Currently it seems to be impossible to play two or more
sounds in sequence seamlessly. Or to find out the
length of a sample.
It might also be useful if sounds could be generated
easily (generate a few hundred samples, maybe have it
cloned with a fast C routine to create a longer sound).
This would enable "C64-quality" sound effects with very
little code or data.
And does the panning function delay the other channel?
Or merely adjust the volume?
See also the request about SDL_sound or whatever.