On Mon, 10 Apr 1995 13:02:18 -0400 
 der Mouse <mouse@Collatz.McRCIM.McGill.EDU> wrote:

 > You did say "isa".  If you mean the bus used in many x86 family
 > machines, you should probably take it to port-i386 rather than
 > current-users.  Machine independence is relative; in the context of
 > current-users, nothing x86-specific can be machine independent.

Not necessarily...The i386-family isn't the *only* machine out there with 
ISA...Otherwise, why would there be `/sys/dev/isa' ? (or eisa, or pci...)

 > Extremely unlikely to happen, because it is bound to break

Absolutely un-true.  With your example of the sparc, simply translate the 
in-core BSD disklabel into a SunOS disklabel and write it.  The reverse 
already happens at read time.  While it is true that the BSD disklabel 
carries other information, there is likely spare space in the SunOS 
label *somewhere*...you just keep the extra information there, and grab 
it as appropriate at read time.  If there is no spare space, then what's 
the difference compared to what you already have?

 > compatability with native OS disk labels on most platforms.  For
 > example, NetBSD/sparc uses SunOS-compatible disk labels, but a pack
 > with an fdisk-style partition table on it for NetBSD/i386 will perforce
 > not have a valid SunOS disk label for NetBSD/sparc.  You would have to
 > special-case every port to know about every other port's disk labeling
 > schemes.

Not every port.  You bring it out into the MI layer so that everyone can 
use it...

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