[Portaudio] last call for V18 bug reports (C++//)

[Phil Burk]

Hello Pieter,

Thanks for the comments.

> I should have tested that CVS-snapshot much earlier, (sorry)...
> The only problem I now find when I compile pa_sgi is a
> C++ comment (//):

Right! I can easily scan the entire tree and kill any // comments. Thanks
for thinking of this.

> ok: pa_tests/patest_saw     (but it does not exit gracefully)

Crashes? Clicks? Laptop falls off table?

> ok: pa_tests/patest_record
> ..  pa_tests/patest_sine  gives coredump (bufSize 64 = too small?)

The implementation should be able to handle small buffers. If the user
buffer is too small for the host buffer than it should fit N users buffers
in a host buffer. When it is time to fill each host buffer, then PA should
call the user callback N times.

This allows the app to specify the smallest possible buffer size which could
allow very small latency on some platforms. If I recall, the Apple CoreAudio
folks claim that "theoretically" the latency under OS X could go as low as
one sample. So we want to allow small user buffers.

> Most tests thus still run ok.
> Or at least don't run worse than with V17.

Great! That is the most important thing. It would be a bit scary to throw in
a large change at this late stage. So I am inclined to use the current SGI
implementation. But let me know if there is an simple, safe, secure, robust
fix for 'pa_tests/patest_sine.c".

Phil Burk