[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

DebConf12 Debian mobile BoF notes

Hi all,

During DebConf12 we had a BoF discussing Debian mobile stuff:


Thanks to the DebConf video team, the session was recorded:


Here are some notes from the session:

== Agenda

* We are here to keep Debian relevant for mobile devices.
* Changes since last year?
* What now?

== Changes since last year 

Inside Debian:

EFL mostly updated to recent versions.

FSO upstream worked with us to get their recent release into Debian.

pabs got a Samsung Galaxy S I, OdyX got a random Chinese tablet.

Debian mobile email list, IRC opened, not much work done.

Outside Debian:

GPU driver reverse engineering projects for ARM Mali, Qualcomm Andreno,


Android Linux kernel stuff slowly merging.

KDE community released Plasma Active. http://plasma-active.org/

KDE community working on Vivaldi Tablet. http://makeplaylive.com/

OpenMoko/Replicant/FSO/SHR synergies and expansion of hardware support. 


GTA04 became available:


Canaima (Debian derivative) working with hardware vendors to produce
Canaima Mobile.

Mozilla started Boot to Gecko / Firefox OS. 

MeeGo community was closed, Maemo is being de-funded, various projects
split out from Maemo/MeeGo as a result; Jolla, Mer, Nemo, Cordia Tab.

Tizen was released (Samsung/Intel, EFL, git repos indicate Debian/Ubuntu

WebOS has been freed (not all yet?). http://openwebosproject.org/

== Discussion

Mobile is important to keep Debian relevant.
How do we deal with Linux drivers and binary blobs?
We need to pick a UI which is supported long term by upstreams and

What about Debian chroots (or LXC) on Android kernels? that would solve
the problem of all the underlying headaches.

We could take cyanogen mod to steal hardware support from them and build
WebOS on top or any other stack. It might not be easy due to power
management differences between android and Linux, interesting bugs might
appear, but it would be nice to try it.

Meego was proven to be hard to follow due to ABI changes.

Tizen and WebOS are reasonable designed platforms, but we don't know
what we get or we don't get => we need investigation work. (send a
report to mailing list)

Need folks to start investigating the various platforms and report back
on how useful they are.

FSO, how is it? It is better aligned to Debian philosophy and binary
blob free. There is the feeling this platform is unmaintained and not
maintained long term, but it is expanding beyond the devices of OpenMoko
Inc. It is also a forum we can actively contribute, unlike vendor-driven
projects like Android.

E17, is active and part of Tizen and SHR, it can be suitable and
interesting for the purpose.

Getting stuff into upstream kernels takes a long time, kernel support
cycle is a problem for Debian. Still even an issue for OpenMoko GTA02.

We could provide out of Debian Linux kernels (maybe via PPA-like). The
Debian OpenMoko team did this for GTA02 and it proved suboptimal since
the upstream community (OpenMoko Inc, SHR etc) wasn't using the latest
upstream kernels nor focusing on getting patches merged there.

Embedding Debian appliances (eg FreedomBox) in an existing framework is
a pragmatic way to speed time-to-market whether using Windows (via
Virtualbox?) or Android in container.

Venezuela might be investing in development of a Debian based mobile OS
(Canaima). Partnership between govts of Venezuela & China. Vtelca
(http://www.vtelca.gob.ve/) & Orinoquia (http://orinoquia.com.ve/) have
relationships with Chinese manufacturers like ZTE and Huawei.



Attachment: signature.asc
Description: This is a digitally signed message part

Reply to: