Goodbye Mozilla
chrislord.net
chrislord.net
The worst part of the current tech-environment is that companies are dead years before they even know it. The next platform is already tearing apart your current one, while its in its infancy. And the entrenched players in the next space will be incredibly hard to unseat.
By the time you recognize the threat, attempt to turn the ship to respond, you are already deep down aware that there is very little chance of a happy ending. This results in convoluted late-stage product strategies that have 3+ directions since management knows no single one of them are viable.
This is a deeply uncomfortable truth that keeps me up at night.
Firefox had an opportunity to become what Chrome for Android eventually became; the better 3rd party browser. But they made 2 critical mistakes (as an outside observer):
1. They didn't take mobile (and specifically Android) very seriously at first. Fennec felt like a side-project at best. 2. XUL didn't work very well on Android, and it took them a long time to make the switch to native.
I believe things could have turned out very differently had they done the right thing from the beginning.
1) Most people experience most of the "web" on their phones via a non-browser. This may be by clicking on a shared link within the FB/Twitter/WhatEver app and seeing it inside an integrated WebView, or it may simply be viewing content normally available via the web through the lens of an app presenting the "same" content "natively". Either way, the point here is that browsers have had less of a role.
2) For a comparatively new OS (Android is ~10 years old, modern Desktop OSes being 30+) people browsing the web directly are going to stick to the default. They're going to downloads apps for things their phone doesn't already do; for webpages, it already has an app for that. Android Browser is the "Blue e" here.
The above may not be the case for the typical HN crowd, but it's quite naïve to think the majority want and act the same as we.
I am hopeful that this trend is changing - that slowly dedicated mobile web browsers are becoming a more viable option - but certainly this was not the case when Fennec was launched. The best we could hope for was for Chrome/Safari to gain ground over "apps". Mozilla's offerings never stood any real chance of being super successful in this early climate and tbh FirefoxOS, ill fated as it was, was probably the least worst chance they had to do so.
They should have pursued commercial strategies like becoming the default browser for Samsung and others. That would have got them around the "default" problem. Sell the "differentiation", build a kick-ass browser for Android; that would have given them a chance. They didn't even try it.
I really want uBlock and uMatrix like on the desktop. When I tried from within FF on my phone neither were supported.
Yes, as well as many other extensions. I'm using Self-Destructing Cookies, Desktop by Default, Smarter Scrolling and Android Text Reflow in addition to uBlock, thus tweaking my mobile browsing experience to something much better than the stock browser.
We seem to be heading towards a browser engine monoculture, and i am not sure i like that even if said engine is open source.
- UBlock Origin
- NoScript
- HTTPS Everywhere
- Self-Destructing Cookies
- Better Privacy (remove Flash cookies and delete Local Storage)
- Canvas Blocker
I've found most plugins that work on desktop Firefox work on Android Firefox.
Firefox is at a similar position here because on mobile nobody cares about a browser with slightly more functionalities than the builtin one.
And yet Firefox, for a period of time, had trumped IE on Windows. And of course, Chrome trumps IE on Windows today.
Besides, the "age of the OS" doesn't matter -- the age of the users does, and younger users tend to be even more savvy and favorable of third party browsers.
Android's browser was never in a position quite that bad.
I disagree, which is why I cited Chrome for Android as an example. It was not the default for a while (a couple of years at least?) but was still fairly popular. I remember doing webdev at the time and getting bug reports for Chrome. I also have received bug reports for Chrome for iOS.
I agree that it's hard to beat a good default, but the original Android Browser was so bad that it opened the door for competition.
I believe they succeeded thanks to their huge resources, their popularity and their good insight, but which of these factors was the main one?
Rule of law or better ideas? Meet billions of dollars. We all know who wins ... justice delayed long enough is justice denied.
I would chalk it up to the massive amount of experience among the early team members. Chrome was started by ex-Firefox people. The initial tech lead for Chrome, Ben Goodger, was the former tech lead for Firefox. Chrome Extensions were developed by the guy who invented GreaseMonkey. V8 was designed & led by Lars Bak, a veteran of Beta, Self, StrongTalk, and eventually tech lead of Java HotSpot. Google literally hired all of the world experts in the technologies involved and had them put together a new browser.
