HNHacker News
TopNewBestAskShowJobs

zbraniecki

1,111 karma · joined October 18, 2013

Zibi Braniecki, engineer at Mozilla. Working on l10n/i18n stack, ECMA402 and performance.
submissionscomments
zbraniecki··on How you can contribute to Firefox 57 success
> That's a lot of ifs... And again these are very technical aspects, not strategic/product ones.

Correct. Lots of ifs. And on top of that, ifs from a regular engineer which means I know even less about all the factors in play than I should to predict.

> Mozilla has been putting 110% of their effort on a Desktop browser recently. Fine, their choice, but I can't help but think that their are cornering themselves in a niche and are missing where the new tech users will be.

My personal take is that we have to start somewhere. We can try to come up with a completely new technology, and we're trying to - see Servo. We can make Gecko the best web engine on the planet. Or we can try to build amazing products on top of an ok engine.

I believe we have to start somewhere, and stating by bringing Gecko to excellence is not a bad choice.

> As you can guess, I really feel it's a shame that gecko is becoming really good, just not used in a context that matters.

yeah. I know :)

zbraniecki··on How you can contribute to Firefox 57 success
If you don't bother to understand the factors, then yes, it's very easy :) Unfortunately, making opinions without learning about the factors in play is called "ignorance" and usually is frowned upon.
zbraniecki··on How you can contribute to Firefox 57 success
I do not have numbers for how many users use extensions, but my gut feeling is that it's much lower than it was in 2011. We've been moving to e10s over last months enabling it for users who don't have legacy addons (so, they either don't have addons or they have e10s modern addons).

I found a status update from the e10s meeting from Jun 2nd - https://wiki.mozilla.org/E10s/Status/June2

It indicates that over 60% of beta users had e10s activated which can only be activated for users who don't use old extensions. Based on the telemetry data it seems that only 3% were not eligible. I'm not involved in e10s project so I don't know the details and updates, but I believe that the number of users affected is actually quite low (which doesn't mean irrelevant, just that majority of users won't be affected).

> As for the opt outs, this may be true for most things, but I am not able to opt entirely out of Pocket's "Recommended by Pocket" trending adverts

In the home tab? The icon on the top-right corner allows you to uncheck the "Recommended by Pocket" and "extensions.pocket.enabled" about:config option allows you turn off pocket completely.

Now, that we own Pocket, I hope we'll merge it more seamlessly into Firefox Accounts because I feel like the sole fact that it uses it's own brand name makes it feel like Firefox comes with a separate piece of software bundled.

zbraniecki··on How you can contribute to Firefox 57 success
I can't even express how awesome it feels to hear it. Particularly, because the "duplicate tab context menu" landed just yesterday and I happened to write the final patch [0] :)

[0] https://bugzilla.mozilla.org/show_bug.cgi?id=455722

zbraniecki··on How you can contribute to Firefox 57 success
I understand that this is how you see our actions.

Working on Mozilla, I do not agree with you on multiple points:

> If attracting contributors was such a priority for mozilla, you would think that so many screw-the-users-we-are-doing-this-anyways decisions wouldnt come to pass.

It's way more complex than you present it. Working on a big project, with a solid legacy, will make you face dramatic decisions where each outcome is going to cost. For example, most users, due to their vested interest in one outcome, and lack of understanding of the cost of the other (they're not going to spend countless hours maintaining legacy APIs), would say "Keep legacy extension system forever". But as a module owner of Gecko, you understand the tremendous cost of doing it. The decision you're making is way more complex and multi-factored that the decision those users made when they voiced their opinion.

Trying to say that "if you care about being open source project, do what users tell you to" is imho only possible if you don't understand the complexity.

See, for example, https://en.wikipedia.org/wiki/The_Innovator%27s_Dilemma for details.

> its best to not to ignore the feedback users give (note telemetry metrics arnt feedback, especially when interpreted with an agenda, stuff dosent always measure what you think it does).

I'm not sure if I understand what you're saying here. Telemetry tells us what happens. Feedback tells us what users want to happen. Telemetry done right gives us representation of the population, feedback gives us the representation of the sample that voiced their opinion.

