Slashdot Mirror
Why Open Source Phones Still Fail
adeelarshad82 writes "Truly open-development, open-source phones like the Nokia N900 will never hit the mainstream in the US because wireless carriers in the country hate the unexpected, writes PCMag's Sascha Segan. The open-source philosophy is all about unexpected, disruptive ideas bubbling upwards, and that drives network planners nuts. So, you get unsatisfactory hybrids like Google Android, which uses some open-source components but locks third-party developers into a crippled Java sandbox. The bottom line is that while Linux the OS, the kernel, and the memory manager are attractive to phone manufacturers, Linux the philosophy — and users banding together ad hoc to create new things — is anathema to wireless carriers."
"The open-source philosophy is all about unexpected, disruptive ideas bubbling upwards, and that drives network planners nuts."
Open source phones are about being user configurable, extendable and customizable. Wireless carriers like to charge for features, by the feature, and they don't like forking over what you've already paid for. That's pretty hard to do when you don't control one end of the transaction, as others have found out.
No buzzwords or BS about "disruptive ideas bubbling upwards" required.
Palm webOS is also Linux kernel based. That is the proprietary environment based on a Linux kernel, not Android. Android components by Google are distributed under the BSD license, that is the reason there is so much variation between vendors. That was the price to pay to get HTC and the other hardware vendors to jump in the Android bandwagon.
I don't agree with the sentiments of the article. It is true that carriers would like to limit what people can do with the phones but that cat has effectivly been out of the bag for quite a while now. Carriers are content with charging large monthly fees for data plans.
Googles andriod uses java/sandboxing because it protects the phone from potentially "evil" applications.
In terms of radio/carrier network access all phones still use RIL (Radio Interface Layer) to communicate with the business end of the device which is *not* linux or open source so there is little to fear in terms of carrier radio interop.
Whether or not the N900 reaches iPhone numbers is irrelevant to the fact that it will stand in computer history along with the Kaypro II, PDP-11, SORD IS-11, Altair 8080;
I don't care if AT&T likes it or not.
If you actually get your hands on one, you will understand that it feels good to actually own something, and not pay to carry the wireless equivalent of a cable box.
If people in America were "customers" and actually were allowed to decided what they wanted, and not "consumers" to be culled by the wireless carriers, then the N900 would on it's merits be the best selling mobile computer of all times.
Does anyone really like the fact that all you can get from the big wireless carriers is what they want you to have, and not what you want?
Those that go out and buy an N900 will understand.
* Carthago Delenda Est *
There are many reasons to lock shit down.
Fear of teh hax0rs taking down a tower.
Fear of pirates sucking up your bandwidth, and getting all your apps for free.
Fear of zealots circumventing traditional pay schemes by getting voice, data, and other services off network (and thus free).
Fear of the russian mob using the phone hardware to spy on or disrupt other people's communications.
Fear of lawsuits when it gets out that you illegally used copyrighted shit when making the phone's os image.
Fear of people finding out that you rig the fucking battery display to show higher than it is, or that you rig the reception indicator to show full bars when it shouldn't...until you make a call.
Fear of Bob deciding to take his shiny new toy to another network.
While virtually ALL of the reasons center around the company being afraid of people exploiting the company's stupidity, they are still valid concerns - the companies are stupid.
However, TFA is completely incorrect. Companies don't fear the unknown - they know EXACTLY what we'd do with open phones.
No.
It's because they cost hundreds of dollars.
I want an open source phone, I really do, but I can't justify spending 500 on little more than a PDA + phone. I already had a PDA once, hardly used it, and phones that just work as phones are less than a hundred these days. Make an open source phone that's a reasonable price and I'll buy it.
They fail in the mainstream market because there's such a small market for them. The Nokia n900 is a geek's dream, but most people want a phone, not a handheld computer. Most as in 99.99% of the marketplace. And even fewer want a multi-hundred dollar handheld computer/phone. So I'm sure it sells well in the market it was designed for...that .001% of the population that wants a hackable, programmable micro computer that makes calls. So it succeeds where its market is. Saying it fails is like saying the Audi R8 supercar failed. Though, at least that made it into Iron Man.
You could say the iPhone is a failure as well: it only has 1% of the cell phone market. But I think most of the U.S. will disagree with that statement.
There is no control-freak network provider to blame there. Why doesn't open source take over then?
There is a scapegoat for every problem. Microsoft, vendor lock-ins, corporations, bad managers, bad employees, government, society, temporary insanity, depression, depression medication, education, teachers, family, finances...
Not to say we can't perhaps put our finger on real problems that prevent open source from "taking over," but just saying that one can reason and argue for a whole lot of perceived problems that may not actually be the reason.
The Iphone ecosystem is a good example. An example of a phone where I'll have to install anti-virus for my relatives and make sure they are up to date on patches, otherwise their phone will get owned and I'll have to waste a weekend fixing it.
Let's not go there.
The summary almost hints that there do exist popular phone platforms which, while not open source, certainly allowed for quite open development and modification by users for a long time. Many Nokia phones for example.
