Re: I-D Action: draft-ietf-krb-wg-des-die-die-die-02.txt

Simon Josefsson <[email protected]> writes:

> Tom Yu <[email protected]> writes:
>
>> Changes since -01:
>>
>> Deprecate "export strength" RC4-HMAC-EXP.  This requires a normative
>> downref to RFC 4757 (though a trivial one because its purpose is to
>> deprecate part of the downref).
>
> To me, deprecating RC4-HMAC-EXP but not RC4-HMAC implies that we regard
> non-export strength RC4-HMAC as secure.  I think this is unfortunate.

We already address RC4-HMAC in the Security Considerations.

Is RC4-HMAC (non-export) weaker than AES128-CTS-HMAC-SHA1-96?
Probably.  Is there anything that makes it insecure on the level of
single-DES, though?  Keep in mind that RFC 4757 is Informational, so a
standards-conforming implementation isn't required to do anything with
it anyway.

> I'd prefer that we deprecated both RC4-HMAC and RC4-HMAC-EXP.  Second to
> that, that we say nothing at all about RFC 4757 and let readers pass
> their own judgement on it.

There's been text in the Security Considerations about RC4-HMAC since
before my edits.  The document simply neglected to cite RFC 4757
previously.  Are you suggesting that we remove the preexisting text
from the Security Considerations?  Also, RFC 4757 itself recommends
not using RC4-HMAC if stronger enctypes are available.
_______________________________________________
ietf-krb-wg mailing list
[email protected]
https://lists.anl.gov/mailman/listinfo/ietf-krb-wg