This is the most interesting post to appear since the list was re-energized. The weakest elements of comsec are related to matters seldom discussed on crypto fora which are heavily biased toward digital technology. As might be expected on the Internet and its crippling and perhaps fatal dependence on digi-tech isolation from reality. Manifold other means and methods of comsec get short shrift on the Internet yet are the most frequent ways that digital comsec is breached, defeated, embarassed, ridiculed, dumped for legacy methods that require deep familiarity with divers physical and electromagnetic constraints and exploits. Brian Carroll, in case you have missed his copious contributions elsewhere, is more adept than most in addressing the conceits of digi-tech masters by providing leads to these diverse legacies. Don't expect him to rollover and succumb to the usual putdowns of conceited digi-techers. Happily, this noble list is broadminded enough to host his offerings which would be banned on premier digi-tech hideaways from reality. At 12:53 AM 10/19/2013, you wrote:

// analysis, feedback, and pattern matching...

Dude, where's my code? http://phys.org/news/2013-10-dude-code.html

(re: paper is titled "Towards Optimization-Safe Systems: Analyzing the Impact of Undefined Behavior." // Stack)

Birds on repeat: Do playbacks hurt fowl? http://phys.org/news/2013-10-birds-playbacks-fowl.html

(context is of birdcall networks & modem protocols, perhaps this issue of emulation relevant to security protocols, monitoring, attack strategy- or not)

Cuckoos impersonate hawks by matching their 'outfits' http://phys.org/news/2013-10-cuckoos-impersonate-hawks-outfits.html

(crazy cuckoo birds impersonate hawks, exploit of localized overlapping pattern matching. potential comp.security or software systems correlation, beyond poli-sci)