Firefox OS's story from a Mozilla insider not working on the project (2024)
ludovic.hirlimann.net
ludovic.hirlimann.net
I think this is basically correct.
I also think:
- Targeting low end devices first was a huge mistake. The web stack they were building on was simply too inefficient for that to work with the devices at the time. If they'd targeted high end devices then it would likely have worked on low-end devices by the time they were ready to ship it anyway (perhaps this plays into them rushing and having unrealistic timescales)
- It's really unfortunate that Rust wasn't a bit further along during the Firefox OS era. A lot of the Rust UI pieces that are now being developed would make a really strong foundation for something like Firefox OS. And could perhaps also have enabled a better story for "breaking out" of the confines of the web platform.
> [Mozilla] hired someone from the mobile industry to run the company, this led to some culture changes : no more a flat org, but a pyramidal one with middle managers. Culture became way less engineering centric, and started being a bit more top -> down. Focus was now solely on B2G.
when i think about the occasional anti-user decisions Mozilla has been making in the past several years (most recently, AI training policies) i wonder now if it's all residual from this structural shift.
I don't know enough about Mozilla to say, but the OP certainly makes it sound that way.
I think this is part of what Mozilla missed: the key value proposition of FxOS wasn't that it was web-based specifically: it was that it was open in the way that Linux and the web are but Chrome and Android aren't (one might also compare RISC V vs ARM). The promise was that it could bring this openness to graphical operating system (mobile/desktop) environments and installable app development.
The problem with targeting high end first is sales / production risk. A phone based on a new OS is a risky purchase, and it's hard to convince someone to spend $500 on a high end phone without much software available. A $100 low end phone without much software available is an easier sale.
If you can get important software built for your platform, sales are easier, but sales/usage numbers drive developer interest. Also, it helps if the platform offers capabilities developers need; maybe it's my bias, but the number of platforms that launch without everything a messaging app needs was mind bogling --- everyone seems to understand that messaging apps sell phones, but they don't find out the requirements for them until very late.
Apple does just fine without targetting low end (although they do sometimes target mid range), but they live in a different world, and it took decades to build that market placement.
living in California and contributing code to gaia, i could only have a dev phone (not being involved with moz) when i went to an alley in Mexico city known for selling black market (no tax) computers. there i could finally pay usd50 to buy the crumiest bright orange plastic phone to use for dev work.
would i have spent usd200-500 on my hobby to get a good phone months before i had a trip to mexico city? probably. would i pay the usd300 retailers in the us were charging for the crummiest low end phone? hell no.
I guess around 2012 I bought a Firefox OS phone from a Spanish company for I believe 150 euros, which was quite reasonable. FirefoxOS was unusable though. Plus it lacked WhatsApp so I could not use it as a daily driver. In 2025 I would also need to have a banking app on mobile so it’s impossible to use an “alternative” phone OS for me
I mean, skate to where the puck is going, right? It sounds like Firefox OS started development as mobile component prices started to fall in price dramatically. By the time B2G was ready to go to market, the baseline for hardware would be dramatically higher than when they started.
Except they didn't target $100 dollar phone. They were aiming at $20, but giving headrooms to $35. And in about 2 years time they finally believed they made a mistake, and raised the target to ...... $50.
Silicon Valley at the time was all about software. They believed in Bill Gate's vision where all hardware cost will drop to zero and software will subsidise it. And Apple's business model is not sustainable.
It was then I learned Silicon Valley has absolutely zero understanding of anything hardware. Apple was somehow the best and the outliner.
For low income economies, which is where the majority of people are, and the majority of late adopters of smart phones are, a $35 phone is affordable --- or much closer to affordable than the rest of the market.
If you make a smartphone that works at that price point, the addressable market share is huge. And it could be subsidized by carriers in low income countries, too (see the JioPhone). The class of hardware they were targetting could provide a good experience --- similar specs provided good UX on iOS and reasonable to good UX on Windows Phone.
If you make 1 million $35 phones and none of them sell, that's a waste of $35Million (maybe less, because you presumably have some margin, but maybe more, because R&D). If you make 1 million $500 phones and none of them sell, that's a waste of a lot more money. Around this time frame, I bought several $500 phones for $100 each when they failed; the Amazon Fire phone, the Nextbit Robin, the Essential PH-1; all were a great deal for me, and a colossal waste for the hardware team.
