On Sat, Mar 18, 2006 at 04:44:56PM -0800, Jacob Meuser wrote: > On Fri, Mar 17, 2006 at 01:29:05PM -0800, Jacob Meuser wrote: > > On Thu, Mar 09, 2006 at 07:04:08PM +0300, Alexander Yurchenko wrote: > > > hydrogen is a software synthetizer which is able to be used either by > > > itself, emulating a drum machine based on patterns, or via an external > > > MIDI keyboard/sequencer software. > > > > for me, this locks on almost every action on amd64. so did previous > > versions. > > > > anyone else seeing this? > > > > I'm building on i386 now to see what happens there. > > no different there. maybe I'm doing something wrong? for the people > who reported success, did you make any configuration changes or use > any options when starting hydrogen? > > dmesg below, but I really don't think it is hardware related.
turns out it was hardware/driver related. I pulled the SB Live! and used the onboard auvia(4), and hydrogen worked fine. I spent a while trying to figure out why hydrogen wouldn't work with emu(4), fiddling with the OSS ioctls. eventually, I got fed up and just changed hydrogen to use audio(4) instead of ossaudio(3). now hydrogen works on both soundcards. attached is a new version of the port. besides using audio(4) instead of ossaudio(3), I took the version numbers out of WANTLIB (they're not needed) and changed the way the audio device is set. basically just changed the OSS device default from /dev/dsp to /dev/audio instead of hardoding the device to /dev/audio. this way, if someone does have 2 soundcards, they can choose which one to use. oh, I also removed a usleep() from the logging function. occasionally hydrogen would segfault on exit because the logger thread would try to fclose() the logfile, which had already been closed. doesn't seem to cause any problems, and I haven't seen the segfault on close yet. I'm really interested to know how this compares to grange's original port for other people. please let me know! -- <[EMAIL PROTECTED]>
hydrogen-0.9.3-port.tgz
Description: application/tar-gz
