home *** CD-ROM | disk | FTP | other *** search
- X-Gateway-Source-Info: INTERNET
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!news.acns.nwu.edu!network.ucsd.edu!mvb.saic.com!tgv.com!info-multinet
- Date: 29 JUL 92 22:57:54 GMT
- Newsgroups: vmsnet.networks.tcp-ip.multinet
- X-Return-path: <info-multinet-relay@TGV.COM>
- X-RFC822-From: FORREST@RIGEL.TAMU.EDU (Bob Forrest, Academic Computing Services)
- From: FORREST@RIGEL.TAMU.EDU
- Subject: Re: RFC931 support?
- X-Vmsmail-To: SMTP%"info-multinet@tgv.com"
- Organization: The INFO-MULTINET Community
- Message-ID: <238052D629JUL92225754@TGV.COM>
- Nntp-Posting-Host: Mvb.Saic.Com
- Lines: 101
-
- -> From: SMTP%"Adelman@TGV.COM" 29-JUL-1992 14:25:20.18
- -> To: Don.Rainwater@UC.Edu
- -> CC:
- -> Subj: Re: RFC931 support?
- ->
- -> Date: Wed, 29 Jul 92 11:26:22 PDT
- -> From: adelman@TGV.COM (Kenneth Adelman)
- -> Reply-To: Adelman@TGV.COM (Kenneth Adelman)
- -> Message-Id: <920729111738.25e00294@TGV.COM>
- -> Subject: Re: RFC931 support?
- -> In-Reply-To: Your message <1992Jul29.121720.1580@ucbeh.san.uc.edu> dated 29-Jul-1992
- -> To: Don.Rainwater@UC.Edu
- -> cc: info-multinet@TGV.COM
- ->
- -> > Some people (from other sites even) have been asking about support
- -> > for RFC931, which provides for the authentication of a connection between
- -> > two systems. For example, it would allow you to find out who that person
- -> > is that has a telnet connection open to your system from across the
- -> > country.
- ->
- -> > Are there any plans to support this in Multinet?
- ->
- -> Absolutely none. RFC931 is NOT an authentication protocol, it is
- -> a protocol which allows a user on a machine to claim that a particular
- -> connection is owned by a particular user. It doesn't provide any
- -> additional security, and a substantial false sense of security.
-
- There is on-going work on revisions to what was previously known as
- RFC931 and the 'authentication protocol'. The new revisions are known
- as the 'Identity protocol'. I think most people agree with the possible
- problems associated with 'authentication' (there has been a lot of debate
- on the ident mailing list...). However, I think that it has been argued
- fairly well that the upcoming Identity protocol is *quite* useful in
- tracking down forged e-mail and possibly in other forgery attempts.
-
- In a campus environment such as ours, we have a significant problem
- with e-mail forgeries. The platform management is quite trustworthy;
- however, $ TELNET/PORT=SMTP nodename.TAMU.EDU
- kind of bypasses that... Having the ability to provide identification
- to tcp connections would be very worthwhile. While for most applications
- we would not in any way be interested in using the tcp identification
- information in authentication, it would be quite helpful as an
- 'after the fact' method of trying to track down forgers and have student
- affairs (or some other appropriate organization on campus) deal with that
- person.
-
- Don't consider the RFC931 idea as an 'authentication only' scheme and
- blow it off. There are quite legitimate applications that would use the
- information for auditing purposes! (Yes, the auditing information
- would have to be interpreted, but then doesn't *ALL* auditing information
- have to be interpreted? :) .
-
- ->
- -> We'd be happy to implement a protocol which did what RFC931 does
- -> if someone could point out some value to it (other than security
- -> through obscurity), and only if it weren't titled "Authentication
- -> Server".
-
- Auditing is not security through obscurity. Any means of trying to
- track down forgeries would be better than the current method...
-
- Also, the title is changing. The new protocol will not be called the
- "Authentication Server" :)
-
-
- ->
- -> Ken
- ->
- ->
- ->
- ->
-
- How about TGV providing an interface callable from VAX C to return the
- process id associated with a given tcp socket pair between a remote tcp
- host and the host the call is made on? As for 'locking the kernel down'
- being a reason not to implement this, don't --- just return the best
- estimate (we can code other checks to see what the process is doing
- if we really need to). This would be quite helpful!
-
- Or if that doesn't go over, how about giving hints as to how to use
- the nlist(...) calls to obtain the information necessary to determine the
- process id?
-
-
- Thanks.
-
-
- Bob Forrest
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- Programmer/Analyst I HEPnet: FNBIT::UTDAL::RIGEL::FORREST
- Texas A&M University SPAN: UTSPAN::UTADNX::RIGEL::FORREST
- Academic Computing Services THEnet: RIGEL::FORREST
- WERC Room #020 BITnet: FORREST@TAMRIGEL
- Mail Stop 3154 SESquinet: FORREST@RIGEL.TAMU.EDU
- College Station, Texas 77843-3154 Phone: 409-845-9577
- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
-
-
-
-
-