zlacker

[return to "Google is already pushing WEI into Chromium"]
1. tolmas+We[view] [source] 2023-07-26 13:21:25
>>topshe+(OP)
Mozilla should call for Google's removal from the W3C over this implementation of Web Environment Integrity. "But Chrome has 65% market share, what good is the W3C without them?” If Google can take unilateral action to fundamentally change the basic principles of the web, then the W3C is already useless. This will give Google a clear choice: if they want to maintain the idea that the W3C matters, they should withdraw this implementation.

It is unbelievable that over the course of 3 days, the potential future of the web has been put in such dire straits. There's already an existing, far less troubling (while still bad), proposal in the form of Private Access Tokens going through a standards committee that Google chose to ignore. They presented this proposal in the shadiest way possible through a personal GitHub account. They immediately shut down outside contribution and comments. And despite the blowback they are already shoving a full implementation into Chromium.

What we need is real action, and this is the role Mozilla has always presented itself as serving. A "true" disinterested defender of the ideals of the web. Now is the time to prove it. Simply opposing this proposal isn't enough. This is about as clear and basic an attack on what fundamentally differentiates the web from every walled garden as possible. If someone drafted a proposal to the W3C that stated that only existing browsers should be allowed to render web pages, the correct response would not be to "take the stance that you oppose that proposal," it would be to seriously question whether the submitting party should even participate in the group. Make no mistake, that is what is happening now.

◧◩
2. tzs+Yn[view] [source] 2023-07-26 13:57:20
>>tolmas+We
> If Google can take unilateral action to fundamentally change the basic principles of the web, then the W3C is already useless. This will give Google a clear choice: if they want to maintain the idea that the W3C matters, they should withdraw this implementation.

It's pretty generally accepted that the correct way to do web standardization is for proponents of some new thing to implement that thing and deploy it and then once it has been shown to actually work bring a spec to the the standards folks for standardization.

That usually works fairly well, although sometimes if that first pre-standard implementation does too well the original implementor may have trouble replacing theirs with something that follows whatever standard is eventually approved, because there are often significant changes made during the standardization process.

An example of that would be CSS grid layout. That was a Microsoft addition to IE 10, behind a vendor prefix of -ms-. Nearly everyone else liked it and it was standardized but with enough differences from Microsoft's original that you couldn't just remove the -ms- prefixes from your CSS and have it work great with the now standard CSS grid.

It was 4.5 years between the time Microsoft first deployed it in IE 10 and it appearing in other browsers by default (Chrome had it within a year of Microsoft, and Firefox had it about two years after that, but both as an experimental feature the user had the specifically enable). In that 4.5 years enough sites that only cared about IE were using the -ms- form that Microsoft ended up stuck with that on IE 10 and 11 instead of the standard.

◧◩◪
3. deelow+nn1[view] [source] 2023-07-26 17:36:03
>>tzs+Yn
Basically, the ask for forgiveness approach. It's common in large, dysfunctional organizations as well. Sometimes the easiest way to get attention is to break things. Then once enough pain is felt, everyone starts taking interest. Trying to follow a proper change control based process only works when everyone is invested in the process.
◧◩◪◨
4. derefr+iV1[view] [source] 2023-07-26 19:40:35
>>deelow+nn1
No, this is just "doing your own thing in a way that doesn't affect anyone else, and allows you to gather data to cite when designing the standard."
◧◩◪◨⬒
5. troupo+Zq2[view] [source] 2023-07-26 21:54:37
>>derefr+iV1
In the browsers anything that's not behind a flag is immediately relied on by people.

So no, you shouldn't ask for forgiveness and pretend that you're just gathering data.

That's why what Google is routinely doing now (releasing APIs after a very short period in origin trial and without ever reaching consensus) is so dangerous.

[go to top]