OpenMoko: 10 Years After

http://www.vanille.de/openmoko-10-years-after-mickeys-story/

by Kostic • 9 years ago

215 68 9 years ago

rosser• 9 years ago

I still have my Freerunner in a box somewhere. So much promise...

I keep wondering if there's something I could do with it. At this point, though, it's probably more a historical footnote than anything.

DanBC• 9 years ago

Here's a small discussion from 10 years ago: https://news.ycombinator.com/item?id=14388

msla• 9 years ago

Since the main page is a bit slow, here's a mirror:

https://archive.is/8wYHe

eagsalazar22• 9 years ago

This is a project I followed pretty closely at the time and I always felt it was doomed. There were probably 100 reasons it failed that I wasn't aware of but it seemed at the time like people really wanted the open software and platform but OpenMoki got religious about hardware which no one really wanted or cared about. Ultimately this was the route Android went and we see the results.

It is classic for FOSS wonks (I am one!) to forget that most people don't care about the principles of FOSS, they just care about the practical implications of FOSS which aren't always the same thing.

wolfgke3• 9 years ago

I see a contradiction in your arguments:

it seemed at the time like people really wanted the open software and platform

vs.

It is classic for FOSS wonks (I am one!) to forget that most people don't care about the principles of FOSS

eagsalazar22• 9 years ago

How is that a contradiction? My point is that open software has large and immediate positive implications for end users but hardware does not.

Qwertious• 9 years ago

There's the community and then there's paying customers. The former want open software and platform, the latter want a functioning phone. They're both "people", but they're not necessarily the same "people".

problems• 9 years ago

Yeah, Android put the effort into software and let manufacturers deal with the hardware.

It's sad that their approach failed though, I'd really like a modern device with something like OsmocomBB available right about now...

wwweston• 9 years ago

Doesn't close on a bright note:

However, the sad truth is that it looks like there is no business case anymore for a truly open platform based on custom-designed hardware, since people refuse to spend extra money for tweakability, freedom, and security. Despite us living in times where privacy is massively endangered. If anyone out there thinks different and plans a project, please holler and get me on board!

notspanishflu2• 9 years ago

I've hope projects like Halium, UBports, Plasma Mobile, etc end up bringing an open hardware project.

vedanta2• 9 years ago

Yeah, it seems like mobile oses are about as common as Linux distros in 1995. Sailfish, Maru, lineage, among the two you mention. If you have a nexus 5 or one plus one, you're in luck.

I recently tried lineage, ubports and plasma, and though I liked ubports, lineage was a breath of fresh air. Selinux, opengapps, regular privacy notifications, twrp, and privacy and security seem to have not taken a back seat in general. Sign me up.

If there hadn't been contracts 10yrs ago, I probably would have gotten an openmoko. Wanted one badly. If lineage, ubports, and plasma among others can keep developing alternatives to stock android, no reason to see why at least some or all hardware doesn't follow. It's taken AMD the better part of 10 years to have open source drivers that are as good as their closed source drivers on almost all of their GPUs if I'm not mistaken.

lallysingh• 9 years ago

No, that wasn't it. I had the 2nd-gen openmoko phone. It just didn't work. You won't get users until you have something that /does something/. I never even got a phone call made with it. The system didn't have enough of a working spine to make it worth investing time or energy into to get it further.

If you want an open source project to work, it has to already work and then have easy places where people can start making changes. Otherwise, they don't know how much work is needed to get anything done, and don't want to risk weeks of work for minimal, if any, results.

carussell2• 9 years ago

This is one of my most frequent criticisms about open source software. jwz wrote about it almost 20 years ago. (It can even apply to highly polished software like Android, or Signal, or Firefox today, that for one reason or another make things really difficult to work with—if you can work with them at all.)

People only really contribute when they get something out of it. When someone is first beginning to contribute, they especially need to see some kind of payback, some kind of positive reinforcement, right away. For example, if someone were running a web browser, then stopped, added a simple new command to the source, recompiled, and had that same web browser plus their addition, they would be motivated to do this again, and possibly to tackle even larger projects.
We never got there. We never distributed the source code to a working web browser, more importantly, to the web browser that people were actually using. We didn't release the source code to the most-previous-release of Netscape Navigator: instead, we released what we had at the time, which had a number of incomplete features, and lots and lots of bugs. And of course we weren't able to release any Java or crypto code at all.
What we released was a large pile of interesting code, but it didn't much resemble something you could actually use.

"nomo zilla: resignation and postmortem". 1999. https://www.jwz.org/gruntle/nomo.html

digi_owl3• 9 years ago

I think it may be a bit premature.

Just look at what people are doing with RPis and various radio boards.

There was one such project posted on HN recently that had a RPi Zero W hooked up to a touch screen and a mobile radio. All battery powered.

