Re: [TLS] Open Issue: verify_data processing

[Martin Rex <martin.rex@xxxxxxx>]

Steven M. Bellovin wrote:
> 
> On Wed, 18 Oct 2006 18:10:32 +0100, "Ben Laurie" <benl@xxxxxxxxxx> wrote:
> 
> > A pedantic quibble here: this makes the assumption that a hash can
> > always be calculated by chaining partial inputs, which is not a
> > required property of hashes, just one we're used to having.
> >
> Indeed.  In fact, the desirability of this property has been raised at
> some of the NIST hash function workshops -- some people think we will not
> be able to preserve it in a future secure hash function.

It will be interesting to see how they're going to sell such
hash functions to current consumers of PKI-based software.

PKCS#7/CMS as well as Web services security (xmlsig) applies digital
signatures to quite large blobs of data today.  Some of that data
does not fit into memory already today, and one definitely does
not want such a requirement on a busy server.

So any hash function without that property is likely not going
to be suitable for a drop-in replacement of installed base
PKI-based software.

-Martin

_______________________________________________
TLS mailing list
TLS@xxxxxxxxxxxxxx
https://www1.ietf.org/mailman/listinfo/tls