rich@sendai.sendai.ann-arbor.mi.us (K. Richard Magill) (09/07/89)
In article <14090@polyslo.CalPoly.EDU> dorourke@polyslo.CalPoly.EDU (David M. O'Rourke) writes:
Local talk has a limit of 32 nodes [...]
ethertalk doesn't.
hmm, see above about this many nodes, but you're also going to get a
lot of collisions which will make the macs backoff and retry, by that
time someone has already gotten node #1 and probably will respond. [...]
Not over ether.
Talking with some friends they say that they turn appletalk off to
get the best MIDI performance. I'm sure this situation could come
up, but again I think you're hitting extremes that won't be met all
that often. [...]
Presume for the purposes of my argument that we have an ideal LAP.
ie, node numbers are not limitted and that the network bandwidth
verges on infinite.
My point here is that appletalk appears to be full of holes and
doesn't even specify ways to recover from these blunders.
--
rich.desnoyer@apple.com (Peter Desnoyers) (09/08/89)
In article <RICH.89Sep6140332@sendai.sendai.ann-arbor.mi.us> rich@sendai.sendai.ann-arbor.mi.us (K. Richard Magill) writes: > In article <14090@polyslo.CalPoly.EDU> dorourke@polyslo.CalPoly.EDU (David M. O'Rourke) writes: > > Local talk has a limit of 32 nodes [...] > ethertalk doesn't. > Come on now. The original discussion was over the dynamic assignment of local network addresses in the LocalTalk MAC layer. Ethernet is completely irrelevant to the question at hand, as it uses a different means of assigning addresses. (i.e. pre-assigned addresses.) A valid question about Ethernet would be whether the address resolution protocol is safe against the assignment of duplicate AppleTalk node numbers. I don't know the answer to this one. Peter Desnoyers Apple ATG (408) 974-4469