Wikipedia:Reference desk/Archives/Computing/2010 April 1
Computing desk | ||
---|---|---|
< March 31 | << Mar | April | May >> | April 2 > |
Welcome to the Wikipedia Computing Reference Desk Archives |
---|
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages. |
April 1
editSeeing entire Facebook wall at once
editIs there some sort of code I can enter into a web browser that will allow me to see a person's entire facebook wall at once instead of having to tediously click "Older posts" each time to reveal a small portion of the wall? Acceptable (talk) 03:43, 1 April 2010 (UTC)
- I think Facebook uses Javascript on some of its pages. I know that's a bit vague but that's one step closer. Googling may help... Chevymontecarlo. 10:47, 1 April 2010 (UTC)
- If you use the lite.facebook.com interface, then this is done using normal links. This doesn't really help much, other than that it probably makes it a little easier to write something like a Greasemonkey script to pull the information in automatically. 94.168.184.16 (talk) 12:57, 1 April 2010 (UTC)
Large Text File Viewer
editI am running Windows XP Media Center Edition version 2002 with service pack 2 (32-bit of course) and I am looking for a nice piece of software that would let me view very very large text files. I don't have a fixed size. The size ranges from MBs to GBs. The files are generated as part of an experiment I am running and I don't even need to edit them in anyway. I just want to be able to open them and view the simple text files without my computer crashing. I have notepad++ but I would like to know if there is something better out there. Preferably something that is free and small in size. Thanks! -Looking for Wisdom and Insight! (talk) 04:40, 1 April 2010 (UTC)
- Comparison of text editors will let you find the free ones. I tend to use SciTE, though I can't say if it can cope with GB files. Oh, I see that in the "extra features" section of that article there is a heading for "large file support". Handy. Maybe ConTEXT? 213.122.19.48 (talk) 05:04, 1 April 2010 (UTC)
The note for ConTEXT doesn't look so encouraging. I can tell you from experience that Vim plows straight through large files. ¦ Reisio (talk) 09:46, 1 April 2010 (UTC)
- Oh yeah, "a 1GB file gave an Out of Memory error." Missed that. 81.131.12.204 (talk) 10:43, 1 April 2010 (UTC)
- In that case, the problem is trying to load the entire file into memory at once. A smarter program would only load what would fit in the available memory. Note, however, that some operations, like find and replace, would be very slow with such a system, since different chunks of the file would need to be read, changed, and written to temporary paging space. StuRat (talk) 11:47, 1 April 2010 (UTC)
- UltraEdit will handle this - there's a fully-featured 'try before you buy' version. 94.168.184.16 (talk) 12:59, 1 April 2010 (UTC)
- Metapad says it can cope with an unlimited size of file. I've often used it. 78.149.194.146 (talk) 14:17, 3 April 2010 (UTC)
Has Firefox split into two versions?
editMy 3.0.x version still regularly updates itself (at the same time as whinging at me for not getting 3.6.x). It just became 3.0.19. Why are there still updates for 3.0.x? 213.122.19.48 (talk) 04:56, 1 April 2010 (UTC)
- 3.0.19 is a security patch.
- Presumably they don't want to force you to upgrade to a different version, so (at least for a while) they continue making security patches for the older versions.
- It's not that uncommon. APL (talk) 05:05, 1 April 2010 (UTC)
- Similar to how Windows XP and Vista are still supported by Microsoft even though Windows 7 is the new OS —Preceding unsigned comment added by 82.44.54.207 (talk) 10:42, 1 April 2010 (UTC)
- Mozilla will generally give some clue of the support lifecycle for their products (as with many vendors, including Microsoft for example). Although there's no centralised place [1], our article Mozilla Firefox does list the dates. For example Firefox 2 ended in December 2008 (as had been planned) with 2.0.0.20 (had been planned to be 2.0.0.19 but they neglected to fix a critical flaw [2] [3] [4]). Firefox 3 support ended with 3.0.0.19 so while you're okay now, it's strongly recommended you update in the near future. Incidentally Firefox 3.5 support will end in August 2010 [5] so I'd recommend you move to Firefox 3.6. Nil Einne (talk) 07:10, 2 April 2010 (UTC)
Virus-free, except not!
