On Thu, 10 Aug 1995 10:45:51 -0700 (PDT) 
 wrstuden@loki.stanford.edu (Bill Studenmund) wrote:

 > I had in mind a preceeding step where we look to see what OS was used,
 > _then_ we look at the disklabel. Granted this wording is skewed towards
 > ports where UNIX is not the native OS.

Ummm ... You have to know where to look for these signatures (if there 
even _are_ any) too.  So, it's really the same problem as looking for the 
disklabel...

 > > That's what my code did ... it gave preference to the port's `native' 
 > > disklabel format, and then jumped though function pointers looking though 
 > > the other formats until it didn't fail anymore.  If none of them 
 > > succeeded, `no disk label' was reported and a fake label generated.
 > 
 > So basically my suggestion is that something similar to your code
 > gets worked into the kernel. :-)

Well, it'll have to be _written_ again.  :-)  I'm sure I mentioned that I 
didn't have the code anymore ... it was ... less than pretty, and, like I 
said, was thrown together in, like, a day as a proof of concept.  It was 
also an interesting excercise in discovering what sort of issues were 
involved in such a thing ...

--------------------------------------------------------------------------
Jason R. Thorpe                                       thorpej@nas.nasa.gov
NASA Ames Research Center                               Home: 408.866.1912
NAS: M/S 258-6                                          Work: 415.604.0935
Moffett Field, CA 94035                                Pager: 415.428.6939