[Wine]Does a user need to join wine-devel?
mike at navi.cx
Mon Nov 15 15:48:59 CST 2004
On Mon, 2004-11-15 at 21:29, Mark Knecht wrote:
> Sunday afternoon I found that Acid Pro 2.0 will run under
> wine-20040505 but fails under wine-20041019. I have not had an
> opportunity to do regression testing to figure out exactly where it
> started failing but the error message is during install and the
> problem is that the newer Wine doesn't have a 'new enough version of
> Direct X Media', whatever that is.
I suspect we added an implementation of a redistributable DLL that was
previously being installed from MS, and it lacks the correct version
metadata. That's usually the cause of such failures. Try a +loaddll,+ver
trace and see what it's trying to look at.
> I have recently placed a number of new entries to the Wine Bugzilla
> database, but the one that's of most interest to me is this bug:
> This bug is basically that menus are not displayed. There is a small
> .png file attached to that bug that shows the problem.
I've seen this once before, a long time ago. It seems to occur very
rarely, I suspect some unusual API abuse is going on that screws us up.
> This is the bug
> I would most like to find a fix for but how to proceed with debugging
> is beyond me. Anecdotally, this bug reminds me of a menu problem that
> I've heard about being fixed on some versions of Crossover Office. (I
> am a registered CXO user.) It is my hope that wine-20041019 has all of
> the CXO fixes in it for items like this, but I somehow am not so sure.
Depends if it was a bad hack or not. I don't know offhand.
> Previously I had a rather difficult problem. Some details are
> listed in this bug:
> Through regression testing I got down to the actual set of edits that
> caused a problem with a Linux/Alsa program called 'jack_fst'.
> Apparently jack_fst is a 'wine-lib' program (?) that allows Windows
> VST's to run under Wine using Jack and Alsa. All of these programs
> broke with this set of edits:
> At this, with great help from Duane and Walt, a hack was found to
> jack_fst that allows them to run again. However, the Jack folks refuse
> at this time to either investigate or include the hack and no one on
> that side seems to want to debug the problem saying that Wine broke
> something that was working previously. They seem content at the moment
> to just require folks to use wine-20040505. I don't like that solution
> at all.
I can understand their pain. We don't want to break their work.
Unfortunately in this case we had no choice, that change was required
for compatibility with Fedora kernels. I have no clue *why* it'd break
jack_fst, somebody needs to post a strong technical analysis before we
can address it really.
> I have filed a number of other 'bugs' but I wasn't really asking
> for fixes to all of them. Rather I was just putting the data in the
> database in case those apps ever gained steam in the future so that
> someone would know to write me. I have 5-10 other bugs I could put
> into bugzilla but it seemed fairly clear that it wasn't worth the
> effort to rush and do it so I'll enter them over the next few weeks.
Sure, that'd be fine. As you've realised by now we're starved of
manpower. While it'd be great if there were tons of people standing by
waiting to fix problems, in reality that isn't the case.
More information about the wine-users