Talk:Web Environment Integrity

Written like an ad

The first half of this is written like an advert, just repeating selling points from google's press release. Wikipedia:What Wikipedia is not#Wikipedia is not a soapbox or means of promotion 86.13.192.2 (talk) 22:23, 27 July 2023 (UTC)

The article has been updated since and no longer reads as a promotional piece. Anton.bersh (talk) 15:01, 31 August 2023 (UTC)

Thank you Anton. 86.13.192.2 (talk) 09:40, 19 October 2023 (UTC)

Web Environment Integrity API is not a "proposed standard"

Proposed standard is a special term in context of internet/web technologies and it is defined by IETF. It describes APIs/datagrams/etc. which are considered stable and are typically broadly deployed by the time they receive this characterization. WEI is definitely not a proposed standard, it is not even developed within W3C or IETF, it is one developer's individual GitHub repository. Although Google Chrome is close to shipping an initial version of this specification, this does not mean WEI ever advance beyond an experimental stage. Anton.bersh (talk) 16:17, 27 July 2023 (UTC)

The API appears to have been merged into Chromium in the commit titled "[wei] Ensure Origin Trial enables full feature". Gkaf (talk) 00:21, 28 July 2023 (UTC)

That article never specified it was a proposed standard in the W3C or IETF sense, just something that a few Google engineers proposed in their own GitHub repo. PhotographyEdits (talk) 07:32, 28 July 2023 (UTC)

It looks like this proposal has been abandoned, I think. Can somebody who knows the context take a look at this page to see if it needs updating or whatever? I'm feeling far too ill-informed to risk an update myself, sorry! https://github.com/RupertBenWiser/Web-Environment-Integrity Denny de la Haye (talk) 22:09, 2 November 2023 (UTC)

I updated the article to reflect WEI abandonment in [1]. Meta-comment: User:Aaron Liu keeps removing Denny de la Haye's signature from this comment: [2] [3]. DefaultFree (talk) 01:23, 4 November 2023 (UTC)

Please, his signature has always been there. Even if with the undated template it's a signature. I'll just find a WP:3O. On your last reply on your user talk page, even in real life signatures are separate from dates, and so are they on Wikipedia. Aaron Liu (talk) 01:50, 4 November 2023 (UTC)

3O Response: Dennymeta fixed their sig using the current timestamp. Given that there were no intermediate edits this is fine, it is not necessary to "fix" further and certainly should not be modified by anyone other than Dennymeta without a more compelling justification. Polyamorph (talk) 08:49, 4 November 2023 (UTC) Polyamorph (talk) 08:49, 4 November 2023 (UTC)

Hmm. So is it not necessary for timestamps to reflect the time the comment was added? Aaron Liu (talk) 12:57, 4 November 2023 (UTC)

Ideally it should, but the true timestamp is in the edit history and users often modify comments and update their signature - as long as there aren't any intermediate edits it's not a big deal. I can see why you changed it but it's not worth getting into conflict over it. Cheers Polyamorph (talk) 14:22, 4 November 2023 (UTC)

Table

WP:WHENTABLE says: "If there is no obvious benefit to having rows and columns, then a table is probably not the best choice.". I think this table should be converted to prose, for better accessibility. The exact dates of the opposing arguments are not really valuable IMHO, just the content of the critique and who said it should be enough. PhotographyEdits (talk) 12:36, 17 August 2023 (UTC)

The benefit is that you can see all the relevant parts of the quotes at a glance. Aaron Liu (talk) 01:13, 18 August 2023 (UTC)

I'm not convinced that is a benefit within the scope of WP:WHENTABLE. There is not really something that is good to sort or compare in this table. PhotographyEdits (talk) 07:44, 25 August 2023 (UTC)

I concur. I find the chronology useful, but that doesn't require a table. I've taken a shot at converting it to a list, based on the example at MOS:NO-TABLES: [4] DefaultFree (talk) 08:42, 25 August 2023 (UTC)

Looks better! Thanks. I might go even further and just put it as regular prose, but it's okay now. PhotographyEdits (talk) 09:38, 25 August 2023 (UTC)

I concur that the prose looks better than the table. Anton.bersh (talk) 15:14, 31 August 2023 (UTC)

Open Web

@DefaultFree "Open Web" refers specifically to the web standards that make up the web. Thinking "open" as a lowercase adjective doesn't really make sense. A lot of sources online use "Open Web" such as W3C and some use "open web" like Drupal's founder. None use "open Web". Aaron Liu (talk) 14:10, 19 August 2023 (UTC)

That's the Open Web Platform (a named W3C project), not the open Web (a descriptor of the Web). OWP (and similarly, Open Web Foundation) are capitalized because they're proper nouns as official names of a project or organization. It's false that no one uses "open Web"; here are a few Wikipedia articles currently using "open Web", for example:

• Mozilla
• Aaron Swartz
• Liddy Neville
• International Academy of Digital Arts and Sciences

You originally (correctly) asserted that "Open Web" is an incorrect capitalization: [5]