In a sense the Openmoko was too early and too bespoke. Thus driving the cost of entry for would be tinkerers through the roof.

potatolicious• 9 years ago

I think it goes to the other poster's point: "You won't get users until you have something that /does something/"

The RPi was ready to roll out of the gate - they had a working Linux distro from the get-go that actually worked. They also provided really simple instructions on how to get said Linux distro onto a SD card. You can boot the RPi and have it be useful within a matter of minutes.

Openmoko simply never got near enough to working where the community felt like picking it up and running with it.

I don't think Openmoko was premature - it was incomplete. You're not going to get much participation from the community if doing anything useful on your platform first involves fixing/finishing the platform.

petecox2• 9 years ago

The Pi zero phone is a cool hack but I doubt many would opt to use it as a 'daily driver'.

Broadcom, according to the wikipedia page for VideoCore, did however make a couple of Cortex-A9/A7 SoCs that made their way into low end 2012-3 era Android phones.

Perhaps a startup might recycle old Broadcom-based HTC and Samsung models in collaboration with the RPi team.

throw7• 9 years ago

I could be wrong, but my feeling at the time was there was a lack of experience of a strong low-level hardware designer. I remember bug #1024... all I really wanted to use my freerunner daily was making/receiving calls (you know basic phone stuff), something flip phones at the time were absolutely rock-solid at.

paulborza• 9 years ago

I was one of the Google Summer of Code "interns" for OpenMoko back in 2008. I was so excited to get the free OpenMoko device and actually implemented gesture recognition and screen orientation for the device part of the GSoC project. Here's a video of it https://www.youtube.com/watch?v=K2S2rQUETwc

But let me tell you about the device:

1. You couldn't even make a phone call with it. There was a hardware bug that stopped it from making phone calls. It was basically a PDA...

2. If the battery ever ran out, you couldn't start the phone anymore. You had to have a spare Nokia phone that you'd use to charge that battery. Thank god they were using Nokia-compatible batteries. The gist was, never ever to let the phone's battery reach 0%.

3. The development of any UX components was cumbersome. They always insisted in supporting GTK and another platform which I forgot its name. They were such a small team and yet they were building two projects which essentially were doing the same thing. Why?!

Those three were just a few problems. The Freerunner was definitely not a user-friendly phone.

joshumax• 9 years ago

I had issues making calls on my Freerunner, sure, and it certainly wasn't very user friendly (even flashing replicant to it resulted in a very buggy daily driver) but I did enjoy the phone for what it was; a fun platform to hack around with and one of the first (and perhaps almost the only?) mobile device with hackability not only in mind, but centric to its design.

I hope alternatives like the Neo900 make it off the ground, but I'm not holding my breath. Most people just don't seem to care about open hardware when they could be getting the latest, fastest SoC for their $varPhone. Maybe it makes me weird that I'd rather use an open device with a single ARM core at 800Mhz than a closed one that uses the Snapdragon 835...

narag2• 9 years ago

Maybe it makes me weird that I'd rather use an open device with a single ARM core...

IMHO, it's neither a question of being weird nor about the power of the phone. The problem is that choosing an architecture that isn't mass produced, you are doomed to disappear no matter what.

I wouldn't care to use an underpowered phone, provided that it's cheap and widely available, being "cheap" an (important) part of "widely available". A hackable phone would generate interesting apps for a wide market. But you need that there are a lot of people that can buy it and has some problem accessing an Android and its apps.

mattl2• 9 years ago

I worked at the Free Software Foundation at the time. We were lucky to know someone at MIT with access to the tools to fix the tiny hardware bug by soldering something near the SD card slot IIRC.

We also had multiple batteries and multiple beefy micro USB chargers (used later on our G1s) as well as external Nokia battery chargers around the place.

I think I managed to take a call once. I could never send SMS, but I would play a lot of numptyphysics on the MBTA with the combination ballpoint pen/mechanical pencil/stylus I bought.

Sean visited us in the office a couple of times beforehand. I remember showing him some of the HTML tricks Apple were doing for Mobile Safari with viewports and fullscreen things, I don't think anyone ever implemented them. This was pre-App Store too, I think.

I remember being so wary that someone would steal my awesome new phone at first, but that quickly wore off and I went back to my Nokia flip phone until Google had developer G1s (ADP1) with a fancy backplate on them.

narag2• 9 years ago

Somewhat related: I was a little interested in Firefox OS. When I tried to get a device I found that I should either be in a wait list for months, buy a "more or less compatible" phone (it might work or it might not) or get a ZTE (brand that I hadn't heard of until then) with lackluster features at a premium price, compared to a regular Samsung.

