Wine developer frustration (was Re: ntdll: Improve stub of NtQueryEaFile.)

[Henri Verbeet]

On 15 June 2015 at 18:31, Jeremy White <jwhite at codeweavers.com> wrote:
> With that said, I think this is a hard problem; I think it boils down to
> a request for Alexandre to be 'nicer', with varying shades of just what
> that means, and it's been discussed enough through the years that it's
> hard to discuss it constructively.
>
For what it's worth, most requests seem to be either about lowering
the acceptance threshold in some way (e.g. wrt. authorship, code
quality, amount of tests, etc.) or providing more hand-holding. I
think the majority of the request in the first category are just bad
ideas. The second category are usually things that would be nice in
theory, but for which people seriously underestimate the amount of
work that would be involved. For things in the second category in
particular it becomes an issue of weighing "How much useful work can I
do in the time it's going to take me to help this person." vs. "How
much useful work is this person likely to produce after I've helped
him/her." Some people just need a little help to get used to the
process. Some people might become competent developers once they e.g.
have another 5 years of experience. Some people simply never will.

Perhaps another interesting thing to note is that the people that tend
to become most frustrated are the ones that come to the project with
the approach of "I need to fix this one specific application/bug right
now". They tend to run into all the hard DIB-engine-level bugs, and
since they need to fix it right now building up some experience and
trust isn't an option, and they pretty much get stuck in that
position. Arguably that doesn't have so much to do with the
development model as it does with available developer capacity to fix
incoming bugs.

> So far, the only one of those I read in this thread is to ask bugzilla
> workers to stop closing bugs that mention wine-staging with such rapid
> hostility, particularly if the bug is clearly a bug in mainline Wine.
> There seems to be consensus on that.  Have we now done that?  Is there
> any further action we need to take?
>
I think it was more general than that, about discussing bugzilla
policy in general. The plan seems to be to do that at WineConf,
although personally I think it's one of those things that can happen
just as well on the mailing list.

On 15 June 2015 at 22:14, Stefan Dösinger <stefandoesinger at gmail.com> wrote:
> Also I believe many people are afraid to say "this looks OK to me" because they think it might lower their Julliard Rank if the patch contains a stupid mistake. Reviewing is harder than writing patches.
>
For me personally, it's more that there are various levels of review
you can do. If I explicitly ack a patch to Alexandre, unless stated
otherwise, that means it's something I'd personally commit. Aside from
just reading/checking the patch, that implies that I've applied it,
ran the tests on it, ran the tests on Windows if it touches the tests
(because, the testbot and D3D), and so on. For D3D, I care enough to
do that, but for most other areas I don't, and I might just point out
issues if they happen to catch my eye. These days, I have enough to do
that I don't even read a lot of patches though. (On that subject, if
you want people to actually read and reply to your patches, try using
git send-email, or at least sending them inline.)

On 15 June 2015 at 22:30, Marc Bessières <marc.bessieres at mykolab.com> wrote:
> I agree, I think AJ and the core Wine developers are overloaded and maybe
> close to burn out.
> So they don't have the energy to build a welcoming developer community.
> As wineHQ is not a welcoming as it could be, the community does not grow, so
> they are not many new developers that can help the share the workload.
> So the core Wine developers are overloaded and we sadly close the circle :(
I suppose so, on some level anyway. At the same time, while I may
personally feel a bit thin-spread at the moment, I'm pretty sure
Alexandre is currently giving people more feedback than he ever has
before. So I'm not so sure I necessarily see that correlation. I also
have the impression that compared to e.g. 5-10 years ago the average
skill level of potential new contributors has dropped a bit, and
perhaps that generally speaking simply fewer people are interested in
"proper" (i.e., not involving browsers, phones or tablets) software
development.