Loud minority can sway the discussion in any way (WebVR on Linux is the most important thing for Mozilla to work on!). Telemetry gives a different picture (3% of our users are on Linux, and 0.01% of them uses WebVR at all).

You can also see it in this thread. I wrote a blog post about ways to contribute to an open source project. You'd guess the majority of the conversation should be about technicalities involved in contributing discussed among those who want to contribute but have questions. Instead, majority is about whether Mozilla is doing what the given person wants it to do on a very high meta-level (I want extensions to work this way, or I want data to be collected that way).

There's a substantial difference between those two - one is about a case where an engineer working on Firefox UI filed a bug about switching the way we react to clicks on a button but had no time to fix it. Maybe it also affects you? Maybe you want to go and write a couple lines of JS to fix it? Cool! Let's talk. The latter is "I have an opinion on privacy policies at Mozilla, and I voiced it together with a number of other individuals who, when asked, say that there should be always less tracking. We brought our pitchforks and loudly demanded it. Now it seems that Mozilla is, after holding a public discussion for months, adding the telemetry which is not in line with what we told them to do so it means they ignored us! Outrageous!"

I'm not saying that we've always done the right thing (I believe we made very high level mistakes on multiple projects, but I also believe I don't have the full knowledge that went behind those decisions, so tt's easy for me to say). I'm saying that this post is for people who may want to contribute to Mozilla but didn't know how. Sort of "www.whatcanidoformozilla.com" or "#contribute-newbies" style. And "the feedback" is about high level complex decisions that Mozilla is doing that are not in line with the persons opinion. Much more of a "mozilla.governance" or "mozilla.upper-management.decisions" style.

It's easy to have high-level opinions if you don't aim to understand the full complexity of the topic at hand. I can right now tell a government it should stop a war. Any war. Always. I can tell the government that I'd prefer the taxes to be lower. Or higher. I can go to Google and say that I'd prefer it they made Android an open source project again.

That's very easy, because I'm not bored with all the details, I'm ignorant when it comes to understanding all the factors, and I just want the outcome. But saying that those entities "don't listen", just because they won't do what I say they should is imho a wrong conclusion.

> unless you agree with what they have already decided prepare to be ignored

I'm sorry you have this experience with Mozilla. I believe that your voice is heard more often than you think, and judging if your voice was heard solely on if the decision matches your opinion is wrong.

> It seems to be the general conclusion inside mozilla is that Chrome is winning because of x, we need more x, for any value of x

I do not believe you to be correct here. Firefox team is working on making Firefox the best Firefox possible. Which means that we're building the product we believe to be the best and that involves taking what's best in Safari, Chrome, Edge, Vivaldi etc. if that fits into our idea of a web browser.

People love to see patterns where there are none (see Apophenia) so for years I've heard that each change we make is "copying" someone else. If we switch the back button to the left "we copied Edge", if we switch to the right "we copied Chrome". If we switch tabs to curved, we copied someone, if we switch to rectangular, we copies someone else and so on. If we add something that other browsers don't do, there's generally silence, until someone starts claiming that we add unnecessary shit that bloats the browser :)

I tend to believe that the reason for that is there are a lot of people and for each decision a small subset of them that is not happy voices their opinion. And if the crowd is large enough, every opinion will be unpopular with some minority which will start vocally opposing it. So if you look for feedback, it'll be always there, and predominantly negative for every decision that is a matter of preference.

Now, there are things that aren't a matter of preference in a general sense - performance, stability, security etc. I believe that V8 gained a lot of users by doing those things well, and I don't think there's anything wrong with us aiming to match or bypass V8.

> I wish that firefox cared more about being firefox, and less about becomming a worse version of chrome.

I hope that 57 will fit that description for you :)

> [1] https://groups.google.com/forum/#!topic/mozilla.governance/8.... feedback here is virtually 100% negative, but I gaurentee this change is going in anyways.

I don't know where it'll go, but I believe that the decision which doesn't agree with feedback does not equate not hearing the feedback, or the feedback not mattering. I hope you understand that it's a complex, multi-factored problem that is not easly solvable with simple solutions.

