[lsb-discuss] statfs specification addition: review

Theodore Tso tytso at mit.edu
Thu Dec 21 15:59:00 PST 2006


On Thu, Dec 21, 2006 at 01:18:25PM -0800, Wichmann, Mats D wrote:
> 
> >Sure, assuming that we've also added the xattr interfaces into the LSB.
> 
> which we haven't... not xattr, not libacl... are things like
> that mature enough to go in now?

Probably, although we should ask Andreas Gruenbacher out of courtesy
before we declare the ABI as cast into concrete.  It hasn't really
changed in quite a while, though, and looking at the interfaces, and I
can't see any reasons why they would change.  

The one potential worry I could see is that the ACL interfaces were
taken from an abandoned Posix 1003.1e draft (draft 17).  Since it's an
abandoned draft, we can't point to it, and so we would have to crib
the function definitions from the ACL man pages (unless we can get
permission from the IEEE to use the original 1003.1e draft 17 text).
The current libacl man pages are derived from the FreeBSD version
which were rewritten by Robert Watson such they aren't _obviously_
lifted from the Posix.1e draft, but if you compare the two there are
some striking similarities in the general structure of the function
definitions.  After all, there are only so many ways to describe a
function's behavior, but try telling someone like David Boise that.  :-)

Presumably there shouldn't be any problems with the IEEE getting
snitty about things, but it's one of those things that we might want
to look into ahead of time so we don't run into trouble later.

So do we have application writers asking for these interfaces?

						- Ted




More information about the lsb-discuss mailing list