On Fri, Jan 25, 2008 at 06:00:12PM +0100, Tim van der Molen wrote: > On Fri, 18 Jan 2008 22:23:36 +0100, Deanna Phillips wrote: > > As you may recall, a long-overdue update to esound was backed > > out at the last minute before the 4.2 release. The thread is > > here: > > > > http://marc.info/?l=openbsd-ports&m=118541595130336&w=2 > > > > I've just tried the 2.38 update on auich(4) and can't reproduce > > the problems we once had. > > > > 2.38 also fixes some bugs with socket creation and cleanup that > > are causing micropolis to skip some sounds. This bug can be > > reproduced by throwing many files at esdplay at once with > > something like 'for file in *.wav ; do esdplay $file& ; done', > > or running the Firebomb scenario in Micropolis and watching > > stderr. > > > > esd should be able to mix those sounds (that's what it's for), > > but with the version in tree it gets very confused and skips > > some, fails to open the audio device or resample, etc. > > > > So. Here's Martynas's old diff pulled from cvs. > > > > Anyone want to try again? > > Using the same laptop as I did in the thread referenced above, I still > have the behaviour I described at > <http://marc.info/?l=openbsd-ports&m=118616452718942&w=2>.
esd should use AUMODE_PLAY_ALL. it should have always been using that, but AUMODE_PLAY_ALL was broken for emulated formats for a really long time. it's been fixed for a while now though. *anything* ported from OSS to audio(4) should use AUMODE_PLAY_ALL. AFAIK, audio(4) is the only (relatively) common audio system that even has a mode that tries to enforce "realtime" by eating "late" data. -- [EMAIL PROTECTED] SDF Public Access UNIX System - http://sdf.lonestar.org
