Hippo, birdie, two ewes...

...to the lovely and multi-talented jilara!!! Have a great one!!

Don't do that, then.

According to this paper [pdf], AES encryption may be much easier to crack using linear cryptography than is commonly thought. The paper is, however, non-constructive: it makes a plausible case but doesn't actually demonstrate a solution.

The attack's runtime is comparable to performing 64^w encryptions where w is the (unknown) minimum Hamming weight in certain binary linear error-correcting codes (BLECCs) associated with AES-256. If w < 43 then our attack is faster than exhaustive key search; probably w < 10. (Also there should be ciphertext-only attacks if the plaintext is natural English.)

It also gives a construction for a family of encryption algorithms that avoid the problem. The attack does rely on having a large number of plaintext/ciphertext pairs encrypted with the same key, or an even larger amount of ciphertext with known statistics.

The workaround for now would appear to be to use a different random or pseudorandom key for each document (and of course to compress each document before you encrypt it, and use CBC mode encryption). You're left with a key-management problem, but managing that many ID/key pairs is no worse than managing all the document IDs in a hash-based version control system like git. Similarly, systems that use one-time random session keys should be safe as long as they're using compression and CBC mode, and renegotiate their keys frequently.

(From cryptome.)

Baby wars?

Imperial walker stroller

From Thingamababy by way of BoingBoing and Gizmodo.

( I find your cuteness... disturbing. )

Tracks

Did a little work on "I Wanna Be a Webmaster" and "TEOTW". Just tweaking, really; "Wannabe" might still want some work on the shaker part. Then I went over to the Baycon 2007 directory to check out the concert. Clean except for a couple of nasty noise spikes. So I think we have a new version of "High Barratry" [ogg] [mp3]. Check it out.