Rather than Chrome being the story of huge resources, I see Chrome as an example of what a small team of highly-skilled, highly-motivated experts can do.
Edit: I found a full list of team members and description of their background from a blog post at launch:
https://s.niallkennedy.com/blog/2008/09/google-chrome-team.h...
https://www.niallkennedy.com/blog/2008/09/google-chrome.html
It really was about 20 people, and they all have extensive resumes working on products like Firefox, HotSpot, GMail, Google Gears, or other Google products.
I'd actually switched over to Chrome before joining Google, and told all my family to switch, and it was apparent at launch that it was just a better browser. Once you're at the point where your users tell their friends to switch to your product, there's basically nothing your competition can do to stop you other than massively improve their own product, because you're getting large and exponentially-growing amounts of free advertising.
How could Mozilla compete against Google (or even Apple) to pull this talent?
Again: Big money wins (exception being those organizations that are both big and inept).
"The process involved coming up with an idea, presenting it and getting approval to run with it. You would then repeat this approval process at various stages during development. It was, however, very hard to get approval for enough resources (both time and people) to finesse an idea long enough to make it obviously a good or bad idea. That aside, I found it very demoralising to not have the opportunity to write code that people could use."
The early Chrome people indicated a similar frustration with upper management once Firefox got popular.
Expense approval is less clear, because it's not obvious which things your immediate manager can approve and which need to be kicked higher up the chain. But in my experience asking your manager works pretty well. It's possibly I've been lucky with managers.
> Rather than Chrome being the story of huge resources, I see Chrome as an example of what a small team of highly-skilled, highly-motivated experts can do.
I think this is a false dichotomy; the huge resources are what allowed them to hire all of those experts.
It appears it was management and/or politics that stopped other companies from doing what Google did.
Note that excludes the V8 team, which I believe was a decent number itself.
And for a number of years now, Chrome has had both a huge team _and_ huge marketing spend.
Thanks to the Web applications, POSIX became irrelevant for desktops, ability to run a web browser is enough regardless of the kernel.
Google can switch the kernel in ChromeOS for whatever they feel like and no one will notice.
The web also nurtured the GNU/Linux server. Apache on Linux slaughtered IIS on Windows and itself got killed by Nginx on Linux in turn. The web giveth, the web taketh away...
The majority of web stacks, regardless of language, run on Linux.
All the programming languages I care and use for web applications have zero dependencies on OS specific APIs, some of which I can even run bare metal or on an unikernel, with zero code changes.
The cloud concept, Web APIs and serverless executions, the OS of server running my web application is irrelevant, thanks to rich runtimes.
As for Apache and IIS, they are pretty much alive outside HN bubble.
https://news.netcraft.com/archives/category/web-server-surve...
Meanwhile, the online ad dollars never paid for the remaining staff/product left behind to maintain web sites, so slowly those online entities have been shrinking to skeleton crews, too, or deploying beyond panicky revenue models that no user wants to endure, making matters even worse.
So, for 20 years now, it's been a constant zigzagging journey for publishers. Page impressions, lead generation, rich media (e.g video), 'viewable' ads...
And in 2 years it will be "the next big thing", all the while more and more pubs shutter their doors and content continues to be narrowly focused, delivered by social.
Hell, not just the publishers fading out... I'm seriously worried about how this content evolution is affecting our personal biases and cognitive abilities.
This is why people like Richard Stallman who don't compromise on principles are critically important.
Principles are fine and dandy until you have to commit suicide to maintain them. Given the good Mozilla does as a going concern, becoming "that one weird browser that can't play your videos" would just result in more people using Chrome and Edge.
This kinda reminds me of the GCC/Clang thing. If I recall right, Clang got a huge influx of attention, in part, due to Stallman's unwillingness to allow GCC to export its AST.
Perhaps we need a rule for this. "The tech community interprets developer obstinance as damage, and routes around it."
This is one of a number of reasons, yes.
I don't think the industry consortium is really even that relevant here; it was proposed to the W3C in Feb 2012 jointly by Google, Microsoft, and Netflix, and already had a fairly well fleshed out proposal. It was shipped in Chrome, with Netflix supporting it, little over a year later. I'm fairly certain those three companies would've gone ahead with it regardless of what the outcome of the W3C proposal was.
