Greg A. Woods <woods@most.weird.com> writes:

> I think I still lean towards a virtual filesystem as my first choice and
> a sysctl style interface (assuming it can be made much more dynamic than
> it is today) as my second choice; even for an SES front end on a real
> SCSI bus.  At the same time though I think most of the fancy stuff
> should be kept in user land.

Amen to that, which is why I believe a non-polling ioctl interface with
at most 4 ioctls (how many objects? gimme the map..? what's object N's
status? Set object N's status...) is the simplest and most portable
mechanism. But of course, I *would* think that...

-matt