Some choice quotes:
> If the lock screen pops up while reading a webpage, you'll need to reload the page again. If you start the stopwatch and leave the app, the stopwatch stops. You can set the e-mail app to check for mail every five minutes, but there is never any free memory, so the mail check never runs. There isn't even anything to keep the crucial alarm process alive. If the phone is busy when an alarm is supposed to go off, it just doesn't go off. An alarm you can't trust to work 100 percent of the time is useless.
> Typing on the Cloud FX is pure agony, and this is one of those really crippling deal breakers that makes the Cloud FX a bad phone at any price. The keyboard doesn't support multitouch, so you if press "Q" and "P" at the same time it splits the difference between the touch points and enters "Y." This means you can only ever carefully hunt and peck at a slow rate. It's very easy for the Cloud FX to slow down, so if anything is going on while you're trying to type, buttons will take a few moments to register. There's also no auto correction and no copy/paste.
> Navigating webpages is a nightmare. The problem isn't just that the phone is slow, it's that scrolling is nearly impossible. A lot of times the phone is busy, and scrolling doesn't do anything. When it does scroll, you'll find the rest of the page often isn't loaded and you'll get a gray screen that takes a few seconds to be loaded into memory. The other problem is that scrolling is often interpreted as pressing on a link. This, combined with the speed of the device and the often-frozen scrolling, means the Cloud FX is frequently doing things you don't want—and doing them very slowly.
> The performance of the Cloud FX really cannot be understated. Screen taps sometimes take seconds to register. Firefox OS has a recent apps screen, but there is never any free memory, so nothing other than the current app is ever open. During particularly slow freak-outs, the screen will just turn black. If the phone falls asleep, or the alarm pops up, or a phone call comes in, your app closes and you lose your progress. Even something as simple as opening a folder of apps has a load time measured in seconds.
> Getting devices out to the developing world will be a big focus for many OEMs, since the market is just so huge. There are billions of people out there that still need access to the Internet. We know we can't expect something so cheap to hold a candle to the high-end devices we normally review, but all we ask is that the device be executed well and that the decisions that went into the product make sense. The Cloud FX doesn't make any sense, though, even at $35. It doesn't seem like anyone set out to choose the best software for this device, and that was really the crippling decision. Low-end smartphones need a low-end-appropriate operating system, and Firefox OS isn't up to that task. We get the feeling there are relatively good, sub-$50 devices out there—but this isn't one of them.
If the core failure is decision making, having a more generous hardware target might have helped, but some other issue would likely have come up.
...there wasn't enough free memory to store a 64 bit timestamp of when the stopwatch was started?
> You can set the e-mail app to check for mail every five minutes, but there is never any free memory, so the mail check never runs. There isn't even anything to keep the crucial alarm process alive. If the phone is busy when an alarm is supposed to go off, it just doesn't go off. An alarm you can't trust to work 100 percent of the time is useless.
Correct me if I'm wrong, but I think all of this stuff worked on iOS 1 in the original iPhone. It didn't have multi-tasking for most things, but preinstalled apps like the Clock and Mail were special.
This sounds to me more like unfinished software than hardware limitations. Of course, I can imagine that hardware limitations would make finishing software harder—but Apple clearly did it.
Also, they forced app developers right from the start (well from iPhone OS 2 because 1 didn't have third party apps) to expect their app to be killed at any time, and to save state constantly so that they could resume exactly where the user left off. Sounds like Firefox neglected to do this which is really awful when targeting such limited devices. It's one of the little things hat could have made it more workable.
And, I don't think there was a way for a third party Mail app to e.g. fetch new mail in the background. You essentially had to use Apple Mail.
And no, third-party apps could not run in the background in iOS 2, indeed. That's why a lot of people still jailbroke (Cydia etc)
What do you mean, "Rust UI"? In FirefoxOS, the idea was that all applications were developed solely with web technologies.
More Rust would have been nice, but would only have made sense with the support for compiling towards wasm (or asm.js, at the time).
Rust would have been used for implementing the web technologies, not by end users. I would note that many of the "web" technologies did not exist before Firefox OS (and Chrome OS) started adding them in a flurry, so really almost anything was on the table. The key thing was that they were open and specified.
And nowadays Firefox is the only of the "major" browsers that doesn't even support PWA anymore.
FireFoxOS could have won the battle by just existing. 5% of the smartphone market would be worth much more than only 5% of the browser market.
"Just existing" isn't enough to get you to 5% of the market. At the time Mozilla started investing in B2G, it wasn't clear if the also-rans of the mobile OS marketplace were chasing, between them, 1% share or 10% share (now, it's clear that you're chasing 1% share).
