6 ms·

Quantum I think was too incremental and conservative, trying to integrate Servo into Gecko. They’ve ended up making a desktop browser which is as good as Chrom

Quantum I think was too incremental and conservative, trying to integrate Servo into Gecko. They’ve ended up making a desktop browser which is as good as Chrome, but not good enough to demand attention. Mobile, where parallelism should be most beneficial, has lagged behind, and parallel layout, which is where the real prize is, has disappeared over the horizon. Servo has been relegated to WebVR experiments which are speculative at best, at a point when VR is only available to a truly tiny percentage of the population, and should it be successful will probably be dictated by market share anyway. Do Quantum by all means, but follow five other paths to make Servo a success. Servo was designed to be embeddable, why for instance not attempt competition for Electron. Even if it only implemented a subset of the spec it could be a big success. This is Mozilla’s lifeblood, and it just seems attention is elsewhere.

kodablah 8y ago
> Servo was designed to be embeddable, why for instance not attempt competition for Electron. It should be noted there have been side projects to back nodejs with spidermonkey [0] and have electron APIs backed with gecko [1]. I agree with you and the thing I want more than anything is a cross platform browser engine embeddable with a supported C API that's not Chromium. Servo was on its way, but work has definitely slowed. But I acknowledge that even though I would build my own browser UI on top of an embeddable gecko engine, it probably won't affect adoption that much to be worth the effort. 0 - https://github.com/mozilla/spidernode https://github.com/mozilla/spidernode 1 - https://github.com/mozilla/positron https://github.com/mozilla/positron
pcwalton 8y ago
If you've been following #servo lately, you'll see that we're starting a reboot of parallel layout right now (which I'm leading, more or less). It just so happens that I'm also juggling a million other things simultaneously. For example, I'm working on power usage improvements on macOS that should help address some of the complaints upthread. Also there's the entirely new GPU font and SVG renderer (which is a novel technique). Also there's making sure WebRender ships. Etc.
- tjr225 8y ago
  Nice! Looking forward to better power management - I'm a daily user in spite of this.
- Eridrus 8y ago
  Is any of this work coming to Mobile? I agree with the GP that this is where Firefox has the biggest opening, but it seems like there's very little investment in Firefox on Android.
  - pcwalton 8y ago
    I'm not responsible for or in control of product plans, but we are testing and tuning much of the work for mobile, yes.
- 43920 8y ago
  I know it's probably really early to ask this, but do you have a rough estimate of when the power usage improvements will happen? Battery life is the main thing stopping me from using Firefox more, so I'm really looking forward to this being fixed.
  - agumonkey 8y ago
    I had a ~power related issue using Quantum. It "killed" my cpu (temp rising by 10s of degrees so fast I didn't have time to throttle before kernel thermal shutdown). I guess it's so parallel and so gpu intensive that my whole system goes into fever. ps: a x201 core i5 520m
    - mort96 8y ago
      What? Your system doesn't handle running at max CPU and GPU usahe even for what I presume is seconds before thermal shutdown? That seems like an extremely serious issue with your system.
      - agumonkey 8y ago
Sylos 8y ago
Servo is nowhere near ready. It's been a few months since I've seen actual results, but I'd be surprised, if it nowadays completed ACID3 (which was state of the art for browsers in 2009). The thing is that rewriting a browser engine doesn't just happen over night. No bigger browser engine has been written from scratch since the previous millennium. Also, no browser engine implements all currently specified webstandards. So, Mozilla would have to full-pelt develop Gecko to be able to keep up with Blink in webstandards, and then also develop Servo at more than full-pelt to be able to catch up in finite time. That just doesn't add up, which is why Servo was specifically started as a research project. It might never be completed. They have been able to integrate various components from Servo directly into Gecko, therefore sharing the development work which they didn't really plan with, so that's one tiny reason to still hope for it to ever catch up, but yeah, just don't be too optimistic. Having said all that, the integration of components from Servo into Gecko isn't done yet. WebRender is still missing, which should bring another good performance boost. It's sort of been overhyped with artificially complicated CSS animations [0], so the real world effect isn't that big, but it's still very much noticeable. [0]: https://www.youtube.com/watch?v=u0hYIRQRiws https://www.youtube.com/watch?v=u0hYIRQRiws Aside from that, on Android, Mozilla is building a new framework for building Android browsers in general, called "android-components", which also includes a much cleaner integration of Gecko on Android, called "GeckoView". They've already rebased Firefox Focus onto GeckoView and are actively working on rebasing Firefox for Android (Fennec) onto it, too, which is internally called "Fenix". This should also significantly reduce the lagginess and speed up their development in the long-run. I guess, what I'm mainly trying to say is that there's definitely still things happening.
