From owner-freebsd-ports Tue Jun 27 01:14:42 1995 Return-Path: ports-owner Received: (from majordom@localhost) by freefall.cdrom.com (8.6.10/8.6.6) id BAA14982 for ports-outgoing; Tue, 27 Jun 1995 01:14:42 -0700 Received: from silvia.HIP.Berkeley.EDU (silvia.HIP.Berkeley.EDU [136.152.64.181]) by freefall.cdrom.com (8.6.10/8.6.6) with ESMTP id BAA14974 ; Tue, 27 Jun 1995 01:14:37 -0700 Received: (from asami@localhost) by silvia.HIP.Berkeley.EDU (8.6.11/8.6.9) id BAA01199; Tue, 27 Jun 1995 01:14:35 -0700 Date: Tue, 27 Jun 1995 01:14:35 -0700 Message-Id: <199506270814.BAA01199@silvia.HIP.Berkeley.EDU> To: me@freebsd.org CC: ports@freebsd.org In-reply-to: (me@tartufo.pcs.dec.com) Subject: Re: xemacs port in combination with ispell port From: asami@cs.berkeley.edu (Satoshi Asami) Sender: ports-owner@freebsd.org Precedence: bulk * If I'm not severely mistaken this won't work so easily. There might be * site-lisp code that is dependent on the editor being used - I.e. settings * for special xemacs or fsfmacs features. I agree that it's not *very* probable * that both xemacs and fsfmacs are insalled on the same machine, but one never * knows and would have to ensure that at least no files get overwritten with * versions from another type of editor. Point well taken. Well, we (the ports) aren't going to put anything into the site-lisp directory except for stuff like ispell (the editor ports don't put anything in there by default, right?). So unless we are going to ship some ports that depend on a particular variety/version of editor, and we ship multiple of them, we should be ok. On the other hand, individual site admins might want to put things in that directory as well, and the ambitious ones that put a lot of stuff (who also have both xemacs and mule, e.g.) might run into trouble. But I think those people can recompile the relevant editors without our "shared site-lisp hack" (we should provide an easy way to disable them). I think it's better to optimize the common cases, i.e., to make things like ispell work for everyone. But I only have emacs and mule, which are quite compatible (the latest version of mule follows the latest version of emacs quite closely) so I might be mis-judging the severity of the problem. Michael, what do you think? Satoshi