As this postmortem points out (and the linked graphics developer's testimony discusses as well), Mozilla was heavily focused on the smartphone OS development, to the point of letting its existing browser competency atrophy. And despite all of that effort, the result can barely even be qualified as an also-ran.
In the counterfactual world where Mozilla didn't focus on trying to build a smartphone OS but instead pushed more for quality in its desktop and mobile browsers, or even invested in ecosystems like Node.js or Electron built on SpiderMonkey and Gecko respectively (which dovetails nicely with their everything-in-JS vision), their relevance would have been far greater, I think.
There was brief interest by the community, and a few applications were written to take advantage of it, but it didn't last.
It wouldn't. Another reason for Firefox OS's failure was their insistence of doing everything in Javascript. Do not bet against Javascript mentality. They quite literally wants to write most part of OS in JS.
A few developer actually blogged about this once Mozilla announced they will discontinue the development of Firefox OS.
Most of the JS code was the applications (including the homescreen).
Personally it was the only reason I ever gave Firefox OS a try.
ZTE started selling the ZTE Open[0] on ebay direct to buyers, super easy for me to obtain even in the US. The phone had miserable specs but still ran fairly well in my opinion, to the point of using it as my daily phone for months.
Eventually it was let down by most of the early Firefox phones not getting the updates to add in features like MMS support.
It was a sad day when they called it quits, and what remains in KaiOS just reminds me we could have had a strong competitor in the budget market.
Based on my understanding of Alex Russell's info, that's not how the mobile market works. Low-end device performance in 2023 was at least 8 years behind flagship phones.
FFOS was doomed because the pixel explosion wasn't just stretching the GPUs of that time, it had also made CPU rendering a serious bottleneck, a fact the Android team also struggled to accept. Mozilla lacked the resources to do anything about it, although I didn't get the impression they actually understood the problem at all (as shown by it not being mentioned here), while the Googlers resorted to that most Googly of activities: arguing via walls of text that try to deny the obvious truth.
I have long maintained that what Mozilla should have done was a Firefox OS based e-reader. The mania that exists for all things e-ink in the tech industry is real, and if you combined that with a proper open platform you could leverage it into being a self sustaining business.
If thousands of vendors had made something Android 2.X level for the entry level market and the technical people who actually want what Google engineers wanted the ecosystem by volume would have been mostly FirefoxOS instead of mostly out of date Android and the ecosystem by revenue still Apple's.
Aside from all else, and what about the very responsive Windows Phone 7 or similarly flawed WebOS? Let alone all the feature phones, Blackberry etc. This was an enormously competitive era, and Google had to do some serious fighting to secure the future of Android at that time.
Phone manufacturers do not exist to serve the bidding of the software builders, they actually have to make money, and doing that requires looking at least superficially competitive. The primary way users experience that is through the graphical interface on the device, hence why my then employer made incredible amounts of money selling game builds for all these other devices to demonstrate how close to iPhone equivalent they were.
It's no accident that the Android golden age began with fusing the UI concepts of WebOS that worked with a rendering pipeline that started to make sense for the hardware it was on.
Mozilla ignored the autonomy of the parties that should be in an ecosystem and tried to make them wait for Mozilla's choices and implementations on things. I bet they had more than enough people on every one of their waiting lists for being involved but they discouraged actual involvement.
BBM was much more mass market than your attempt at revisionist history would suggest.
Unfortunately there's very few KaiOS phones on the market anymore. Nokia stopped making them and reverted back to Series 30 and 40. I wonder why.
Actually this is the reason that I, at the time working at Mozilla on the rendering team and helping with Firefox OS, proposed will-change. To solved this exact problem. With it, I managed to get the home screen to be perfectly smooth using pure web technologies and no hard coding on low end mobile hardware. Here's the original proposal: https://lists.w3.org/Archives/Public/www-style/2013Nov/0414....
Coincidentally I was profiling and tuning the keyboard pop-open for 256 MB devices at the time. Profiling and tuning the trade-offs of when to render and animate the graphical layers. Knowing that these trade-offs would be very different and incorrect for a 64 MB device.
B2G management was driven by a mix of "the web is the platform" ideology[0] and a willingness to promise whatever carriers and OEMs wanted to close a distribution deal regardless of whether it was feasible or made any sense.
