> I would still like to see a proposed plan, in some detail, of how you
> think it should work.  I don't believe that anyone has said that there
> are not limitations in the current scheme.  Other schemes also have their
> limitations.  If you make a detailed proposal, then we can evaluate the
> prospect of changing The Way It Works.  Without a detailed proposal, you
> will find a lot of resistance to change because it has worked this way
> for a long time and only caused minor problems/confusion/annoyance (and
> the occasional operator error due to confusion).  It will also require
> re-educating a lot of people if we do change it, so it can't be changed
> without real, tangible, defensable reason.
> 
> So, I repeat...  Give us a proposed better way of doing things.  I'm not
> asking for code--just a detailed description...

Three things:

1) Make sure the scsi name-to-controller/target/lun mapping scheme is
clearly documented in an easy-to-find place ("man scsi" would be
nice).  Make sure that the documentation describes how to hard-wire
particular names to particular controller/target/luns.

2) It Sure Would Be Nice If you could wire in a few scsi devices and
wildcard the rest...say wire in sd0, sd1, and sd2 and let sd3..sdN
match any other scsi devices that were found.  

3) If the mount commands for local file systems had an "-o label=xxx"
option, this could be included in fstab, and mount would refuse to
mount that device if it did not have label xxx.  Fsck could also check
the label and refuse to write to the device if it doesn't have that
label.

Keith