Path: utzoo!news-server.csri.toronto.edu!cs.utexas.edu!uwm.edu!ogicse!morgan
From: morgan@ogicse.ogi.edu (Clark O. Morgan)
Newsgroups: comp.mail.mush
Subject: Re: Spell_checker for MUSH v7.2.0
Message-ID: <18409@ogicse.ogi.edu>
Date: 10 Mar 91 00:34:27 GMT
References: <1991Feb27.230011.17092@brolga.cc.uq.oz.au> <91067.125623QQ11@LIVERPOOL.AC.UK>
Organization: Oregon Graduate Institute (formerly OGC), Beaverton, OR
Lines: 31

In article <91067.125623QQ11@LIVERPOOL.AC.UK> QQ11@LIVERPOOL.AC.UK (Alan Thew) writes:
> [ about a method of invoking "ispell" from to spell check a message ].
>
>This works very well with ispell, mush 7.1.2 and HP-UX 7.0.

Works on BSD4.2 Symmetry as well.  Okay, here's a request for a future
enhancement.  It would be nice if Mush acquired the following new
escape sequence:

   ~L [<program>]

   Semantics:  ~L extracts the Subject: line and body of the message
   currently being prepared by the user and places them in a temp file.
   This temp file is passed as a command line argument to <program>.  If
   <program> is omitted, the program string in the variable $spell is
   used instead.  When <program> exits, the Subject: line and message
   body are extracted from the temp file and appropriately pieced back
   into the original message file.  The temp file is deleted.

I make this request because the current method of using "~v ispell" has
the drawback of passing the entire message (including all headers) to
ispell.  And of course, ispell checks the headers for spelling errors
and of course, ispell doesn't like most of the data included in a
standard RFC-822 message (particularly message addresses).  Sooooo, the
user gets to wade through an initial spurt of misspellings every time
the program is invoked.

BTW, as always, I am a great fan of Mush.  This is an enhancement request.
This is _not_ a complaint.
-- 
Clark O. Morgan     morgan@cse.ogi.edu     ...!uunet!ogicse!morgan