to fix the problem. diff I think this is template trouble because it is a recent problem yet it affects previous versions of my talk page going back for years. --Bob K31416 (talk) 17:03, 28 June 2015 (UTC)
Thanks. Apparently there was a screwup in software development at Wikipedia which caused these issues with the templates. For years these weren't issues. Unfortunately, although these measures will work with current versions of Talk pages, they don't affect the history. The historic versions are vulnerable to adverse changes in Wikipedia software at any time. Interesting that the historic versions of pages aren't really permanent, because they depend on any future changes in the software at Wikipedia, as in the present case that we discussed. --Bob K31416 (talk) 14:35, 29 June 2015 (UTC)
"Misplaced" notification message
Latest comment: 9 years ago10 comments5 people in discussion
So, I just got a notification message that I'm about 100% positive was not meant for me. Does this happen often, or is it unusual? In either case, is there anything I should do on my end?... Thanks in advance. --IJBall (contribs • talk) 18:02, 28 June 2015 (UTC)
I mean, when I checked the diff, it wasn't my user name that had been pinged. (Now, I believe my username was mentioned somewhere else on the Talk page in question – but the 'ping' did not seem to be meant for me...) --IJBall (contribs • talk) 18:29, 28 June 2015 (UTC)
Yes, it wasn't a ping but a full scale transclusion of Dennis Bratland's userpage. Your username is mentioned in the userpage under "Article alerts" and hence a ping notification was sent.- NQ(talk)18:45, 28 June 2015 (UTC)
Heh. I didn't even notice that! (I wonder how many others got "pinged" on that!...) Anyway, good to know – at least I'll know what to look for if that ever happens again. Thanks to both of you! --IJBall (contribs • talk) 18:48, 28 June 2015 (UTC)
There are safeguards to reduce the number of accidental pings but sometimes they also prevent intended pings. mw:Manual:Echo#Technical details says: "If the number of detected to-be-pinged users exceeds 20, no notifications will be delivered." PrimeHunter (talk) 19:43, 28 June 2015 (UTC)
Changed pages on my watchlist
Latest comment: 9 years ago20 comments6 people in discussion
Hello! Hm, a minute ago all pages on my watchlist turned into the "visited" status while I haven't clicked the "mark all pages as visited" button, which I've also hidden using some JavaScript. What's going on? I really need to know which pages need reviewing as I have hundreds of unreviewed article modifications in about 20 days worth of a backlog. — Dsimic (talk | contribs) 18:07, 28 June 2015 (UTC)
There's something wrong for sure, as even the article that was modified in the last minute or so, after the all-around change on my watchlist, isn't currently marked as modified. — Dsimic (talk | contribs) 18:11, 28 June 2015 (UTC)
Yup. If you disable the "Display green collapsible arrows and green bullets for changed pages in your Watchlist, History and Recent changes" gadget in preferences, it displays fine (with bolded text). - NQ(talk)18:13, 28 June 2015 (UTC)
I don't have that option enabled. However, the green bullets are back, but displaying modified pages in bold still doesn't work. — Dsimic (talk | contribs) 18:15, 28 June 2015 (UTC)
I'm right in the middle of moving all watchlist related styles from common/skin CSS to a default gadget. That may cause some missing styles for up to five minutes, but it should then be OK again. -- [[User:Edokter]] {{talk}}18:17, 28 June 2015 (UTC)
I opted to simplify all the watchlist options and gadgets. The default (gadget on) is green bullets, but when switched off, I planned to restore the watchlist to software default using bold titles, thereby negating the need for the "no bold" option. However, that resulted in a style flash, so I restored the no-bold as default. I may re-add that option somehow. -- [[User:Edokter]] {{talk}}18:27, 28 June 2015 (UTC)
Well, something doesn't work as it should. If I untick "Display green collapsible arrows and green bullets for changed pages in your Watchlist, History and Recent changes" in my preferences, the result is quite bad – no green bullets and no bold titles. How can we have bold titles back, together with green bullets? I can't work without those. — Dsimic (talk | contribs) 18:35, 28 June 2015 (UTC)
They both have the exact same meaning. I intended to reinstate the bold titles when the gadget is turned off, but I have to find a solution for the flash styling. However, if there is a demand for both to be controlled separately, I'll restore that option. -- [[User:Edokter]] {{talk}}18:46, 28 June 2015 (UTC)
It would be the best to have separate controls for the green bullets and bold titles. I really find it difficult to do anything without the bold titles. — Dsimic (talk | contribs) 18:49, 28 June 2015 (UTC)
It might be a while before Edokter can get this sorted; you might wanna add the following to your Special:MyPage/vector.css:
I'd rather not have two options, oh well. At least all the code is (somewhat) consolidated. Just putting the final touches in and I'm done. -- [[User:Edokter]] {{talk}}19:00, 28 June 2015 (UTC)
Latest tech news from the Wikimedia technical community. Please tell other users about these changes. Not all changes will affect you. Translations are available.
