[ from advogato, entry #35 ]

take two. galeon crashed. twice. hmm.

linux-wlan-ng

The aforementioned auto-magic kernel build detection turned into a minor enema of the build/makefile system. We now automagically pick up kernel version, build CFLAGS, modversions crap, kernel pcmcia detection, and all sorts of other fun stuff.

I'm holding off on a release of 0.1.15-pre6 until I get some promised ARM fixes. This will hopefully be the final -pre, as I don't have the time to tackle the interrupt-driven cmdcompletion anytime soon. Which incidentally is a big scary hairball. I'd also like to test out the new locking code in a controlled environment. I only have one report so far, and it worked better, but still has issues.

freesci

Updated the adlib emulator to the latest-and-greatest, which should hopefully fix some of the lingering quirks. The relative volumes aren't quite right yet, but they're better. I even managed to preserve the 12-voice code yet again. One of these days I'll finish extending that to the full sixteen voices.. but first, need to find out why the pcm subsystem (and/or emulator) is crashing under Win32. I also need to finish the re-write of the ALSA pcmout driver, as the async model doesn't quite work. I was hoping to avoid threading it, but c'est la vie.

[censored] driver

It works on the embedded target now -- at a paltry 96kbps peak. As opposed to ~95kBps on a PC. :) If we turn down the debug output, it stops working, presumably due to bus timing problems. Looks like we'll have to get a new revision of the hardware before we can make it sing.