My own personal "this project is hopelessly doomed" moment was roughly 48 hours before the 1.4 code freeze when I was contacted by a colleague asking how feature X was coming along. Feature X was needed to fix Bug Y, and Bug Y was in the critical bug list for the code freeze. But it turned out the B2G project managers had decided to manually track their critical bug list outside of Bugzilla, and didn't make any effort to track the transitive dependencies of those bugs, or notify anyone that they should be working on these. "ngmi" hadn't been popularized yet but that was very much my feeling at that moment. Active development ceased by the end of the year.
[0] which I think could have panned out at least technically, given sufficient resources, time, etc
That always struck me as one of the more impressive things that Jobs achieved with the original iPhone - getting buy-in from Cingular to do things Apple's way and not Cingular's. Pre-iPod Apple probably couldn't have done it.
It cannot be overstated what a gamechanger it was to have a recurring revenue stream incrementing with each new user. Beside Blackberry, no cellphone company was able to finance maintenance based on its userbase AND hardware sales, they all only worked (and mostly still work) based on hardware-sale alone.
Normal folks just want to use phone to help them with their lives, and as such if they can’t find apps which helps them, they won’t use that phone as their daily driver.
They don’t care if it was built with rust or js or java or anything obscure.
I do wonder, though, did Mozilla ever try to convince Facebook to build a FirefoxOS build of WhatsApp?
Yes, they tried.
From what I know, a big difficulty in convincing them was that FFOS didn't have actual native applications, it was running everything on the Firefox Gecko Engine (on top of Linux). WhatsApp would have had to develop their application as a WebApp in Gecko, to run on quite low-spec hardware, which would have been a huge undertaking at that time.
(FYI, the first FFOS device was launched in 2013, Facebook acquired WhatsApp in 2014)
The problem was never technical, only business. WhatsApp was basically saying: "We'll build an app if we think we can get XY millions new users". FxOS was just not shipping enough to reach the bar. KaiOS shipped in India with Jio, Facebook saw the traffic coming from KaiOS devices using the basic FB app and decided to move forward with a WhatsApp app.
You're right, the userbase was the deciding factor, but also in context of the required effort to support the development/maintenance.
To my knowledge WhatsApp did a cost-estimation for FFOS-support back then and concluded that they need to see more than just sales projections to start working...
The reason why the phones even landed in South America (and Spain) was because Telefonica heavily invested into Firefox-OS and lobbied several HW-vendors to provide devices for it.
The idea for Firefox-OS to carve-out its niche was to go lower in spec than Android could, providing a solid entry-level Smartphone experience at a price an Android device could not match just due to Hardware spec alone.
Google understood the threat, and indeed they couldn't lower the bar that low anymore and have Android provide a decent experience on a device with 1GHz single-core CPU and 512MB RAM.
As a fast measure, Google created "Android Go" as a stripped-down Android variant for lower-spec hardware, exactly to respond to Firefox-OS.
--> The fact that WhatsApp was missing on Firefox-OS was the final deathblow, yes. But even before launch of the device, the wave of "Android Go" devices matching/undercutting Firefox-OS in price was already building up, and carriers not confident in Firefox-OS were starting to order those devices.
Then, at the moment the Firefox-OS device faced its major setback after launch, that wave already came crashing down into the procurement departments of all the carriers, and so the gap to threaten Android was closed.
Just to reiterate the scale: Google made a "Go" variant of Android and ALL it's GMS-applications just to close this gap. After the threat was gone, "Android Go" was discontinued again
> After the threat was gone, "Android Go" was discontinued again
Sometimes the invisible hand of the market is just a middle finger pushing on the scale.
To me it was money well spent. About £50 (CTE Open???) for a firefox phone and it covered all I needed at that time. Calls, Text, Games.. and allowed html5 projects, which I created to do things with my home server. At the time, it was just a WIN-WIN as I could not be asked to create Android apps.
I was sad when Firefox no longer supported it. Evenutally, I finally jumped in the Android world. I dont think my reasons were entirely about Firefox, but because I was changing my roles and resposibilities at work and had to make/answer more calls that ever. My Pay as you go scheme was not suitable anymore so decided to buy contract.
I was proud holding off for as long as I could but in todays world where I need WhatsApp, MSTeams, VPN, etc.. I doubt I would have stayed on FirefoxOS if still going strong today. Alternatively I could just ask for a Company Phone. LOL.
Overall, the Firefox phone was nice. A bit quirky in places but for the hardware at the time (compared to what was on the market) it was pretty cool.
Some people laughed at me with their IPhone of Galaxy.. but I didn't care.