When you read a page on the mobile site, you can now see a link to go to its talk page. [88]
The code coloring tool has changed. You can now use many more languages. It now also works on mobile. [89][90][91]
You can look at a new site to learn how to reuse data from Wikimedia sites. [92]
Problems
JavaScript was broken on some wikis due to a code error. VisualEditor and other tools that use JavaScript were broken. [93]
Changes this week
The new version of MediaWiki will be on test wikis and MediaWiki.org from June 30. It will be on non-Wikipedia wikis from July 1. It will be on all Wikipedias from July 2 (calendar).
You can now use a tool in VisualEditor to add and edit code in color. [94]
When you edit a sentence in the Translate tool, it can show you older translations. They help you save time if they look alike. The older translations should now work better. If you see problems, you should report them. [95]
If you are an admin or have other special rights, you now need a strong password. [96]
Bot and JavaScript coders: The old continuation mode of the API for action=query doesn't work any more. [97]
Meetings
You can join the next meeting with the Editing team. During the meeting, you can tell developers which bugs are the most important. The meeting will be on June 30 at 19:00 (UTC). See how to join.
Future
When you create a new account, it will also create one on Meta-Wiki and mediawiki.org. [98]
Latest comment: 9 years ago3 comments2 people in discussion
Perhaps one of you can easily fix this--I can't. If you click on the German link for Theodor Julius Jaffé you get led to Carl Seffner (strange--but Seffner made his cenotaph). If you click on the English link in the German article for Jaffe, you to get directed to the right place. I don't understand the first thing about Wikidate, but you all are the smart people. Thanks. 207.93.13.150 (talk) 17:12, 29 June 2015 (UTC)
Add "mark as read" action to hover-popups in edit histories and watchlists
Latest comment: 9 years ago9 comments3 people in discussion
(I'm guessing this would be uncontroversial enough to count as a technical matter.) In the hover-popup in edit histories and watchlists (actions popups or actions user popups etc.) how about adding a "mark as read" action? Sometimes you already know, without actually visiting the article, that you want the article to "unbold" in future refreshes of your watchlist.
Also, will require downloading the full article, because that entry point is the only way that will count as a 'read'. And there is no way to unread right now at all.. —TheDJ (talk • contribs) 06:43, 30 June 2015 (UTC)
How do you know it's the only way? Obviously the whole point here is to avoid the actual download of the article. EEng (talk) 10:40, 30 June 2015 (UTC)
Because I know what the php code for this looks like. It's basically just a timestamp for each entry in your watch, that is updated from the code that constructs the page for users. There are no other interfaces to this timestamp right now. —TheDJ (talk • contribs) 14:15, 30 June 2015 (UTC)
See, you should have more faith! Goody. Now all we need is someone to modify the dropdown... EEng (talk) 11:36, 1 July 2015 (UTC)
Unmatched end tag being reported for Watchlist page
Latest comment: 9 years ago5 comments2 people in discussion
It seems there is an extra closing paragraph tag being inserted on the Special:Watchlist page when there is at least one watchlist-message still pending/rendering (i.e. has not been hidden or dismissed manually yet). My console reports the extra end tag occurring at line 64 fwiw...
Latest comment: 9 years ago11 comments6 people in discussion
I now have green arrows and bullet points in my watchlist for unviewed pages, and blue arrows and bullet points in my watchlist for viewed changes. I have two requests:
Change the blue to gray.
Make the arrows and bullet points larger so that they are more easily distinguishable.
