> I think you're lacking a bit in historical perspective. For quite a > long time, *most* of the actually useable apps we had came from other > unices. When I first came here (1997?), the best soundfile editors I > could find were DAP, MiXViews, and Snd. IIRC, *all* of those were > developed on SGI or some other non-linux system. > > And it wasn't really so long ago that we had no sound at > all. Considering that Linus originally thought that Linux would never > run on anything very different from the particular x86 hardware he > had, I don't think we should fault him too much for agreeing to an > expedient general solution at a time when there was nothing but a few > unrelated drivers for a few soundcards.
What was, was. Leave it alone, let it be. Our lesson of the day is that OSS is past its prime (it arguably never had one, but that is beside the point), and as such should be replaced by a better counterpart, in this case Alsa. Defending OSS's historical merits (or lack there of) has nothing to do with OSS's current state and its effect on the LAD community. It might have provided useful things to Linux (and did), but if we here on LAD are not prepared to do away with the primary stalling mechanism of the further Linux audio revolution, then who will? >I would say lack of *desire* to do anything else on the part of all but a >very small group... most of whom are here and/or on the alsa lists. So does this mean that we should just build what is barely enough to keep the OS running and leave it at that (and then waste countless days trying to hack the lousily implemented audio architecture trying to convince it to do something it was never designed to do)? Kernel devel people are constantly pushing envelope and working on improving the linux kernel, even though it is already probably the most efficient kernel out there, and even though 70% of the kernel features will not be used by most users. Yet they keep on improving upon it (granted not every time with a positive result, but this is again beside the point)... We here are concerned with linux audio, and even though most of the regular users out there will not notice the difference between a 50ms and 2ms audio latency, we will still make linux do 2ms. Why? Because we can, and because we need it, while at the same time everyone else benefits from it, even if they are not necessarily aware of it. This is what makes [should I say "will make"?] Linux so great... Ico