Being an early adopter is not for everyone, for sure. But being the only adopter is definitely a PITA :)

emidln• 9 years ago

I had a FreeRunner that I used as a daily phone for a few months (until I got a TMobile G1). It most certainly could make phone calls. IIRC, I used a Qt-based environment that was basically a PDA + a dialer. It even had a webkit-based browser. I remember buying a Nokia wall charger so I could charge a second (and third) battery given that the OpenMoko consumed batteries quickly.

My most memorable moment with the FreeRunner was when I finally received my G1. The screen on the G1 was so much brighter, the phone itself was far more responsive, and it had EDGE data service (my area didn't have 3G). I never bothered fixing the usb port that I broke because the G1 was so much better.

cesarb2• 9 years ago

I remember buying a Nokia wall charger so I could charge a second (and third) battery given that the OpenMoko consumed batteries quickly.

If I recall correctly, the OpenMoko didn't have working CPU frequency management; it was always running at the maximum CPU frequency.

I actually wrote something to manage the CPU frequency for the GTA01, but never pushed it upstream; by then, the GTA01 was a dead end, and the GTA02 used IIRC a different chip.

Edit: for the curious, here's the code: http://repo.or.cz/linux-2.6/s3c2410-cpufreq.git/refs

mikepurvis• 9 years ago

I'm grateful to OpenMoko every time I fire up dfu-util to flash something over USB.

lamby• 9 years ago

Ah, nostalgia. I remember attending a big OpenMoko announcement as FOSDEM. Naturally, it being FOSDEM, it was allocated a tiny room which was completely and utterly packed. g

phaedrus• 9 years ago

I bought a Neo Freerunner when it came out. I tried make it my daily phone, and invested a lot of time hacking on it - in the end a wasted effort. There were show-stopper bugs and face-palm-worthy design missteps at every level of the stack.

Its developer community was fragmented among a half-dozen different distros from birth; Openmoko themselves forked their own distro so that even if you wanted to run the "official" one, you had two incompatible choices - and neither the two "official" distros, nor any of the user-developed ones, supported all of the hardware + apps necessary to actually use this thing as a phone. (I.e. one might have audio but not support making calls; another might support making calls but no audio volume; a third might lose all your contacts, etc.) If we could have taken the union of features from all these distros, it could have had a great smartphone OS; that just goes to show what gratuitous fragmentation cost us.

As for hardware, we were screwed from both sides by the hardware companies (flaky drivers) and Openmoko's own circuit design. Due to driver/card incompatibility, I never found a high capacity SD card which would work completely reliably in the phone; only the cramped 512MB SD card which came with the phone worked reliably. Mind you most of the time the problem would show up after you spent a few hours compiling & setting up Gentoo or whatever on a new 8GB SD card only to have the 3rd boot fail with a corrupted card.

And in the circuit design side, from the specific problems the phone's board had, I get the feeling Openmoko's engineers were computer engineers but not very knowledgeable in audio electronics. They failed to properly separate and route the audio ground, nor to give a good range of audio volumes, and connections for the mixer chip were a mess. Actually, electrical noise was a problem in other areas, as well; accessing the SD card generated noise or RF leakage which would cause the GPS unit to lose it's lock. (So good luck loading a map while navigating.) Interference in bandwidth (albeit just in contention for buses, not actually noise) was also responsible for the limitation of not being able to use the SD card and talk to the video processor at full rate at the same time - so playing video from the card was not workable, either.

I had some fun with the device, learned a little bit, and ultimately rebuilt it into a larger case with a larger battery and an amplifier for the earpiece. However, I would have been better served if I had done all these things on a development board, rather than being tricked into buying a prototype development board shaped like a phone. The experience led to me "missing the boat" on getting into mobile development; it took a long time for me to get another smartphone, because I had blown my money and my efforts on the freerunner. For the next couple years whenever I would consider getting another smartphone my then-wife (now ex) would counter "yeah that's what you said when you wanted to buy the freerunner - and you have that and you don't even use it!"

vanous• 9 years ago

Thank you for the summary, it perfectly pictures my experience too. I had spent do much time and effort (did the audio cap fix, baseband update and few more)... So much that even now phone calls with my brother (also freerunner ex-user) still start with "do you have an echo".

I used freerunner for quite some time as secondary phone, developed an app or two in python/enlightenment (efl), wrote much of the SHR wiki manual and ended up not even considering Firefox phone or Ubuntu phone at all...

Rjevski3• 9 years ago

As usual here's another victim of the free software problem - too much emphasis on freedom and not enough on actual functionality.

I don't care how free your phone is, if it can't make phone calls I don't want it, nor does anyone else.

sargun• 9 years ago

OpenMoko holds a special place in my heart. It was the first project that I contributed to in any meaningful fashion. I wrote a couple patches to the Python implementation of FSO. It taught me a ton about how GSM, embedded devices, and power saving worked. I used it as an IoT device, long before IoT was a common thing.