I do keep my eye on other Open Source/Free phones. Honestly, if there is a performant phone that is affordable and runs GNOME smoothly (and nice experience) I honestly would consider it.. as long as the apps mentioned above can run.
That said I think Mozilla were right to try. The phone ecosystem we live in today is locked down and tightly controlled by tech giants. Leveraging web tech to bridge that gap made sense, even if it had drawbacks. But it was a competitive environment where they needed to nail absolutely everything first time and just couldn't pull it off.
Tech Bloggers couldn't stand that a startup would launch on Apple (not even Android) and they couldn't download them and write about them. They made a lot of noise about this. But it was always nonsense.
I think the average non-technical user uses a whole lot more apps than what you give them "credit" for. The technology industry has pushed for more and more everyday services to be app-based. And most of these apps work only on iOS and Android.
In the 2020s every serious app developer from Epic to Spotify to Facebook to Tinder is desperately concerned by the platform policies of Apple/Google, and would gladly support Windows phone as first class platform. And so would consumers - enterprises would love to have their AD groups work neatly on employee devices and many individual buyers would love to have more open freedom too.
Absolutely right.
It gave up too early.
And Mozilla gave up on B2G too early. KaiOS has been quite a big hit and it's still making money. In time, the hardware came down to meet the requirements of what wasn't a really lightweight OS when it was launched in 2013 or so.
WP7 and WP8 weren't amazing (although three of us liked it), but they offered tremendous value on the low end. You could get a $100 WP8 phone that was more responsive than a $250 Android phone. Microsoft was telling people that most WP8 devices would be upgradable to WM10, then they removed low memory devices from the list, then they released WM10 in such a state that upgrading anything was a bad idea.
In particular, (trident) Edge was so awful, it was almost unusable. Mobile IE had a terrible rendering engine, but at least the UI was responsive. Edge queued inputs, so when a page was busy loading, it wouldn't respond to the back or stop buttons or navigating to a new page until the page finished loading... Not sure what the stop button is there for when it waits to stop until after everything is done.
When you are driving developers away by launching a different development model every release, users are going to be driven to the browser, and if it sucks too, what are they going to do?
WM10 had Continuum for high end phones which is neat, but it neglected the low end market where WP7 and WP8 shone.
It wasn't until the final build of WM10 that they got performance to be good on my test phone. (Edge still sucked) But that was way too late. They needed to have WM10 come out of the gate as good or better than WP8. AFAIK, WM10 never got bigger installed base numbers compared to WP8 even though many WP8 devices were upgradable to WM10.
Perhaps moving phones into the overall Windows org wasn't the right choice for the product. Almost certainly eliminating dedicated Test engineers reduced the quality of the product.
In an alternate reality where Microsoft provided a usable WM10 base install, tested and fixed WP8 -> WM10 upgrades before release so apps didn't fail on upgrade, had an option of a reasonable browser (letting android or chrome into the store or at least offering the old mobile ie), things would have been a lot better. If Intel hadn't pulled the plug on Atom for phones days before the Continuum demo, it might have been buy a high end x86 Windows Phone and run win32 apps on it when you're at your desk... Maybe WM10 would have still failed, but it would have put up a good fight.
MS were obviously caught on the back foot. WP7 was based heavily on Windows Mobile because that was all they had ready to go at the time. Then they rebooted the OS entirely with WP8 (based on NT, should have always been that way), removing backwards compatibility for existing apps and phones. Personally I gave up on Windows Phone right there and then: they'd obsoleted a phone I bought (IIRC) a little over a year previously.
You might be right that as time went on they might have made it a success. Certainly, they had the resources to keep throwing money at it until they did. But those early missteps really hurt them. You need passionate evangelists for your product, I’m only half joking when I say they should have swallowed the cost and upgraded every WP7 owners phone for free.
That’s not true; I developed a WP7 Silverlight app that worked with just minor bugfixes through Windows Mobile 10.
What they did do was bungle the marketing on it, so the public conciousness mixed up “WP7 devices can’t run WP8 apps” and “WP8 devices can’t run WP7 apps”.
WP7 itself was a throwing away of compatability with Windows Mobile apps; reportedly, that loss of goodwill is why Google refused to support the platform. WM6.5 apps would have been out of place on WP7, but if Microsoft was still yelling about developers x3, maybe they would have made that transition possible --- require a new frontend for cohesive UX, but allow for core libraries from older apps to run as-is, instead of forcing the new apis.