- dman 8y ago
  Maybe we need simpler standards? Ones that are implementable without spending a billion dollars.
  - richjdsmith 8y ago
    I'm sure the firefox team would agree with you wholeheartedly. Chrome is the new IE though. Making whatever specs they need for Google Sites - standards be damned.
modzu 8y ago
ff mobile supports addons. this is HUGE!!!
morpheuskafka 8y ago
I wish Mozilla would stop doing so much non-development stuff. Now that Thunderbird is moving on its own, FFox is their only product and yet they waste a lot of time sending out emails about various petitions and similar. They should definitely support pro-Internet petitions as a company (as does Google and others), but IDK why they are now running essentially a mailing list infrastructure for them.
- dralley 8y ago
  The Mozilla Foundation and the Mozilla Corporation are separate entities.
  - Karunamon 8y ago
    One funds the other, does it not?
- michaelmrose 8y ago
  Do you mean is you disagree with what they do politically and so you are going to sit here and pretend you wish they didn't waste resources on non technical issues. This isn't terribly charitable.
  - krageon 8y ago
    Even though I broadly agree with them politically, I still think companies should be apolitical. All these politics involved in things that aren't inherently political are tiring.
asituop 8y ago
I really can't agree more with you. WebVR is not a good investment, when you are not even on par in terms of performance for displaying a web page, the last thing you look into is WebVR. This is one of the main problem at Mozilla, they often get influenced by the trend or they spread funds over many different projects ( that range from VR to Museum Arts, grants for LGBT, harassment prevention, Bioinformatics, etc. those are very noble/interesting causes too but you cannot save the whole world at once ). But I don't know maybe this is good for marketing and it allows them to get even more donation / sponsoring? But I'm really not convinced... I think what can really save Mozilla is WebRender/Servo. The best thing that happened to FF is surely not Pocket (once again, huge investment and most people don't use it) but Quantum and this comes from the Servo project. As you said, they could also make a killer electron alternative and it would be very fitted to Servo as you can start with a subset of the spec and focus on performance/footprint, but their lastest attempt is from what I know this : https://github.com/paulrouget/servoshell/blob/master/README.md https://github.com/paulrouget/servoshell/blob/master/README.... which is almost 1 year old
- sfink 8y ago
  It's understandable, but don't mix the Mozilla Foundation with Mozilla Corporation. The former wholly owns the latter, but they do rather different flavors of work/outreach and they're not really competing over the same one of resources.
  - pas 8y ago
    > they're not really competing over the same one of resources. They compete for the same money, no?
pas 8y ago
Firefox and Gecko has a lot of tech debt, and it's big. It'd be too risky to just try to throw big chunks of it out and hope that the replacement will be great. Though the momentum seems lost, the WebRender integration is still ongoing. ( https://mozillagfx.wordpress.com/2018/11/29/webrender-newsletter-32/ https://mozillagfx.wordpress.com/2018/11/29/webrender-newsle... ) And it's big, it's complex, it's hairy, the web is a mess. So, it's slow and full of terrors. (Shaders, caching, multi-threading, mobile, multi-core, C++ and Rust interfacing, etc.) The next thing seems to be Fission, which is the continuation of the e10s (electrolysis) content process model into full/more Site Isolation. ( https://wiki.mozilla.org/Project_Fission https://wiki.mozilla.org/Project_Fission - but it seems hard as fuck: https://mail.mozilla.org/pipermail/firefox-dev/2018-July/006628.html https://mail.mozilla.org/pipermail/firefox-dev/2018-July/006... nowadays they seem to be stripping out the old Gecko XUL stuff and replacing it with something newer and better: https://mail.mozilla.org/pipermail/firefox-dev/2018-November/006903.html https://mail.mozilla.org/pipermail/firefox-dev/2018-November... )

