Hi Andy, well I've ordered your book to learn PD, so I'm at least that committed! :-)  My overarching goal is to make tools for composing, improvising, and sequencing, so at this point, I'm punting on signals. First I want to support everything in Rick Taubes's Notes From the Metalevel book, along with a bunch of my own ideas, for algorithmic and computer assisted composition/improvisation. But I am hoping to get signals into play eventually. Which makes this a good time for a question... 

In Max, my understanding is that if Overdrive is on and Scheduler in Interrupt is selected,  we actually get only two threads: low priority thread running who knows when, and high priority thread alternating between a pass of eating up the scheduler queue (up to the limit) and then a pass of the audio vector render. So given that I'm forcing all messages to run in the high priority thread (unless the user asks for danger-mode Any), I *think* I should have no problem calling into S7 and sharing data from both dsp perform routines and regular event scheduled messages. I have not, however, actually done this yet. Can anyone tell me what the situation in PD is there? Or point me at good docs/resources on how the threading and scheduling works in PD?

Thanks
iain


On Fri, Oct 23, 2020 at 12:23 PM Andy Farnell <padawan12@obiwannabe.co.uk> wrote:
My last love affair with parentheses was in Racket.
Although my only ever somewhat successful musical adventures
in this style were with Roger Dannenberg's CLM/Nyquist

Lets see what happens if Iain brings S7 over.

Iain, how are you thinking of handling signals
 (or not)?

andy

On Fri, Oct 23, 2020 at 11:31:48AM +0200, Kjetil Matheussen wrote:
> On Thu, Oct 22, 2020 at 10:26 AM Andy Farnell
> <padawan12@obiwannabe.co.uk> wrote:
> >
> > Would love to see Scheme come to Pd.
> >
> There's also k_guile from 2004, that still compiles:
> https://github.com/kmatheussen/k_guile
> Don't know if it still works though.
>
> Guess there are other alternatives as well, but they all probably
> suffer a bit from gc non-rt performance. To fix this I later extended
> the code from k_guile to run in a separate thread to avoid cpu spikes,
> but this code was only added to the snd sound editor (when snd was
> compiled as a pd external), and that code was probably removed from
> snd when switching from guile to s7.
_______________________________________________
Pd-list@lists.iem.at mailing list
UNSUBSCRIBE and account-management -> https://lists.puredata.info/listinfo/pd-list