editOn my USB drive I currently have a naughty little application contained in a single .exe file that passes virus scans by both Norton and Avira, yet upon launching it (it's a legit app that has been spiked) two trojans pop up in my temporary files directory and are promptly zapped by both Norton and Avira (I've tested this on two machines). My question to you good sirs is: what sort of devilish programming can produce an app that passes virus scans and then immediately creates trojans upon execution? I find this fascinating. 59.46.38.107 (talk) 05:24, 1 April 2010 (UTC)
- It's simply impossible for an AV to catch everything, they can only check for known identifiers. ¦ Reisio (talk) 09:49, 1 April 2010 (UTC)
The security company that makes the antivirus has to be aware of that malware, virus or trojan before it can create something to combat it. It's kind of like a cat-and-mouse game between hackers and security companies - the hackers come up with a new virus, the security companies come up with a update that combats that. Then the hackers create something that bypasses that update, and so on. Chevymontecarlo. 10:37, 1 April 2010 (UTC)
I would say don't always rely on your AV to keep you safe. Even if it's kept updated and you scan your computer regularly, you need to have a bit of common sense as well to avoid any problems (i.e. avoid downloading stuff from dodgy websites or sites you don't know or trust). It may take a couple of months before a security company makes an update to its software to combat a new malware program. Chevymontecarlo. 10:45, 1 April 2010 (UTC)
- As has been pointed out, most virus detection works by checking for "exact match" with known malware. Some advanced programs can check "fingerprints" which are more flexible than an exact-match. Such fingerprints typically include individual code-fragment matches, especially if the bad part is cleanly separated, i.e. in a linked library or even just conveniently separated elsewhere in the program code block. But if the "clever" malware program just munges up his code - sometimes, even changing a single character in the source-code - these exact-match fingerprints break down completely. Ideally, we could detect viruses without requiring any "exact match" - allowing protection mechanisms to adapt to new implementations of malware. I knew a guy who was doing research into malware detection by analyzing a program's intent and capabilities - i.e., noticing that a piece of code would be used to scan the entire hard-disk when a program's declared intent doesn't require this. That would indicate suspicious behavior. Here's one report, Using specification-based intrusion detection for automated response, (2004). Nimur (talk) 14:20, 1 April 2010 (UTC)
- ... and the "devilish programming" will just be some self-modifying code that creates the code for the trojans as it runs. There is no way that the anti-virus software can detect the code for the trojan because it doesn't exist until the spiked software is run, but if you report the spiking to Norton and Avira they might include the signature of the code that creates the trojan in their next update. Dbfirs 17:29, 3 April 2010 (UTC)
- As has been pointed out, most virus detection works by checking for "exact match" with known malware. Some advanced programs can check "fingerprints" which are more flexible than an exact-match. Such fingerprints typically include individual code-fragment matches, especially if the bad part is cleanly separated, i.e. in a linked library or even just conveniently separated elsewhere in the program code block. But if the "clever" malware program just munges up his code - sometimes, even changing a single character in the source-code - these exact-match fingerprints break down completely. Ideally, we could detect viruses without requiring any "exact match" - allowing protection mechanisms to adapt to new implementations of malware. I knew a guy who was doing research into malware detection by analyzing a program's intent and capabilities - i.e., noticing that a piece of code would be used to scan the entire hard-disk when a program's declared intent doesn't require this. That would indicate suspicious behavior. Here's one report, Using specification-based intrusion detection for automated response, (2004). Nimur (talk) 14:20, 1 April 2010 (UTC)
How login super user
editHi. I want to know how is the super user in microsoft windows.And how to login.
