A truly open Linux phone with GPS debuts

location based services

THERE HAVE BEEN a lot of phones claiming to be 'Linux phones' and those that do run a Linux kernel, but they all miss the point of Linux: to be open. FIC is about to change that in a big way with a truly open phone, the OpenMoko.
This is the first phone in a long time to get us really interested in what it is, what it isn't, and the philosophy behind it. The philosophy is the thing that makes Linux great, and not in the sense of window-dressing or half-hearted openness, it is really open. It runs the latest kernel, 2.6.18 as of a few weeks ago, and you can get software from a repository with apt-get.
The OpenMoko is meant to be a fully mobile Linux machine that happens to look like a phone. The point is simple, where others have a Linux kernel with a locked proprietary stack on top of it, this one is open, top to bottom. You can use your own tools on it, compile your own kernel. and bang on the bare metal if you are into that sort of thing. Everything barring a few small drivers is GPL'ed.
There is no vendor lock-in, the only proprietary parts are the radio and the GPS unit, but they are fully exposed. The radio must remain closed for legal reasons, but it is available as 16 virtual serial ports that you can do whatever you want with. You hit the GSM module with a similar set of commands. If you like the software they include, great, use it, tweak it and have fun. If not, write your own.
If you want your phone to do something it does not, you can get the software you want from any repository that you can reach on the net. FIC will have a certified one with apps that will work but you are more than welcome to point it at any other you chose, or even to make your own. Remember, this is all about being open.
Let's step back and look at what the phone itself is before we get into the software that runs on it. The hardware itself is a Samsung 2410 266MHz ARM9 with a 2.8-inch VGA touch screen. There are only two buttons on the phone, the rest is handled by the touch screen, a microSD slot, Bluetooth 2.0, and USB for connectivity and charging. It also has two 1W stereo speakers so you can repurpose it to an MP3 player or anything else you would like.
The initial version will come with 128MB of flash and 128MB of DRAM. There is the potential for a version with 1G of flash, but with a slot, do you really need it? OpenMoko comes with a 12mw battery for somewhere around three hours of talk time, but there will undoubtedly be more options if it takes off. It also has a Globallocator GPS unit and the phone bits are TI quad band GSM. The only thing lacking is Wi-Fi and that is planned for the next gen hardware.
What does it look like? Sleek and simple, something I am a fan of, and really oval, something I am not nearly as sure about. Either way it is nothing objectionable, and clean is usually better than fussy.

With the hardware out of the way, that brings us back to the software, philosophy and the question of why. Why is the most important one, why would any vendor want to adopt a commodity part that any of their competitors can pick up just as easily? Why would you want to write for this? Why would you want to buy it?
For vendors it is simple, lowered cost. Each generation of cell phone is mostly new, there are huge software costs to write the most basic things for any new phone. In six months it is replaced again, and the cycle repeats. It is a huge pain for vendors of all sorts, not to mention expensive.
If the hardware was common, the vendors could reuse the software and concentrate on what makes them money, fashion and services. Fashion is the simple one, phones sell on the skin, not nearly so much on what is below it. Common hardware lets vendors concentrate time and money on that part, wrapping it around the commodity electronics.
The services market is much more important to the carriers, that is where they make the do or die margins. Ringtones bring in billions of dollars, and for every new phone you have to write a ringtone downloader, manager, player and billing system. When the hardware is outdated in 6 months, you do it all again. If you stop and think about it, you need to do this for every bit of functionality, a daunting task at best.
The OpenMoko is aimed directly at this problem. You write it all for Linux once, and then you recompile for every new phone. If the APIs stay constant you may not even have to do that. The vendors end up putting their dev dollars where it makes them money and differentiates them from their competition, not toward reinventing the wheel.

What about the community? Why would they care? Cheap open hardware in a market where it doesn't exist now, what more do you need? It is open from top to bottom with only a few drivers here and there closed, and it runs the latest tools, frameworks, and software. If you don't like what is there, you can roll your own.
This phone could very well be a hackers paradise. There is a full package manager, so if you want a web server, go get Apache. If you want mapping software for the GPS, you type apt-get and off you go. Games? Sure. Services? Sure.
As an end user, the appeal is obvious, and I don't just mean a quad band GPS phone with tons of accessories and a GPS for $350. It is open and not locked down, you can make it your own and get what you want. Instead of the carriers dictating, they can offer, and if they are the best, they will get your money. If not, you can chose another repository and off you go.

On top of downloading apps, the repositories and the app manager can keep things up to date for you. Not only can you download the data, you can also set it to patch or upgrade all of your wares for you automatically. If you download an IM client for mom, set it to stay up to date and you won't have to be her tech support nearly as often.
Before you cringe at the $350 number, it comes with a full range of software and is a fully functional phone that you can modify. One of the more interesting concepts is the text communications app. It combines SMS and email into one program because from a phone, is there really a difference? Write your message, pick a recipient and send. If you don't like it, go get another one, or write your own, see the point of open now?
It also comes with accessories like a car mounting kit, straps, chargers and everything you need in the box. There is none of the buy it and get a doorstop until you spend another $200 on branded bits, this is a complete kit.
The concept of OpenMoko was driven by Sean Moss-Pultz, a self confessed geek. His idea was to create an ecosystem similar to that of desktop Linux for the mobile community. The more people that have one and write for it, the better it will be for everyone.
His bosses at FIC want to make and sell lots of hardware, so a community producing killer apps is just the thing for them. After a little persuading the OpenMoko was born, and in my opinion, it has huge potential. There are not really any downsides for any of the players, open means just that. If you can make a better mousetrap, you are more than welcome to. If FIC can ferment a meritocracy, they will have accomplished their goals.
The initial run of OpenMokos will be small and out soon, this month in fact. They are meant to be seeded to developers and those wanting to port apps before the mass market launch in January. If you are one of those wanting to get your feet wet, contact FIC for the details.
So there is now an open GPL phone that won't say no to you. It has GPS for location and is reasonably priced. OpenMoko could be good for all the players from users to carriers, and if it takes off, it's usefulness could grow exponentially. I want one. µ