The Routing Over Large Clouds Mailing List Archive by date

[vjs@mica.denver.sgi.com (Vernon Schryver)]

> From: Juha Heinanen <jh@lohi.dat.tele.fi>                        
> 
> i find steve's comment about whose interest nhrp is totally
> inappropriate to a technical mailing list such as ion.  if steve doesn't
> like nhrp or anything else that ion is working on, i don't understand
> why he subscribes to this list.

For a long time, at least a decade, TCP/IP network stuff has been more
than 90% politics and less than 10% technical.  Over that last 10
years, as the nature of the IETF has shifted, the ratio has grown to
more than 95:5.  If you do not pay attention to the political (by which
I mean all non-technical) reasons why people support, oppose, or care
about what they do, they you have no hope of meeting your own goals,
even in the profoundly unlikely case that your concerns are purely
technical.

There is just not that much that is technically hard or intellectually
profound about network stuff.  The hard stuff comes from "navigating
the business issues," such as the technical, cost, price, and sales
consequences of the sizes of the buffers in ATM switches (now, 5 years
ago, or 5 years from now) the ATM services offered or not offered by
the PTT's, or the long term consequences of the design choices made by
router vendors on applications and protocols, and not to mention the
feedback back into the size of buffers, PTT services, and router vendor
choices.

Steve Deering's judgment might have been phrased a little less baldly.
He might have said something like "the clothes are transparent" instead
of blurting out the fact of the emperor's out-of-shape and aging body.
Still, if participants do not advert to the personal motives of all of
the players, from his "telecommunications companies" to the switch
vendors to the likely (and often obvious) interests and prejudices of
all individuals (including Steve Deering and Juha Heinanen), ATM will
continue to go around and around in circles.

Yes, from where I sit, ATM and IP/ATM has done a random walk in the
last few years.  There are been more motion than distance travelled.
Yes, of course, the reasons for that are mostly what I loosely labelled
"political".


Vernon Schryver,  vjs@sgi.com