On Wed, Dec 19, 2012 at 7:59 PM, Tom R. [email protected] wrote:
we need to patch the script that looks for it.
This is vaguely similar to what happens under OpenBSD - PortAudio is
installed and working with other applications but Gnuradio doesn’t
find it. I think your detection method needs work, but I haven’t
looked at how you’re doing it.
In /usr/local/lib I have
libportaudio.a
libportaudio.la
libportaudio.so.1.0
This is a standard (for OpenBSD) installation and works with Audacity,
Fldigi, etc.
Trying to use Jack audio as an alternative is giving shm errors but I
haven’t followed them up. I could probably write a simple audio sink
easier than figuring out Jack audio’s problem. I’ve written easily
half a dozen programs that use the native Audio for recording. Don’t
know much about Linux sound though. And yes, having multiple sound
layers is common in Linux, FreeeBSD and OpenBSD. I decided to learn
the native, bottom layer, which in OpenBSD (and probably NetBSD) is
just called Audio. Supposedly based on Sun audio.
There’s a main parameter struct containing 2 member structs (for
record and play), and an ioctl that writes them to the sound card.
After that you can open() the sound card and read and write the same
as if it were a binary file. Very simple.
Alan