GSOC student (re)introduction

Armin Krezović krezovic.armin at gmail.com
Sat Apr 30 14:52:11 UTC 2016


On 29.04.2016 09:16, Pekka Paalanen wrote:
> On Thu, 28 Apr 2016 04:19:32 +0200
> Armin Krezovic <krezovic.armin at gmail.com> wrote:
> 
>> Hello everyone,
>>
>> As you may have seen, I was chosen to participate in GSOC
>> and work on your project.
>>
>> I do need to apologize for not contacting you sooner, but
>> it was a busy week here (midterms week), so I didn't have
>> much time to think about anything else. I hope you understand.
>>
>> I'll skip the introduction part, as I've already introduced
>> myself on this mailing list and also in my project proposal,
>> which is publicly available.
>>
>> I do want to note that the email address used last time was
>> different. From now on, this address will be used because
>> it's the one I used to register for GSOC.
> 
> Hi Armin,
> 
> I'll try to remember that. You should use that as your author/reviewer
> identity too, I often end up looking up email addresses from commits.
> 

Yes, that's the idea.

>> I must admit that I do not fully understand the part I'll be
>> working on just yet (it's not the weston code, but the DRM
>> part itself), but I am confident I can master that part as
>> well before the coding starts. So, expect some newbie questions
>> during the Community bonding period :).
> 
> I wonder if there really will be much DRM part. DRM backend already
> handles hotplug, and I think most of the culprits are elsewhere.
> 

A quick look at the compositor code gave me a similar answer. However,
the DRM backend might need some rework to support extended screens. I'm
not there yet, it's just a guess at the moment.

> There seem to be lots of fairly trivial patches floating around these
> days, you are more than welcome to hone your reviewing skills on them.
> That would be excellent bonding period activity.
> 

That's a great idea. I'll bear that in mind and watch for patches that
I may be qualified enough to review :)

> Also, now that this years' GSoC selections are over, I think we could
> open the tasks reserved for GSoC applicants free for all once again. So
> if there is anything you'd like to tackle as an intro, feel free. :-)
> 

I may take a look at some trivial tasks. No promises though :)

>> Also, as I've noted earlier, I also want to use this Community
>> bonding period to discuss with everyone involved in the project
>> on what would be the best way to implement the feature I will
>> be working on when the coding period starts. I did suggest two
>> different ways in my project proposal, but I am open for more
>> suggestions. I'll send a separate mail about that as soon as I
>> familiarize myself with the code that's currently there (I can
>> manage that before next Wednesday).
> 
> Very good.
> 
> One of my first thoughts is how will you be able to conveniently test
> output removing and hot-plugging. Surely fiddling with cables while
> running the DRM-backend will be awkward, and if you need to debug, even
> more so. Yet, you need that before you can start finding the pain
> points.

I have a laptop and can emulate display unplug by simply closing the lid.
Then, I can use a different laptop to debug a local session via screen/tmux
and ssh. That's just one of the ideas.

> 
> How about taking either the x11 or the wayland backends as the testbed
> for the Weston core changes, and writing a test plugin (either similar
> to, or extend tests/weston-test.c) that exposes a new protocol extension
> you can use to add and remove output windows from the backend by using
> a command line utility? [*]
> 
> That way it would be easy to tell the backend to create and remove
> outputs at will, and you could later modify it for use in 'make check'
> tests. The DRM-backend does not really lend itself easily for that kind
> of output fiddling, that's why choosing x11 or wayland (and headless)
> backends.
> 
> Actually, if you choose the wayland backend, we might be able to avoid
> patching the support into headless if we run weston/wayland under
> weston/headless in the test suite.
> 

This is even better idea than mine. I'll keep it in mind.

>> Last, I want to thank whoever it was that choose project mentors
>> for the great choice of mentors! I saw the welcome mail, but
>> didn't get the chance to reply to it. So, I want to use this
>> message to thank Pekka for the nice welcome mail. It's an honor
>> to be the first student to be selected for the Wayland project.
> 
> You're welcome!
> 
> 
> Thanks,
> pq
> 
> [*] Yes, I really did just propose that. The point is that it will be a
> test plugin, just like our existing test plugin that too exposes
> interfaces we would never expose at production runtime, and will never
> be installed with weston.
> 

Armin.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <https://lists.freedesktop.org/archives/wayland-devel/attachments/20160430/52360ed2/attachment.sig>


More information about the wayland-devel mailing list