Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!think!think.com
From: rlk@think.com (Robert Krawitz)
Newsgroups: gnu.emacs.bug
Subject: Re: rmail.el should obey bash's MAILPATH
Message-ID: <32484@news.Think.COM>
Date: 28 Dec 89 15:42:00 GMT
References: <8912261600.AA01743@ouareau.IRO.UMontreal.CA>
Sender: news@Think.COM
Reply-To: rlk@think.com (Robert Krawitz)
Distribution: gnu
Organization: Thinking Machines Corp., Cambridge MA
Lines: 32
In-reply-to: mcgill-vision!iros1!pinard@EDDIE.MIT.EDU (Francois Pinard)

In article <8912261600.AA01743@ouareau.IRO.UMontreal.CA>, mcgill-vision!iros1!pinard@EDDIE (Francois Pinard) writes:
]When rmail-primary-inbox-list is nil, rmail.el(rmail) dynamically
]constructs it using rmail-spool-directory, then $LOGNAME or $USER.
]Under bash, the user inbox is described by $MAILPATH.  I believe rmail
]should first check for the existence of $MAILPATH first.

Does any program but bash actually use this convention?

]On this site, primary inbox names depend upon the NFS location of the
]home directory.  So, the value of rmail-spool-directory cannot be made
]unique for all users in paths.el.  If one manages to have MAILPATH
]properly set, everything else should ideally fall in place.

1)  What do mail/Mail/mh users do?  None of these programs, from
straight out of the can, use MAILPATH to the best of my knowledge.

2)  This is simple enough to do, in any event.  Simply put the following
in default-profile.el:

(setq rmail-spool-directory (or (getenv "MAILPATH") rmail-spool-directory))

(you may wish to set rmail-primary-inbox-list or whatnot).

Putting everyone's favorite site-specific hack into rmail (or any other
emacs lisp program) is a bad idea.  It increases the amount of clutter,
which makes the code harder to maintain and configure.  The whole point
of having an extension language like emacs lisp is to enable individuals
and sites to do significant customizations themselves.
-- 
ames >>>>>>>>>  |	Robert Krawitz <rlk@think.com>	245 First St.
bloom-beacon >  |think!rlk	(postmaster)		Cambridge, MA  02142
harvard >>>>>>  .	Thinking Machines Corp.		(617)876-1111