[thomasfortes]

Both are the new IE, one push features without caring about the rest of the ecosystem and the other refuses to implement standards without caring about the rest of the ecosystem.

The end result is that the web right now has stuff that works only on Chrome and stuff that works everywhere besides Safari.

And the fact that iOS users can't change their browser forces developers that want their projects to reach the maximum amount of users to have Safari as a baseline instead of the current standard, making Safari a de facto standard.

So yeah, I think that both are the new IE, just in different stages of the life of IE.

[tarsinge]

> the other refuses to implement standards without caring about the rest of the ecosystem

There is standard, and standard as previously Chrome only feature that Firefox felt pressured to implement and was then a posteriori made into a standard.

[mannerheim]

Pretty much every browser except Safari supports WebGL2

[thamer]

The WebKit commit is over 8 months old already: "Enable WebGL2 by default", Sep 14, 2020 – from https://trac.webkit.org/changeset/267027/webkit

CanIUse reports that it's only available in Safari Technology Preview for now, which is a macOS program: https://www.caniuse.com/webgl2

With WWDC 2021 only ~9 days away, one can hope that they'll announce its support on all platforms in the next major release. If it was already "good enough" last September…

[mannerheim]

I was aware it was in Safari Technology Preview, but I wouldn't really count a feature in a beta release as being available. And who knows when it will be available on iOS.

[jpttsn]

Honest question: Do we have any scroll vs. marquee type situations today?

Because (Unpopular): I believe the standard should primarily cover how the overlapping functionality works, and refrain from limiting or prescribing the extent of functionality.

Comparing: If I build a HTTP API, I don’t have to support the DELETE verb for any endpoints. I can support ENCHANT if I want magic that other servers don’t have. But if I use GET, the endpoint handler should be idempotent. That’s the kind of standard I appreciate.

I don’t see any realistic win-win otherwise. Either you hold Chrome back from implementing new crap, or you force Safari to implement stuff they don’t want to. The efficient number of browser vendors seems to be small, so I think the standard body has just overplayed it’s hand.

[jefftk]

> Because (Unpopular): I believe the standard should primarily cover how the overlapping functionality works, and refrain from limiting or prescribing the extent of functionality ... the standard body has just overplayed it’s hand.

But that is how web standards work already? Vendors are not prohibited from adding additional functionality.

[jpttsn]

A lot of comments say Safari is not implementing web standards, by virtue of not implementing e.g. push notifications.

[jefftk]

Those comments are correct: there are many web standards that Safari has either decided not to implement, or has not gotten around to yet. A browser doesn't have to implement a new feature just because it has been standardized.