1314jb@houxf.UUCP (J.BOKOR) (12/10/84)
I am amazed at the naivete of those who underestimate the capacity of the NSA. They have a budget which is kept secret from the *president* himself. They won't tell him because they're afraid Stockman will find out. You better believe they're listening to everything, including encrypted messages especially. Crypto-freaks out there back me up on this...the NSA has even been known to certify encryption methods as "safe" for use by the general community, which they have themselves cracked so that *they* can then monitor the "secret" public traffic. "Never underestimate the power of a schnook." - Boris Badunov, World's Greatest Nogoodnik
stekas@hou2g.UUCP (J.STEKAS) (12/11/84)
> I am amazed at the naivete of those who underestimate the capacity > of the NSA. ... You better believe they're listening > to everything, including encrypted messages especially. To monitor "everything", the NSA would have to be doing keyword searches on about 10^12 bps of data during the busy hours of each business day. Figuring that each byte of data needs at least 8 clock cycles to process, that translates into more than 10^6 MIPS worth of processing. That's about 1000 Crays and we haven't even begun to talk about encrypted data (and won't all the interesting stuff be encryted) or the network that ties them to the telephone network. Considering what it cost the Bell System to put together a network which could recognize the numbers 0-9 and close the right relays, it is impossible to believe that the NSA could be doing speach recognition on EVERY line. Jim
1314jb@houxf.UUCP (J.BOKOR) (12/11/84)
>> I am amazed at the naivete of those who underestimate the capacity >> of the NSA. ... You better believe they're listening >> to everything, including encrypted messages especially. >...that translates into more than 10^6 MIPS worth of processing. That's >about 1000 Crays and we haven't even begun to talk about encrypted data... >...it is impossible to believe that the NSA could be doing speach (sic) >recognition on EVERY line. Oh, come on, grow up! (-:) First of all, 1000 Crays is a gross everestimate of the expense of carrying out such surveillance. Be creative! This is a *highly* parallel problem. You want a tremendous number of identical pattern recognition machines. Think VLSI. I concede that they can't handle the speech traffic, but don't be so quick to dismiss the possibility of large scale electronic data surveillance. I also concede that they don't listen to absolutely *EVERYTHING*, they have to be selective. Also, don't think that the NSA are the only ones out there listening, there's also the Soviets and the Japanese to think about, just to name two.
gjk@talcott.UUCP (Greg J Kuperberg) (12/12/84)
> I am amazed at the naivete of those who underestimate the capacity > of the NSA. ... You better believe they're listening > to everything, including encrypted messages especially. Yeah, I heard they really go for the ROT13 jokes... In fact, I found a NSA agent hiding under my bed just the other day. --- Greg Kuperberg harvard!talcott!gjk "Strange women lying in ponds distributing swords is no basis for a system of government." -Monty Python
jsq@ut-sally.UUCP (John Quarterman) (12/12/84)
To get some idea of what NSA is really like, try this book:
The Puzzle Palace (A Report on America's Most Secret Agency)
James Bamford
Houghton Mifflin Company Boston
Copyright (c) 1982 by V. James Bamford
ISBN 0-395-31286-8
--
John Quarterman, CS Dept., University of Texas, Austin, Texas 78712 USA
jsq@ut-sally.ARPA, jsq@ut-sally.UUCP, {ihnp4,seismo,ctvax}!ut-sally!jsqbaba@spar.UUCP (Baba ROM DOS) (12/14/84)
> >> I am amazed at the naivete of those who underestimate the capacity > >> of the NSA. ... You better believe they're listening > >> to everything, including encrypted messages especially. > > >...that translates into more than 10^6 MIPS worth of processing. That's > >about 1000 Crays and we haven't even begun to talk about encrypted data... > > Oh, come on, grow up! (-:) First of all, 1000 Crays is a gross > everestimate of the expense of carrying out such surveillance. > Be creative! This is a *highly* parallel problem. You want a > tremendous number of identical pattern recognition machines. Think VLSI. Well, I always liked the story that the *real* reason that IBM shut down it's Josephson junction project was that it had delivered the specified number of systems to the NSA and was fulfilling the final clause in the contract. Baba
