Re: Reacting to previous messages

[Jon Crowcroft <J.Crowcroft@cs.ucl.ac.uk>]

In message <199906020838.BAA09345@flipper.cisco.com>, Liming Wei typed:

 >>> i am very concerend that cisco seem to think that IETF WG proposals
 >>> that dont fit in with their "easiest path to product lines" are so
 >>> worth supressing.
 >>I thought SM represents the "easiest path to product lines", because
 >>it is so "simple" ?

dont be silly. its not even called that anymore. i didnt claim it was
easiest to product anyhow - it is easy than the COMBINATION of all the
existing family of protocols you need, for SOME classes of
applciation. given these other sub-optimal things already exzist, then
clearly anythign new is harder....please stop asking me to explain the
obvious to people who know this already.
 
 >>Also, where is that official announcement from cisco ?

IF you read this list (SM) you'd see that a LOT of seperate cisco
engfineers have sent messages like daves. no one said anythign about
"officla.

 >>> proposal. we have not had a single ISP say that this will de-rail
 >>> their curent PIM deployment - if one does say,m then that too would
 >>> constitute possible (although more debatable) grounds for ceasing and
 >>> disisting.

 >>As to whether SM will de-rail the current PIM work, it is not that it
 >>will make a dent on PIM, it is the amount of attrition from
 >>non-results for most people. You might check how many companies are
 >>deploying PIM (or maybe easier to check how many without PIM), and how
 >>many companies are serious about SM including the authors
 >>companies/institutions.

sure - not just companies -we in our little academic net wiorh only
1.5 M hosts have deployed PIM - stop making it "PIM versus the world"

why do you think we want to "dent" pim.....? you have a bizarre idea
of our motivation. why are you (and lots of other cisco folks) so
defensive - what have yo uto lose if we are right or wrong - either
way, yo uget to have more products, features and sales...if we are
wrong, then the market will buy more boxes, and if we are right, the
market for multicast will expand too....

 >>BTW, the name "Simple Multicast" is not exactly a representation of
 >>the protocol machinery. E.g. CBT describes a property of its
 >>algorithm, so is the name PIM. But "Simple Multicast" is a
 >>substance-free name that makes RPF-flooding machinery look complex.

SM is not a routng protocol - like most the rest of multicast SM is a
change to the addressing system.

your comment about "substance free" is equally applicable to your
email - please make a technical comment or stop bothering us.
please get up to date on the discussion before saying anymore too.

 cheers possibly

   jon