Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!uunet!unisoft!greywolf From: greywolf@unisoft.UUCP (The Grey Wolf) Newsgroups: comp.unix.internals Subject: Re: (was slashes, now NFS devices) Message-ID: <3425@unisoft.UUCP> Date: 7 Mar 91 23:43:13 GMT References: <15236@smoke.brl.mil> <123382@uunet.UU.NET> <1991Feb22.141910.17013@decuac.dec.com> <14363@ulysses.att.com> <468@appserv.Eng.Sun.COM> by peter@ficc.ferranti.com (Peter da Silva) & In article <468@appserv.Eng.Sun.COM> lm@slovax.Eng.Sun.COM (Larry McVoy) writes: & > Nobody, including RFS, has ever come up with networked devices in any & > sort of general fashion. & & Intel's OpenNET seems to do a pretty good job of it. Consider (bridge is & running System V/386, xds13 is running Xenix/286): & & % net name & //xds13 & % stty -a < //bridge/dev/console #if (the net name "directory" is not freely configurable) AAAAAUUUUUUUGGGGGGGGGHHHHHHHHH! NONONONONONONONONO! Tell me that this is not becoming COMMON! I think I'd go quietly quite mad (sane?) were that the case! The "//" notation can have some bad effects if your program has a tendency to expand ~$user/bin (or however you have home directory expansion, tilde just seems to be the most common), especially if $user == "root". Why don't they just standardise on /..., or /net, or allow arbitrary mount points...? The Apollo workstations use this scheme; and some things break there, too. "So don't write your programs like that!" That's a cop-out. "//" is an inherently bad design. #else As long as it's configurable that's not bad. #endif & -- & Peter da Silva. `-_-' peter@ferranti.com & +1 713 274 5180. 'U` "Have you hugged your wolf today?" -- # On the 'Net: Why are more and more fourth-level wizard(-wannabe)s trying # to invoke ninth-level magic instead of taking the time to climb the other # (quite essential) thirteen or fourteen levels so they can do this properly? # ...!{uunet,ucbvax}!unisoft!greywolf