Getting users to make decisions in favor of openness is an unsolved problem and it's why Richard Stallman, while laudable for acting on principle, has increasingly little impact. Rather than hectoring Mozilla for not committing suicide, your efforts should be directed to figuring out how to get ordinary people to make different buying decisions so groups like Mozilla aren't faced with the decision between standing up for principles and having users.
https://addons.mozilla.org/firefox/addon/resurrect-pages-isu...
While I understand that we should be calling it a DDOS, it isn't an 'attack'; so that seems a little severe.
Can't we just go back to calling it 'slashdotted'? =)
I'm still sore about that.
One of the phones, orange, I think the first or second model, is still in use. My wife now uses it on our annual trip to home country, and it still works fine as a phone. It's somewhat safer to use it, than the iPhone she has here in US - simply for not being stolen. It works just fine years after development. Most surprisingly though... the battery life was at 50%, a year after we haven't used it. (unless my son secretly charged it, and used it...)
I kind of wish for Firefox OS to come back. I liked the idea of apps being web apps, though it was a bit blurry what's downloaded and what's not, but with today's prevalent mobile internet that should not be a big deal (just a bit more pricey)
You sacrifice the best things about a web OS if you wrap it in a traditional OS UI.
Letting "the web optimize" your dialer, though, is perhaps the most dangerous idea I've heard in a while. Core applications under no circumstances should be optimized by the web because that's how you get malware masquerading as those core applications. No bueno.
I think the implication is that if FirefoxOS had focused on making it not awful, they would have done really well.
It'd be awesome to have a phone for which app development is all done in HTML / CSS / JS while still remaining fast and efficient.
Apple - you don't need native apps you can build great web apps. Developers and users rebel, a native SDK is introduced 9 months later.
Blackberry - you can build apps using Adobe based web technology. You get the picture...
Instead of web apps, Google stuck the world with a sub optimal Java implementation that they spent years trying to optimize.
Today they see Google Dominating because of Pure marketing, so they want to chase that dragon... They are killing everything Mozilla stood for in the process.
The most likely I'd see there would have been a "Certified for reflashing" platform.
Andy Rubin
the definition of open: "mkdir android ; cd android ; repo init -u git://android.git.kernel.org/platform/manifest.git ; repo sync ; make"
https://mobile.twitter.com/arubin/status/27808662429?lang=en
Product differentiation. Android is a crowded hardware market where everyone looks the same.
Chromebooks are great devices for a lot of people who simply don't need a full operating system. Look at how Chromebooks are successful in (US) schools. This is the perfect market for Mozilla.
I went to my public library the other day and they have computers to search for books, find info, etc, and the computers run Windows, but the actual app they use is browser based (IE)! Unbelievable. Mozilla could have the perfect product for them.
Web based anything has huge overhead compared to native code.
Yet they targeted ultra low-end hardware for emerging markets...
The benefits of web-based apps are the low barrier to entry and freedom from platform lock-in (openness).
They needed to target rich open source loving users with ultra high-end hardware that could cope with the overhead, and use Freedom as the USP.
Not sure that would've worked, but it would've fit their OS better.
One thing I can say - at least FirefoxOS was 1000x better than Ubuntu Phone.
Having worked there for 11 1/2 years, i can tell you this is not really accurate (it was until ~2007).
It certainly may appear that way to the outside, however :)
I like the consistent experience between the desktop and mobile versions - eliminating the need to download dedicated apps for each adblock-like service you want on your phone and the dependencies that come with it (root etc).
Safari on my near 2 year old iPhone 6s scores 89.
Mobile devices are approaching desktop performance for web, but I have never seen anything approaching this from Firefox on Android.
Its not so much that I actively seek to leave when that happens as much as I become more receptive to other opportunities. And recruiters and other companies are very good at luring candidates, so it all adds up.
Is the aim to do the actual work on the client?
Slightly related if anyone knows: is speech recognition a priority for the folks at Gnu/FSF?
If you create a decent Siri/Alexa clone and license it under GPL, I'm sure GNU would be happy to adopt it. :)
That said, a quick web search dug up these:
- https://voice.mozilla.org/ to collect training data.