kodablah 8y ago

> Servo was designed to be embeddable, why for instance not attempt competition for Electron. It should be noted there have been side projects to back nodejs with spidermonkey [0] and have electron APIs backed with gecko [1]. I agree with you and the thing I want more than anything is a cross platform browser engine embeddable with a supported C API that's not Chromium. Servo was on its way, but work has definitely slowed. But I acknowledge that even though I would build my own browser UI on top of an embeddable gecko engine, it probably won't affect adoption that much to be worth the effort. 0 - https://github.com/mozilla/spidernode https://github.com/mozilla/spidernode 1 - https://github.com/mozilla/positron https://github.com/mozilla/positron

pcwalton 8y ago

If you've been following #servo lately, you'll see that we're starting a reboot of parallel layout right now (which I'm leading, more or less). It just so happens that I'm also juggling a million other things simultaneously. For example, I'm working on power usage improvements on macOS that should help address some of the complaints upthread. Also there's the entirely new GPU font and SVG renderer (which is a novel technique). Also there's making sure WebRender ships. Etc.

tjr225 8y ago
Nice! Looking forward to better power management - I'm a daily user in spite of this.
Eridrus 8y ago
Is any of this work coming to Mobile? I agree with the GP that this is where Firefox has the biggest opening, but it seems like there's very little investment in Firefox on Android.
- pcwalton 8y ago
  I'm not responsible for or in control of product plans, but we are testing and tuning much of the work for mobile, yes.
43920 8y ago
I know it's probably really early to ask this, but do you have a rough estimate of when the power usage improvements will happen? Battery life is the main thing stopping me from using Firefox more, so I'm really looking forward to this being fixed.
- agumonkey 8y ago
  I had a ~power related issue using Quantum. It "killed" my cpu (temp rising by 10s of degrees so fast I didn't have time to throttle before kernel thermal shutdown). I guess it's so parallel and so gpu intensive that my whole system goes into fever. ps: a x201 core i5 520m
  - mort96 8y ago
    What? Your system doesn't handle running at max CPU and GPU usahe even for what I presume is seconds before thermal shutdown? That seems like an extremely serious issue with your system.
    - agumonkey 8y ago

Sylos 8y ago

Servo is nowhere near ready. It's been a few months since I've seen actual results, but I'd be surprised, if it nowadays completed ACID3 (which was state of the art for browsers in 2009). The thing is that rewriting a browser engine doesn't just happen over night. No bigger browser engine has been written from scratch since the previous millennium. Also, no browser engine implements all currently specified webstandards. So, Mozilla would have to full-pelt develop Gecko to be able to keep up with Blink in webstandards, and then also develop Servo at more than full-pelt to be able to catch up in finite time. That just doesn't add up, which is why Servo was specifically started as a research project. It might never be completed. They have been able to integrate various components from Servo directly into Gecko, therefore sharing the development work which they didn't really plan with, so that's one tiny reason to still hope for it to ever catch up, but yeah, just don't be too optimistic. Having said all that, the integration of components from Servo into Gecko isn't done yet. WebRender is still missing, which should bring another good performance boost. It's sort of been overhyped with artificially complicated CSS animations [0], so the real world effect isn't that big, but it's still very much noticeable. [0]: https://www.youtube.com/watch?v=u0hYIRQRiws https://www.youtube.com/watch?v=u0hYIRQRiws Aside from that, on Android, Mozilla is building a new framework for building Android browsers in general, called "android-components", which also includes a much cleaner integration of Gecko on Android, called "GeckoView". They've already rebased Firefox Focus onto GeckoView and are actively working on rebasing Firefox for Android (Fennec) onto it, too, which is internally called "Fenix". This should also significantly reduce the lagginess and speed up their development in the long-run. I guess, what I'm mainly trying to say is that there's definitely still things happening.

