Subject: Re: conditional symbolic links
To: None <>
From: Bill Sommerfeld <>
List: current-users
Date: 11/24/1993 10:07:19
   i don't believe that they should be done at the VFS level, because
   you are now installing un-intended semantics in foreign filesystems.

   Adam Glass

I agree here; doing this "wrong" at the VFS layer might get in the way
of (e.g.,) an AFS-to-NFS translator which interpolated @sys on behalf
of the client.

Clearly the "right" way to do this is to implement a common
"@variable"-expander, invoked from certain VFS's (UFS & NFS certainly;
probably not isofs or msdos) when they detect an '@' in an entryname.

I think John Brezak meant "uname -n" when he said "uname -h".

					- Bill