home *** CD-ROM | disk | FTP | other *** search
- X-Gateway-Source-Info: INTERNET
- Path: sparky!uunet!wupost!usc!noiro.acs.uci.edu!unogate!mvb.saic.com!tgv.com!info-multinet
- Date: 30 JUL 92 01:00:02 GMT
- Newsgroups: vmsnet.networks.tcp-ip.multinet
- X-Return-path: <info-multinet-relay@TGV.COM>
- X-RFC822-From: "Ned Freed (Postmaster)" <NED@SIGURD.INNOSOFT.COM>
- From: "Ned Freed " <NED@SIGURD.INNOSOFT.COM>
- Subject: Re: RFC931 support?
- X-Envelope-to: Adelman@TGV.COM, info-multinet@tgv.com
- X-VMS-To: IN%"Adelman@TGV.COM"
- X-VMS-Cc: IN%"info-multinet@tgv.com"
- Organization: The INFO-MULTINET Community
- Message-ID: <23805E5230JUL92010002@TGV.COM>
- Nntp-Posting-Host: Mvb.Saic.Com
- Lines: 49
-
- RFC931 is obsolete. There is work underway to replace it with a somewhat
- modified protocol. The new protocol is called the "identification protocol".
-
- This functionality is _extremely_ useful in several situations. One that's
- near and dear to my heart is logging the identity associated with an
- incoming SMTP connection. This can be used to track many of the more obvious
- forms of mail spoofing back to their source. Sure, it is nowhere near secure,
- but it makes mail forging a matter requiring some technical expertise.
- Currently no technical expertise whatsoever is required.
-
- This sort of protocol is also very useful with informal interactive
- applications like Internet Relay Chat and TALK. It is also useful for
- tracking usage of things like Gopher.
-
- Nobody claims that any of this provides any sort of security or authentication.
- It certainly does not. But what it does provide is a reasonable mechanism
- for tracking certain types of usage in an informal manner. Often this sort
- of tracking is all that's needed, especially when more elaborate and
- provably secure schemes either have substantial cost (either computationally
- or financially, take your pick) or have very substantial restrictions
- (as in export).
-
- The identification protocol specification, once finished, will become a
- proposed standard. In other words, there is significant interest in this
- protocol and deployment of it has already reached a point where a standard
- for it is needed.
-
- Finally, there is a competing protocol of sorts -- the identification
- MIB. This essentially embeds the functionality of the identification
- protocol (plus a bit more stuff) inside of SNMP. It is actually an
- interesting tradeoff for several reasons. First of all, the identification
- protocol is TCP-based, and that means more overhead in terms of setting up
- and tearing down connections. (SNMP is UDP based and hence there's no
- connection setup involved.) However, SNMP imposes its own sort of overhead
- (ASN.1 encode/decode, multiplexed agents, etc.). And insofar as client
- code goes the TCP protocol approach is a hell of a lot simpler to do.
- But then SNMP is starting to acquire a reasonable security infrastructure...
-
- Anyway, there are a bunch of issues here, not the least of which is the fact
- that the development of the identification protocol specification has led
- to some really heated debate. (Readers of the IETF mailing list know what
- I'm talking about here!) But the bottom line is that this protocol is
- going to be popular (actually, it is already), it is going to be heavily used,
- it is very useful, and the problems with the protocol name have been
- eliminated. If TGV could see their way clear to implementing it it would
- save me the trouble of doing it myself ;-)
-
- Ned
-