Hello! While the gray bullet points might make the watchlist look better and more usable, making them larger would make no sense IMHO. That's the standard bullet point size used all around Wikipedia. — Dsimic (talk | contribs) 05:51, 30 June 2015 (UTC)
I need to lean toward my monitor to make the distinction between the bullet and the arrow. Perhaps there could be a user preferences setting for the size of the icon, or the icons could be different in color as well as shape, for example the green bullet could be changed to a purple bullet? --Pine✉06:07, 30 June 2015 (UTC)
Do you use bold titles for changed pages? If not, please try ticking that preferences option, bold titles make a world of difference in watchlist usability. :) — Dsimic (talk | contribs) 06:14, 30 June 2015 (UTC)
What browser do you use Pine ? I know there is a bug with Safari for instance, where these things become tiny. Also the changes are English Wikipedia specific and managed by User:Edokter. —TheDJ (talk • contribs) 06:38, 30 June 2015 (UTC)
I'm glad you like the concept. The default watchlist has had green bullets ( / ) for unvisited pages for over a year now, so I extended them to the enhanced watchlist as well. So that's where the green comes from. The bullets have the same deminsions as the default list bullets (5x13). The arrows ( / ) are blue to indicate they are clickable. That concept was introduced in CategoryTree (for example, see Category:Contents). The size of the arrows are the same as the default grey ones used in the interface. The space where they appear are constrained to 15x15. The problem TheDJ ponts out should no affect Safari, as the images are rendered using background images, not using the list-style-image CSS property. -- [[User:Edokter]] {{talk}}16:49, 30 June 2015 (UTC)
Which fact (that the green bullets are images) I'm grateful for, because I use Safari, and the normal (non-image) bullets are *tiny* in Safari.
All of these features are handled locally, by the volunteer community rather than by the WMF. (This, plus the fact that I probably did something wrong here, is why I can't get these green bullets everywhere.) Edokter is the person you want to talk to. Whatamidoing (WMF) (talk) 18:08, 1 July 2015 (UTC)
Shouldn't be. One of the things about CSS is that (except inside quoted strings) spaces and newlines are interchangeable, and where one space is permitted, more than one is also permitted. So a blank line, which is two newlines, is syntactically equivalent to a single space. --Redrose64 (talk) 21:22, 1 July 2015 (UTC)
June 17 Labs outage effect on still missing June 17 pageview stats
Latest comment: 9 years ago1 comment1 person in discussion
Latest comment: 9 years ago6 comments3 people in discussion
A reader reported (via OTRS ticket:2015062710013531 ) problems while attempting to print Tau protein. I confirmed that it was not a print to PDF problem as has cropped up recently, but a direct print to a printer from the article. The problem appears to be the beginning of the article. I don't yet know the browser, I will respond to the individual that they should post more information here. — Preceding unsigned comment added by Sphilbrick (talk • contribs) 17:13, 30 June 2015 (UTC)
Fixed, but I thought links did not have to have exact case. Maybe I misunderstood. The reader referred to "Tau Protein" so I searched for that term, got a hit, and missed that it lowered the case on the second word.--S Philbrick(Talk)17:48, 30 June 2015 (UTC)
Who should look into this? My first thought was to post something at the article talk page, but if I understand your response the issue is with the info box, so if anything, it should go to the editors of the info box. Is this correct?
What should be said to them? Barring a better answer I intend to direct them to this discussion so they can see what the problem is but I do not know how to summarize it well.
I asked the reader to try sending first to PDF and then printing which worked. I now wondered why this worked, so I tried it again myself and noticed that the PDF does not include the info box. I tried to other articles and in neither case did the info box show up in the PDF. Is this known?--S Philbrick(Talk)19:39, 30 June 2015 (UTC)
Autocorrect on search bars
Latest comment: 9 years ago5 comments4 people in discussion
Apologies if this is the wrong place to bring this up.
Could you please add autocorrect="off" to the search fields in Wikipedia? Usually people come to look up proper nouns, and iOS devices constantly try to "correct" them. It's maddening. — Preceding unsigned comment added by LastFootnote (talk • contribs) 15:28, 30 June 2015 (UTC)
Latest comment: 9 years ago4 comments3 people in discussion
The "Learn more" button on the above advert - I'm presuming its for EU users only - doesn't work on my tablet computer, Android 4.0.4. Can someone get the spanners out please, or add a plain link to it. - X201 (talk) 21:17, 30 June 2015 (UTC)
It appears not to be for only EU users, but it really should be, IMO. The "Learn more" does appear to be a plain link, but it only covers the text itself, and not the entire button. --Yair rand (talk) 21:21, 30 June 2015 (UTC)
How to render measurement for best results
Latest comment: 9 years ago4 comments4 people in discussion
I have a problem with, and a suggestion for, how measurements are treated in Wikipedia. I was directed here.
The basic measurement policy is to cite the measurement in one unit and put another in parentheses. So we say, "the speed limit is 60 MPH (130 kph)".
