The roadmap/goals for post-MozFest 2014 pique my interest. The following especially:
"Users will be able to read from sensors and control motors, LEDs, solenoids, slave boards, etc"
That will make it easier for people to start playing with electronics from the browser on their Raspberry Pi. Chrome has a serial API, which I think is only accessible to chrome extensions - https://developer.chrome.com/apps/serial. If Firefox OS on the Pi comes with no-limits unlike Chrome, then it should be fun to play with.
Very cool that more operating systems are supporting the Raspberry Pi. However I think targeting the Pi to completely run a quad copter/drone is a bad idea. Unless Firefox OS has some special realtime control extensions that allow you to stop the kernel from pre-empting program execution, putting a mission critical control loop in a user space program is going to cause problems.
I like the idea. I AM confused though.
Currently Fx OS doesn't provide a decent (imap) mail client and seems to have issues with services/background stuff (think.. IM client).
If things like the ones described in the post are the goal (media server, low level hardware access), does that mean modifications to the core system? Is - and that's my real question - the lack of the mentioned features merely "not yet written" vs "impossible to do right now and blocking on core/system changes"?
it's definitely in the not yet written category. Background services is something hard to design right and we don't think any of the other mobile operating systems got it right (think: battery, data plan drain, privacy, security, user control).
There are multiple ideas on how to approach that as more and more systems in FxOS would benefit from it (background music player, IM, various synchronization services), and I expect us to come with first approach relatively soon.
by focusing on old, cheap and slow hardware, they're losing out on 3rd-party dev talent that would be interested to run FxOS on more modern hardware and letting the optimization for old devices trickle down later. i would love to easily run it on my BeagleBone Black and contribute apps i make back to their market.
> by focusing on old, cheap and slow hardware, they're losing out on 3rd-party dev talent that would be interested to run FxOS on more modern hardware and letting the optimization for old devices trickle down later.
By focusing on old, cheap, and slow hardware they ensure that the optimization will happen at all.
Cool! Please can you buy me one of those expensive computers so I can start now?
Here the thing, if you start focusing on the cutting edge, that's where development tends to stay. By starting with limited resources, it means that the software becomes more powerful as hardware gets cheaper.
hmm.....strange that firefox is being put up on a v6 arch. The current generation of boards (such as firefly) is based on ARMv7-A. a much more powerful and matured arch. Next, v8 is coming up with 64 bit support.
Not strange at all. Raspberry Pi is vastly more popular than any other competitor, perhaps even all competitors combined. It doesn't really matter that it's slower or that it uses an old instruction set.
There are nearly 4 million Pi's sold now vs. a few hundred thousand BeagleBone Black boards (which use ARMv7). If you only have limited resources to get your OS running on one embedded hobbyist board, I think the choice to use the Pi makes sense.
"Users will be able to read from sensors and control motors, LEDs, solenoids, slave boards, etc"
That will make it easier for people to start playing with electronics from the browser on their Raspberry Pi. Chrome has a serial API, which I think is only accessible to chrome extensions - https://developer.chrome.com/apps/serial. If Firefox OS on the Pi comes with no-limits unlike Chrome, then it should be fun to play with.
Looking forward to it ~!