[meta-freescale] Call for maintainers (was [RFC PATCH] fsl-commity-bsp: Add meta-qt5)

Otavio Salvador otavio at ossystems.com.br
Mon Mar 24 09:55:09 PDT 2014


On Mon, Mar 24, 2014 at 12:58 PM, Eric Nelson
<eric.nelson at boundarydevices.com> wrote:
> Hi Otavio,
> Since we have diverged off the topic again, I changed the
> subject line to invite more folks to chime in.
> On 03/24/2014 07:26 AM, Otavio Salvador wrote:
>> On Mon, Mar 24, 2014 at 10:08 AM, Daiane.Angolini at freescale.com
>> <Daiane.Angolini at freescale.com> wrote:
>>>> I am not sure we ought to include it or not. I see valid points for both
>>>> and I will focus my answer in the cons:
>>>>   * Documentation: our Release Notes, User Guide and FAQ are still
>>>> uncomplete. Add new stuff will only make it worse as those will also
>>>> need
>>>> to be documented.
>>>>   * Tests: We see limited tests using the images we have and people does
>>>> not
>>>> provide much feedback when we open the test form. More things will only
>>>> complicate it more.
>>> Until here I understand. And I agree.
>>>>   * Size: more metadata means more updates and maintenance. We need more
>>>> people helping the metadata maintenance before extend it.
>>> We already have almost 100% of the boards with maintainers. What do you
>>> mean, when you
>>> say we need more maintenance before extend the metadata?
>>> I still don´t have a clear idea if I like or dislike the idea of
>>> downloading more meta layers by default.
>>> I´m still deciding ;)
>> We need to split the maintainership areas:
>> * boards: Yes, most of board has someone committed nowadays and this
>> is awesome. This help us to get tests and feedback from those boards
>> and more people to help to address board specific things. However this
>> does not address the rest of meta-fsl-arm ...
>> * core BSP support: here we have some people working. You and me has
>> been doing most of work until now and Lauren has been started to
>> contribute more to this since Freescale started to work in the
>> 3.10.17-1.0.0 BSP. Here we need a lot of help and it does affect
>> /every/ board we use/add/maintain ...
> These are areas where both carrots and sticks might help. It seems
> very reasonable to ask each board maintainer to update and test
> against some set of target images within a certain time period.

Part of this has been addressed by the Test Form, which has been done
and maintained mainly by Daiane which try to make clear what we intend
to support during the release. However I agree we may need a more
frequent test to fully address this and here I am out of ideas. So do
you have any idea how we can improve on this regard?

>> * technology specific support: here is where we are beginning. This
>> involves a lot of commitment and time. Here is where Qt5 and Chromium
>> are going to be covered. Currently we have no one really committed to
>> either and we cannot supply something in FSL Community BSP which is
>> not /maintained/ and for maintained I mean someone watching it and
>> doing the rework/improvements/commitments need for it to keep working
>> when Qt5 5.3 or Chromium 40 is released.
> I think you're under-stating the issue here, since there are many
> more bits than just Qt5 and Chromium.

Of course; the idea here was to elucidate two more known and often
asked examples...

> Just off the top of my head:
> - Is anybody testing DirectFB on a regular basis? (seems a question
> on the ML today),

This is supported by Freescale so I assume they are doing those tests.

> - Up until the recent addition of , there weren't any easy-buttons for
> testing the gstreamer plugins in a non-X environment, and

Here again we need to split this:

Freescale 0.10 plugins: this is official material from Freescale and I
expect those to be tested by them. For example gplay makes it more or
less easy to use this.

gstreamer-imx (1.0): Carlos has been doing an awesome job here but he
cannot do all himself and as a community we ought to help here. We
need to think how to improve this and I am open for ideas here...

packagegroup-fsl-*, fsl-image-machine-test: Rogerio Nunes has started
an images rework which provides a much better base  for test, reuse
and long term maintenance...

... however I know that you have this background information so this
point is not clear to me. What you mean here exactly?

> - There's Wayland activity, but AFAIK, no image available to take
> advantage of it.

It is still not integrated in meta-fsl-arm. The patches are being
worked out and we ought to have it integrated soon, however, as
DirectFB this is something Freescale officially supports so we can
rely on them for tests.

>> So it is a extensive topic; I hope it is clear now why I see we need
>> more people involved to grow.
> I feel like I've thrown a couple of bricks in the last week (Chromium
> and Qt5), but I hope they can be constructive.
> I do think that the questions you raise are worthy of a separate set of
> discussions (or at least a separate e-mail chain).
> Since there are a lot of different, interested parties with different
> agendas, I wonder whether a different (more interactive) forum might
> be a better match.
> Conference call? Google Hangout meeting?

Yes; I think a Google Hangout would fit perfectly fine here. Some
people has suggested it privately in past but our community was still
too small for it to make sense. I think we are big enough for it to be
worth it and I am supportive to the idea.

What others think?

> I'm sure it's too late to schedule anything at FTF, but perhaps not.
> I'm also sure that there will be those interested in the topic who
> won't be attending.

Sure. I will send an e-mail about FTF later today and I think it is a
great place for us to meet and discuss some of those points
face-to-face as the FTF and the Google Hangout allow more interactive
discussion but this all needs to be summarised here, in the mailing
list, for a final decision.


Otavio Salvador                             O.S. Systems
http://www.ossystems.com.br        http://code.ossystems.com.br
Mobile: +55 (53) 9981-7854            Mobile: +1 (347) 903-9750

More information about the meta-freescale mailing list