Path: utzoo!mnetor!uunet!lll-winken!lll-tis!ames!pasteur!ucbvax!ncsa.ncsa.uiuc.EDU!gaige
From: gaige@ncsa.ncsa.uiuc.EDU (Gaige Paulsen)
Newsgroups: comp.protocols.appletalk
Subject: Re: problem with NCSA TELNET?
Message-ID: <8805010350.AA01046@zaphod.ncsa.uiuc.edu>
Date: 1 May 88 03:50:03 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 33



I don't know what Pacer is referring to, but I would like to answer a couple of
the questions posed relative to the 2.1 version of NCSA Telnet for the Macintosh

When NCSA Telnet terminates, it releases the AppleTalk port with which it is
listening for IP packets (port 72 for those who care).  If it didn't do this
then you could not run NCSA Telnet, quit, and run NCSA Telnet again without
getting the Appletalk Listener Installation error.

As for giving up the socket, I believe that the offending programs are reversed,
but I do not have much personal experience with pacer.  To my knowledge, pacer
installs a listener at boot time via an init which never gives up the socket
which must be listened to for IP packets (hence the need to increase the 
system heap size in older versions of the software).  NOTE: I am not certain
about this for current versions and therefore this should not be taken as the
final word, but it does correspond directly to the symptom of NCSA Telnet not
being able to get the socket when Pacer has been used.

By the way, NCSA Telnet does work with the CITI drivers which have a control
panel entry with which the socket listener may be temporarily uninstalled
(although CITI now provides a version of NCSA Telnet which has been modified
to run directly over their drivers).

I am very interrested in hearing any more comments on this subject as we are
quickly nearing the 2.2 release of NCSA Telnet sometime early this summer.


Thanks for your comments,
Gaige B. Paulsen
National Center for Supercomputing Applications


Disclaimer, I don't need no stink'n disclaimer.......