TweetNaCl: how cr.yp.to’s developers got carried away by the carry bit

blog.skylable.com

15 pointsedwintorok7 comments

Open articleView on HN

Funny how small is the world, I reported them the same bug on February 13th, you might have been first though :)

FWIW I reported mine on 2014-04-22. Was yours about the sizeof(u32) issue, or the carry bit?

It was only the carry bit [1] and the saved space mention from the ChangeLog.

[1] http://seb.dbzteam.org/blog/2014/04/28/tweetnacl_arithmetic_...

Looks like TweetNaCl could use a bugtracker or a mailing list :)

cr.yp.to is Daniel Bernstein. Was TweetNaCl written by Bernstein? I thought not.

5 minutes later:

Yeah, I should have read this more carefully before jumping in to comment. Sorry. Previous comment preserved as a monument to my dumbness.

5 seconds later:

Dammit, Edwin caught me, too. :)

The paper on TweetNaCl lists him as an author, and its a subdomain of cr.yp.to: http://tweetnacl.cr.yp.to/papers.html http://cr.yp.to/papers.html#tweetnacl

The bugreport was actually acknowledged by Peter Schwabe.

Did djb actually write every line of TweetNaCl? Probably not.

I actually didn't know he wrote any line of it; I feel like I remember him writing about the person who actually wrote it. But I think I'm wrong about that too. It's a neat post you wrote.

Reply on news.ycombinator.com