iiit-ih@cybaswan.UUCP (i.izikowitz) (09/12/88)
Is there any way of implementing the broadcast of messages
in a local area network which conforms to the OSI standard
reference model?
Would this be possible with a modification of the 7-layer model to
deal with messages of this type?
Note: By "broadcast" I mean completely unacknowledged!
-----------------------------------------------+------------------------------
Ivan Izikowitz, | Phone (+44) 792 295213
Networking and Systems Engineer, | Fax (+44) 792 295532
Institute for Industrial Information Techology,| Telex 48149
Innovation Centre, University of Wales, +------+ JANET: iiit-ih@uk.ac.swan.pyr
Swansea SA2 8PP | UUCP: ..!ukc!cybaswan.UUCP!iiit-ih
----------------------------------------+-------------------------------------
My views are not necessarily those of my employers!mm@itivax.UUCP (Michael McFarland) (09/16/88)
In article <82@cybaswan.UUCP>, iiit-ih@cybaswan.UUCP (i.izikowitz) writes: > > Is there any way of implementing the broadcast of messages > in a local area network which conforms to the OSI standard > reference model? > > Would this be possible with a modification of the 7-layer model to > deal with messages of this type? > > Note: By "broadcast" I mean completely unacknowledged! > > Not currently. (ISO lets you go connectionless all the way up to Presentation.) But the question is being addressed by the ANSI T5.1 committee. (As an addendum to the 7498? DP 9545?). If you want more info about the committee, send me a message. Spike McFarland ITI PO Box 1485 Ann Arbor, MI, USA 48106 (313) 769-4306 { These views are not necessarily those of my employer. }
collin@hpindda.HP.COM (Collin Park) (09/20/88)
In article <82@cybaswan.UUCP>, iiit-ih@cybaswan.UUCP (i.izikowitz) writes: > > Is there any way of implementing the broadcast of messages > in a local area network which conforms to the OSI standard > reference model? I am assuming that you are here referring to getting reliable data transfer amongst "N" application-entities, but using LAN broadcast packets to minimise overhead. If this is what you intend, then the response of mm@itivax.UUCP (Michael McFarland) is correct. But if all you are interested in here is the use of broadcast unacknowledged LAN packets to implement unacknowledged data transfer, then the following might work: The connectionless network protocol (ISO 8473) currently contains an "inactive subset" (signalled by a single-octet network-layer header, i think containing zero). If all the AEs wanting to communicate have the same set of selectors (T,S,P), then you could broadcast packets with that set of selectors in the appropriate headers and the "inactive subset" network header. > > Would this be possible with a modification of the 7-layer model to > deal with messages of this type? I believe the MPDT (multi point data transfer, formerly "N-way" project within ISO/IEC JTC1/SC21/WG1 is addressing any potential modifications to ISO 7498. The last I heard (about a year ago), they were trying to get objectives from the "using" working groups (e.g., Connection oriented or connectionless? Do you need to save bandwidth or time or overhead or <whatever>, and in what priority order? etc etc). If you have opinions on this, I would recommend that you get involved with your national body to put your opinions forward, whether this be through the SC21/WG1 or one of the "using" groups, e.g., SC21/WG5 or WG6. > > Note: By "broadcast" I mean completely unacknowledged! > > mm@itivax.UUCP (Michael McFarland) writes: > Not currently. (ISO lets you go connectionless all the way up to > Presentation.) But the question is being addressed by the ANSI T5.1 ^^^^^^^^^ I think you mean "ANSI X3 T5.1" here. This is the sort of projection of JTC1 SC21/WG1 in the US. (There is a set of US "T" committees, e.g., T1D1.2, which are disjoint from the X3 subcommittees.) > committee. (As an addendum to the 7498? DP 9545?). > ... my opinions do not necessarily reflect those of my employer or anybody important. -collin park