Re: [dane] I-D Action: draft-ietf-dane-smime-04.txt

[Paul Wouters <[e-mail]>]

On Thu, 9 Jan 2014, Viktor Dukhovni wrote:

>> Should anyone insist of hmac-sha256, though, base32 would be necessary.
>> But I've yet to see any credible claims that sha1's vulnerabilities
>> affect hmac-sha1, so there shouldn't be a reason for sha2 or sha3.
>
> Indeed, but if one really wants to avoid HMAC-SHA1 because it is
> now unfashionable, one can use SHA2-224 to get a 56-byte label.

SHA2-224 would have my preference, as SHA1 is on its way out FIPS-wise
and it is just easing not having to maintain SHA1 exceptions to the
"disallow sha1" code paths.

> A sensible administrator will keep the unhashed input names in a
> configuration system that generates the corresponding DNS entries.

Yes, base32 is as unradable as a hash for the admin eye.

I think I'm fine with using sha2-224, if it saves us the hassle of doing
label splitting. But still a little worried about hashing various
character sets.

Paul