The loss of compatibility wasn't even the biggest issue here - it was the loss of functionality. WM6.5 had a huge enterprise feature set, only rivaled by Blackberry (and probably Nokia to an extent). Then came WP7 and had...literally nothing. They didn't even support Exchange ActiveSync in the beginning.
This was enough for a not insignificant number of devs to drop out of the already small market and lower the already limited app supply even more.
I like to think that WP7 was a smart dumb phone. I had a great UI, that I loved. And just enough smart stuff to make me happy. Maps. Unified messaging. Cortana wasnt completely enshittified yet either. Small package (L520) and I loved it.
WP8 tried to incorporate Windows 8 design features, and everything got worse. Vendors pulled out of the unified messaging. Cortana would send my maps to the advertisers address rather than closest requested. And the L525 that ran it actually went backwards in specs.
I would still, today, go back to a supported, bugfixed, WP7 in a heartbeat. And I bet there are dozens! of people just like me.
I also upgraded to Lumia 550 as my Windows Mobile 10 device and it wasn't bad either though the platform was close to its demise by that time. I eventually gave it to my Dad and got a Galaxy J7 in exchange which became my entry point to Android world.
But I tend to share the view that WP was, with the Nokia WP8 devices, finally getting somewhere. The fact not everything had to go through Silverlight to get on the screen was a big help!
There was a lot of noise in the industry back in that era that suggested this whole fight was far more intense and nastier than many would imagine (and I have to be careful to not be any more specific than that), but ultimately I believe Android only reached an acceptable state because of the need to defend from iOS and Windows Phone/RT.
From the perspective of a European carrier back in the day: This would probably have been true in the early iOS/Android days, but by then we were wayy in the "There's an app for that" era. All metrics we had back then showed potential app availability being the tie-breaker and the "WP has no apps" perception the biggest reason for consumers to avoid it.
And conversely, to have apps you need developer buy-in, and to get that you need to demonstrate potential for wide adoption or be so intriguing that devs want to build for your platform regardless.
This is where starting low-end hurt. As noted by others, this made it difficult for Firefox OS to perform well in users’ hands, but additionally the users willing to take a risk on a new platform generally aren’t usually the ones who are interested in low-end hardware. People buying cheap phones primarily want a high value proven workhorses, and anything running an experimental OS is the opposite of that.
As a result, buzz around the project was minimal and quickly dissipated. Devs weren’t excited about it and didn’t talk about it and thus didn’t develop for it.
Still the best mobile device I've ever owned.
I was at Mozilla during this time, and I remember a huge blocker was WhatsApp. They were not interested in porting, but at the time it was essentially the killer app in the same emerging markets that Firefox OS was targeting.
If the web client would have existed back then it might have helped
They didn't need to and couldn't have expected to. The real problem was that to get to a successful product, you needed to be willing to lose money for a long time. Android and iPhone were years ahead in terms of polish and optimization, and were not exactly sitting still. Firefox OS was going to be bad for a long time before it could ever be good, and did not have the luxury of its competitors being in the same state any longer (people don't remember how bad Android used to be). These are necessary growing pains for any incredibly complex system: mythical-man-month stuff that should be familiar to anyone in software development.
At the time, Mozilla had just renegotiated their search deal for $900 million over 3 years [0], with guaranteed payments that did not depend on actual search traffic. That was a lot more money than the project had ever seen before, and meant there was a limited window in which Mozilla could neglect the desktop browser in the hopes that the mobile OS would take off. In hindsight it proved to be too limited.
When Firefox OS was finally killed, management told us [1] something along the lines of, "The growth expectations we had going in to the project turned out to be unrealistic." I think it is okay to stop doing something when it is not working. This is just the inverse of Charlie Munger's advice to, "Figure out what works and do it." One wonders what could have been done with more realistic expectations, though.
[0] https://arstechnica.com/information-technology/2011/12/firef...
[1] I was also there, although outside of fixing a few media-handling bugs, I never worked on B2G directly.
Great little device, genuinely.
EDIT: It was a ZTE Open
I have to check if I still have it around...
One funny detail I noticed recently was that its battery was made by BYD, back then an unremarkable brand. So its operating system stalled, but at least the maker of another component became immensely successful since then.
Though I am very unhappy to see that its bounce-and-stretch animation at the end of scrolling is appearing everywhere :/ It was uncomfortable and weird looking then, and it still is now.
I have a Nokia 6300 4G, it works fine for what I want: A small dual SIM phone with good call quality and a 4G router when away from home. Any apps are a bonus though Meta recently withdrew the WhatsApp one.
