[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
WG Last Call: AH Transforms to Proposed Standard
Ref: Your note of Tue, 21 May 1996 12:13:04 PDT
I missed the "quick survey" (I was away from my email),
anyway I want to repeat my opinion on this issue.
Based on the message I sent to this list on May 8
I support using HMAC-MD5 for AH and ESP+AH.
(I do this with my hat of "implementer" and the recognition of the
need for best possible performance rather than with my more conservative hat
of "pure cryptographer").
I recommend adding the following paragraph (or any correct English-variant
of it) to the documents defining HMAC-MD5 as the default transform:
The currently known cryptanalytic results on MD5 do not indicate
a practical weakness of MD5 for its use with the HMAC construction.
Due to this fact and the performance advantages of MD5 over other
alternatives (e.g., SHA-1) this document defines MD5 as the basic hash
function to be used with HMAC. However, implementers need to be aware
that future cryptographic developments may call for the replacement of
MD5 with other hash functions. In particular, implementers are strongly
encouraged to provide support for SHA-1 with HMAC in addition to the
required support for MD5. This will facilitate a future migration to SHA-1
without jeopardizing interoperability.
Hugo
Follow-Ups: