Lyvim Xaphir wrote: > On Wed, 2008-08-13 at 20:48 +0200, Mihai Moldovan wrote: > >> Hi Ly vim: pong :) >> >> _______________________________________________ >> oss-devel mailing list >> oss-devel@mailman.opensound.com >> http://mailman.opensound.com/mailman/listinfo/oss-devel >> >> > > A hearty "ping" back to you. Glad to be here. I've been an unsung fan > of oss, Hannu, Dev and all the devs I havent seen yet, since the mid > 90's. Thanks to everyone that helped me get my Aureal card working > under Linux back in the day. > > I'd like to see Hannu and Dev stay with this thing, so I'm here to help > any way I can, including trying to scratch up fundage. Recent talk on > Hannu's blog has had me worried. > No reason to be worried. I will continue working on OSS for at least few years. However it's getting apparent that some day working on OSS alone will not be profitable any more. We need to find some other products that give additional revenue.
Another thing is that the current sound hardware on the market is extremely lame. All systems have now HDaudio that is unbelievably complicated spec. The controller side is OK but the codec spec is far too liberal. It gives computer manufacturers to do whatever they want. The result is that every single PC will require changes to the driver before they work. There is something called UAA (Universal Audio Architecture) requirements that the PCs should met to get the "Vista compatible" sticker. However in practice this means that OSS should work exactly in the same way than the UAA driver. Unfortunately nobody outside Microsot knows what the exact algorithm is. The only alternative PCI (consumer) sound card that still exists is SB X-Fi. However writing a fully featured driver for it would take as much time than writing a complete operating system from scratch USB too is very complicated. Additional problem is that every single operating systems have their own USB stacks that are completely incompatible with each other. So getting USB audio to work under more than one OS at the same time is mission impossible. FireWire is lame too because no operating system (supported by OSS) has documented/supported interface for firewire drivers. Firewire could be supported (under Linux) using the oss_userdev driver but even that will require significant amount of work. The final problem is that practically all current OSS applications are seriously broken. They (mis)use features like non-blocking I/O incorrectly. It will be necessary to fix all of them before OSS can be successful. However in many cases it's almost impossible to get the applications to even build. If I get something fixed then it's impossible to find who maintains the code or at least they don't care to respond. So I have decided to try to shift development of OSS to a true community project. There are many "software based" features I would like to add. We have some commitments to Sun (our main sponsor). We do custom application and driver development on contract. However I feel that I should focus on completing the OSS documentation in the near future rather than doing the actual hacking. However my motivation to hack any new low level drivers is zero or negative. I have done it for 17 years and it's pretty much enough for one person. I can do it if some company manages to develop hardware that makes perfect sense. Otherwise help from the community is highly welcome. I see this even more important than financial funding (which is welcome too). Best regards, Hannu _______________________________________________ oss-devel mailing list oss-devel@mailman.opensound.com http://mailman.opensound.com/mailman/listinfo/oss-devel
