Hi all, my name's Stefano and I'm an Italian university student in computer science and engineering. I'm developing, in my spare time, a kind of modular wrapper for sound effect plugins (something like imlib2 or gstreamer but for sound processing). Recently I just accidentally knew about BeOS, tried Haiku with QEMU on Linux and I found it beautiful and very interesting for what I'm gonna do for a couple of reasons: good real-time capability and the fact that it is a young project where a lot of things are yet to be done and where I see a big interest in making things better (that's not always true for open source/free software, believe me). As I already pointed out on the linux audio developers' mailing list, such a generalized framework for sound processing can bring three important improvements: 1. interoperability among different standards (obviously) 2. possibility to experiment with new processing paradigm (obvious), which coupled with fourier transforms / wavelets / whathever can speed up a great deal when processing networks or chains of effects (far less obvious) 3. maybe push forward the idea that sound processing can be done with objects which can be represented by simple data (math formulas, circuits, associative neural networks, etc.) instead of dynamically loaded plugins. There I didn't find much mind opening probably (maybe this was not the right moment since they're going to release and promote LV2) but anyway I think that the idea is quite cool, so if you're interested just let me know. At the moment I'm writing it in strict ANSI C, limiting OS-dependent stuff as much as possible (only use readdir, opendir and libltdl, so it probably will just work out of the box for Haiku without modifications). Regards, Stefano