EDIT: I just read that review you linked and it suffers the same problem - a total lack of perspective. The only critical error I can see is the lack of internal battery to keep time synced, which will be a major setback for browsing because of authentication and whatnot. Other than that, the sorts of people spending 35$ on a phone do not care about the three different materials, the thickness, the shitty camera, etc. I've seen the most horrifying, hackneyed together internet connectivity solutions in China that to mock a 35$ phone for being "too thick" is to be just... out of touch.
The lesson here, and where we failed, is that if you don't have control over the full stack you're shipping (hardware and software) and can't manage the QA and sign off process before shipping, you expose yourself to bad surprises.
Anyway, crappy hardware cares more about software efficiency. Firefox OS struggled massively with performance, to the point of absurdity; targeting low-end hardware with that is an idea that only ever made sense on paper. Supposedly it was just a temporary problem, optimizations were coming… eventually...
It does when those Bay Area hackers are the ones who decide whether or not they will port their app to your phone.
The mistake is rather that for closed protocols you are dependent on the Bay Area hacker to port the apps instead of "everyone can implement the open protocol".
GP's point is that targeting these markets was a mistake. Their strategy was fundamentally flawed from the start.
> Other than that, the sorts of people spending 35$ on a phone do not care about the three different materials, the thickness, the shitty camera, etc. I've seen the most horrifying, hackneyed together internet connectivity solutions in China that to mock a 35$ phone for being "too thick" is to be just... out of touch.
Again, GP's point was that the idea of selling a $35 phone to China was what killed Firefox OS in the first place.
If Mozilla was targeting the market for cheap, crappy phones, they were also targeting the search market on cheap, crappy phones. Unfortunately Yahoo, Bing and Google, the companies who comprise nearly all of Mozilla's revenues, need access to extraordinarily rich markets to justify their bids. It's not clear to me that they'd be substantially better off if the Mozilla phone was considered a success.
There's also government social welfare contracts to consider, or good faith agreements with a government that allows you to corner a more profitable market (you provided five hundred thousand affordable phones to our people, so we'll let you lay fiber through our country / get through our Great Firewall / whatever).
Yea, but you don't "bet the company" on it if the opportunity is smaller.
> Mayo says it took the focus off of Firefox. “It was close to a bet-the-farm effort,” he said.
Not so. Android decided to do cheap and lean, too. I did some price comparisons in emerging markets back in 2015, and everywhere that Firefox OS tried to play, there was an Android alternative available at a comparable price. More details are here:
https://www.technologyreview.com/s/537661/firefox-maker-batt...
Firefox OS was doomed the day they partnered with Telefónica. It's a massive juggernaut which phagocytes every startup they touch, assimilate their talent, then destroy the rest.
For those who don't know Telefónica, it's something akin to T-Mobile or AT&T (maybe even bigger, not sure). They had to change the name and branding of their phone division, a massive feat in both cash and brand recognition a few years ago to clean up their shitty reputation because people hated them so much.
In the tech side that HN may be interested in, they earn money by winning government contracts and developing crappy tech like a regular consultancy. They have a very high tech R+D division which I'm sure does wonderful work but their projects don't seem to surface.
In summary, if you see something "developed by Telefonica" it should turn your alarms on.
It does make it hard to learn from the experience though. Maybe the lesson is about learning how to quit at the right time.
I'm a Mozilla employee and I was very deeply involved in Firefox OS work.
On the engineering level, Telefonica was very helpful and working with them was actually a good experience. (can't say the same about all partners :().
Hindsight 20/20 is a powerful bias and it's easy now to talk about why FxOS wasn't meant to be.
All I can say, is that we tried our best and designed a pretty usable device in the end - just week ago I launched my Z3 with latest nightly of Firefox OS 3.0 (from April 2016) and damn, it's actually pretty nice and super smooth, and I'm speaking as a user of the Pixel.
I deeply believe that there's room for an open source operating system and that web stack is a good platform to base it on. The legacy of Firefox OS in form of tons of spec work and engine improvements (Firefox memory usage!) makes it much easier to do this now and I hope someone will.
But anyone building a new mobile OS will be facing the elephant in the room: will you get the top 5 apps to run or be ported on your OS? Without that, no way you can succeed. For FxOS, the unwillingness of Whatsapp to build an official app was more damaging that all the other reasons I read about.
OSes are commodity vehicules for apps, which are rightly what users care about. This gives a lot of power to prominent apps, which are often silos themselves - so maybe the fight to give is to displace these!
I've also known Telefonica engineers and they are some of the best. Maybe my bitterness against the behemoth clouded my words and I apologize for that.
The problem, as Lance said, is their inability to execute anything that doesn't provide them a direct, noticeable impact on sales (said to me by a Telefonica exec)
What I'm truly sorry for is the fact that the Firefox OS flopped. As an engineer, I know that you can build an awesome product but the execs may trash it if it doesn't go in the expected direction. And, with Telefonica, that happens too often.
It does have it's glitches (TV guide screen UI is very unresponsive) but I have few complaints. The Youtube app probably gets some usage as well as screen mirroring from Android/Windows PC.
