And also note in what Nick said in the referenced mail: it would be
nice if someone actually _cared_ about the async nature. The fact
that it has been broken from the start, and nobody noticed is a strong
hint that currently there isn't anybody who does.
Maybe fuse will be the first one to actually care, and then I'll
bother with putting a lot of effort into it. But until someone cares,
nobody will bother, and that's how it should be. That's very much in
line with the evoultionary nature of kernel developments: unused
features will just get broken and eventually removed.
Miklos
--