pure transit server

[Russ Allbery]

Jeffrey M Vinocur <jeff at litech.org> writes:

> Question first, from a theoretical standpoint -- do we care about
> resistance to attack?

I can't picture an attack based on hash collisions that couldn't be
performed equally well by just injecting new messages with the same
message ID as the messages one wants to play games with, so no, I don't
think so.  Unless I'm missing something.

> We probably _don't_ want a cryptographically-strong hash
> function...unless we're worried about malicious Message-ID generation or
> something.

MD5 isn't really all that slow, compared to other hash functions.  It also
has the nice property of generating 128 bits of hash; hash functions that
only generate 32 or probably even 64 bits of hash seem unlikely to be good
enough for our purposes.

If anyone is going to look into hash functions, I highly recommend
starting with <http://burtleburtle.net/bob/hash/>, and note that his
LOOKUP2 hash function is already in INN as hash_lookup2(); I'm using it
for the generic hash table implementation.

-- 
Russ Allbery (rra at stanford.edu)             <http://www.eyrie.org/~eagle/>