Hi,
BlueZ just got accepted as an organization for this year's Google
Summer of Code.
There is a small list of project ideas here[1], for the next days we
need to polish it and add more content. Any help is appreciated.
We are also looking for mentors, if this idea interests you, please
contact me off-list or go to #bluez-gsoc at freenode, this is also
where most of the discussion regarding GSoC will take place, in case
you just want to follow it.
As always, patches and suggestions are welcome ;-)
And thanks to everyone who contributed ideas to that list.
Cheers,
--
Vinicius Gomes
INdT - Instituto Nokia de Tecnologia
[1] http://hostisdown.org/gsoc/
On Mon, 2010-03-22 at 18:14 +0200, Luiz Augusto von Dentz wrote:
> Hi Bastien,
>
> On Mon, Mar 22, 2010 at 2:13 PM, Bastien Nocera <[email protected]> wrote:
> > Hey,
> >
> > On Thu, 2010-03-18 at 22:50 -0300, Vinicius Gomes wrote:
> >> Hi,
> >>
> >> BlueZ just got accepted as an organization for this year's Google
> >> Summer of Code.
> >>
> >> There is a small list of project ideas here[1], for the next days we
> >> need to polish it and add more content. Any help is appreciated.
> >>
> >> We are also looking for mentors, if this idea interests you, please
> >> contact me off-list or go to #bluez-gsoc at freenode, this is also
> >> where most of the discussion regarding GSoC will take place, in case
> >> you just want to follow it.
> >>
> >> As always, patches and suggestions are welcome ;-)
> >
> > A couple of comments:
> > - Is AVRCP 1.4 support really a full Summer's work? Could be with
> > PulseAudio or application integration?
>
> Some avrcp 1.4 needs some integration so I guess it is ok.
>
> > - The "any" adapter work is a waste of resources, as I've mentioned in
> > the past on the list.
>
> Well if it eliminates one round trip for most application I guess it
> is worth adding, and we might add other tricks as storage and API
> simplifications on top of it.
One round-trip at the cost of more code on your side. I'm really not
sure it's worth it.
> > - The VDP support probably needs kernel-level hacking as well, to avoid
> > being constrained to one of the supported frameworks, and create a
> > user-space driver for the device (using CUSE?)
>
> It could be considerate, but if that is a good idea to vdp it would
> also be for a2dp, there is also the problem that avdtp is currently
> implemented in bluetoothd so I wonder if we should start changing
> things or just implement vdp on top of the existing avdtp code.
Right, except that for sound we have a "broker" application in
PulseAudio. There's no such thing for V4L devices. In any case, it
should probably be a separate application that would take the data from
bluetoothd and make it available as a video device through v4l.
The VDP-using webcam I have uses MPEG-4 as the video codec, and you
probably don't want to have to reimplement that, so the broker app could
probably use GStreamer to make the video available.
> > I'd be happy to do testing and some integration work for both the VDP
> > and DUN works.
> >
> > I could also add:
> > - Port gnome-user-share to obexd, instead of obex-data-server (easy)
> > - Port osso-gwobex test suite and extend it (medium)
> > (Those 2 could be combined into one project)
> > - Printer (initial) setup integration in gnome-bluetooth (medium)
>
> I guess those are too gnome related for being in BlueZ, expect maybe
> gwobex, of course we could help/co/mentor projects under gnome
> organization.
Or vice-versa :)
Given that porting gnome-user-share to obexd shouldn't be a long winded
task, porting the osso-gwobex test suite would nicely complete it.
The printer integration work could probably be re-used by other
bluetoothd-setup front-ends like KDE.
> > - Extended hostname support (using xdg-hostname) (medium)
> > - Kernel driver and user-space changes for Sixaxis joypad (including
> > rumble). The bdaddr assignment would be done in the kernel, with a
> > helper in bluetoothd, extra functionality would be added in the kernel
> > for the other features (medium)
>
> Great ideas and those fits perfectly as bluez organization projects.
