I'm sorry but this comment rings very hollow to me. I have seen this "hypothetically they could have done more" type of sentiment repeated so much on HN and it's not helpful nor is it a meaningful criticism of Google. Every browser has some non-standard features. That isn't new, Google isn't the first one to do it, they certainly won't be the last. Do they have a choice to not do that? Sure, but nobody chooses to do it because it makes it harder to actually iterate on things. The "behavior" is widespread and every vendor is already making excuses for it.
There is a real problem here and it's absolutely related to the fact that Google is only incentivized to develop/test on their own browser, but that's really orthogonal to other browsers being slow or not being able to improve performance of a polyfill.
Oh come on. Google is never going to have an incentive to do better if we just excuse them for their bad behaviour, especially in cases like this where claiming any equivalence is simply ludicrous. One must be wilfully ignorant to act as though their ambitions are "orthogonal" to other browsers purely due to their own faults.
Google did not have to publicly ship their pre-standard experimental "v0" web components implementation so early in Chrome. They chose to do so regardless of what other browser vendors expressed.
Likewise they did not have to make YouTube use them so soon, thus forcing other browsers to rely on a polyfill that could not feasibly be made remotely performant compared to just implementing web components more quickly, wink wink.
They chose to do these things the way they did. They wanted to ship it on their timeline, and to hell with what other browsers wanted to spend time on first instead. They wanted to look like they were heroes for pushing the web forward, while in reality they were just holding other vendors and APIs back to get the one they valued the most done first, no matter how much of a mess they caused in the process (the transition from v0 to v1 was hardly quick or painless). And that's just web components.
When is the last time you saw Firefox ship such a major web API in such a non-final and un-vetted state, and then use one of the largest web properties on Earth to get others to prioritize it as they wished? Or even Apple, for that matter?
It's flat out ridiculous to try equating the vendors in this manner. They don't have the market dominance or even the same force of apologists burying the lede on their bad behaviour.
I'm sure the other browsers also have their own Project Fugus underway too, where they're just shipping a slurry of new APIs regardless of whether anyone else will ever implement them? Or is that the others' fault somehow too, because they should also be trying to fragment the web as much as possible as quickly as possible?
If we collectively just want Chromium to be the only engine because we value rapidly iterating on new APIs more than anything, then let's at least be honest about it.
Look, I hear what you're saying but it all just sounds like hypotheticals to me. If you want to take that approach, hypothetically the other vendors could have chose to standardize the feature, it could have become standard, and all the other browsers could have implemented it and it wouldn't be a problem. But that didn't happen. And I have seen plenty of other features that were gated behind Moz or Webkit prefixes.
I'm not trying to be a downer here. Realistically, there will always be a browser out in front that is going to iterate on features faster than the others. That's normal as long as you have more than one browser. If we want to criticize Google for practicing anti-competitive behavior then let's do that, but it just really doesn't make sense to me to put "they shipped a feature that someone else didn't" in that category.
It's not that you're being a downer, you're just not presenting things in a fair or accurate manner.
Google didn't merely "ship a feature someone else didn't". They keep shipping new features as they wish, whether others even agree. They have in fact accelerated that attitude with Project Fugu. Some of them are quite complex or consequential APIs.
They do not deserve a free pass for it because Mozilla once shipped one or two relative minor features before Chrome did, or Apple added some weird CSS visual property for their latest iPhones without consensus. We're talking orders of magnitude of difference here.
The problem isn't who is first to ship. It's the casual disregard for even reaching consensus on the basics before shipping something, the sheer rate of output, the interop issues left in the wake, and the anti-competitive tactics being applied. Those are not "hypotheticals" in the slightest.
What is hypothetical is acting like anyone else could just magically compete on those terms. Microsoft couldn't keep up with them. Opera couldn't. No new engine has even come close to breaking into a general market yet, though a couple are feverishly trying.
Is that really ok with us? If so, then let's just be honest. Let's just say "new APIs are more important to us than engine diversity, and we don't mind Google being as evil as possible to kill the other engines off." As a webcompat worker, I'd love to see that honesty.
Do you think it's abusive of Linux to offer APIs beyond what is standardized in POSIX, breaking comparability with other nix-like OS such as FreeBSD? Or is it abusive that clang constantly adds features that gcc lacks, making programs that use clang no longer compile with gcc?
average disingenuous HNer making comparisons between compilers and web browsers. Everyone involved knows Shadow DOM V0 was a rush job, as was the YouTube redesign that used it (it had major perf issues even in chrome when it came out). The standardized ShadowDOM v1 is better in every way and works in all browsers. It's pretty clear that Google wanted V0 to spread as far as possible so they could force it to become a standard, as removing it would "break the web". Shadow DOM, regardless of version, isn't critical for a product like YouTube. The "web" is only the "web" if parties involved play fair, even just a bit, otherwise it's back to IE6.
Does Linux also happen to control a major piece of hardware pretty much everyone uses, and used it to force BSD to adopt some new driver system they wanted to prioritize, which ended up being completely different by the time the dust had settled, and held back other important advancements in the meantime?
That is, it all depends on the full context. Competition is fine, but not anti-competitive behaviour. I maintain that we have had too much of the latter from Google, and that it is only increasing as folks intentionally look the other way and try to boil arguments away to mere deflections and other apologia.
There is a real problem here and it's absolutely related to the fact that Google is only incentivized to develop/test on their own browser, but that's really orthogonal to other browsers being slow or not being able to improve performance of a polyfill.