Comcast. The word you're looking for is Comcast.
A more recent example would be the merger of the despised Time Warner Cable with Charter, now renamed to Spectrum.
Yeah, I remember when Firefox OS first came out, and when I heard of Mozilla's strategy of starting in the third world and then moving up, I knew Firefox OS was doomed.
Why? Because it reminded me of something I heard about car platforms, years ago. Back in 2002, GM introduced a pair of cars in the European market both based on the same platform: the Vauxhall Vectra and the Saab 9-3 (not the first generation of either nameplate, but the first to share a platform). GM made sure to introduce the Saab first. Why? Because Saab was a premium brand and Vauxhall wasn't. GM knew that "Saab built on Vauxhall technology" wouldn't fly -- people would see it as Saab degrading themselves by raiding a mid-level car for parts. On the other hand, "Vauxhall built on Saab technology" would go over very well -- people would see it as Vauxhall incorporating elements of a premium car.
The moral of the story is: you don't release a budget model and work your way up; you release a premium model and then bring it downmarket to cultivate the image of bringing luxury technology to the masses.
Your comment actually reminded me of an old Top Gear episode [1][2] where they described how Saab were instructed by GM to change only the body and the badges of a Vauxhall Cavalier when producing the Saab 900, but they ignored them and modified so much that the two cars only shared 1/3rd of the parts [2, 03:59]. The same thing happened again with the Vauxhall Vectra and the Saab 9-3, but to an even greater extent.
It certainly sounds like Saab didn't want to be associated with GM or Vauxhall or use any of their inferior technology or parts.
[1] http://www.topgear.com/videos/jeremy-clarkson/tribute-saab-p...
[2] http://www.topgear.com/videos/jeremy-clarkson/tribute-saab-p...
80% as good for 20% of the cost wins. You just have to be willing to accept selling for 20% of the cost.
The problem Mozilla has is that their hardware is not going to be any better/cheaper than somebody else's because it is somebody else's, and they can't undercut Android on price because Android is free.
What might have been an interesting strategy is to create a Firefox-branded version of Android that ships Mozilla equivalents to the Google apps.
The main problem with Firefox OS and also Canonical's Ubuntu phone was that they treated the lowend market of certain countries like second or third class citizens, assuming the cheap price would make up for the fact that there were no apps and the performance was terrible. What those companies didn't count on (but frankly should have), was that Android would just get cheaper and cheaper and the those countries would be fine with running two or three year old hardware (especially since apps frequently support old versions), rather than running a second-class OS.
I would always ask Mozilla what they would do when Android simply got as cheap, and I never got an answer that was in any way encouraging. And the answer was, they would accept defeat and give up.
A friend of mine had a Firefox OS phone at one point that she got for free at a conference, and I just remember at the time it was very difficult to even purchase a low end model outside of Europe at the time.
Part of it is carrier lock-in, but another huge factor is that you can't just install Firefox OS onto any old mobile device. Back in the late 90s/early 2000s, you could install any Linux on any PC/x86 system. PC was and still is a solid standard. The mainline kernel does a really good job of supporting the PC, both old and new (with really old support getting dropped and forked into things like the Linux/386 project for people who still use 386/486 embedded systems).
There is nothing similar for the ARM market. There is no standard. Almost no phones support Device Tree config (and definitely none of the flagships). With Windows and Linux, you install your base, you might need to install some drivers/modules and then you're done. ARM is a spec sold to SoC manufactures and then vendors attach whatever the fuck they want to whatever random pins. There's no unified boot loader. No standard way of determining devices. No standard platform. ARM isn't a platform. It's just an architecture.
http://penguindreams.org/blog/android-fragmentation/
There are plenty of x86 chips out there that aren't in PCs (The Wonderswan and the PS4 are both great examples of totally-not-a-PC), but they are the exceptions.
Ubuntu Mobile could have changed all this with the Edge. We may have finally had some standardized phones for open source and alternative operating systems. As it stands, every rom you see out there has to be customized for the hacked-n-patched together kernels of each and every single phone. It's not always an issue of binary drivers and no standard Kernel ABI, as some of these vendors patch things deep in the kernel in totally hacky, non-upstreamble ways for their kludged together hardware. Combined with binary blobs, closed baseband firmware, etc. etc. it's just a mess.
Currently there is no, convenient hackable mobile device platform, and that's keeping us from the same Linux revolution on embedded devices.
I'm surprised that Google and in particular Apple aren't coming under scrutiny from competition watchdogs as Microsoft did with IE in the early 2000s.
Competition watchdogs in almost every other first world country, however, are investigating Google for various antitrust claims.
Apple doesn't get a lot of scrutiny from competition watchdogs because they aren't a monopoly anywhere. Heck, outside the US, the iPhone has a pretty tiny market share.
I don't find patents or lobbyists particularly agreeable - but for an organization of a certain size to go without them in the 21st century is likely a very expensive strategic error.
In 2001 if you wanted to switch off Windows you could buy BeOS, Linux, or OS2 and run it on the same computer that you had windows on. Back then I did just that, switching from Windows to BeOS when BeOS was made free, I did my university work on BeOS running on a Toshiba laptop that came with Windows.