jhull@spp2.UUCP (12/14/84)
In an earlier article, somebody wrote: >> I am amazed at the naivete of those who underestimate the capacity >> of the NSA. ... You better believe they're listening >> to everything, including encrypted messages especially. >>...that translates into more than 10^6 MIPS worth of processing. That's >>about 1000 Crays and we haven't even begun to talk about encrypted data... >First of all, 1000 Crays is a gross overestimate ... >I also concede that they don't listen to absolutely >*EVERYTHING*, they have to be selective. The basic procedure is: 1) Listen to everything you can. 2) Place each data source in a category by how interesting it is, i.e., how often it will be scanned fully in the future 3) Use the capacity released by 2 to scan something else. Make the scanning frequency assignments of 2 such that every data source gets scanned at least once and at least sometime after that Here, we are only referring to digital traffic. Analog traffic is something else entirely, but the basic procedure is similar. >You can bet they stopped those practices after the "kremvax" affair.... > > Piet Beertema, CWI, Amsterdam >... something a public as "news" ... >(2) For other stuff, ... under the relevant federal court >decisions interpreting the federal wiretap statute, ... > + Donald E. Eastlake, III You can ABSOLUTELY count on monitoring continuing today and probably forever, regardless of what the courts say. I don't know what "those practices" are. -- Blessed Be, jhull@spp2.UUCP Jeff Hull trwspp!spp2!jhull@trwrb.UUCP 13817 Yukon Ave. Hawthorne, CA 90250
dhc@petrus.UUCP (12/15/84)
and that's the list of NSA keywords. Amazing! -- David H. Copp
rjc@ubu.UUCP (Richard Caley) (12/15/84)
> From > flame!qtlon!ukc!mcvax!seismo!hao!hplabs!tektronix!uw-beaver!cornell! > vax135!ho uxm!houxf!1314jb Thu Jan 1 00:00:00 1970 > From: 1314jb@houxf.UUCP (J.BOKOR) > > You better believe they're listening to everything, including > encrypted messages especially. Hey! I'm moveing all my foul pinko gay subversive commie plots to net.jokes :-) <- Just in case. -- -------------------------------------------------------------------------- "In the beginning was a flame ...... " Paul Kantner. .......... mcvax!ukc!flame!ubu!rjc [ Any opinions in the above crawled in while I wasn't looking ]
avolio@grendel.UUCP (Frederick M. Avolio) (12/15/84)
Look, this is getting silly... You daily read stuff posted on USENET and
ARPAnet bboards from the DoD. But you're not worried about the *Army*
reading your news, are you? And probably many of you know whether there
happens to be a network host located at NSA. In fact, if you know, it is
probably because it is no secret. The reason that it would be there is the
same reason why folks at the BRL(for example) post to unix-wizards, etc.
Because they happen to be doing similar jobs (programming C, O/S stuff,
U*IX-hacking, etc) as many of us.
What is it that makes people think they are saying or writing anything
of such great interest??? :-)
--
Fred Avolio, DEC -- U{LTR,N}IX Support
301/731-4100 x4227
UUCP: {seismo,decvax}!grendel!avolio
ARPA: grendel!avolio@seismo.ARPAsean@ukma.UUCP (Sean Casey) (12/16/84)
They don't call it FLAME for nuthin.
ron@brl-tgr.ARPA (Ron Natalie <ron>) (12/17/84)
> The reason that it would be there is the > same reason why folks at the BRL(for example) post to unix-wizards, etc. > Because they happen to be doing similar jobs (programming C, O/S stuff, > U*IX-hacking, etc) as many of us. > As a matter of fact UNIX-WIZARDS, INFO-UNIX, INFO-MICRO, INFO-APPLE, INFO-CPM, MSG-GROUP, TCP-IP-DIGEST, INFO-MUSIC, INFO-C, INFO-PASCAL, UNIX-SOURCES are all distributed from BRL. -Ron