5 ms·

At this point I have enough confidence in Google's developers to consider bad performance in Firefox not an accident.

by iotku 5y ago

watermelon0 5y ago
It's much more believable that they are developing with Chromium in mind, and testing primarily on it. I've seen many webapps that are completely unrelated to Google, that are more performant on Chromium browsers, just because most developers are using them for development. It's also easier to justify spending time fixing bugs/improving performance if it affects the majority of users, instead of spending effort for 4% of users.
- dralley 5y ago
  Google has, in the past, made YouTube reliant on non-standard browser APIs and shipped a polyfill for other browsers that had terrible performance. It's not just that they passively focus on Chromium, they will deliberately trash performance on every other browser without a second thought.
  - diegocg 5y ago
    Google didn't have much of a choice. It's either a polyfill or not supporting Firefox at all. Or having two different frontends, or stopping feature development to make Firefox developers happy. Which are terrible choices. And going beyond of what standards offer is pretty much how web evolution had always happened.
    - dralley 5y ago
      What? Of course they had a choice. Google built their redesign on top of a prototype set of APIs that didn't end up being standardized. That's their own fault, and it should be their problem. It's not just Firefox devs, it's everyone else apart from Chrome, because Chrome had gone and implemented said APIs that were never standardized, so they didn't need a polyfill. And I would bet money that ShadowDOM v0 would have been removed from the Chrome codebase earlier if not for the fact that Youtube was using it - they wouldn't have forced themselves to use the same polyfill as everyone else. Shipping a new, standards incompliant frontend was a choice, not an immutable fact of nature. There was nothing wrong with the previous frontend apart from the fact that nobody is going to get promoted for not shipping the new thing. And since the consequences of shipping impact everyone except for Google, who cares, let's ship the garbage UI anyway. Don't make excuses for this kind of behavior. Google absolutely has the resources to do these things properly but they didn't.
      - md8z 5y ago
        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.
        wisniewskit 5y ago
        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.
  - floatboth 5y ago
    It didn't actually have bad performance. I never had performance problems on YouTube, and I even made my own apps with Polymer / Web Components v0 (before v1 became a thing) purely in Firefox (never even tested in Chromium), it all worked pretty well.
- Symbiote 5y ago
  None of that is an accident, at a company with Google's experience and expertise with web browsers and apps.
Drdrdrq 5y ago
It would be interesting to profile these sites and make PRs for Firefox to fix any possible issues, then see how long it takes for Google to find new ways to make them slow again.
the8472 5y ago
https://twitter.com/johnath/status/1116871231792455686 https://twitter.com/johnath/status/1116871231792455686

At this point I have enough confidence in Google's developers to consider bad performance in Firefox not an accident.

by iotku 5y ago

At this point I have enough confidence in Google's developers to consider bad performance in Firefox not an accident.

It's much more believable that they are developing with Chromium in mind, and testing primarily on it. I've seen many webapps that are completely unrelated to Google, that are more performant on Chromium browsers, just because most developers are using them for development. It's also easier to justify spending time fixing bugs/improving performance if it affects the majority of users, instead of spending effort for 4% of users.

dralley 5y ago
Google has, in the past, made YouTube reliant on non-standard browser APIs and shipped a polyfill for other browsers that had terrible performance. It's not just that they passively focus on Chromium, they will deliberately trash performance on every other browser without a second thought.
- diegocg 5y ago
  Google didn't have much of a choice. It's either a polyfill or not supporting Firefox at all. Or having two different frontends, or stopping feature development to make Firefox developers happy. Which are terrible choices. And going beyond of what standards offer is pretty much how web evolution had always happened.
  - dralley 5y ago
    What? Of course they had a choice. Google built their redesign on top of a prototype set of APIs that didn't end up being standardized. That's their own fault, and it should be their problem. It's not just Firefox devs, it's everyone else apart from Chrome, because Chrome had gone and implemented said APIs that were never standardized, so they didn't need a polyfill. And I would bet money that ShadowDOM v0 would have been removed from the Chrome codebase earlier if not for the fact that Youtube was using it - they wouldn't have forced themselves to use the same polyfill as everyone else. Shipping a new, standards incompliant frontend was a choice, not an immutable fact of nature. There was nothing wrong with the previous frontend apart from the fact that nobody is going to get promoted for not shipping the new thing. And since the consequences of shipping impact everyone except for Google, who cares, let's ship the garbage UI anyway. Don't make excuses for this kind of behavior. Google absolutely has the resources to do these things properly but they didn't.
    - md8z 5y ago
      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.
      - wisniewskit 5y ago
        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.
- floatboth 5y ago
  It didn't actually have bad performance. I never had performance problems on YouTube, and I even made my own apps with Polymer / Web Components v0 (before v1 became a thing) purely in Firefox (never even tested in Chromium), it all worked pretty well.
Symbiote 5y ago
None of that is an accident, at a company with Google's experience and expertise with web browsers and apps.

Drdrdrq 5y ago

It would be interesting to profile these sites and make PRs for Firefox to fix any possible issues, then see how long it takes for Google to find new ways to make them slow again.

the8472 5y ago

https://twitter.com/johnath/status/1116871231792455686 https://twitter.com/johnath/status/1116871231792455686