Just to make it clear - I do not vote for this or that way, these are just MHO.
> > > I'd like to see a transparent way of access to audio cds[...] > A very silly solution, why doesn't the > > > kernel handle this? > > > > Possible reasons: [...] > > 2. because it is easily done using user-level tools so adding this > > to kernel just bloats it without any obvious advantage > Each program is implementing this in his own obscure way. Many new users have > no clue as to how play CDS with xmms, have no idea you actually can type > audiocd:// in konqueror, and complain that audio CDS in general do not work. > If it does not belong in kernel (I've heard that argument for supermount as > well, but ok), As we already discussed - it is impossible to do what supermount currently does (i.e. harmless ejecting of busy media) without kernel modification in some way. So it does belong to kernel. Of course, there may be alternative implementation but IMHO supermount is the most simple one that combines both relative safety even for rw media with small amount of generic code changes (after current patches to VFS have been backed out). > than lets up someone creates a libaudiofs which all programs > will use (but I doubt this will let me: 'cd /mnt/cdrom; cp track*.* ~/' ?) I must admit I have near to zero knowledge about audio CD recording but if your example involves anything more than simple block reading - then no, it does not belong to kernel. -andrey