- an implementation of the DeepSpeech algorithm, with a goal of running inference client side. (https://github.com/mozilla/DeepSpeech)
I'd expect the company to be pretty democratic. but every month you hear a story like that.
specially shocking: "What business a not-for-profit company, based primarily on doing open-source, web-based engineering has making physical, commercial products is questionable, but it failed long before that could be considered." ...and that was right after their failed mobile phone line failure!
That just isn't done elsewhere. You criticize internally and form a front to the outside. Meanwhile, Mozilla employees cry for their own CEO to be fired. That's just insane and wouldn't fly elsewhere.
Mozilla Corporation (of which I was CEO) was the subsidiary of Mozilla Foundation, not vice versa. So in no case were those few Mozilla Foundation employees who tweeted against me, the people who had me as "their own CEO".
Are you stupid, or dishonest? Pick one. Your posts are BS, and crucial logical contradictions are not "nuance".
It's first and foremost an open-source community, and some people are paid by Mozilla Corp. or Mozilla Foundation to dedicate their entire working time to that community. Allowing open dissent has always been one of the community values, so long as it's otherwise respectful.
https://www.mozilla.org/en-US/about/governance/policies/part...
At least as of a couple of years ago when I left, there were no internal guidelines above and beyond these.
Socially, of course, you could shoot yourself in the foot just like anywhere else, but we did enjoy a great deal of intellectual freedom and were expected to be vocal if we thought something should be different. I can't possibly view that as a bad thing.
The Eich situation was and still is touchy, and I won't go much deeper into that other than to echo Fabrice's comment that it was originally driven by a handful of Foundation employees that weren't part of his company (Mozilla Corporation is a different corporate entity than Mozilla Foundation) and to add that within Corp there was a diverse range of opinions, both privately and publicly expressed.
As for "wouldn't fly elsewhere," there's never been any particular ethical reason why an employee shouldn't publicly speak their mind short of spilling trade secrets. The reason has always been fear of being fired. Influence via fear has some pretty sharp downsides, and I'm happy to see the current trend of people submitting to it less.
How so? I think most people have gotten the memo that if you want to keep your job you will toe the line and only say things that couldn't possibly offend anyone. Not only did influence via fear win, it has become so entrenched that it isn't even visible anymore.
But hanging internal disagreements out for the competition? Stupid. There's no upsides.
there's never been any particular ethical reason why an employee shouldn't publicly speak their mind short of spilling trade secrets
It's not about ethics. The public perception of a company matters. Hanging internal disagreements out there does not help so you're just hurting yourself.
Here, I totally agree. Mozilla grew a dysfunctional pattern of participants (more likely to be employees than not in my experience, but Mozilla hired most of the active contributors) stabbing projects, individuals, and other sub-groups in the back, under cover of "being open". This was inevitable given the framing in the "open vs. transparent" document:
https://wiki.mozilla.org/Working_open#Open_vs._Transparent
Even now, Mozilla punches itself in the face too often, with punchers (and sometimes punchees!) claiming it's all for the best.
Taking care to give colleagues a chance to interact over a nascent or less-than-clear technical controversy, before blogging or tweeting, is not being "closed". It is standard peer review with scalability via layers-of-the-onion socializing combined with the "hermeneutic spiral".
Shooting first, fast, and in public in a large community with competitors and press listening is "open" in a vacuous sense, but it has the downside risks you note.
Maybe it's optimistic confirmation bias, but it sure seems like people are less afraid to speak publicly online.
And the same happens in politics and economy: everybody also thinks that been a citizen makes then elected to say how the economy and the politics should be lead. But their efforts stops there, in talking. No hands on works, neither in politics, neither in the company they are complaining. Mostly because as they don't know the core business that much, they can't get the big picture and start thinking in a binary way.
This has upsides too, I'm fairly confident Mozilla wouldn't be able to comply with a secret court order.
https://brendaneich.com/2014/01/trust-but-verify/
Our plan then (while we were still at Mozilla), and my advice now, is to avoid blind trust. Use systems where you can both trust and verify that your data is not being secretly surveilled.
Especially in hindsight.