Anyhoo, I respect your perspective, I'm sorry this is your experience. I hope my personal opinion helped you see that there are other ways to see it. At the same time, I'm just an engineer on the project, so I can't speak for the high-level decisions as much as I can talk about technical details of the platform and contributing to it. Which I wrote a blog post about. :)

zbraniecki··on How you can contribute to Firefox 57 success
Yeah, I see what you're saying.

I think there's a deeper debate that Mitchell is running over the last few years at Mozilla, and that's how to find a balance between "your voice is heard" and "someone is making a decision".

If you voiced your opinion, and the decision has been made that is not in line with it, it's very easy to feel that "my voice didn't matter. Mozilla doesn't want my opinion".

I believe that it's a fallacy of communication.

So, in total. Sure, as the project grows and priorities change, we get to a more structured way of making decisions. It's getting harder for a casual contributor to steer the decision in a completely different direction. But I do not believe that it's any different from Fedora, Wikipedia, Gnome, KDE, Wordpress or any other open source tech. When you're a 10-person github project, it's easy to say that anyone can file an issue and change the course of it. When you're a multihundred million users project with 20 years of legacy code, and thousands of contributors, it is harder to see the connection between your opinion and the result.

But I do believe that if you gain trust in the community, and use the right forums (and I realize that finding the right forum to voice your opinion is not easy), you do have an impact. Even if the final decision is not in line with it, which will happen, unless you and the decision owners are always 100% correct.

zbraniecki··on How you can contribute to Firefox 57 success
Hi Fabrice!

> At the same time I don't see the situation improving

I actually do, but as you know, I'm an optimist by nature :)

The path that could lead to Mozilla tech opening for the case you're describing may look like this:

* Fx 57 is successful * Gecko successfully transitions to be a modern Web engine with its own advantages over other engines * The new features such as Rust and leapfrogging advancements like Stylo/WebRender etc. add up to make people want to use Gecko/Servo over other engines in their products * We're successful at building a node/electron like wrapper for Gecko/Servo * The innovation begins

I don't think it's certain, and I don't know if it'll happen, but I do now that there are people tinkering with things like GeckoView and electron like wrappers, with target for next year. It's not guaranteed, but I do hope it'll happen.

zbraniecki··on How you can contribute to Firefox 57 success
Hi! I'm the author of the post.

So, it's not "asking for help", it's informing how you can help. I'm also not "begging for contributors" or asking to help with "company goals", I'm suggesting ways to help with our current major release. If we're successful, it should make the whole Web better thanks to bringing more competition and choice to the user.

Major open source projects are actually fun to work with for many people. They're very challenging not just because of problems you solve, but also because of the massive project culture and technical infrastructure in place that you learn in the process. That learning helps people grow as engineers. Also the feeling that you made an impact on the lives of millions of users is quite awesome. For others, they're already Firefox users and it's just about fixing a bug that was itching them.

So, generally speaking, there are reasons to help major open source projects, even if they can afford to hire full time engineers as well.

Two side notes:

1) We're way, way smaller than all our competitors. The number of engineers that Google, Apple and Microsoft can hire to work on their browser engines is significantly higher than what Mozilla can do. We can't compete and never will be able to. That's why, beside of our culture and Mozilla Manifesto, we rely on our ability to attract contributors. 2) It's a chicken and egg. Keeping the project really open (not technically open as in - we'll dump the source code to github every once in a while), requires continuous effort. Having a healthy community of contributors is important for our project culture to stay open. If people stop contributing, some things will naturally decay into less open state making it harder in the future for new contributors to come in. And I'm not even talking about code only. Mozilla's marketing, engineering, privacy, security, UX and other projects are working in the open.

zbraniecki··on How you can contribute to Firefox 57 success
We also need help in making it practically anonymous :)
zbraniecki··on How you can contribute to Firefox 57 success
Building SpiderMonkey is not hard at all today: https://developer.mozilla.org/en-US/docs/Mozilla/Projects/Sp...

The only issue is that you have to clone whole mozilla-central even if you only work with small part of it. But once you pull it, the building and hacking on it is quite easy :)

zbraniecki··on How you can contribute to Firefox 57 success
The plan is to support it in 58: https://github.com/prefiks/u2f4moz/issues/59#issuecomment-32...
zbraniecki··on How you can contribute to Firefox 57 success
I'm sorry you had this experience.

