On Thu, 28 Mar 2002, James Myles wrote: > General question: > What key size do you use,
My key is 1024-bits long. > and how long is it valid for? AFAIK, expire dates depend on usage and environment. I have mine set to never expire. I'm not sure of the guidelines for these values. I suspect they're in a state of flux due to Bernstein's paper: > Interesting recent Bugtraq posting: > ><http://online.securityfocus.com/cgi-bin/archive.pl?id=1&start=2002-03-25&end=2002-03-31&threads=0&mid=00cc01c1d2d4$ad3bc640$c33a080a@LUCKYVAIO> > > Thoughts or comments? I generally trust Bruce Schneier when it comes to security related information. His past articles have been uncompromising, yet balanced reviews of security technology and practises. His article: http://www.counterpane.com/crypto-gram-0203.html#6 was also mentioned in the bugtraq post. Its worth emphasising that Bernstein's result is asymptotic. It's kinda like if keys were infinitely big, then his algorithm would be much better (crack a key three times bigger in the same time) than the current algorithms. Bruce points out two specific weaknesses: 1. He hasn't taken into account the interconnects between elements of the cluster. This makes design and building the device difficult to very difficult, but not impossible. 2. Real life keys are not infinitely big. For realistic keys, his algorithm could even be slower than the more conventional factorising approaches (such as NFS). Cryptography and cryptoanalysis are art-forms. Often the algorithms involved have been verified (or at least haven't been broken yet), but often the weakest link is the algorithm's implementation or people's use of the software. Example: Enigma would have been unbreakable if the German's hadn't been so lax with their encryption practises (BTW, AFAIK British Admiralty codes were broken more or less from the word "go") The art in this is the ability to see the flaw along the chain from keyboard to remote site and spot any flaws. This comes with practice (or so people would have you believe). I would trust "experts in the field" (or at least the verifiable expert ones). So, when Bruce feels he needs to revoke his key and issue a new one, I'll follow suite. [just my 2p-worth ;] Paul. ------------------------------------------------------------------------------- Particle Physics (Theory & Experimental) Groups Paul Millar Department of Physics and Astronomy [EMAIL PROTECTED] University of Glasgow [EMAIL PROTECTED] Glasgow, G12 8QQ, Scotland http://www.astro.gla.ac.uk/users/paulm +44 (0)141 330 4717 A54C A9FC 6A77 1664 2E4E 90E3 FFD2 704B BF0F 03E9 ------------------------------------------------------------------------------- -------------------------------------------------------------------- http://www.lug.org.uk http://www.linuxportal.co.uk http://www.linuxjob.co.uk http://www.linuxshop.co.uk --------------------------------------------------------------------
