For music/audio stuff you can do the dsp stuff with c and then
communicate with another process written in a higher level language,
i.e python over OSC. The LAD irc crew have been discussing using this
idea for a new sequencer/daw thing.
Hmm this is interesting as I have been considering doing something
similar as well. I was wondering how much of a difference having a GUI
built in a non realtime style, either in an interpreted language or
similar setup, how this might affect things for audio use.
I was considering using OSC style communication between a backend
written in C/C++ and a GUI written in a seperate language.
Seablade