Dean Kusler wrote:
I'm sure many of y'all have already seen this, but I hadn't seen it
posted to the list yet. Preliminary details on Bruce Schneier's blog.
http://www.schneier.com/blog/archives/2005/02/sha1_broken.html
Of course, collision attacks probably aren't that important for
Monotone's use of SHA-1, since you'd need to have a malicious person
with write access to your database, in which case you're probably
screwed anyways.
you'd also need an attacker wealthy enough to scrape up 2^69 work units,
in which case you're also probably screwed anyways.
but we should, as a matter of courtesy, update our standard hash to
something stronger soon. as soon as there's a community consensus about
which hash in particular represents "something stronger".
it's worth noting that any competing system which uses digital
signatures as a security mechanism is equally weakened by this exciting
development.
-graydon
|
| 
