Path: utzoo!attcan!uunet!ogicse!uwm.edu!zaphod.mps.ohio-state.edu!rpi!uupsi!njin!spcvxa!terry
From: terry@spcvxa.spc.edu (Terry Kennedy)
Newsgroups: comp.os.rsts
Subject: Re: TCP/IP for RSTS (was: Finger V1.1 bugfix - minor)
Message-ID: <52.26075dbb@spcvxa.spc.edu>
Date: 21 Mar 90 10:55:55 GMT
References: <49.2604a23f@spcvxa.spc.edu> <16786@hercules.csl.sri.com>
Organization: St. Peter's College, US
Lines: 22

In article <16786@hercules.csl.sri.com>, kph@dustbin.cisco.com (Kevin Paul Herbert) writes:
> Finger would be really useful if someone had TCP/IP for RSTS. I know
> that a few years ago, somebody asked me a bunch of questions about how
> to implement TCP/IP for RSTS, but don't know how far they got.

  Well, for outbound Finger requests from RSTS/E, if you have any of the
TCP-aware operating systems on the same DECnet (Ultrix, VMS, TOPS-10/20),
the RSTS Finger code will transparently route via that node, so it isn't
a limitation for outbound requests.

  For inbound requests, you can do the poor man's routing hack with mul-
tiple @'s, like: finger terry@spc11b.spc.edu@spcvxa.spc.edu, which will
use the TCP link to a TCP-aware host and then go via DECnet to the RSTS/E
system.

  Also, Finger is used on a number of BITNET machines, and in that mode
the 11's are directly FIngerable from the net, and vice versa.

  I'm not saying that TCP/IP for RSTS is a bad idea (far from it!), but I
do want to point out that Finger is useful without it.

	Terry Kennedy