home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.protocols.ibm
- Path: sparky!uunet!spool.mu.edu!umn.edu!csus.edu!netcom.com!netcomsv!dlb!dave
- From: dave@dlb.com (Dave Buck)
- Subject: Re: SDLC Point-to-Point Full Duplex
- Message-ID: <1992Dec19.001256.13187@dlb.com>
- Keywords: SDLC
- Reply-To: dave@dlb.com (Dave Buck)
- Organization: D.L.Buck & Associates, Inc.; San Jose, Calif.
- References: <BzAwJq.9nM@world.std.com>
- Date: Sat, 19 Dec 1992 00:12:56 GMT
- Lines: 22
-
- In article <BzAwJq.9nM@world.std.com> mdivya@world.std.com (Marigowda Divya) writes:
- >This is regarding SDLC Point-to-Point Full Duplex. Question is,
- >if the primary sends a Poll (eg: RR-P) followed by
- >couple of I-frames, should the secondary stick with the original
- >Nr (at the time the poll was received), or should it keep updating
- >the Nr; correspondingly, should the primary do the same?
-
- In fact numerous implementations including IBM's do seem to change the
- Nr value from I frame to adjacent I frame or RR. But your question is
- "should they", not "do they" ... that is certainly unclear from the
- IBM specs. I would be so bold as to state that a good implementation
- should accept such behavior from a peer, but provide an option for
- this behavior for local transmissions.
-
- >If this were to be true, then isn't it possible, that within one
- >POLL/FINAL exchange, any number of I-frames can be exchanged
- >between the Primary and the Secondary.
-
- Yes. Which makes it look very much like an HDLC implementation.
- --
- Dave Buck dave@dlb.com {amdahl,ames,daver,netcomsv,sun,zygot}!dlb!dave
- D. L. Buck and Associates, Inc.; San Jose, California; (408) 972-2825
-