Wine developer frustration (was Re: ntdll: Improve stub of NtQueryEaFile.)
kyle.auble at zoho.com
Sat Jun 20 17:26:34 CDT 2015
On Sat, Jun 20, 2015 at 12:33 PM, Josh Dubois wrote:
> Michael said near the start of the thread that "access on the official
> winehq infrastructure" was a problem for some people. Maybe commit
> access to a second git branch would be a start.
From personal experience, I don't think it's just the git repo. One
time, when I didn't feel up to working on Wine code but still wanted to
help some in my free time, I started looking into bug triage. I had
Bugzilla and AppDB accounts, a subscription to the bugs maillist, had
already practiced testing on developmental build of Wine, and read what
instructions I could find on the wiki and Dan Kegel's page.
In the end though, beyond just adding more comments to the bottom of
the page, I couldn't figure out how anyone earns the rights necessary
to fix tags or recategorize bugs. I even tried emailing the Bugzilla
admin to ask but never heard back. Honestly, beyond some of the
developers and site admins, whom I assume have elevated rights across
WineHQ, I still don't know who processes the bug reports.
> With a second branch, more patches could be accepted, but committed *in
> the branch where they fit best*, and maybe there would be less
> contention. Ideally there would be shared goals both to allow some
> experimentation in the 'staging' branch and eventually to get all
> features in a 'polished' state and into 'master.'
Isn't the whole issue with wine-staging patches that they're more
ad-hoc/hackish then we want for Wine proper? Correct me if I'm wrong,
but if you take an ad-hoc fix that suddenly makes something work,
then replace the programmatic output with debug info, don't you have a
test case? I'm just wondering if maybe our thinking about wine &
wine-staging is (to quote The Dude) too uptight. Perhaps instead of
seeing it as:
Diablo fan makes one-off fix ->
Pessimists shake heads & optimists start pushing that rock uphill ->
Maybe once in a while, it becomes part of the Wine *program*
... we should see it as more ...
Diablo fan makes one-off fix ->
Someone distills the effective knowledge into a test ->
It moves into the Wine *test-suite* <-
Wine-staging or AppDB points out the original patch as a stop-gap.
On Sat, Jun 20, 2015 at 10:41 AM, Jeremy White wrote:
> I guess our challenge is to find a way to make the connection to the
> positive engagement you did eventually receive.
Since I've worked on other parts of WineHQ, not Wine itself, my
experience might be different, but I've never really had problems with
negative engagement so much as no engagement at all. That's not
universally true: you were very helpful when I was starting to work on
the website some, Jeremy. Plus Andre H and Dimi P were really good
about keeping in touch with thoughts while I redid the wiki.
Otherwise though, it seems sometimes like almost everything that isn't
a direct comment on a patch just falls through the cracks on the
mail-list. I get the impression IRC might be a little better in that
regard, but sometimes it's not easy to stay on IRC. When I was working
on the wiki, I actually didn't even have internet access at my flat.
Even then, it might be misleading to view it as primarily a
communication issue. I think it's more a symptom of the coordination
problems I mentioned in my first email.
More information about the wine-devel