[CM] s7: Snd in CM, CM in Snd anyway?

[Rick Taube]

another way would be to load a scheme only verision of cm into a s7. bill and i actually ran some experiments a year or so ago on using pure scheme, unfortunately at that time there was a huge speed differene running pure scheme vs the current method of using FFI calls to C++  to accomplish most things. also even a pure scheme CM will need some sort of object system, however crude (the current way is to is use struct like ‘objects’ ).  
it is also be possible to create a version of CM that used juce and but without using its GUI widgets but the code base would probably have tons of #ifdefs  in them. i actually had that at one point, but it didnt seem that anyone was using it and I can only support so many things.
and finally, the last way would be to add goops to s7, then you could load either CM2 or CM3 into it, however slow or fast it might be
On Jun 10, 2015, at 5:03 PM, Bill Schottstaedt <bil at ccrma.Stanford.EDU> wrote:

> If I remember right, the cm+clm process would write a sound
> file (via with-sound or whatever), then use X window properties
> (via sndctrl?) to tell Snd about it.  A motif-based Snd could
> still use this path, but I'd have to resurrect a lot of ancient
> code.  If you were happy with the no-gui Snd, I think it could
> be loaded into cm+clm as a giant library.  If you want the
> Snd graphics, then we have Juce colliding with Motif or Gtk,
> so we're back at the original case.  s7 watches stdin, so maybe
> we could use pipes or something: cm writes test.snd, then
> writes "(open-sound \"test.snd\")" to the pipe -- it will
> work the first time!
> 
> _______________________________________________
> Cmdist mailing list
> Cmdist at ccrma.stanford.edu
> http://ccrma-mail.stanford.edu/mailman/listinfo/cmdist
>