Brian: Okay, here's the deal. Everyone who files a release-critical bug on 3.3.x, you send me their names, and I'll kill them. That way, I figure we can all get 4.0.1 a little sooner, and start filing release-critical bugs on that :)
Thanks for the news. Will cause much rejoicing around the office.
Heart troubles to cap off the mess of the weekend. Carefully staggered around the office, leaning on things most of the day. By afternoon, I was able to stop leaning on things for support, and people stopped saying things to me like, "You really look like shit, you know that, Dancer?"
Oh...and of course, the timeless classic that I keep hearing: "Your health is more important than anything. You should go home." Yeah...until you need something done urgently...then my health can take second-place, right?
Cooking up a better communications model between MIDI drivers and the mixer. Heck, I can actually use it to maintain multiple input streams into the mixer, but that would be expensive memory-wise (but possibly useful...under the current system you can't cancel a sound. Once it's mixed it's mixed. There's also no clear determination about when a sound is finished [Though I have figured a way around that]...on the plus side, memory use is a constant, and there's no real limit to the number of sounds you can mix simultaneously, other than the CPU which is not heavily occupied by a mix)
Adding streams to be mixed would allow us to cancel a stream any time, or possibly even indulge in submixing at differing volumes....but since more memory would be used (buffers per stream), we would have to place some kind of upper boundary on the number we can permit. Hrm.
BTW, no, I'm not any kind of an audiophile. Just a hacker and a geek.