The Problem: If there are more than a few such figures, the parenthetical figures become tedious. They're not that helpful to the (say) metric reader, and they're distracting to the native reader.
A good example is: Track Gauge. No one really cares how many millimeters apart the rails are. No reader is well served by the parenthetical conversions.
Parenthetical conversions were a reasonable choice when Wikipedia was young, but these days it's just not technically necessary.
The Solution: Let the reader choose which to see. Use CSS to render the reader's preferred units. If the browser and/or technical environment somehow preclude that, fall back to current behavior.
For example, let the reader's login hold a measurement unit preference. Use that in rendering the page, and include the converted measurement such that most browsers will render it as a "tool tip", displayed when the mouse hovers over the figure. At the top of the page, include a link to choose units for this page only, just for now, in case the reader wants to try viewing the page from a different measurement perspective.
There are some details to work out. On the Track Gauge page linked to above, the primary measures are mostly US, but under "Medium gauge" there are references to "1,000 mm (3 ft 3 3/8 in) gauge railways" because the gauge was specified metrically. It might be that the reader's options are "US/metric/natural". There's also prior art we can build on.
My suggestion is really two parts, one policy and one technical. We're starting here with what's technically feasible in Wikipedia. If we can arrive at a workable answer, then we can take it over to the policy group as a concrete proposal.
Although I do know something about it, HTML/CSS isn't my area of expertise, and I don't know much about how wiki markup is converted into what the brawser actually sees. What's the best way to carry this out? Jklowden (talk) 22:54, 30 June 2015 (UTC)
Lest we mislead anyone, 60 MPH is not 130 kph but close to 100 kph. The conversion is quite close to: MPH = 0.6 x kph. Akld guy (talk) 05:28, 1 July 2015 (UTC)
I like the suggestion; after all, we already do the same thing for date formats outside of article space. And I assume that the {{convert}} template would be the enabler here. While I predict a giant brouhaha if this gets to the policy stage, there are a few tricky details which I believe impact the technical design of this: for example, there would need to be some kind of opt-out provision for text where parenthetical conversions are legitimately encyclopedic; one example that comes to mind is this paragraph. Next (and maybe this is what you meant by providing a "natural" option), there are customary units in certain contexts where a unit could be converted but shouldn't always be. For example, the speed of sailing vessels is typically expressed in knots, which can be converted to a more widely-understood measurement, but that would be annoying for readers familiar with the area. I agree that most readers don't know what a knot is and would benefit from a more familiar measurement, but I'd hate to lose that for myself. Regards, Orange Suede Sofa (talk) 17:10, 1 July 2015 (UTC)
@Calidum: If you want to just prevent them from turning green (so they'll all be dark blue), you can go to Special:Preferences#mw-prefsection-gadgets, scroll down to the "Watchlist" section, and uncheck Display green collapsible arrows and green bullets for changed pages in your Watchlist, History and Recent changes. --Ahecht (TALK PAGE) 03:29, 1 July 2015 (UTC)
Redirect templates for all countries lacking their own full Wikiproject? (e.g. {{WPLIBERIA}} acting as WPAFRICA|Liberia=yes automatically)
Latest comment: 9 years ago4 comments4 people in discussion
I'm concerned that there are a lot of Africa articles about specific countries which are just filed as "WikiProject Africa" on their Talk pages, since people might not be familiar with how to code in sub-projects. It would be way faster if I could just type {{WPLIBERIA}} onto a page about Liberia, rather than type out an Africa code and then have to put in full sub-codes.
Could we ensure that all countries that don't have a functional shortcut like {{WPIRELAND}}, {{WPJAPAN}}, {{WPBRAZIL}} have a functioning code, even if it just acts as a parallel of {{WPAFRICA|Liberia=yes}}? That way if someone makes an article for "Fooistan" and in good faith goes to the Talk page and puts down {{WPFOOISTAN}} doesn't get a redlink, and thus have to either go puzzle out what the full code is (and the formats aren't exactly consistent so you can't just guess), just put the WP for its continent down, or leave it redlink or blank. That way too, if a full Wikiproject for Fooistan is created in the future, the templates are already correct, we'd just change their destination to the smaller country-specific Wikiproject instead of the continent-wide one, and we don't need to go hunt down individual articles to change their Project.