DefaultFree (talk) 22:13, 19 August 2023 (UTC)

I'm talking about sources online, not Wikipedia articles. The Open Web Platform is also equivalent to the "Open Web". "Open Web" does not equal the Web, it's specifically the open-source or open-standard technologies that underlay it. "Web" has also been gradually decapitalized as it became more and more commonplace like something we take for granted and exists everywhere. No matter "Open Web", "open web" is definintly correct. [6] mixes "the Open Web" and "an open web". This article uses "the" so it should also be capitalized. Aaron Liu (talk) 22:25, 19 August 2023 (UTC)

OK, sources online, not Wikipedia article text. Here are some credible external sources online using "open Web":

• W3C
• World Wide Web Foundation
• Pew Research Center

Here's the WMF using "open Web" in a blog post: [7]

Your original edit was correct in uncapitalizing 'open'. DefaultFree (talk) 22:47, 19 August 2023 (UTC)

Hmm, so I guess there is no standard to change the capitalization to anything. Also I don't see where your first link uses "open Web" Aaron Liu (talk) 18:54, 20 August 2023 (UTC)

Ah, I was viewing it with JavaScript turned off. With JavaScript turned on, the instances of "open Web" are initially hidden behind collapsible containers. Try expanding the sections labeled "immersive Web". DefaultFree (talk) 22:42, 20 August 2023 (UTC)

Ah, freaking JavaScript. Aaron Liu (talk) 22:47, 20 August 2023 (UTC)

If the reason for using that variant is to avoid unnecessary capitalization then we should use the all lowercase version instead. The WMF also uses it (it's also their name for the tag on the "Diff" blog) and there are a lot more mentions of it online than "open Web". For example you can see that in Google ngram. Aaron Liu (talk) 01:56, 22 August 2023 (UTC)

I still contend that "open Web" is correct, and "open web" is incorrect, because 'Web' is a proper noun. But if we agree on 'open', then the remaining issue is a "capitalization of Internet" situation, and we should apply the same approach here as is generally done with 'internet' vs 'Internet'. The most recent MoS 'Internet' discussion I can find is Wikipedia talk:Manual of Style/Archive 223#Capitalisation of "Internet" (the global interconnected network of today), which appears to be inconclusive. (Also, I had no idea that the Google Ngram dataset was case sensitive, cool!) DefaultFree (talk) 02:13, 22 August 2023 (UTC)

The Internet and the Web are different things. Assuming it does weigh on "Web" capitalization, whether to use "Open" or "open" was also inconclusive. I don't really agree on "open", I want to either fully capitalize or don't capitalize. (unfortunately while Google Trends is also case sensitive that's quite useless as most people search in lowercase, uncool) Aaron Liu (talk) 02:45, 22 August 2023 (UTC)

It looks like you opened Wikipedia:Redirects for discussion/Log/2023 August 22#Capitalization of Web to continue this discussion; linking to it here for any other editors wishing to follow the conversation there. DefaultFree (talk) 02:54, 22 August 2023 (UTC)

It doesn't talk about the capitalization of web just that the redirect is useless so I didn't see a point Aaron Liu (talk) 03:01, 22 August 2023 (UTC)

(I meant to include this in my original reply:) "open Web" only surpassed "open web" at the end of two years: 2001 and 2003. 3 if you count 2009. Aaron Liu (talk) 03:03, 22 August 2023 (UTC)

Yeah that was a pretty elegant and nice solution. Aaron Liu (talk) 22:06, 22 August 2023 (UTC)

Copyedits

@DefaultFree I agree that the Mozilla phrasing might be very confusing, but I do want to make clear that the principle Mozilla's talking about is that everyone can use the Web. For the restricting browsers part all three sources say that. Vivaldi says that in the section "So, what is the issue?". EFF says that in the section "WEI helps websites block disfavored browsers". FSF says that throughout the entire article including "It will be used by governments to ensure that only their officially "approved" (read: backdoored) browsers are able to access the Internet; it will be used by corporations like Netflix to further Digital Restrictions Management (DRM); it will be used by Google to deny access to their services unless you are using a browser that gels with their profit margin." I agree with the W3C change. Aaron Liu (talk) 16:05, 25 August 2023 (UTC)

I'm not opposed to describing the nature of Mozilla's driving principle, as long as the resulting copy is grammatical and meaningful, and representative of the source material without WP:SYNTHing. Regarding claimed it would significantly limit the browsers that could be used, not every source makes that precise claim. Vivaldi says "there is no guarantee". EFF claims that the limiting would be unsuccessful because the "small percentage" of holdback clients would render WEI unusable by websites. Perhaps you could use individualized quotes from each source to describe the specific claim that the specific source is making. DefaultFree (talk) 17:05, 25 August 2023 (UTC)

