Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Posting-Version: Notesfiles; site hpcnof.UUCP Path: utzoo!linus!decvax!decwrl!Glacier!oliveb!hplabs!hpfcdc!hpfcla!hpcnof!dat From: dat@hpcnof.UUCP Newsgroups: net.news.group Subject: Re: Orphaned Response Message-ID: <53200001@hpcnof.UUCP> Date: Sat, 26-Oct-85 18:32:00 EST Article-I.D.: hpcnof.53200001 Posted: Sat Oct 26 18:32:00 1985 Date-Received: Fri, 1-Nov-85 00:15:25 EST References: <10523@ucbvax.UUCP> Organization: 26 Oct 85 14:32:00 MST Lines: 42 So...what happens if you do something like this: % ln /notes/mod.laser-lovers /notes/fa.laser-lovers (or whatever the exact syntax is) The idea being that since the notes are kept merely as a set of files each with a number indicating it's ''unique'' message id on the local system, you could have the 'old dogs' reading the fa.* groups, and the new people coming onto the system reading the mod.* groups. This would also allow a very smooth transition, because you could keep the directory links for infinite time without ANY overhead at all (except perhaps a single inode). What would this do to the notes / readnews systems?? Also, how hard is it to have a 'translation' table on the gateway of a machine: other machines ----> translation ----> local in the <---- table <---- machine world so that entries like "translate fa.sun-spots to mod.comp.sun" would work both incoming and outgoing. (note that this would also allow systems with greater than 16 (!!) character filenames to have MNEMONIC names instead of CRYPTIC names. They'd just pity their neighbours..) This is perhaps what was refered to as the 'alias' table?? Leaping into the fray! -- Dave Taylor Hewlett Packard, CNO ...ihnp4!hpfcla!d_taylor or ...hplabs!hpcnou!dat