Right now if you have an iPhone or Android phone you have very little option to switch operating system on your device.
Also, what got MS in trouble was their contracts with hardware vendors. Companies like Compaq were required to pay MS for a Windows license even on computers they sold without Windows installed, and that was an abuse of MS' power. If Google is charging companies for Android even for phone they sell without Android installed, that would be analogous.
[0] "While Android continues to grow its share of the global smartphone market, it is actually losing share in the U.S. Kantar’s data shows Android phones making up 55 percent of U.S. smartphone sales compared with more than 60 percent in the same period a year ago." http://www.recode.net/2017/1/11/14239038/iphone-apple-androi...
And yeah, sure, you could run OS/2 or even BeOS, but these never reached statistically significant penetration.
Just as today you can run Sailfish, or Cyanogen/Lineage, etc... the latter supposedly has a market share of 3% or so. Better than BeOS, Linux, etc... in 2001.
> Thank God we nipped that in the bud.
Alternatively, ƒacebook, but that's probably too subtle (and a minor currency anyway).
Furthermore, many exploits don't care if their running on a JIT or on a scripting engine - that's an implementation detail. And if you really cared about security, you could design a verifiable native subset (as several systems have done) - webassembly shows that even slow standardization processes that need much more complicated cross-device compatibility can move meaningfully in that direction.
Disallowing non-first party JITs completely for security reasons is at best delusional, but more likely intentionally anti-competitive. Minor security bonuses are just gravy.
I imagine you never used a mainframe OS.
No, I don't. But all my friends wanted an iPhone and they're not programmers.
I think most people didn't know or even consider one could develop apps for phones at the time. Or why would someone want apps on a phone. Or even what an "app" is. They only knew "programs" on Windows.
People bought what was available. There were starting to appear some apps for Nokia phones but it wasn't widely popular. It was also a hassle to download / update.
The wow factor of the UI and interface design more than made up for it though, but there is a famous interview where Balmer laughs at the iPhone because it was expensive, had no keyboard, and no applications whereas Microsoft was selling millions of Windows phones a year with all of those.
And to counter your anectode - pretty much none of my friends wanted an iPhone and they're not programmers either ;P
In fact people wrote apps for rooted iPhones before iOS came out with official app support and an Appstore.
It was crazy how after apps came to the iphone suddenly people did know and there was even a little cachet to it.
Because what the globe needs is yet another company pandering to tech elites and ignoring the huge numbers of people out there of modest means? I mean seriously, Apple does all this stuff you're saying, pulling in devs with the shiny stuff, and targeting the SF yuppie... and this helps the people in Spain and India how? Not to mention that you're talking about the market segment already occupied by Android. Mozilla was trying to cater to those left behind.
Mozilla's raison d'etre isn't to push libre software per se. It's to enable the masses; libre software is merely an asset in that mission.
Read again what I wrote: the point was to get the phones into the hands of hackers, so they would write apps and create that ecosystem to ultimately benefit people elsewhere. (If there's a place with a more active hacker culture than the Bay Area, let's bring the phones there.)
I agree with amyjess' point made elsewhere. A quality phone is the means by which you get people interested in a platform, particularly developers. I think that's what's happening with Android and iPhone, in that the first phones are expensive, but the cost is spiraling ever downward. I just would rather see a freer alternative succeed, and hoped FirefoxOS would be it.
> iPhone ... but the cost is spiraling ever downward
The iPhone 1, the one that got people excited, was US$400. The iPhone 7 is US$650 (for the bottom model). Inflation in the US has not been 50%+ for the intervening decade.
You misunderstand--I desperately wanted FirefoxOS phones to succeed, precisely because an Open Web and free software is important, and ultimately a phone that can run any arbitrary software or OS. I know FirefoxOS wasn't that, but it seemed to be a step in the right direction. But if a phone is a lousy application platform and if developers can't be enticed to hack on them, then the question of freedom is moot.
By focusing on low-end phones and not Nexus- or Galaxy-class handsets, they lost sight of the fact that smartphones are aspirational devices. People want the best phones they can get, with high performance and high-resolution screens. They don't care about libre software or the Open Web if the devices on which it runs are perceived as inferior. (Everyone I know uses either an iPhone or a reasonably performant Android phone, and I don't run in wealthy or tech circles much these days.)
Look at the runaway success of the iPhone. It wasn't cheap, but because it was clearly a quality device, both regular people and developers clamored for them. (Remember the hue and cry when Apple initially said they wouldn't provide a market for third-party apps?)
If Mozilla really wanted to spur interest in FirefoxOS among developers, they should have set up kiosks at every Fry's Electronics in the greater S.F. Bay area (on the doorsteps of Apple, Google, Facebook) to sell decent, unlocked phones directly. People would have lined up to buy them just to hack on them. I would have seriously considered buying one as an alternative to a Nexus phone, if one had been available.
1: Ars Technical eviscerates a $35 FirefoxOS phone: http://arstechnica.com/gadgets/2014/10/testing-a-3