Re: [gentoo-project] OT - Tinderbox question

[Rich Freeman]

On Thu, May 15, 2014 at 7:21 AM, hasufell <hasufell@g.o> wrote:
> Sergey Popov:
>> And yes, we need tinderbox. But, c'mon. stop talking loudly on ML and
>> get things done if you can.
>>
>
> This is like working on patches while upstream already said "not
> interested".
>
> Since QA doesn't think it's their job to run a tinderbox, I will work
> with those people who actually care about it, instead of QA.

Perhaps you should let QA speak for what it thinks its job is?  Note
that random posts in random bugs by random members of QA isn't the
same as QA saying something.

In any case, anybody who wants to run a tinderbox can run one.  They
can even work on a project to set up an "official" one.  You don't
have to be in the QA team to do QA either - you just don't have any
special authority if you do so.  Gentoo projects are allowed to
compete.

Honestly, this sort of argument can come across a bit like somebody
who has never even submitted a patch asking for cvs access, being told
to go contribute positively and they'll eventually get cvs access, and
then whining about how Gentoo doesn't want any new developers.  Don't
get me wrong - I realize that you do in fact contribute quite a bit
which is why you DO have commit access, so it isn't a great analogy.
My point is more that the QA team couldn't hold somebody back from
running a tinderbox even if they made it their sole mission in life.

I do things for Gentoo because I find it interesting, and I want to
give back since I benefit from it.  Sure, having commit access can
make it easier, but there are many who contribute to Gentoo without
even that.  You don't need to have a special job title to contribute
in almost any way.  That's the beauty of FOSS.

If you want to run a tinderbox, do it!  If you want to create a
tinderbox project and get others to help you with it, do it!  If you
get something running and want an infra box to run it on, bring it up,
though it would first make sense to get things working at a reduced
scale.  We can always find sponsors, and we get monetary donations
that could potentially be used to pay for a tinderbox.  What doesn't
make sense is to go out and spend a lot of money on hardware and then
find out that nobody has the time to do anything with it.  As has been
pointed out many times in the past, the CPU-hours is the least of your
problems when running a tinderbox.  If you can work out the process
for collecting and dealing with logs/etc then getting somebody to give
you plenty of cores to run it on should be easy.

Rich