> This branch is 400945 commits ahead of, 15894 commits behind mozilla-b2g/gecko-b2g:b2g48_v2_6
and it seems they're still using bugzilla, so they're obviously hard-core
Here are the list of their 74 supported devices https://www.kaiostech.com/explore/devices/
Browser, phone, email/contacts, maps, payment, authentication, file manager, networking (NFC, Bluetooth, Wi-Fi, cellular).
Seems like with the progression of web APIs and WASM, most of this functionality can live in the browser.
They're honestly pretty horrible. Everything is just so much effort on them. Nothing makes sense. And I'm no dummy. I worked for Nokia for a while on prototypes in the mid-2000s, which is the kind of era these Kai phones hark back to.
How can a FOSS project based around Linux and Firefox be picked up and closed? That ought to be impossible -- this isn't BSD-licensed stuff.
The front-end part of FxOS was always under an Apache 2 license, much to the dismay of some employees, but that was forced by commercial partners to allow them to differentiate while keeping their changes private.
I had a terrible experience, nothing worked correctly, basic, fundamental apps such as phone, contact, alarms, had lag spikes of over a second for any action and randomly crashed, it had no CJK support (kind of a big deal for me, and like 20% of the world)...
Some apps worked remarkably well, but having Youtube work better than messages is kind of pointless given the sVGA screen size. And the app ecosystem was basically dead. I tried developing a bit for it (apps are a simple html+css+js package), but at the time publishing to their app store required integration of their advertising solution, which absolutely sucked (in 6 month of use, I was never served any other ad than "hot MILFs in my area"), which meant that neither commercial nor OSS apps would target the platfom. IIRC side-loading sort-of worked depending on phones, which means it doesn't work for the average user.
Oh and sometimes the Kai Store app would send advertisement through push notifications.
I loved KaiOS on paper, but using it was a chore. A 2010-era "dumb" phone could do most of what a KaiOS phone can, but better, except for connecting to 3g/4g networks, and older networks are deprecated in many regions, which means they're not an option anymore.
Maybe they got better over time, but what a huge disappointment that was.
OEMs are not going to release the driver source code. They simply will not, at best they do random code dumps of whatever kernel version they were using at the time.
This is still true today and even Google tried to make it better with Project Treble. If Google cannot fix this even after all these years with total market dominance then literally Mozilla couldn’t do a single thing about it
Mozilla could have nudged it in the right direction. They didn't try. Not even a step forwards. The entire value proposition from non-technical standpoint was "we're not Google".
Others are trying with much less leverage or resources and actually progressing. Firefox OS was born, lived and died without making a dent.
— Firefox OS would've worked
— Mozilla leadership has always been interested in chasing fads
— Letting smart, motivated people work on cool shit is necessary to create the future.
cries in Bell Labs
I think the fundamental problem then (and still a problem today) is that Mozilla only ever had the budget to do Desktop XOR Mobile. Not both.
They basically mortgaged Desktop Firefox to build Firefox OS, but when things didn't pan out was quickly as they liked, they had to revert back to Desktop to save the cash cow.
After that experience, upper management became averse to mobile, to the extent that it was (and probably still is) very difficult to get resources allocated to mobile implementations of Gecko features that are supposed to be cross-platform. In practice many of those features are never implemented with mobile in mind.
I wish there was a (good) third option. I actually liked Windows Phone and was sad it was canned.
So we decided to start a competitor to Android to win in an even more difficult market against both Apple AND Google because we could not win the easier market.
It was never going to beat iOS, but it could've had Google's (and at the time Microsoft's lunch) for cheaper phones.
On the browser front, Google only had a third of the market at that point. Firefox was losing users, but the situation wasn't nearly as dire as it is right now.
Chrome might have only had a third of the market, but anybody around at the time could see that this was only a matter of time. Chrome was just better in so many ways. Mozilla needed to buckle down and focus all their engineer on the browser.
It does not mean "you can buy it" (purchasable) or "you can get hold of them" (available) or "you can order them" (orderable).
It means accessible to people with sensorimotor disabilities.
However this is not what "easily accessible" means.
There's no point in saying "I didn't mean that" unless you also explain what the intent was.
* The build process was atop Android, and pulled in a massive bulk of dependencies (open and closed), in various ways, and was effectively unauditable.
* I was turned off by it looking like they might be selling out the home screen in a way that threatened privacy&security. https://techcrunch.com/2014/02/05/firefox-launcher/
* I wanted a trustworthy, sustainable, open platform, and unfortunately it didn't look like Firefox OS was headed in that direction.