But I've heard that US carriers didn't really want to offer them in unlocked state, and Nokia wouldn't castrate its products; so the carriers went with RAZR... (and look where Motorola is now)
So this really seems like your local problem. Since Nokia almost completed open sourcing of Symbian and more than 50% of smartphones run that OS, I'd even say that the article is quite irrelevant on the larger scale.
One that hath name thou can not otter
Then do it.
Get the FCC approve your devices for use.
Get any sort of decent battery life out of a mesh network with no towers while still maintaining access to the PSTN and emergency services.
Sell the device at a profit.
It's so easy why didn't I think of it?
They don't. You can write native code for Android phones. You just need a small java wrapper nothing more.
"Truly open-development, open-source phones like the Nokia N900..."
are you kidding me???
what is "Truly open-development, open-source" about a platform that has
* proprietary power management (bme)
* no docs for the gsm modem interface (and no source code for the apps using it)
* proprietary powervr graphics drivers
* proprietary osso-dsp-modules
read also:
https://bugs.maemo.org/show_bug.cgi?id=1584
http://wiki.maemo.org/Why_the_closed_packages
i'm not so much pissed by proprietary applications as i can replace the rootfs by a free and open source one what pisses me off is the undocumented hardware used and lacking communication with upstream kernel development.
dont call this device "truly open"-blah... it is definitely NOT.
there are a few devices that strive to be as open as a linux phone should be:
openmoko tried and indeed even though the calypso is undocumented they provided a implementation of how to interface it and thanks to it one can use all of its hardware without binary blobs - NOT POSSIBLE ON THE N900!!!
then there is the FLOW by gizmoforyou which uses a gumstix overo as the base and added a telit modem for which you can download the FULL DOCS from their website - hey guys at nokia, this is the kind of modem you should have picked if you wanted your device to be called "truly open"!
the modem used in the n900 uses ISI for which no reference interpretation in oss exists.
is it only me or did the slashdot crowd forget what "truly open" means and is now all over a device that is open on the top but not if one wants to really start messing around with it?
Heh, I don't think you realize what "two times lower population density" in Finland means in context of cellphone carrier.
It means much higher costs per customer. A need for more infrastructure just to cover vast, almost empty areas.
And they still have better service % lower prices. Heck, they even passed a law defining fast broadband access as a right... (and, no doubt, large part of it will be provided wirelessly
One that hath name thou can not otter
>
Get any sort of decent battery life out of a mesh network with no towers while still maintaining access to the PSTN and emergency services.
If it isn't a "public" phone system but was more like a large voip network they might be able to find loopholes.
Anything can be found funny, from a certain point of view.
This is not my position, this is reality...
Worse economic position (at least when they were starting to invest in their communication network), much more costly to build and operate...and they still beat you. By a long shot
But hey, I see where you're coming from; "bad, commy" gov interventions, regulated market, etc. (why do people like you can't get over the idea that governments are simply a reflection of...society itself? If the latter seems to be functioning decently, so will the former)
One that hath name thou can not otter
One of the stipulations that Google managed to have placed in the FCC license for commercial 4G LTE spectrum is open device access, which is absent in current wireless spectrum licenses. They did this by getting approval for a clause that if a certain minimum bid for the spectrum was met, that that open device access rule would go into effect, then they bid that amount, and then proceeded to let Verizon outbid them, ensuring that clause would go into effect. Carriers may have been able to get away with this type of draconian control over their networks in the past, but it seems that's coming to an end with the shift to 4G LTE already underway. With this open device access regulation, actual user-accessible open source handsets may finally be able to see widespread use.
"I like systems, their application excepted", George Sand (French)
/puts of flame proof long johns/ You want to know why Linux hasn't had a snowball's chance in hell at retail? It is actually quite simple: You can't shop for Linux devices at Walmart without playing the paperweight roulette, which scares the living hell out of consumers!
If you really want Linux to have that critical breakout, then get the heads of all the major distros together, have them shake the living hell out of Linus and the other kernel devs, and nobody is allowed to leave the building until an agreed upon standard is written and approved to where you can just put a "Linux 32/64" folder on the driver CD and be done with it!
There are PLENTY of shops like mine who would LOVE to sell Linux machines, there are plenty of mainstream customers that could use Linux security, but I can't sell it and they won't buy it. Why? Because you can't answer these questions-which wireless USB cards on sale at Best Buy work in distro foo? Printer? Sound cards? Can you give me a 100% guarantee that my customers can shop at Walmart/Best Buy/Staples and have ZERO chance of getting a paperweight? You can't, because Linux and the driver situation is all fucked up. The kernel developers should be worried about the kernel and NOT maintaining fricking printer drivers!
With Windows I can say "see this pretty little flag on the box? See how it says "certified for Windows 7"? Yeah, that's you. Just look for that and you are good". It takes a customer all of 5 seconds to look at the box and shop with confidence. same thing with OSX, just look for the little Apple and the "10.whatever" and if it lines up with what you got? Hooray, you're all set to go. With Linux you get the "fun" of trawling forums before you can even buy a damned thing (which if you believe mainstream customers are gonna research before purchase I got some swampland in AR to sell you) and Deity help you if the "driver" which usually needs some serious fricking tweaking and CLI foo to get going was written for firmware A and you got firmware F, because guess what? Enjoy your paperweight!
