construct-stylesheets VS topics

> With adoptedStyleSheets they're objecting to making backwards incompatible changes

Which would not be bacwards incompatible if they hadn't shipped something that wasn't agreed on in the first place.

Again, slowly: they literally shipped that to production despite loud and explicit objecitons from both Firefox and Safari they shipped it to production. When asked to hide it back behind the flag, "but backwards incompatible change, the framework we're developing is already depending on it"

And since you're quoting rniwa, here's the relevant quote https://github.com/WICG/construct-stylesheets/issues/45#issu...:

--- start quote ---

I feel like I’ve put so much time & energy into making this feature something sane & useful, and all you did was basically to dismiss many of my feedbacks and go with whatever you like and just ship it. And now you’re saying you can’t make changes because you shipped it?

I’m sorry but that’s just not how standards work.

--- end quote ---

> With WebMIDI they're saying they want to do an announcement before making the change.

Indeed. Once again: because they shipped an API that neither Safari nor Mozilla supported. Now that this issue has surfaced (no thanks to Chrome), they can't just roll it back or fix it because people already rely ono this behaviour, which the implicitly acknowledge.

It's not true, to w3c's surprising credit.

What Google does, is publish a "draft" which is as far from a standard as their authors are from the Moon. This gives Chrome the leeway to call it an "emerging standard" and just ship it. It doesn't care if there are objections, or that other browser vendors will not implement it. It's now a "standard" in Google's dictionary.

For something to become a W3C standard even in the present world, you need a consensus and at least two independent implementations. None of that exists for stuff Google pushes out (hardware APIs, web transport, constructible stylesheets [1], the list goes on...).

The correct name for those is Chrome-only non-standards.

[1] These one isn't even a draft. It is.... "a collection of interesting ideas" in a working group https://wicg.github.io/construct-stylesheets/ Shipped by default in Chrome, of course

I literally provided just some of the examples. Those are easily verifiable.

Web Transport is shipped by default. What was the input from other browser?

Here's the timeline for HID: https://github.com/mozilla/standards-positions/issues/459#is...

Constructible Stylesheets: the spec contained a trivially reproducible race condition, the API was badly specified. Google shipped against any objections and refused to bring it back under the flag. Full discussion here: https://github.com/WICG/construct-stylesheets/issues/45. Shipped in Chrome https://github.com/WICG/construct-stylesheets/issues/45#issu... (may be hidden on mobile) despite multiple unresolved issues. Two years later Chrome did add a better API that people originally requested, other issues potentially remain.

> If there are examples of 'Apple ignoring standards' actually meaning Chrome-only features please tell me one.

Easy.

The most obvious/glaring one is WebHID. Enjoy the timeline: https://github.com/mozilla/standards-positions/issues/459

It's not just HID, of course. All/most of the hardware APIs are considered harmful by both Safari and Mozilla. Chrome is shipping them enabled by default, and there's no end to clueless developers maoning about this and calling Safari (mostly) and Firefox (from time to time) too slow in "moving the web forward". Needless to say that all those non-standards are pushed forward by Chrome.

The less obvious one is Constructable Stylesheets.

The spec had an obvious flaw that could lead to easily reproducible deadlocks. And that is on top with other issues with design, API naming etc. A team within Google (lit-html) wanted this feature, so Chrome shipped it against clear objections from both Safari and Firefox. And then refused to move the feature back under a flag because "0.8% of page views in Chrome" were suddenly using this feature. And proceded to gaslight other browsers' developers https://github.com/WICG/construct-stylesheets/issues/45. See e.g. a response to that https://github.com/WICG/construct-stylesheets/issues/45#issu... Of course there's now a "looking ahead" that wants to do exactly what Safari and Mozilla wanted to do in the first place: https://web.dev/constructable-stylesheets/#looking-ahead

In general, Chrome pushes 40 to over 100 new Web APIs with each release (that is, every two months). How many of them are actual standards that had actual input from other browser developers? In how many Chrome actually listened and implemented suggestions? https://web-confluence.appspot.com/#!/confluence

> The negatives are often theoretical

They are not theoretical. Too bad webapicontroversy.com has been shut down (it looked like this [1]), but you can scroll down to "defer" and "considered harmful" in Mozilla's positions here: [2]

There are more, of course, but they are not visible unless you're willing to follow thousands of issues across hundreds of GitHub repositories. One that springs to mind is, of course Constructible Stylesheets. Mozilla and Safari: the spec describes an algorithm that leads to deadlock in trivial code, we wont implement it until this is fixed. [3] Chrome: ship it, because lit-html (developed by Google) wants it and is already using it. And then procedes to gaslight people and misrepresent their positions (cant' find the relevant link, but at this point I can't find the will to dive into the cesspool).

