On 2010-10-08 09:32, David Henningsson wrote:
On 2010-10-06 23:09, Ray Rashif wrote:
Background: https://bugs.archlinux.org/task/21031
Dieter's workflow involves the following:
* jack
* rosegarden
* fluidsynth
For a week or so, Dieter did not touch these. He had:
* jack-0.118.0
* fluidsynth-1.1.1
* rosegarden-10.04
In that week, among the kernel and other system software updates a
(bleeding-edge) Linux system typically has, he ended up with:
* fluidsynth-1.1.2
Suddenly, Dieter could not resume his workflow. Fluidsynth's playback
ports refused to show up in jackd's connections.
Hmm. I'm on Ubuntu Maverick here. If I run FS 1.1.2 with the following
line (as specified in the background link) :
fluidsynth -r 48000 -i -a jack /usr/share/sounds/sf2/FluidR3_GM.SF2
The audio port does show up in qjackctl. The midi in does not, but that
is correct. For them to show up, you must also specify "-m jack", like
this:
fluidsynth -r 48000 -i -a jack -m jack /usr/share/sounds/sf2/FluidR3_GM.SF2
However, when I try that here, fluidsynth does not start properly. I'll
see if I can trace this down further.
Okay, fluidsynth is quitting because there is nothing keeping it alive
(no shell and no midi files to play). So to fix this, you should not
supply the -i switch.
However, when you don't supply the -m jack switch, it actually keeps
running. This is because of a race condition in the alsa-seq driver that
I just fixed (r376).
I'm a little curious if this is intended behaviour, anybody who knows?
If it is, we should perhaps add an information message saying "No shell
and no midi files to play - quitting.".
// David
_______________________________________________
fluid-dev mailing list
[email protected]
http://lists.nongnu.org/mailman/listinfo/fluid-dev
