Then it just comes down to how sophisticated the test for overlapping content is. In this case, it sounds like it wasn't very sophisticated -- but it was probably still good enough to be an effective optimization on lots of sites. Worth improving, definitely. (3/?)
-
-
But I don't think failing to make a particular premature optimization (i.e., optimization without yet seeing the need for it) implies that somebody's code is bad. Graphics on the web has tons of performance cliffs; lots of folks teach these days about how to avoid Chrome's. 4/4
1 reply 0 retweets 3 likes -
Replying to @davidbaron @WithinRafael
Yes, but the accusation was that this could have been caused by nothing except malice, which seems ridiculous. It seems like a very straightforward case of a slow path being triggered accidentally, I can think of a million innocent reasons I might want a placeholder div.
3 replies 0 retweets 0 likes -
it's not the triggering of the performance cliff that's the problem here. That's fine, that happens all the time, that's because people aren't testing in Edge. The problem is not fixing it after it was reported.
2 replies 0 retweets 2 likes -
I think I disagree, the party with the bug should fix it - in this case Edge failed to handle valid and reasonable markup correctly, but rather than fix it asked for other people to workaround the bug. The website *did* change to accommodate them, and you still see a problem.
1 reply 0 retweets 0 likes -
Edge handled the markup correctly, they just didn't optimizer it well. Here's a browser engineer saying why edge's optimization makes sense https://news.ycombinator.com/item?id=18703568 …
1 reply 0 retweets 1 like -
Replying to @ManishEarth @taviso and
The original post seems to indicate that YouTube ignored them and gave no feedback. If this is false then it's fine. Of course it was eventually fixed, Google properties get redesigned every few months.
1 reply 0 retweets 1 like -
Replying to @ManishEarth @taviso and
Ultimately, browsers are chock full of optimizations like this. When you start doing weird enough things you hit performance cliffs. Web devs are aware of these and notice these when testing.
1 reply 0 retweets 2 likes -
Replying to @ManishEarth @taviso and
I guarantee you a website as complex as YouTube would have hit and fixed some cliffs in Chrome too. By changing the website, not changing Chrome. Changing browsers to fix this stuff is often hard because the more complex the heuristics the less useful the optimization.
1 reply 0 retweets 2 likes -
Replying to @ManishEarth @taviso and
If YouTube had agreed to fix it, or even disagreed with a reason ("we need this div for important things", which let's edge work with them on heuristics) it would have made sense. They didn't, they declined to fix it with zero feedback. The rest was incompetence, this is malice.
1 reply 0 retweets 1 like
I'm sorry you find explanations of malice and incompetence so easy to accept. I hope your colleagues are more understanding when your actions cause unintended results than you are of theirs. Let's just leave it at that.
-
-
Okay, I guess I'm being harsh on the malice thing. I can see some valid reasons for this to crop up, I got a bit angry is all. Sorry.
1 reply 0 retweets 0 likes -
Replying to @ManishEarth @taviso and
However, you're incorrect in calling it Edge's bug. This is very much Youtube's bug, and they made a mistake in not fixing it. Whether or not that mistake was caused by malice is unimportant to me, it has the same effect.
1 reply 0 retweets 1 like - 3 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.