Did anyone else notice the CAIDA skitter poster in the background of George Bush's speech at the NSA?

k claffy kc at caida.org
Tue Feb 7 21:02:07 UTC 2006


[warning NOT operationally relevant, just need to clarify]


re http://news.yahoo.com/photos/ss/events/ts/122805nsaspying/im:/060125/480/dcev10301252131;_ylt=Ag51RnYLYcMpHtd_Cq9ZJCNiWscF;_ylu=X3oDMTA3dmhrOGVvBHNlYwNzc20-

thanx to those who forwarded this, it was news to us.  note that the
AScore image has been used often by government agencies over the past 5
years, and its use doesn't imply any relationship.  but in interest of 
full disclosure (though i believe it's entirely unrelated to the photo-prop): 
in 2005 NSA did supplement (about $100k, and we hope more this year) an
NSF grant, in order to keep skitter (macroscopic topology measurement)
on life support.  Joel's DARPA and NSF URLs are from skitter.infancy --
before NSA's rescue maneuver, skitter hadn't had any earmarked funding
in over 5 years (Cisco has also helped fund some analysis, but funding
raw measurement is hard).  the NSA funding allowed us to keep it going
one more year, while we try to ascertain its role, if any, in some future,
if any, community-oriented measurement infrastructure:

	http://www.caida.org/outreach/papers/2005/conmi/

(comments/feedback More than welcome.)

in any event, there is nothing classified or covert in the skitter project
or measurement, and in particular we have never made any data available
to NSA that isn't available to any other researchers who can stomach
filling out the required web forms:
http://www.caida.org/tools/measurement/skitter/research.xml 
and in case of commercial use, join caida as a member, though we
kind of relaxed that requirement when all the commercially interested
customers were non-profits according to wall street's (apparently
even less validated) measurement methodology.

we have different access control policies for passive (tapped) data,
since that data belongs to the owner of the tapped link, so the owner
controls what happens to it.  (afaik, DHS is doing the most to advance
this ball on scaling sensitive data sharing ( http://www.predict.org )
in a way that ISPs support, but it's also severely underfunded within 
USG and thus slow to get off the ground.)  

not to drive this into the ground (unless it helps), but it bears 
recognizing/remembering that NSA has two missions, only one of which 
seems to get media attention: 
	http://www.nsa.gov/about/about00003.cfm
publically available macroscopic Internet topology data falls into the 
first.  anyone who thinks NSA needs caida for its SIGINT mission has 
never been brought up to speed on the capabilities of either.. 
(caida is not currently getting any commercial data (network upgrade -> 
funding new monitor, building new monitor, installing new monitor, etc.  
same cooperative trick, different decade.  we don't have it down yet..)

and finally, while we appreciate the international affinity for ASporn, 
our real goal (all along) is to improve the integrity of empirically 
grounded Internet science, e.g.,
	http://www.caida.org/analysis/topology/rank_as/
	http://as-rank.caida.org/
which additional funding is certainly necessary but not sufficient to
improve.  and no, no existing USG agency has financially embraced the
(open, anyway) Internet measurement mission.  which is arguably not
the worst case scenario, but i [re]read brin's 'transparent society'
[ http://www.davidbrin.com/tschp1.html ] last month, and it reads much 
more eerily with 8 years of empirical data to support his analysis. eeps.

sorry for latency, length, tangents, and operational sub-relevance,
but felt a need to clarify, 
k



More information about the NANOG mailing list