Why "carnivore" type systems can't be (entirely) open source

[Roland Perry]

 In message <010201c0949c$895482c0$3e0a989e@eloka>, Owen Lewis 
 <oml@eloka.demon.co.uk> writes
>
>At the outset, ISTR that data via GSM was as unusual as via modem?

 Analogue cellphones used normal modems, albeit in the rather unusual 
 format for then - a PCMCIA card. The best data rate to choose was 4,800, 
 which actually worked better than some of the slower rates. 9600 was out 
 of the question.

 When Digital arrived, there were slightly different "modem" cards, which 
 were really rate adapters. But most early data was done by SMS and even 
 schemes which sent emails by concatenating several SMS messages.

>The early
>adopters were those keenies who had been using custom tailored acoustic
>couplers for years (probably running at 300 bps).

 I never did use an acoustic modem with a mobile of either kind. The 
 first message I sent by "rate adapter" modem card on a GSM phone was in 
 October 1994; I still have the email - it was to a cc:mail gateway run 
 by CompuServe. Having proved the technology (the people I was 
 demonstrating it to were rather amazed that the phone, PDA, software and 
 modem were all straight out of the box) after that it was always "just 
 money" to buy the toys and fund the airtime.

>Even now, datacom via GSM
>is less than ideal. 9600 suks OK.

 Yes, it's not that good, which is why I use a 28.8K GSM card these days 
 (Nokia Data Card with airtime from Orange).
-- 
             Roland Perry | tel: +44 1733 207705 | roland@linx.org
              Interim CEO | fax: +44 1733 207729 | http://www.linx.net
London Internet Exchange | mbl: +44 7050 604080 |       /contact/roland