- From: Phillip M. Hallam-Baker <hallam@dxal18.cern.ch>
- Date: Thu, 23 Mar 1995 19:04:56 +0900
- To: Dave Kristol <dmk@allegra.att.com>, http-wg%cuckoo.hpl.hp.com@hplb.hpl.hp.com
- Cc: hallam@dxal18.cern.ch
>FWIW, I've implemented John Franks's scheme in my server using base16. >I think that biased my thinking. Phillip is right that MD5 is 128 bits, >and nothing need be said about encodings, except for transport. Ah well thats OK then. so long as everyone realises I'm right we can put whatever people like in the spec so long as it is clear. In my original equations I think I simply assumed that the base-16 encoding was orthogonal to the hash function, the last thing to be done. That said we should rewrite the spec using the functions D for digest and H for hex (uppercase). And make it explicit that we use H(D( x )) each time H appears presently. My implementation will then be slightly non standard in using the uppercase for generation etc but accepting a digest passed using lowercase hex characters :-) >2) If 16 binary bytes should be used for each of H(A1) and H(A2) in ><digest>, or for H(A1) in <message-digest>, why bother with the ':'s? I have never understood that one. My original scheme didn't use them. Same with the usernames. I could never understand the argument on : vs @ since the only reason its there is to stop the fields running together in a possibly ambiguous fashion. Phill.
Received on Thursday, 23 March 1995 10:17:41 UTC