It would create too much confusion if it's unclear whether a template corresponds to a WikiProject. {{WPAFRICA}} accepts several parameters. Should {{WPLIBERIA}} pass all parameters on to {{WPAFRICA}}? Does {{WPLIBERIA|importance=Top}} mean top importance for Liberia or all of Africa? {{WPAFRICA}} has both Liberia-importance and importance. If importance=Top is changed to Liberia-importance=Top when {{WPLIBERIA}} calls {{WPAFRICA}} then do we invent a new parameter name for Africa importance in {{WPLIBERIA}}? And many articles cover more than one country and have to use a single WikiProject banner to avoid the banner being repeated for each country which is just a wrapper. For example, Talk:Ebola virus epidemic in West Africa currently says {{WikiProject Africa|class=B|importance=High|Guinea=y|Guinea-importance=High|Liberia=y|Liberia-importance=High|Sierra Leone=y|Sierra Leone-importance=High|Nigeria=y|Nigeria-importance=Mid}}. PrimeHunter (talk) 13:12, 2 July 2015 (UTC)
If {{WPLIBERIA}} is set up as a true redirect (like Template:WPBOTSWANA is), it will recognise exactly the same parameters as the template that it redirects to. If however it is set up as a wrapper (like {{WikiProject Trains in Japan}} is), the recognised parameters can be a subset of those recognised by the main template, but every one that should be recognised needs to be explicitly passed through. --Redrose64 (talk) 17:57, 2 July 2015 (UTC)
– Paine 17:21, 11 May 2015 (UTC)
Resolved
Done — 
, Its done , thanks dear X201 
. –Davey2010Talk 14:30, 25 May 2015 (UTC)
Stale
Work in progress; comments welcome
Fixed—cyberpowerChat:Online 09:21, 2 June 2015 (UTC)
I didn't see [39] then [40]. If you cannot see, in regular view, List of commercially important fish species as sortable, then the problem is your end. Nothing is broken this end. That table is sortable to us. I'm sorry to have agitated you. Anna Frodesiak (talk) 13:46, 1 June 2015 (UTC)
. Regards, JoeHebda (talk) 11:37, 4 June 2015 (UTC)
![[a 2]](/wiki/File:Bad_Title_Example.png){kind=small}
Like--ukexpat (talk) 15:38, 10 June 2015 (UTC)
JohnBlackburne: A bit of context, this new extension was imported from the German Wikipedia. The Germans also have a map template that I can't seem to import the JSON file for, so I left it alone for now. Yurik pointed out on my talk page that, in the absence of the should-have-happened-ages-ago interwiki transclusion feature, the best solution for the quickly oncoming onslaught of template sprawl is the creation of a bot that copies all of the templates' content every once in a while from a master template somewhere: either on MediaWiki or on Meta. I am afraid that it is unlikely that someone will be found to do it fast enough, before the organic growth Wikipedia is so famous for smothers us all in the heat death of yet another thousand modules and templates on the various wikis all maintained independently but doing the same damn thing. ResMar 03:50, 13 June 2015 (UTC)
Comment: to readers... we still have the working counter services Supercount and Wikichecker! (Wikichecker may take a few refreshes to load) — MusikAnimal talk 15:00, 17 June 2015 (UTC)
Resolved
/ 
) for unvisited pages for over a year now, so I extended them to the enhanced watchlist as well. So that's where the green comes from. The bullets have the same deminsions as the default list bullets (5x13). The arrows ( 
/ 
) are blue to indicate they are clickable. That concept was introduced in CategoryTree (for example, see Category:Contents). The size of the arrows are the same as the default grey ones used in the interface. The space where they appear are constrained to 15x15. The problem TheDJ ponts out should no affect Safari, as the images are rendered using background images, not using the list-style-image CSS property. 
![[16]](https://commons.wikimedia.org/wiki/File:Saint-S%C3%A9verin_Sanctuary,_Paris,_France_-_Diliff.jpg){kind=link}
![[17]](https://commons.wikimedia.org/wiki/File:Saint-S%C3%A9verin_Nave,_Paris,_France_-_Diliff.jpg){kind=link}
![[18]](https://commons.wikimedia.org/wiki/File:Saint-S%C3%A9verin_Organ,_Paris,_France_-_Diliff.jpg){kind=link}
![[19]](https://en.wiki.x.io/wiki/File:United_States_change_frame_1870-07-15.png){kind=link}
![[20]](https://commons.wikimedia.org/wiki/File:United_States_change_frame_1870-07-15.png){kind=link}
![[84]](https://en.wiki.x.io/wiki/File:Laurajeanedenlandcover.jpg){kind=link}