Sorry for the late reply. Since I appear to be inable to produce better prose on Mozilla than the quote how about also having a quote for that? ...Mozilla opposed it, stating "Any browser, server, or publisher that implements common standards is automatically part of the Web."
I'm not sure how to quote Vivaldi in a way that accurately sums up the section. The best I can do is summarize it as "feared that attestation providers would not be trustworthy"
I think FSF does pretty clearly say that it would significantly limit.
That's not what EFF is saying by the small percentage part. What it's saying is that the WEI standard proposes to also randomly select a small percentage of its users to not be attested but it sees no reason that Google wouldn't make this percentage too small and websites won't ignore them for more profit. I think I can summarize that like this. Aaron Liu (talk) 00:07, 26 August 2023 (UTC)

Regarding the proposed Mozilla copy, unless readers infer additional context, it's unclear what the quote "Any browser, server, or publisher that implements common standards is automatically part of the Web" has to do with WEI or the WEI proposal. The current text makes a statement about WEI in relation to the Web; this proposed text just makes a statement about the Web, with no explicit connection to WEI. DefaultFree (talk) 01:17, 26 August 2023 (UTC)

Full draft: ...Mozilla opposed it, stating "Any browser, server, or publisher that implements common standards is automatically part of the Web... Mechanisms that attempt to restrict these choices are harmful to the openness of the Web ecosystem and are not good for users."
...Vivaldi opposed it as "simply dangerous" and feared that attestation providers would not be trustworthy.
...the Free Software Foundation opposed it as "an all-out attack on the free Internet" and claimed it would significantly limit the browsers that could be used.
...the Electronic Frontier Foundation opposed it as "a bad idea that Google should not pursue" and rejected its proposal of selecting a "small percentage" of random users to simulate behavior without WEI in order to prevent websites from using it as a whitelist. The EFF claimed that "[m]any websites will consider that “small percentage” of users an acceptable price to pay" and feared Google would set the percentage extremely low to attest more human ad clicks.
P.S: I think we should link to Mozilla instead of its foundation. The writeup says "Mozilla", the repository owner is "Mozilla" and the Foundation simply finances it and doesn't do any other stuff AFAIK. Aaron Liu (talk) 14:02, 26 August 2023 (UTC)

This looks better. Some remaining issues:

• The ellipsis in the Mozilla quote needs square brackets; see MOS:...
• EFF never says they "rejected" Google's holdback proposal, and the word choice might be overstating their authority in the decision-making process (I doubt Google will yield to their criticism).
• I'm not sure 'whitelist' is accurate, as whitelists are usually lists of explicitly approved entities, and this would be based on client behavior rather than client identity. The source doesn't use this wording.
• 'to attest more human ad clicks' - source doesn't say human clicks, and WEI attestations don't assert anything about the humanity of the agent driving the client.

I'm fine with retargeting the Mozilla link, and I think we should also think about updating Brave to Brendan Eich, as his statement was issued on his personal Twitter account, rather than an official Brave channel. It might be better if we could find an official statement and switch to that, since it's not clear to me that Eich in his individual capacity qualifies for inclusion in the list.

Additionally, while I don't object to this edit on this basis, I'm a bit confused about what principles you would like to govern quotations in this article. This edit significantly expands on a quotation that you previously minimized in this diff, on the basis of WP:OQ, and the table vs list vs prose distinction doesn't bear on the recommendations listed there. DefaultFree (talk) 05:42, 27 August 2023 (UTC)

• That MOS part says "occasionally"; it isn't a requirement and is mostly just used when the quoted part has ellipsis. I will add spaces according to that though.
• how about "opposed"?
• Maybe. How about ...in order to prevent websites from blocking unattested users?
• EFF says benefit Google’s ad fraud department by authenticating more ad clicks. Ad fraud is tricking ad services into thinking there are more valid (human) clicks than there actually are. The WEI explainer also emphasizing attesting humans a lot: This trust may assume that the client environment ... is transparent about whether or not a human is using it is in the first paragraph and the first example is attesting human ad clicks. I don't see what else "authenticating" ad clicks can mean.

Sure, I think "Brave Software CEO Brendan Eich" would be good. However it's also kinda official as he also made this change to Brave's documentation that they will remove WEI. I similarly doubt if that should be included.
I was going to say that I can't find Brave's statement but I just saw the link to Brave's blog and this official statement. I can't really find much to elaborate on though, how about On August 1, Brave Software announced they will not include WEI in their web browser.

The quotes before and after the edit have the same meaning in that context. If you're talking about the Mozilla quote, that's because the quote has additional meaning (on which principles) that I can't paraphrase well. OQ: A summary or paraphrase of a quotation is often better where the original wording could be improved. Aaron Liu (talk) 15:46, 27 August 2023 (UTC)

I think square brackets are always more clear, but after a closer reading of MOS:... I admit it is optional and a matter of local consensus. "Opposed" and "blocking" work for me. Human-ness is 100% WP:SYNTH. DefaultFree (talk) 16:22, 27 August 2023 (UTC)

Fine. feared Google would set the percentage extremely low to combat ad fraud. Aaron Liu (talk) 18:01, 27 August 2023 (UTC)