So then if everyone concedes that Linux audio is a mess, why hasn't anyone fixed it? The situation has been the same since I started using Linux in 2007.
Sure, there aren't as many problems now, but I still occasionally have problems with programs basically taking over the sound card, stopping Pulseaudio from working. Stuff like that which gets annoying.
PulseAudio is great, but it's more of a hack than anything else. Why do we need two sound architectures to achieve one thing? And if PulseAudio can stand in for OSS and Esound, then why can't it just stand in for ALSA?