Not many people know this, but OpenMoko was largely rolled by Dash Navigation -- https://en.wikipedia.org/wiki/Openmoko#Dash_Express. This gave Sean and Co. the cash and ability to develop the OpenMoko, but unfortunately, Dash failed, because as a CNAV device, it was too little, too late as compared to the ultra-converged cell phone. I wish that they would have found other entities to back the project, but the market just wasn't there yet.

feborges• 9 years ago

I remember John "Maddog" Hall speaking about it in Brazil. I was just a kid at the time.

mattl2• 9 years ago

Yeah, Maddog had something he was running that was supporting them with Android.

https://en.wikipedia.org/wiki/Koolu

k__• 9 years ago

I remember 2007, when I started my bachelors degree some senior students had one and were so proud.

Half a year later, nobody talked about it anymore and it was iPhone all the way.

acidtrucks• 9 years ago

I've always wanted to run openmoko. If I could run it on modern phones and replace android, I totally would.

nickpsecurity• 9 years ago

Has anyone tried to approach the Shenzhen companies doing quality knockoffs of iPhones and Galaxies about doing a FOSS phone with OpenMoko features? I mean, they clearly can build one. Question is whether they would for a reasonable amount of money.

kasbah• 9 years ago

The problem is these would be gonkai phones (see link). Within the western legal framework of "intellectual property" you can't open source what you don't own. The impedance miss-match between the two working methodologies makes any conversion a lot of work, e.g. the Fernvale effort to make a open hardware _dumb_ phone platform: https://www.bunniestudios.com/blog/?p=4297

odammit• 9 years ago

I bought one of these way back when. It was the model with the hole in it for hooking to a lanyard or something else absurd...

It was so dumb and jittery I microwaved it a little and then swatted it off my roof with a tennis racket.

Was fun, would destroy again 10/10

scarhill• 9 years ago

I also still have my Freerunner in a drawer somewhere. I bought the phone early on and then ended up starting the android-on-freerunner project[1], when Koolu, the company that had begun an Android port, abandoned the project. I used the Cupcake version as a daily driver for quite a while, but it was never anywhere near as stable as a "real" phone.

1 - https://gitlab.com/android-on-freerunner/android-on-freerunn...

uberneo• 9 years ago

I still remember how happy i was holding the first Opensource Linux phone in my hands back in 2008 . It was fun to compile a basic c program using gcc . I remember my first hack of my life #1024 http://neofundas.blogspot.ie/2009/09/1024-hardware-fixdeep-s... Also remember when at 6.30 AM in morning you have to type "ps -ef | grep -i alarm" , find the process id and then kill -9 id to stop the alarm and that too using the stylus. Definitely these commands will make sure that you will wake up 100% , a perfect alarm :)

Nursie2• 9 years ago

I loved my freerunner but the article hit the nail on the head - there seemed to be a whole bunch of half-finished software stacks for it, rather than ome complete one.

The next one, which was definitely going to be the real future of Openmoko, was always a few months away, and as soon as it got close you'd find out the team had ditched the framework and were already working on the next next one.

As a result, as someone who had wanted to jump in and start experimenting with writing application code, there wasn't really a platform to do that on. So after a few months of frustratingly dropped calls, or silent calls and unsent text messages, I went out and bought a small, shiny feature phone and didn't look at another smartphone until the nokia N900 showed up.

digi_owl3• 9 years ago

That seems to be an ongoing problem with FOSS.

Because nobody is there to crack whips about quality targets etc, devs keep wondering off path as they run out of shiny features to implement, or find themselves going "I could do X if only...", and thus a new codebase is birthed...

majewsky• 9 years ago

Or rather, it makes you wonder if the devs ever used the software themselves. In my experience, above all else, dogfooding is what gets software into a productive state.

Nursie2• 9 years ago

The major problem in this case (IMHO) was that there was someone to crack the whip, and they didn't do it!

The project was run by a company who had employees and at least one manager, but nobody was setting a direction for those employees to move in, and making the hard decisions (like "no, we're sticking with this stack to get something stable and production ready, even if it is flawed").

btcnnewsdaily• 9 years ago

I remember these, definitely had it's unique features about them at the time.

toyg• 9 years ago

I'm genuinely surprised to find out, in this thread, that so many people actually bought an Openmoko device. I only remember it being a perennial work-in-progress, and when they eventually started actually selling something, it was so painfully obsolete that I couldn't get myself to consider it (unlike the N900).

cesarb2• 9 years ago

It's sampling bias: the sort of people who would buy an Openmoko device even before it has working software are also the sort of people who would browse HN regularly.

wink• 9 years ago

Or even more so, comment on this thread.

I wasn't going to, because I didn't have one - but one of my coworkers at the time had one, and we absolutely made fun of him that he had to first get a new kernel to be able to either send a text message or make a call (don't remember which of the two, the other one didn't work at all).

OpenMoko: 10 Years After

Related Stories