Just make it simple guys. Remember KISS? Make it so hardware manufacturers can put Linux drivers on the CD and a penguin on the box without having to keep an assload of driver developers on hand just to try to keep up with the shifting sand that is Linux right now. Make it so ANY customer WITHOUT needing to do research or put in a metric crapload of CLI commands can simply walk into Best Buy and put a device in their cart and know 100% that it will work on Linux. If the "inferior" Windows and OSx can do that, then surely you guys can too...right?
ACs don't waste your time replying, your posts are never seen by me.
What I meant by the kernel was the kernel should have the basic I/O-file systems, networking, etc and that the drivers for device foo belong in the ABI. And as for "windows drivers causing BSODs"? Yeah, that was true in 1998, but working PC repair I have had to deal with some seriously shitty hardware and since XP came out finding a BSOD due to 3rd party drivers is EXTREMELY rare. 99 times out of 100 the device simply doesn't work. And in 99 out of 100 of those cases a simple uninstall/reinstall fixes it right up.
While I think your idea of a "retailLinux" is intriguing, and if I didn't suck ass at website design I have some ideas that could mitigate it somewhat, I think ultimately in the end Linux is doomed to stay just where it is at, and it is NOT because of some conspiracy. sadly, it all comes down to politics and the "SCoN!" (Source Code or Nothing!) brigade. They will NEVER allow a stable ABI, or an easy way to just slap drivers on a CD and ship it, because "Gasp!" you might actually get a few vendors that don't release their source code for RMS to rummage through.
To see how militant the SCoN! brigade is, just look at the Anti-TiVo clause in GPL V3. Here you have the defacto leader of the GNU movement rewriting sections to specifically target a SINGLE company he doesn't like, and lets be honest here, okay? If the TiVo was easily "hacked" to run unsigned source code, how long do you think it would be before the net was flooded with "Free TiVo!" code and/or easy to use ways to copy any and all content off of said TiVo? I would say about a week, and TiVo would go bye bye, but RMS don't care about TiVo or any other busines for that matter.
And THAT is why ultimately Linux is doomed to a niche at retail. It will cost serious money for advertising, fixing the problems I outlined in my previous post, making inroads with retail stores like Walmart, all that takes money. I'm sure there are plenty of companies that would be happy to meet with RMS and try to come to some sort of compromise, but to the SCoN! there is NO compromise, ever. To them GNU is NOT an OS, but a religion, a way of life if you will. They do NOT care if it ever becomes more than a niche, as long as their beliefs are upheld.
And to ultimately make serious inroads on the desktop you will HAVE TO make shopping for Linux as easy as shopping for Win7 and OSX, which means you HAVE TO be able to put drivers on CDs and penguins on the box. But because the SCoN! has so much power within Linux that will simply never happen, because then companies might be able to release binary drivers like nVidia does, but without having to blow serious mountains of cash like nVidia does on driver development. The SCoN! would rather things be hard and Linux be a niche than to compromise, now or ever. That is the problem with zealotry, it always gets more extreme, never less.
And I apologize for the length of this rant, but I have been hoping and trying different distros and waiting for Linux to finally get to where I can sell it without going bankrupt since WinME came out and stank up the joint (remember ME? Shudder) but every time the after sale support ends up costing me MORE than a Windows license, and in the end it is all because of paperweight roulette.
What Linux needs is guys like me, the mom & pop shops, to sell your OS and make getting support for it as easy as dropping a windows box off at Worst Buy. And we LIKE your OS, as none of us enjoy cleaning porn bugs off an infected Windows box. But until I can sell a Linux PC without having my gut tied in knots waiting on the customer to get pissed at me because they lost playing paperweight roulette it just ain't worth the pain. Sorry, no sale.
ACs don't waste your time replying, your posts are never seen by me.
... so you're one of the guys screwing up communications on the amateur bands, just for your fun. Thanks. Thanks a lot. And thanks for caring about someone other than yourself. Would you corrupt others' Internet communications as readily?
(n.b.: This type of illegal CB operation is especially bad because the illegal "channels" used are in the portion of the amateur 10m band used for international narrowband, weak-signal work -- usually in Morse code, and often at the threshold of audibility in a 250 Hz bandwidth. Since the transmission modes were different, the illegal operators often can not hear the communications they are disrupting; further, since the "freebanders" use wider, single sideband transmissions, a single illegal transmission can interfere with dozens of narrowband signals at once. Since this band is capable of worldwide communication at certain points in the sunspot cycle, the interference can quite literally be global in nature.)
By the way, the world has changed. In the UK, an amateur radio licence is now free, valid for the lifetime of the user, and available online. If you're worried about the licence examination (but you're a geek, so technical matters are no problem for you -- right?) there are clubs that will hire the room, give you the study book, and teach you the exam material, all for £45. So if you want to talk to the world, why not just follow existing international standards and agreements, and get an amateur radio license?