Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!rutgers!apple!bbn!jr@bbn.com
From: jr@bbn.com (John Robinson)
Newsgroups: comp.emacs
Subject: Re: (call-process-region)
Message-ID: <36060@bbn.COM>
Date: 15 Feb 89 22:01:59 GMT
References: <1190002@hpcllcm.HP.COM>
Sender: news@bbn.COM
Reply-To: jr@bbn.com (John Robinson)
Organization: BBN Systems and Technologies Corporation, Cambridge MA
Lines: 20
In-reply-to: pratap@hpcllcm.HP.COM (Pratap Subrahmanyam)

In article <1190002@hpcllcm.HP.COM>, pratap@hpcllcm (Pratap Subrahmanyam) writes:
>I was looking at the code foe (shell-command-on-region) today and I find that
>there are ways to actually delete the region with the output from the shell
>command, but there is no way to not see the shell command output. I had to 
>go around and call (call-process-region) instead. 

>My question to the creators these defuns, why is that (shell-command-region)
>doesn't let you discard its output whereas (call-process-region) does. After
>all one seems to a logical subset of the other.

Hmmm.  It's as easy to type ^W after the shell command runs as to type
some prefix-arg before.  Or you could add a >&/dev/null to your shell
command.  (shell-command-region), I think, covers the common cases,
namely filtering some stuff and (re)running a shell command on
something, which you would sant for interactive use.
(call-process-region) is the more general function underlying the ones
designed to be interactive.
--
/jr
jr@bbn.com or bbn!jr