This was honestly a realistic risk, not because of the link you shared but because over the course of development, the success of FFOS became increasingly dependant on the commitment of carriers, and the main interest of those carriers was to create a OS under their control...
Trying to develop software for pre-iPhone phones seemed to be a mess of barriers to ISVs, like "BREW" and carrier approval.
It makes sense that carriers still want the phone locked down as much as possible, so they can extract more money from their captive customer.
(Also, it's still an expensive chore to get Pixels on which a carrier hasn't locked the bootloader against alternative firmware like GrapheneOS. Lawmakers/regulators should say that carriers may only provide voice and data service, for a fee, and aren't allowed to use their position to extract money other ways.)
The downside for me was that it didn't feel premium.
Compared to the 2012 HTC One S that I had back then, which was very slim, with aluminum finish and great screen, it didn't stood the chance.
Even during B2G and Firefox OS Mozilla still believe Google were their Friends. And will somehow continue to fund them, because Google is a search company, not a Chrome / Android company.
B2G announced in 2011, but I remember some idea / hints of it started floating around before then. Probably somewhere along the way of developing PDF.js if I remember correctly.
Dont bet against Javascript. Anything could be done in Javascript should be done in Javascript. Practically speaking that means apart from Kernel and drivers all things should be javascript. Even when it is easier for it to be done in C++.
Hardware is cheap. Software is expensive. Moore's Law meant hardware cost will half every 2 years and eventually so cheap it will be subsidised by software. That is why by aiming at low cost smartphone they meant $35 dollar smartphones. I think this thought disease was actually started by Bill Gate. That was the time I learned Silicon Valley have absolutely no idea about anything supply chain and hardware. Apple being an absolute outliner and exception.
Firefox was slow the day Chrome was born. Because everything is comparative. And e10s ( Electrolysis ) took 7 years before it was shipped on Desktop. It was originally targeted for Firefox 4. And took another 1 - 2 years for additional tuning and ironing out rough edges. As the article suggested. Everything at the time was mobile and B2G. It took a while before they realise Desktop Firefox, their cash cow wasn't competing and need some attention. ( It was too late )
I dont remember if Memshrink was started because of B2G. But Firefox at the time definitely have memory issues. Memshrink is ultimately is what leads to Firefox being the best browser to handle thousands of tabs. And About:memory was born out of it.
B2G as an idea wasn't wrong per se. But it was wrong timing and execution. Mozilla as a company was way too idealistic. From technological, marketing, strategy to execution. Ultimately as the article suggest, it was a management issue.
And if Mozilla have taken those lessons and learned from it in 2016. I believe they could still have had Firefox OS windows of opportunity in 2021. When somehow both Google and Apple became evil. Instead they wasted money on AI and many other initiatives.
I believe the downfall of Mozilla was when they won the war against IE. And that perhaps planted the seed they could really achieve everything. To this day I am not sure if they realised they won because of the Mozilla community being their evangelist ( Not being a negative thing at the time of 00s ) and perhaps more importantly. Microsoft IE wasn't even competing.
I still remember the New York Times Ads. Good Old Days.
> They hired someone from the mobile industry to run the company, this led to some culture changes : no more a flat org, but a pyramidal one with middle managers. Culture became way less engineering centric, and started being a bit more top -> down.
> Besides becoming more like a normal company, the new CEO grew the size of our teams, added project managers, Sales people, to make sure B2G would reach a huge audience.
As an engineer, we all want to think engineers are all that is needed. However sales and management are needed. Too many layers are a real problem, but it doesnt say much when its coming from people at the bottom. They can't see what is happening above them.
Oh they could and the results show it. The new people made a lot more money including the Google deal. While destroying Firefox
And this is a common management mistake: "We have the money, we need to grow." Meta have basically been doing that for years now, launching all kinds of projects that were mostly failures. But their ad income from FB and IG is more than enough to compensate for that. Google are in a similar position. When was the last time they actually launched a truly successful new product?
Maybe it would be better to let it be. In the case of Mozilla, focus on Firefox on all of the different platforms. On the other hand, not trying to grow is a career risk for management. If you are not implementing a growth strategy, why exactly are we paying you millions per year?
I have been in such companies, it becomes frustrating when your voice is not heard. Then you start to focus more on office politics and less on your work.
Imagine every application developer thought this way! For every app I want to use, I would have a different device running a different OS?
Supply chain attacks eradicated overnight.