[Mesa-dev] [PATCH v2] docs: Document GitLab merge request process (email alternative)

[Erik Faye-Lund]

On Thu, 2018-12-13 at 18:07 +0100, Axel Davy wrote:
> On 13/12/2018 17:57, Mathias Fröhlich wrote:
> > Hi,
> > Initially it seemed to me that I am about the only one sticking
> > with mailing lists.
> > And I personally feel like a too small contributor to really try to
> > influence your
> > decisions too much. But these recent hand full of mails all tell me
> > that I am not
> > that alone. I personally did contribute to several projects during
> > the past years.
> > All that only in part time, thus it had to be *very* efficient for
> > myself. And that is
> > something that I achieved by a consistent 'interface' to all those
> > projects. Just
> > my widely used and highly convenient mail client. So, all that
> > worked in a sufficiently
> > efficient way because I could combine this kind of 'work' even with
> > my private mail
> > that I could handle in between with that single 'interface'. So
> > going to any web site
> > there is already a detour and having multiple of them for each such
> > project gives an
> > even longer detour. Okay today it's mostly mesa that is left as
> > well as a communication
> > middle end used in vizsim applications. But going away too far away
> > from a mailing list
> > will be mostly a loss of efficiency for me.
> > As I said, my two cents, that should not keep you all from doing
> > changes that finally
> > increase your all efficiency ...
> > 
> > best
> > 
> > Mathias
> > 
> > 
> > 
> 
> Hi,
> 
> 
> I have to add my voice here as well.
> 
> Even though I do not feel able to give review for most of the mesa
> code 
> base,
> I appreciate to have all patches in the mailing list in my mail
> client.
> 
>  From time to time, I give feedback for some set of patches, for
> example 
> when I see patches related to dri3 or that could impact Nine.
> 
> It also enables me to get an overview of all the recent works and
> new 
> features Nine could use.
> 
> I feel like if most patches go through MR without getting a mail 
> feedback, I would not be able to do those as efficiently.
> 
> I would appreciate if I could *flag* some files or directories, and
> when 
> a MR impacts those (for example dri3 files, gallium interface,
> gallium 
> Nine, etc),
> I could get an automated mail with a summary of the MR, in order to 
> encourage me to look at it.
> 

AFAIU, the proposal is to always send the cover-letter of the series to
the mailing-list with a link to the PR. Shouldn't that show you the
summary (with a list of changed files) anyway, so it should be as easy
as it currently is to figure out what to review?