[5018] in Kerberos
Re: Secure telnet/PPP/Kerberos/STEL/... (was Re: STEL: Secure TELnet -- Call for Beta Testers)
daemon@ATHENA.MIT.EDU (eichin@MIT.EDU)
Sat Apr 22 00:29:53 1995
Date: Sat, 22 Apr 95 00:19:05 -0400
To: Ed Phillips <flaregun@udel.edu>
Cc: kerberos@MIT.EDU
In-Reply-To: "[5015] in Kerberos"
From: eichin@MIT.EDU
> hehehe... last time I saw any telnet code, _it_ was a rat's
> nest! I think it all could use some cleaning up ;-)
Sure -- but with r*, it is the *protocol* that should be taken out and
shot. rsh needs to connect back to the client for stderr, which is
Just Wrong; rlogin uses BSD "pseudo-oob" data which is why rlogin -x
has never handled interrupts or resizes gracefully if you weren't on a
VAX running raw 4.3BSD; neither do any kind of negotiation.
ftp is perhaps a better example of a case where the protocol matters
more than the code -- the only kerberized (or gssapi'zed :-) ftp's
that I know of are standard command line clients, but given either a
Web Browser or a client with a user interface, like ncftp, why would
anyone want to use the standard clients, if they had a choice about it?
_Mark_
