Re: [Xen-devel] [PATCH v5 0/1] cameraif: add ABI for para-virtual camera

[George Dunlap <george.dunlap@xxxxxxxxxx>]

On 3/12/19 9:07 AM, Jan Beulich wrote:
>>>> On 12.03.19 at 09:48, <jgross@xxxxxxxx> wrote:
>> On 12/03/2019 09:19, Oleksandr Andrushchenko wrote:
>>> From: Oleksandr Andrushchenko <oleksandr_andrushchenko@xxxxxxxx>
>>>
>>> Hello!
>>>
>>> At the moment Xen [1] already supports some virtual multimedia
>>> features [2] such as virtual display, sound. It supports keyboards,
>>> pointers and multi-touch devices all allowing Xen to be used in
>>> automotive appliances, In-Vehicle Infotainment (IVI) systems
>>> and many more.
>>>
>>> Frontend implementation is available at [3] and the corresponding
>>> backend at [4]. These are work in progress, but frontend already
>>> passes v4l2-compliance test for V4L2 drivers. libxl preliminary
>>> changes are available at [5].
>>>
>>> This work adds a new Xen para-virtualized protocol for a virtual
>>> camera device which extends multimedia capabilities of Xen even
>>> farther: video conferencing, IVI, high definition maps etc.
>>>
>>> The initial goal is to support most needed functionality with the
>>> final idea to make it possible to extend the protocol if need be:
>>>
>>> 1. Provide means for base virtual device configuration:
>>>  - pixel formats
>>>  - resolutions
>>>  - frame rates
>>> 2. Support basic camera controls:
>>>  - contrast
>>>  - brightness
>>>  - hue
>>>  - saturation
>>> 3. Support streaming control
>>
>> So since the first post in July 2018 there has been no reaction from
>> Konrad to this interface. I guess he has plenty of other things to do.
> 
> Having gone through all the versions' threads (just their titles) I can't
> find any explicit ping to him. Yes, five versions should have been
> enough to draw attention, but then again this may have indicated to
> him that things are still too much in flux.
> 
>> Maybe it would be a good idea to add someone else as a maintainer for
>> the "PUBLIC I/O INTERFACES AND PV DRIVERS DESIGNS" section in
>> MAINTAINERS to avoid such stalls in the future?
> 
> Well, iirc he had volunteered himself for that role, so I guess the
> preferred action in such a case would be for him to also step back if
> his other duties no longer permit him fulfilling the maintainer role here.
> Without the specific MAINTAINERS entry, as in the old days, THE
> REST would assume responsibility again, which personally I'd prefer
> over adding a second individual to the section. Unless someone else
> (like you) volunteered again.

The "unless" between "adding a second individual to the section" and
"someone else... volunteered again" indicates that they're in contrast
somehow; but I don't understand what contrast you mean.  We certainly
can't *assign* anyone to that responsibility, so the only way someone
would get their name down there is to volunteer.

Or are you making a distinction between "stepping up because they
thought it needed to be done" vs "stepping up because they think it's an
interesting thing to do and are jumping at the opportunity"?

I generally think having one or two people *specifically* responsible
for things is better than having a reasonably large group of people (THE
REST) responsible for a reasonably large number of things.  So while I
certainly think it makes sense to have it revert to THE REST in the case
that nobody steps up, I think it would be better if someone actually did
step up.

 -George

_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxxx
https://lists.xenproject.org/mailman/listinfo/xen-devel