As a FreeBSD user, I have been wondering for years what the whole ALSA deal is about. All it really meant to me was that it made it very difficult for developers to write portable audio software, so that basically translates to “you can’t use a new version of Rosegarden unless you use Linux”. Thankfully, most important audio/synthesis applications work with OSS, so I can run snd, pd, csound, and other applications on FreeBSD, just as NetBSD or maybe even Solaris users are.

I’ve never had any problems with multiple audio channels on FreeBSD, or used a sound server, or anything like that, and I don’t get the hype behind “low latency” when that means your buffers are so small that mp3’s hickup at the slightest problem… Now, I see the benefit of that for synthesis applications like Csound or Puredata, but some kind of clever method is needed to switch between these types of applications.

You seem to really know your stuff, and I hope that people see the light and start using OSS again. (Then maybe it wouldn’t be so hard to port applications like Rosegarden, Ardour, etc… to other operating systems)

I am grateful that you offered a BSD licensed version to the BSD community. Given that at least on FreeBSD, an implementation of OSS is used, it would seem a logical step to incorporate more of the 4.0 code and/or API in the audio sources. I also agree completely that a well documented API is more important than being open source. (especially considering that in many cases the BSD folks can’t even use the alsa code at a kernel level due to the GPL being incompatible with any other open source license, the BSD license in this case — given that, how are third party developers supposed to document something which makes them risk infringing on others’ copyrights?). So I don’t expect good ALSA emulation on other operating systems than Linux to happen very soon, if ever. (by all accounts from the perspective of a user of an operating system that is not distributed under the GPL, ALSA does seem to be quite proprietary)

Ah if FreeBSD supported MIDI keyboards, I would never need Windows again…

P.S. Not that I’ve dealt with any of ALSA’s mixer utilities, but as a Scheme and Lisp programmer, I will tell you that Lisp and Scheme are probably far easier to deal with than the ALSA mixer tools (and infinitely easier to learn than C/C++/etc..)

It’s amazing how the things you’ve said specifically in this blog entry still fall on deaf ears. People are still repeating the “binary only” nonsense, even after you have confirmed is no longer the case.

Regards

At first we blamed PulseAudio because it’s generally weird and screwed up in Ubuntu, but that wasn’t it. Turns out doing “sudo alsa force-reload” got sound back on the laptop’s speakers once the headphones killed sound completely.

The sound hardware was the “snd_hda_intel” ALSA module and using the SigmaTel 9228 chipset. The issues were likely limited to that, but we don’t know for sure.

OSS4.1 completely fixed it, as installed per:

https://help.ubuntu.com/community/OpenSound

Kewl.