I'm also happy to say that we actually did improve our build chain. I have a build env for Firefox on Mac, Windows and Linux and it all works quite smoothly. The only issue is that full compilation takes time (30 minutes on my laptop).

Fortunately we also now have Artifact Builds which allow you to hack on the front-end JS code without having to rebuild the C++ backend [0].

[0] https://developer.mozilla.org/en-US/docs/Mozilla/Developer_g...

zbraniecki··on Ask HN: What's it like to work at Mozilla these days?
I've been working, in various forms, on Mozilla for 17 years now. I've joined the project as a volunteer back in 2000, and ended up working full time since 2010. I worked in multiple roles during that time, mostly as an engineer, but spent some time working on contributor relationships, contributor tech and some market analysis even. I've worked remotely for most of that time from Europe, but am working from San Francisco for almost 4 years now.

My personal experience has been very positive. I'm working full time on projects that impact the shape of the Web, I have good work-life balance and incredible number of highly challenging problems to solve.

I've been with Mozilla since they were a 40 person startup, through the crazy days of releasing Firefox 1.0, sudden growth during Firefox OS days (I spent 2 years working on it!), and now I work on Firefox 57 and beyond. It's been a great, if bumpy, ride and I can't stop enjoying it :)

I've seen bad culture fits, I've seen clusters of bad management and projects that "felt weird". People who got into Mozilla through those usually ended up leaving, but I'd say it's a vast minority of cases.

And wrt. "sinkin ship" - It depends on us. Working for a non-profit has some downsides. We need to stay open, inclusive, some things happen slower, we get much more "negative feedback" from people who feel more entitled to judge our work etc. But it also comes with some huge benefits. I never optimized anything at Mozilla for profit, our leadership sees money as a tool, and we can focus on what we solve with the tool. Some of my long-standing projects are way too "altruistic" to ever exist at Google, Apple or Facebook, because they don't impact the revenue stream. We can and use the Mozilla Manifesto to reason about decisions all the time, and thanks to all of that - the Web community really supports us. So, if we can get amazing products, we may not have the marketing budget of big companies, but we have the support of the global community. And releasing amazing products is something that depends on us, not any external factor. Right now we're pushing for Firefox 57 and I do believe we're on the right track to succeed (yes, I disagree with Andreas :)).

My take is that if you have the "NGO" muscle in you, enjoy spending a lot of time on big problems, like the open-source community model of work (which we drift towards with most projects, although there are pockets that are slow to open up) and are not money-driven (not because they pay bad, just because you can get better money as a hired gun at some corpo ;)), Mozilla is worth giving a try :)

zbraniecki··on Google Fires Employee Behind Controversial Diversity Memo
a) I'm not sure what you base your believe in. My gut feelings tells me the opposite. If you polled a representative sample of employees you'd get very different results in the 2nd chart

but that's not important because:

b) The idea that average people are good at understanding the consequences of such document is inherently flawed. It's my problem with a lot of aspects of American culture (I'm an European living in SV) - regular people are not in position to evaluate complex systems.

Every time I see an article about, say, a shooting, that ends with "Joe Black, who lives nearby, believes that police should do X and Y to stop the increase in crime", my blood boils. Joe Black is in a terrible position to understand the crime dynamic in his city, to understand how the police operates and what does and doesn't work in the pursuit of lowering crime rate.

Now, coming back to the topic, I'd argue that 95% of Googlers are completely not qualified to respond to the second question. At all.

There's a bunch of people in the HR department, with degree in sociology, group psychology, organization behavior and design, maybe people who studied micro-discrimination, gender imbalances etc. that are qualified to answer it. But the idea that you can poll employees to learn about whether this document is harmful is completely nuts.

Sorry for venting :)

p.s. But yeah... while writing this comment I saw an article about the poll asking Americans if the US should attack North Korea [0]. I honestly believe that the culture of doing that, while pretends to stay close to the roots of the country of freedom, where citizens decide about their country, is in fact terribly harmful because it builds conviction in people that they're in the right position to make such judgements on topics they have insanely little knowledge about.

