Hey Spencer!
Indeed, updating RtAudio will allow us to more easily provide ASIO
support and sort out the Linux audio landscape in perhaps a more
sensible way, which I am very pleased about. Probably the most
immediate manifestation will be in the form of a separate ASIO binary
aimed at "power users"; down the line I envision command line flags to
select which audio backend you want to use, within a single binary.
Yes, this makes sense to me. For one thing this scenario already proved itself with the "unofficial" ASIO version that was there already. it'd be a good idea to also consider a ASIO version of The Mini, in this context. There was talk about this on the list in the past, searching for ASIO should get you some of Stephen Sinclair's notes on compiling it in.
By the way, do you have links to mailing list or forum posts related
to the Linux audio confusion? I must have missed or forgotten that
thread, and it would be good to understand more specifically whats
leading to the confusion.
Actually it was quite simple; there was a .deb package that a lot of new users took. Quite a sensible choice, of course. That package was for Jack, which is also a sensible choice... but then it turned out that quite a few of those new users were unaware of how to use Jack and didn't yet know how to start the server, etc. This led a disappointingly silent first experience.
I'd like to suggest that we try for a binary that first check whether there is a Jack server running and connects to it if there is. If not it'd fall back on ALSA. With no ALSA available it could take OSS as a last resort, though I wonder how likely that situation is in practice. If that could/would work we would probably be done with all of those questions for a very long time and it'd be very convenient too. I suppose flags should over-ride this, for example if Jack is running and using soundcard A and you want ChucK to use ALSA on soundcard B for some reason.
Does that make sense? Does it sound realistic?
Yours,
Kas.