hmm... seems to me that if you put your cxtxsxdx or whatever names in
/dev/scsi/dumbnames or something, and then make symlinks to them in the /dev
directory with names like sda1 or whatever your favorite names are you get
the best of both worlds. You can mount with your /dev/sda1, do e2fsck, or
whatever on it just as normal. If you do an ls -l on it, you see exactly
what device it is, and the names won't change if you plug a new device in.
Then, the only time you would have to futz with the "ugly" names would be
when you installed a new device (got to make a new link for it) or if you
had an emergency boot where you had to specify the root device on the kernel
command line.
Perhaps that naming scheme isn't the best, but it is short and gets the job
done, though I don't know where the c,t,s, and d letters come from. I
assume they stand for "controller, ID, LUN, and partition" but see no
logical connection between those letters and their meanings...
Evan Jeffrey
ejeffrey@eliot82.wustl.edu