Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!usc!zaphod.mps.ohio-state.edu!wuarchive!udel!haven!adm!smoke!gwyn
From: gwyn@smoke.BRL.MIL (Doug Gwyn)
Newsgroups: comp.unix.questions
Subject: Re: Rlogin and control-S on SGI sysV machines.
Keywords: XON
Message-ID: <12216@smoke.BRL.MIL>
Date: 23 Feb 90 09:04:53 GMT
References: <595@celia.UUCP>
Reply-To: gwyn@brl.arpa (Doug Gwyn)
Organization: Ballistic Research Lab (BRL), APG, MD.
Lines: 21

In article <595@celia.UUCP> celia!keith@tis.llnl.gov (Keith Goldfarb) writes:
>By manually setting ixon mode on the local
>tty, I can make the ^S be interpreted locally, and hence its
>effect is immediate.

Of course, and in fact if you require DC3/DC1 flow control you should
LEAVE stty ixon -ixany in effect on your local system at all times!

>However, programs that use it meaningfully
>never get it (unless it is preceded by a ^V, but that is a 
>hassle).  My question is, how can I get the best of both worlds --
>a ^S that works when I type it, and that is heard by the remote
>host?

You're asking the wrong question.  The right question is, how in the
world can an application require that DC3 or DC1 be used "meaningfully"
when so many terminals require their use for in-band flow control?

I think you've been misled by the default that some versions of EMACS
have of expecting you to type ^S to search forward.  You should rebind
that to some more useful key.  Around here we use ^\ instead.