- The superuser in Windows is usually known as Administrator. F (talk) 09:38, 1 April 2010 (UTC)
Yes, if you're the only person that uses your computer, and you only have one account, that is the Administrator account. As to 'how to login', you would just click on your user icon when you first boot up, and then enter a password if one has been set. Chevymontecarlo. 10:35, 1 April 2010 (UTC)
- It's not necessarily that simple. Depending on the version of windows (eg XP or Vista, Pro or Home Basic), you might not see the Administrator on the login screen. You might have to press Ctrl-Alt-Del twice to get the login dialog, at which you type the username Administrator. Or you might have to (eg Vista Home Basic) login as a user with admin privileges, run a command prompt as Administrator, then use NET USER to enable the Administrator account. Mitch Ames (talk) 01:44, 2 April 2010 (UTC)
Distributed computing projects
editI'm thinking of donating some of my spare CPU time to a distributed computing project. Has anyone ever joined one of these projects or are they still part of it, if so what did you think of it? What projects can I join? What are your experiences with certain projects? Thanks. Chevymontecarlo. 10:33, 1 April 2010 (UTC)
- A friend runs the SETI software on his computer when it would otherwise be dormant. This software tries to find signs of alien life by analyzing signals from space. StuRat (talk) 11:16, 1 April 2010 (UTC)
- It's worth mentioning that, when being driven hard, your computer draws more power - so running these programs (the BOINC client supports a number of different projects, with a list on their website - linked from the article) isn't totally free. Whether the cost is significant or not depends on how/when you're running it and how you feel about your electricity bill. :-) 94.168.184.16 (talk) 12:51, 1 April 2010 (UTC)
BOINC, that was it! I remember now. I was listening to a Howstuffworks Techstuff podcast about that a while ago, and I couldn't remember the name of it. Thanks Chevymontecarlo. 13:31, 1 April 2010 (UTC)
- I use World Community Grid, which serves a variety of scientific projects (it uses BOINC). I found it too much of a resource hog on my Mac laptop with 1GB of RAM, but it runs great on my 4GB Linux desktop. (Also it makes me feel better about springing for four cores, because it's about the only thing that ever makes use of them.) I've noticed it interfering with Flash games occasionally, but it's not hard to temporarily disable. Other than that, it has no effect on my day-to-day operations. Paul (Stansifer) 13:39, 1 April 2010 (UTC)
- There's also Folding@Home, which studies protein folding to try and find cures for protein related ailments (e.g. Alzheimer's, prion diseases, etc.). It also has a GPU client that can use your GPU for additional power. —ShadowRanger (talk|stalk) 14:15, 1 April 2010 (UTC)
- I participated in distributed.net years ago. It's apparently still running and is expected to take 660 years for a brute-force decryption of a 72-bit RSA key. Comet Tuttle (talk) 14:39, 1 April 2010 (UTC)
- You probably meant to write "72-bit RC5 key". A 72-bit RSA number can probably be factored on a pocket calculator these days.
decltype
(talk) 02:52, 6 April 2010 (UTC)
- You probably meant to write "72-bit RC5 key". A 72-bit RSA number can probably be factored on a pocket calculator these days.
Cool, thanks guys. I will copy this onto my talk page so I can research further without it getting archived. Chevymontecarlo. 19:12, 1 April 2010 (UTC)
php
editIn php, when there's a url like "test.php?mode=123" where or how is "?mode=123" defined in the php file? —Preceding unsigned comment added by 82.44.54.207 (talk) 10:40, 1 April 2010 (UTC)
- It gets passed as an HTTP GET variable. So the PHP programmer accesses it like this: $mode_variable = $_GET["mode"];, which would fill the variable with the data "123" if the URL was as you had specified. (There are other ways to do it, too. Full info here.) --Mr.98 (talk) 11:12, 1 April 2010 (UTC)
- Thank you, but I don't fully understand. Say the url was "test.php?mode=1" the word "cat" would be displayed. Then if you went to "test.php?mode=2" the word "dog" would be displayed. How is this done in php? —Preceding unsigned comment added by 82.44.54.207 (talk) 16:09, 1 April 2010 (UTC)
- You could trivially set up an array of options. Using a PHP array (details here), define a mapping, like so:
- $options_array = array(1 => "cat", 2 => "dog");
- To get perform the lookup, you'd do the following:
- $mode_int = $_GET["mode"];
- $mode_str = $options_array[$mode_int];
- Or you can combine the operations (if the number is not needed for anything but the lookup) like so:
- $mode_str = $options_array[$_GET["mode"]];
- Does that make sense? —ShadowRanger (talk|stalk) 20:18, 1 April 2010 (UTC)
- YES! Thank you!! — Preceding unsigned comment added by 82.44.54.207 (talk)
- You could trivially set up an array of options. Using a PHP array (details here), define a mapping, like so:
- Thank you, but I don't fully understand. Say the url was "test.php?mode=1" the word "cat" would be displayed. Then if you went to "test.php?mode=2" the word "dog" would be displayed. How is this done in php? —Preceding unsigned comment added by 82.44.54.207 (talk) 16:09, 1 April 2010 (UTC)
Connection detection in Windows 7
editWindows 7 is somehow able to detect whether it has a full internet connection available or not. My first thought was that it looks at the DHCP reply and, if finding no default gateway, will assume that it is in a local network without internet access. However, there seems to be a more sophisticated detection feature, as when I'm using a DSL router (which provides its own LAN IP as default gateway and DNS server addresses in its DHCP reply) that has its upstream connection unplugged or a restrictive firewall setting, I'm still getting the warning message that there is no internet connection available.