[0] http://www.cnn.com/2017/08/08/politics/cnn-poll-north-korea-...

zbraniecki··on Google Fires Employee Behind Controversial Diversity Memo
can you cite the source? Because this sounds very unbelievable to me. I work in a mid-size company structured in many ways similarly to Google and I cannot imagine a representative survey with results already coming up.

Google has close to 70k employees. What was the sample? How selected? How representative?

It sounds very questionable,

zbraniecki··on Languages Which Almost Became CSS
I strongly believe that we're currently in the early stage of the same cycle with localization formats.

The most popular ones, like gettext, xliff, .properties are very limited, while the most sophisticated one - MessageFormat - is stopping a few steps short of having all needed capabilities and is very hard to read/write by hand.

I'm very excited for the next 3-5 years where I believe we'll see more development in this field and hopefully we'll end up with a good equivalent of CSS for Web Localization.

Full disclosure: I work on a project that my team believes to be a good contender for that - Project Fluent ( http://projectfluent.io/ ) I also work within TC39 on ECMA402 on APIs that will make coming up with new localization APIs much easier (PluralRules API, language negotiation and selection etc.).

zbraniecki··on Silicon Valley Women, in Cultural Shift, Frankly Describe Sexual Harassment
This article, and this conversation is about one pathology. You bring up a different one and say "we should treat them equally!" - that's not very different from the "all lives matter" bullshit. Sexism and discrimination of women in tech industry is a massive problem. What you described probably exists in some cases, but is not a) related to the problem at hand and b) orders of magnitude less profound. By bringing it up here and trying to tie those two together you're devaluating the severity of the problem at hand in the same way as if you went to an article about rape and started talking about those promiscous people who like sex.

Those pathetic douchebags were not "primed" by any experiences with women "dating up". They were primed, trained, protected and covered up by the shitty patriarchal society we live in. And it's worth tackling that problem without attempting to dilute it in "maybe she wanted it" sauce.

zbraniecki··on Gnome 3.24 (with Night Light) Released
Why do you think it is a step backwards?
zbraniecki··on Chrome Widevine DRM can no longer be disabled
please, report the bug in bugzilla.mozilla.org and describe your problem. Thank you!
zbraniecki··on Chrome Widevine DRM can no longer be disabled
Thanks for reporting this!

P3 is not that bad tbh. I understand that for each one of us the bug that affects us is the most important one, but working with bugzilla is a bit like data mining. We have 1.3 million bugs reported and we track our respective components, cluster similar bugs and solutions and prioritize what is affecting most people in most crucial ways.

The fact that your bug got triaged and [gfx-noted] is a good sign. It's on their triageboard now and may soon get clustered with other Google website related graphics bugs. You'll get an email when it happens.

Thanks for your help!

zbraniecki··on Chrome Widevine DRM can no longer be disabled
Please, report any problems you encounter. We can't fix things we don't know about :)
zbraniecki··on Chrome Widevine DRM can no longer be disabled
Thank you! Hope you'll enjoy the ride :)
zbraniecki··on Chrome Widevine DRM can no longer be disabled
Thank you so much! It means a lot to me to read so many positive comments.

I've been a volunteer for Mozilla since 2000 and got hired in 2010. I've been through thick and thin, and I deeply believe in Mozilla Manifesto and care about the role we serve in making the Internet a public resource.

https://www.mozilla.org/en-US/about/manifesto/

zbraniecki··on Chrome Widevine DRM can no longer be disabled
did you report bug on bugzilla.mozilla.org ? If not, please do and include all the details you can about your platform.

Graphic driver bugs are nasty :)

Hope we'll be able to address it and remove the one and only reason you can't use Firefox!

zbraniecki··on Chrome Widevine DRM can no longer be disabled
> What do I do if as a normal user

That really depends how "normal" the user is.

If you have basic HTML/JS/CSS skills, you can try to minimize the testcase - https://css-tricks.com/reduced-test-cases/

If you don't, you can always contact the website author and inform him about performance problems and ask him to get in touch with us. He does have access to the whole website and could help us get a temporary account for us to investigate.

If you have more knowledge and curiosity, you can even use our performance devtools to generate a profile. It's actually fairly easy - just start recording in the performance tab and do what you would normally do that is slow.

Then file a bug and attach the profile.