dman 8y ago
Maybe we need simpler standards? Ones that are implementable without spending a billion dollars.
- richjdsmith 8y ago
  I'm sure the firefox team would agree with you wholeheartedly. Chrome is the new IE though. Making whatever specs they need for Google Sites - standards be damned.

modzu 8y ago

ff mobile supports addons. this is HUGE!!!

morpheuskafka 8y ago

I wish Mozilla would stop doing so much non-development stuff. Now that Thunderbird is moving on its own, FFox is their only product and yet they waste a lot of time sending out emails about various petitions and similar. They should definitely support pro-Internet petitions as a company (as does Google and others), but IDK why they are now running essentially a mailing list infrastructure for them.

dralley 8y ago
The Mozilla Foundation and the Mozilla Corporation are separate entities.
- Karunamon 8y ago
  One funds the other, does it not?
michaelmrose 8y ago
Do you mean is you disagree with what they do politically and so you are going to sit here and pretend you wish they didn't waste resources on non technical issues. This isn't terribly charitable.
- krageon 8y ago
  Even though I broadly agree with them politically, I still think companies should be apolitical. All these politics involved in things that aren't inherently political are tiring.

asituop 8y ago

I really can't agree more with you. WebVR is not a good investment, when you are not even on par in terms of performance for displaying a web page, the last thing you look into is WebVR. This is one of the main problem at Mozilla, they often get influenced by the trend or they spread funds over many different projects ( that range from VR to Museum Arts, grants for LGBT, harassment prevention, Bioinformatics, etc. those are very noble/interesting causes too but you cannot save the whole world at once ). But I don't know maybe this is good for marketing and it allows them to get even more donation / sponsoring? But I'm really not convinced... I think what can really save Mozilla is WebRender/Servo. The best thing that happened to FF is surely not Pocket (once again, huge investment and most people don't use it) but Quantum and this comes from the Servo project. As you said, they could also make a killer electron alternative and it would be very fitted to Servo as you can start with a subset of the spec and focus on performance/footprint, but their lastest attempt is from what I know this : https://github.com/paulrouget/servoshell/blob/master/README.md https://github.com/paulrouget/servoshell/blob/master/README.... which is almost 1 year old

sfink 8y ago
It's understandable, but don't mix the Mozilla Foundation with Mozilla Corporation. The former wholly owns the latter, but they do rather different flavors of work/outreach and they're not really competing over the same one of resources.
- pas 8y ago
  > they're not really competing over the same one of resources. They compete for the same money, no?

pas 8y ago

Firefox and Gecko has a lot of tech debt, and it's big. It'd be too risky to just try to throw big chunks of it out and hope that the replacement will be great. Though the momentum seems lost, the WebRender integration is still ongoing. ( https://mozillagfx.wordpress.com/2018/11/29/webrender-newsletter-32/ https://mozillagfx.wordpress.com/2018/11/29/webrender-newsle... ) And it's big, it's complex, it's hairy, the web is a mess. So, it's slow and full of terrors. (Shaders, caching, multi-threading, mobile, multi-core, C++ and Rust interfacing, etc.) The next thing seems to be Fission, which is the continuation of the e10s (electrolysis) content process model into full/more Site Isolation. ( https://wiki.mozilla.org/Project_Fission https://wiki.mozilla.org/Project_Fission - but it seems hard as fuck: https://mail.mozilla.org/pipermail/firefox-dev/2018-July/006628.html https://mail.mozilla.org/pipermail/firefox-dev/2018-July/006... nowadays they seem to be stripping out the old Gecko XUL stuff and replacing it with something newer and better: https://mail.mozilla.org/pipermail/firefox-dev/2018-November/006903.html https://mail.mozilla.org/pipermail/firefox-dev/2018-November... )