My router's logging capabilities are rather limited :-( and I'd like to get rid of that message so as to avoid unneccessary confusion among users (the system has limited internet access via a transparent proxy).
Does anyone know what Microsoft does to determine if there is an internet connection? Do they use ping, or a web request, or something completely different? And which IPs/DNS names are they trying to connect to? -- 78.43.60.58 (talk) 14:09, 1 April 2010 (UTC)
- I've searched around with little success. There are three possibilities I can think of:
- A) There may be a communications protocol (possibly DHCP or a common extension) where the router can inform the machine if it has an internet connection (or indirectly, inform the computer of the router's external IP address, where a null IP address indicates a disconnection). Depending on how sophisticated it is, it may be possible to fool it by chaining routers; if the protocol doesn't change, then the fact that router 1 connects to router 2 might be enough to fool the computer into assuming an internet connection exists, even if router 2 is disconnected.
- B) It tests against a "reliable site" (or sites). I doubt they use this approach, as it is far too susceptible to a single-point-of-failure.
- C) The computer is provided with the DNS lookup addresses from the router (the router itself gets them either through a static configuration, or as part of dynamic connection setup with your ISP). Rather than connecting to statically defined sites, it can simply ping (not necessarily an actual ping, but that's not important) the DNS addresses (there is usually a primary and a secondary). If neither responds, it can conclude you are offline.
- I consider B extremely unlikely. Searching for DHCP documentation (or reading the article more thoroughly than I did) may help determine if that protocol is involved. If DHCP doesn't handle this, the most likely case would be C. After all, if you can't reach your DNS lookup servers, you're effectively disconnected from the internet for most practical purposes, and reporting no connection in this case is mostly correct. —ShadowRanger (talk|stalk) 20:09, 1 April 2010 (UTC)
- It should be possible to test option C with sufficient hardware and technical skills. If you set up a DNS cache on your local network and use it as your primary DNS lookup, then disconnect your router from the internet and you still show as online, then clearly it's fudging it using the DNS. —ShadowRanger (talk|stalk) 20:11, 1 April 2010 (UTC)
- Just realized that you noted that the router is designating itself as the DNS, which necessitates another question: Does the router actually function as a DNS lookup? If it doesn't, then C could still be the case. If it does, then C becomes less likely. —ShadowRanger (talk|stalk) 20:21, 1 April 2010 (UTC)
- A short while ago, I had asked at this very reference desk if it was somehow possible to determine an upstream router's external IP. I was told there no such possibility exists (aside from brute-forcing it with nmap or similar tools). Also, adding another router doesn't change the behavior. So A) is out. Regarding C), I just tested pinging fr.wiki.x.io (I usually don't visit the French wikipedia, so it would be rather unlikely that my computer has its IP in its DNS cache). While the actual ping packets were blocked by the firewall, DNS resolution worked fine. So it seems that Microsoft indeed chose B), which leads me back to my original question: Which domains is it trying to access, and how? -- 78.43.60.58 (talk) 09:57, 2 April 2010 (UTC)
- Addendum: On the router I had plugged in to test A), I can see the following outgoing traffic:
- 224.0.0.22(IGMP)
- 224.0.0.252:5355(UDP)
- Could this be part of the detection mechanism, and if so, does anyone know what it's trying to do / what it expects as an answer? According to Classful network 224.0.0.x are multicast addresses, but what should that tell me? -- 78.43.60.58 (talk) 10:04, 2 April 2010 (UTC)
I haven't done any network tracing in a while, so this was an interesting exercise. I fired up Wireshark and pulled my internet connection. I noted DNS queries to dns.msftncsi.com and figured this might be involved. A quick search reveals this TechNet article on Network Connectivity Status Indicator and Resulting Internet Communication in Windows Vista, which explains it in depth. ---— Gadget850 (Ed) talk 14:20, 2 April 2010 (UTC)
- Spot on! Sweet! I just whitelisted that www.msftncsi.com domain for http traffic (since DNS was already allowed, see above) and ta-da, no more annoying messages! Thanks! -- 78.43.60.58 (talk) 17:20, 2 April 2010 (UTC)
Can I send people a URL that will cause a google doc to be downloaded as a pdf?
