Re: Query Hit bitprint

Quoting mark W <[email protected]> from ml.gnutella.dev-forum:
:It's very informative discussion about bitprint.
:
:>>That's why all the servents I know use SHA1 in query hits,
:>>but offer a X-Thex-URI header during HTTP exchanges, 
:
:servants are not using bitprint in query hit ? 

No, they aren't.
What's the use of sending two hashes that identify a resource?

Of course one could argue something along the lines: "what if SHA1
becomes crackable?  Better have 2 hashes in that case, using the TTH
root hash as a unique identifier?".

But this is a rather moot point.  Because even if the SHA1 was cracked
and it became trivial to produce a matching file given a SHA1, the tuple
(filesize, SHA1) is likely to remain unique.  And even without that,
the combination of SHA1 and TTH can always be used during HTTP exchanges
to disambiguate.  Anyway, if SHA1 gets broken, there will be more
serious problems than that.

I think Shareaza already sends bitprints in their
X-Gnutella-Content-URN headers during HTTP exchanges, which is nice.

Raphael