This will give us an insight into what takes so much time. It may not be enough, but sometimes it is, and in those cases we are able to "peak" into what your Firefox is spending so much CPU cycles on.

It's not perfect, but it's better than nothing :)

zbraniecki··on Chrome Widevine DRM can no longer be disabled
We definitely want JS benchmarks. Please, if you can, file bugs at bugzilla.mozilla.org in Core::JavaScript Engine.
zbraniecki··on Chrome Widevine DRM can no longer be disabled
great point. I haven't read about this experience, but I believe you. I know that there are different kind of animations that trigger different reactions (like, fade in text is different from things shifting on the screen).

Please, track Mozilla progress over the next months if you have the time. We're drafting a workforce to start the UX updates and I hope we'll get voices like yours well represented. I'll make sure to post about it on HN once we begin.

zbraniecki··on Chrome Widevine DRM can no longer be disabled
How would you like Mozilla to be financed?
zbraniecki··on Chrome Widevine DRM can no longer be disabled
Hey, thank you for your input.

We are aware of the importance of those three things. They're all not-trivial problems to solve, but we're working on all three:

1) Performance is an insanely complex problem. We can't easily measure the "real" performance, in a scientific way. We can measure performance of so called "microbenchmarks" which can tell you that, for example, our JavaScript engine can perform X million regular expression computations per second. The X may be higher or lower than what Chakra or V8 can, but how it translates to real world experience is insanely complicated. What we know by now is that JS engine performance is not the problem. At this point all modern engines are fast. What we need to advance are things like performance of DOM, CSS, layout, painting and prioritization (ability to ensure that UI performance cannot be blocked by website's JS computations). We're working hard on several projects this year, under the umbrella of codename "quantum" that is aiming at making Firefox blazingly fast. We're taking lessons and code from Servo and applying it in Firefox. The end result, if we succeed, is that we'll have ability to perform DOM, layout, CSS and painting orders of magnitude faster (even if it's 2x it's still huge!). The other piece of performance, is what's called perceived performance. That one is even more tricky because it's all about what your brain tells you. In cognitive psychology and HCI there are tons of studies that prove that depending on how we use tricks like animation, colors, shapes and gimmicks like focus, we can trick your eye to think that things are faster or slower. Chrome is really good at that and kudos to them! I love watching their UI painting order in 120fps slow-mo just to see all the things they employed. Together with Quantum we're going to look into ways to improve the perceived performance of Firefox. I hope we'll get it right :)

2) On the privacy front we're working with Tor browser and we're also working on several projects aiming at disrupting the current way of browsing the Internet. Read about Firefox+Tor browser and Activity Streams if you're interested.

3) On the QA front, we've made major progress over the last year. That's a huge part of the technical debt that we had to remodel in 2016. We now have much better fuzzers, tests and APIs that makes it easier to write code that will not break. On top of that, multi-process helps us make sure that painting crash doesn't crash your browser, that plugins can't crash your browser and finally that content can't crash your browser. We also designed a whole new language that is significantly better suited to write security critical code than C++ is. It's called Rust and we just started moving pieces of our engine to it. If we can succeed with this, we should end up with a codebase of the unprecedented stability for a project this size and facing third-party code. If you're interested, read about Mozilla Oxidation project.

It is a major bet. We're betting on JS, we're betting on Rust. We're betting on Gecko and our ability to "change the engine mid-flight". I'm excited about the opportunity, but it's going to be a huge effort to pull it off.

If any of this is interesting for you, help us! We're a small team (compared to all other browser engines), but we have a culture that fosters community participation and I believe that we can compete and shape the Web together: https://whatcanidoformozilla.org/

zbraniecki··on Chrome Widevine DRM can no longer be disabled
Hi all.

Just wanted to tell you that a lot of people working on Firefox are here and we really appreciate your word of mouth support and feedback.

This year is going to be huge for us as we're wrapping up the part about reducing technical debt and start gearing toward using our new superpowers to innovate.

We need your help to stay relevant (market share), improve quality (report bugs, write test cases), and improve (lots of code in Firefox is easy to hack on as we move a lot of UI to HTML/JS - like devtools).

← PreviousPage 4 of 6Next →