editIf I'm editing a google doc (e.g. http://docs.google.com/Doc?id=blah_blah_blah), is there some modification I can make to the url (e.g. ...&download=pdf) that will cause the latest version to download as a pdf?
When I share documents with people, I include at the top a list of hyperlinks to all the available viewing options ("click here to edit me," "click here to view me as static html," "click here to request rights to edit me ..."). Thus, it would be nice to add a button, "click here to download me as pdf."
If this pdf thing can't be done, will someone who has a friend working at google tell that person to implement this feature? :)
Thanks, JD Caselaw (talk) 18:35, 1 April 2010 (UTC)
- To produce a URL for downloading as .pdf, go to File → Print (in the google docs toolbar). There is no URL for downloading in other file formats; ask google at (here). -128.59.178.77 (talk) 17:24, 2 April 2010 (UTC)
- That seems a little hard, if i get to work at google i'll try to commit some code for that.
No smoking
editWhat is the name for that symbol that appears if you are trying to copy and paste but there is already a blue area on the screen and you have already released your mouse button, and your mouse won't let you do it again?
When there is a cigarette inside the symbol it means "no smoking".Vchimpanzee · talk · contributions · 21:10, 1 April 2010 (UTC)
Do you mean No symbol? Theresa Knott | token threats 21:12, 1 April 2010 (UTC)
- Yes, thanks. The No symbol article does not refer to its use in computers, and I wouldn't know how to add that information properly.Vchimpanzee · talk · contributions · 21:36, 1 April 2010 (UTC)
- The place to make requests like this for article changes, when you are not comfortable editing the article yourself, is to go to the "discussion" page of the article (click the 'discussion' tab at the top) and then click "new section" and add your request, exactly as you do here to add a new question to the Reference Desk. Comet Tuttle (talk) 23:39, 1 April 2010 (UTC)
- Right. Well, I was hoping for some answers, but I am at a library now and I can probably find a book that has that.Vchimpanzee · talk · contributions · 14:55, 2 April 2010 (UTC)
- No such luck. The books are outdated or incomplete. I could ask a reference librarian. Maybe someone reading this has an answer?Vchimpanzee · talk · contributions · 16:09, 2 April 2010 (UTC)
- I couldn't get anyone to answer me at the library. I forgot to say that here. Okay, this is a challenge I simply must overcome. I have to find answers and add them to the article. Just today I tried to copy and paste and got this symbol.Vchimpanzee · talk · contributions · 18:38, 5 April 2010 (UTC)
- Was it because you copied and pasted by selecting the text and then dragging it? Try, instead, selecting the text to copy, choosing "Copy" from the Edit menu, then clicking where you want to paste the text and finally choosing "Paste" from the Edit menu. I would expect that to always work. Comet Tuttle (talk) 18:47, 6 April 2010 (UTC)
- I always turn the text blue and use CTRL-C. If the computer interpreted something I did as dragging, that might have caused it.Vchimpanzee · talk · contributions · 18:52, 7 April 2010 (UTC)
Downloading videos off of YouTube
editIf I download videos using YouTube converter (or whatever the equivalent of that is), and I want to re-upload it to another video site, will I need some special re-encoding software to upload the video in the same quality that I downloaded it as? If so, can anyone recommend the best software for this? 24.189.90.68 (talk) 21:22, 1 April 2010 (UTC)
- If the other site accepts it in the original YouTube flv format, then you don't need it. And if it requires a different format, you *will* lose quality unless you are reencoding to (and they accept) one of the absurdly wasteful lossless video compression techniques. —ShadowRanger (talk|stalk) 21:47, 1 April 2010 (UTC)