[1] https://user-images.githubusercontent.com/32768/108985355-3f...

[2] https://mozilla.github.io/standards-positions/

[3] https://github.com/WICG/construct-stylesheets/issues/45#issu...

The browser keeps track of he top 5 categories from this list of these 629 topics. https://github.com/patcg-individual-drafts/topics/blob/main/...

Other than when it returns a random topic, the browser only reveals a topic to a site if that site has observed the user on a site with that topic before.

Delete Chrome.

“The intent of the Topics API is to provide callers (including third-party ad-tech or advertising providers on the page that run script) with coarse-grained advertising topics that the page visitor might currently be interested in.”

https://github.com/patcg-individual-drafts/topics#the-api-an...

Those are the top-level categories. Each of them has subcategories which are more granular. Not all of them are public, from what I can tell. Here's an example of some that are. https://github.com/patcg-individual-drafts/topics/blob/main/...

https://github.com/patcg-individual-drafts/topics/blob/main/...

>knows that you are male and sees that you've been recently interested in dresses and panties, and this website happens to be a far-right-leaning activist website, and decides to dox you, or blackmail you, or forward this information to Ron DeSantis's administration for possible criminal prosecution, you're all good with that?

If you want to keep topics a secret you can just block them. Every week of your 5 topics that gets selected there is a 5% chance that a topic in replaced with a random one. If you see a user's topic is /Shopping/Apparel/Women's Clothing/Dresses it could be there by chance. It would also require the site to take out a bunch of ads on these women clothing sites hoping that one of your future website visitors would see your ad.

> You seem to be taking things that are factual, normal, everyday, aspects of the WHATWG working process and trying to imply that chrome is doing something unusual, or untoward with its process here, but it isn't. It's doing what is necessary to make a proposal with WHATWG: have a trial.

And yet, we've seen many such proposals go through this process because Chrome is paying lip service to it. Whatever Google wants it ships. And Google wants this.

As an adjacent (ads- and tracking-related) example: Google's FLoC flopped, hard. So they immediatey shipped the replacement Topics API [1] despite there being no consensus. E.g. Firefox is against [2] (but Chrome presents Firefox's position as "No signal" in the feature status). And despite the fact that its status is literally "individual proposal, not accepted" [3]

Do not assume any good intent on Google's part when it comes to Google's business interests. Their intent is always malicious until proven otherwise. And there have been fewer and fewer cases when they have been proven otherwise.

[1] https://chromestatus.com/feature/5680923054964736

[2] https://github.com/mozilla/standards-positions/issues/622

[3] https://github.com/patcg-individual-drafts/topics

> [0] https://github.com/patcg-individual-drafts/topics/blob/main/...

Nice v1 at the end. We can assume that this list is final and will not be changed?

> nor (by policy) other kinds of sensitive PII.

Yeah, it just exposes interest in family planing, loans, ..., which we do know have absolutely no potential for abuse.

Or given the attempts to outlaw drag there is probably no potential way to use interest in Nail Care or Makeup in a negative fashion, right?

> Yes, if you assume the people who designed this API were idiots

I assume they are getting paid well to play the role.

It's related to Google's Topics API proposal, I guess. This new API automatically categorizes users into pre-defined "topics" that are inferred by the browser through a classifier model (basically matching the hostname with the classifier model). So, in the end, advertisers sent ads for these topics, and users within shall see them).

Don't take my word for it: WordPress treated FLoC as a security concern in 2021: https://make.wordpress.org/core/2021/04/18/proposal-treat-fl...

A good overview of the context: https://digiday.com/media/we-cant-un-floc-ourselves-googles-...

More detail: https://www.eff.org/deeplinks/2021/03/googles-floc-terrible-...

When it comes to Topics, it's essential that there be hands on the wheel at W3C that approach the solidification of e.g. the Topics taxonomy https://github.com/patcg-individual-drafts/topics/blob/main/... from a neutral perspective that takes into account the various ways in which proposed topics could be dangerous, and how strongly to word the specification to prevent it from creeping in increasingly privacy-eroding ways in the future.