Hi Miller
On Tue, 2012-07-03 at 14:11 -0700, Miller Puckette wrote:
Hi all,
Pd version 0.43-3 is available on http://crca.ucsd.edu/~msp/software.htm or via git from sourceforge: git clone git://pure-data.git.sourceforge.net/gitroot/pure-data/pure-data
As always: Many thanks for your efforts!
I'm ready to start hacking on 0.44. The most urgent thing seems to be for me to go back and work on the audio system (and perhaps tweak MIDI a bit more too).
May I ask, what is planned for the audio system, respectively what part needs some work?
My two cents regarding audio: 0.43 introduced a regression (at least on linux, don't know about other OS') in that the behavior was changed so that whenever DSP is turned on, a thread is started, and is stopped again, when DSP is turned off. The implications I experience with this architecture, is that sometimes audio doesn't come back when turning DSP on. Specifically, with the jack backend it sometimes happens, that Pd freezes and I have to restart Pd. With ALSA, when turning DSP off and on again, I sometimes get this error:
snd_pcm_open (output): Device or resource busy
and obviously sound is lost. Usually it helps to turn DSP off and on again to make sound work again. While this is only slightly problematic (though still annoying) in a live situation, it is quite dangerous for installation setups. This could potentially lead to running an installation without sound for the rest of the day.
Keeping the thread talking to the audio back end always running has some advantages:
I'm also interested to hear if others experience similar issues or if it is only me having trouble with 0.43's design. Up to 0.42 everything was fine for me regarding audio.
The thing I'm actually looking forward to doing is to extend textfiles and message boxes and data structures to be able to spit lists of atoms around much more flexibly than now.
Sounds good.
Also pd vanilla needs tilde objects to do interpolating signal snapshots, phase-syncable oscillators,
What does phase-syncable mean? Do you mean, that the phase for [osc~] and [phasor~] will be settable not only on block-boundaries, but also during a block?
Roman