Cheers
Hi Bastien,
On Mon, Mar 22, 2010 at 2:13 PM, Bastien Nocera <[email protected]> wrote:
> Hey,
>
> On Thu, 2010-03-18 at 22:50 -0300, Vinicius Gomes wrote:
>> Hi,
>>
>> BlueZ just got accepted as an organization for this year's Google
>> Summer of Code.
>>
>> There is a small list of project ideas here[1], for the next days we
>> need to polish it and add more content. Any help is appreciated.
>>
>> We are also looking for mentors, if this idea interests you, please
>> contact me off-list or go to #bluez-gsoc at freenode, this is also
>> where most of the discussion regarding GSoC will take place, in case
>> you just want to follow it.
>>
>> As always, patches and suggestions are welcome ;-)
>
> A couple of comments:
> - Is AVRCP 1.4 support really a full Summer's work? Could be with
> PulseAudio or application integration?
Some avrcp 1.4 needs some integration so I guess it is ok.
> - The "any" adapter work is a waste of resources, as I've mentioned in
> the past on the list.
Well if it eliminates one round trip for most application I guess it
is worth adding, and we might add other tricks as storage and API
simplifications on top of it.
> - The VDP support probably needs kernel-level hacking as well, to avoid
> being constrained to one of the supported frameworks, and create a
> user-space driver for the device (using CUSE?)
It could be considerate, but if that is a good idea to vdp it would
also be for a2dp, there is also the problem that avdtp is currently
implemented in bluetoothd so I wonder if we should start changing
things or just implement vdp on top of the existing avdtp code.
> I'd be happy to do testing and some integration work for both the VDP
> and DUN works.
>
> I could also add:
> - Port gnome-user-share to obexd, instead of obex-data-server (easy)
> - Port osso-gwobex test suite and extend it (medium)
> (Those 2 could be combined into one project)
> - Printer (initial) setup integration in gnome-bluetooth (medium)
I guess those are too gnome related for being in BlueZ, expect maybe
gwobex, of course we could help/co/mentor projects under gnome
organization.
> - Extended hostname support (using xdg-hostname) (medium)
> - Kernel driver and user-space changes for Sixaxis joypad (including
> rumble). The bdaddr assignment would be done in the kernel, with a
> helper in bluetoothd, extra functionality would be added in the kernel
> for the other features (medium)
Great ideas and those fits perfectly as bluez organization projects.
--
Luiz Augusto von Dentz
Computer Engineer
Hey,
On Thu, 2010-03-18 at 22:50 -0300, Vinicius Gomes wrote:
> Hi,
>
> BlueZ just got accepted as an organization for this year's Google
> Summer of Code.
>
> There is a small list of project ideas here[1], for the next days we
> need to polish it and add more content. Any help is appreciated.
>
> We are also looking for mentors, if this idea interests you, please
> contact me off-list or go to #bluez-gsoc at freenode, this is also
> where most of the discussion regarding GSoC will take place, in case
> you just want to follow it.
>
> As always, patches and suggestions are welcome ;-)
A couple of comments:
- Is AVRCP 1.4 support really a full Summer's work? Could be with
PulseAudio or application integration?
- The "any" adapter work is a waste of resources, as I've mentioned in
the past on the list.
- The VDP support probably needs kernel-level hacking as well, to avoid
being constrained to one of the supported frameworks, and create a
user-space driver for the device (using CUSE?)
I'd be happy to do testing and some integration work for both the VDP
and DUN works.
I could also add:
- Port gnome-user-share to obexd, instead of obex-data-server (easy)
- Port osso-gwobex test suite and extend it (medium)
(Those 2 could be combined into one project)
- Printer (initial) setup integration in gnome-bluetooth (medium)
- Extended hostname support (using xdg-hostname) (medium)
- Kernel driver and user-space changes for Sixaxis joypad (including
rumble). The bdaddr assignment would be done in the kernel, with a
helper in bluetoothd, extra functionality would be added in the kernel
for the other features (medium)
Cheers