> Sounds great. It all sounded great until it got to the part about
> "federal funding". I, for one, have a problem with my tax dollars
> going towards some professor being able to gawk at another
> professor in a videoconference.
Part of the reason that this was ok was that the NSF was likely to be
picking up the tab for the alternative: a conventional dialed voice
call, at about 3 orders of magnitude higher price than the T3 NSFnet's
actual costs.
People using the NSFnet for AUP compliant voice only saved the
taxpayers money.
About the broader issue of the Internet II project, I have always
suspected that the "retail" market (100 M households) and the R&E
market (1000 largest R&E and industrial institutions) had incompatible
network requirements. The consequence is that no infrastructure could
be optimal for both. It would be a *good*thing* if the providers
could develop two parallel infrastructures: one default free, 100k
routes, typical peak user data rates near ISDN, to optimally support the
retail market.
A second parallel infrastructure only needs to support 1000 (rock
solid stable) routes, with (local) default routes over to the retail
network and sufficient traffic head room to keep the researchers
happy. The parallel network would be a premium service, with charges
per route as well as per connection. It could be trivially built
today (Is being?....) with todays technology, and it could
interconnect many of the existing R&E networks (vBNS, NSI, ESnet,
etc). Since the R&E networks can scale back the routing problem, they
can afford more detailed routing policies and a branchier layer 2,
supporting better global path optimization.
The hard part is getting all of the peering arrangements in place such
that traffic between the premium services of the different providers
does not pass through the congested retail infrastructure.
Internet II and the like are good things. No question!
--MM--
