John Levon writes: > On Thu, Jun 05, 2008 at 07:58:39AM -0500, Mike Gerdts wrote: > > > > Are you worried about fstyp being wrong? If so, wouldn't a fall back to > > > guessing 'ufs' assuage that fear? > > > > So long as libfstyp works it will always be wrong by falling back to > > ufs. I'm somewhat inclined to just given an error message if libfstyp > > can't figure it out. > > That sounds good to me. It even fixes that annoying thing where you > attempt to mount some junk device and get (indirectly) told that it's > not UFS.
Yes, it does sound good. > Would this case also obsolete /etc/default/fs (I don't know offhand if > anything else cares about that?). If this project doesn't obsolete it (and remove the default_fs(4) man page), then I think it should respect the file by using that value as the default file system type in case libfstyp doesn't recognize the file system. Simply leaving this stuff in place and non-functional for administrators to stumble over doesn't sound like the right answer. > What about /etc/dfs/fstypes (once > again, not sure if anyone uses that)? I don't quite understand the relationship with that file. That's used to identify "remote" file systems. (I suspect you're talking about the "use the first line" trickery that [I think] is mostly involved with net booting. In that case, I don't think any sort of change is needed. We shouldn't run libfstyp on remote file systems such as NFS, should we?) -- James Carlson, Solaris Networking <james.d.carlson at sun.com> Sun Microsystems / 35 Network Drive 71.232W Vox +1 781 442 2084 MS UBUR02-212 / Burlington MA 01803-2757 42.496N Fax +1 781 442 1677
