Path: utzoo!utgpu!watserv1!ria!rrivax.rri.uwo.ca!lrb
From: lrb@rrivax.rri.uwo.ca (Lance R. Bailey)
Newsgroups: comp.sys.hp
Subject: Re: CTRL-C and rlogin
Message-ID: <1469@ria.ccs.uwo.ca>
Date: 20 Nov 90 11:21:13 GMT
References: <11382@milton.u.washington.edu>
Sender: news@ria.ccs.uwo.ca
Reply-To: lrb@rrivax.rri.uwo.ca
Organization: Robarts Research Institute -- London Canada Earth
Lines: 33
News-Software: VAX/VMS VNEWS 1.3-4

In article <11382@milton.u.washington.edu>, jdunlap@milton.u.washington.edu (John Dunlap) writes...
>I have an HP-9000 series 500 with HP-UX 5.21 and Wollongong TCP/IP
> ...
>Mostly we are happy, but the response to a control-C is exceedingly
>slow when a voluminous resonse to the screen is arriving:  the 
>output to the screen just keeps on coming.  
> ...
>What steps can I take to cause the unwanted part of the
>response to be discarded?
> 

you can't.
we do close to the same thing here. if a person grabs a tty physically
attached to the wrong computer, then they type the name of the one they
want and a telnet and/or rlogin to the correct machine is done. (ie:
at the login: they type VAX, and at a Username: they type HP)

the problem is that when you hit ^C there has been a lot of stuff already
sent to you and by the time ^C reaches the host, even more. this info
still gets sent (the problem is even worse on a 2400 dial-up.)

i beleive FUSION software has  a priority packet system that sends
certain characters (^C &c.) to the server BEFORE all other queued
packets. this works pretty well.

some IP servers (annex for eg.) also play some tricks to stop the
flow after a ^C is hit.  this is the route i am taking.

Lance R. Bailey, Systems Manager 
================================   box: Robarts Research Institute
email: lrb@rri.uwo.ca                   Clinical Trials Resources Group
  fax: 519.663.3789                     P.O. Box 5015, 100 Perth Dr.
  vox: 519.663.3787 ext. 4108           London, Canada N6A 5K8