Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!samsung!uunet!dptechno!dave
From: dave@dptechno.UUCP (Dave Lee)
Newsgroups: comp.sys.hp
Subject: Re: Something useful for a change
Message-ID: <603@dptechno.UUCP>
Date: 31 Oct 90 22:41:14 GMT
References: <25191@uflorida.cis.ufl.EDU>
Reply-To: dave@dptechno.UUCP (Dave Lee)
Organization: D.P. Technology Corp. Camarillo California
Lines: 32

In article <25191@uflorida.cis.ufl.EDU> bb@reef.cis.ufl.edu (Brian Bartholomew) writes:
>> $ eval `resize`
>
>I would be happier to see the code that resize performs (a termcap
>inquiry and a couple setenv's?) added to the tcsh.  Since there is no
>SIGWINCH to catch, tcsh could just unconditionally perform the code
>each time it printed a prompt.  My understanding is that tcsh is
>already dependent on termcap now, anyway, for editing purposes.  Any
>takers?

Absolutely NOT !

The way resize works is by sending an inquire string to the window/terminal
and waiting for a response.  The window/terminal then sends back some string
that says what the current window size is.
However, if there is already typeahead in the input buffer then resize
gets really confused, or if you type something in echo mode while
resize is sending its inquire string, the window gets intermixed characters
and gets really confused.   

The only way to have this not happen is to set the terminal to raw, noecho 
mode, flush any typeahead, then hope the user doesnt type anything.
Not acceptable. (to me atleast).  Just try putting eval `resize` in your
.login for an xterm and then do a rlogin localhost and type something before
the resize is done (after the login messages).  Yuck.


 

-- 
Dave Lee
uunet!dptechno!dave