> -----Original Message----- > From: Pete Resnick [mailto:[email protected]] > Sent: Friday, January 20, 2012 3:40 PM > To: Murray S. Kucherawy > Cc: Message Abuse Report Format working group > Subject: Re: [marf] DISCUSS on draft-ietf-marf-redaction-04 > > Sorry for not getting back to you earlier; tied up in meetings all > afternoon. But you did hear back from Stephen: This would address his > concerns. This way, you're not saying that H is a good idea for > redaction without explaining the limited meaning of "redaction" that > this document anticipates. I would like to see a "MAY" appear somewhere > in section 3 of your proposed text to indicate that the choice of > algorithm is a protocol option. (E.g., "An implementation MAY choose > one of ROT13, CRC32, MD5, H, HMAC, or any transformation that has a > reasonably low likelihood of collision...blah...blah...blah..."). And I > think you're likely to need references for the example algorithms you > do give, and maybe a a quick line about the features of each, (e.g., > "ROT13 (manually invertible, but visually obscure), CRC32 (invertible > by code, but not simply by a human), ..."). But I think this is > perfectly reasonable.
(Oops, wasn't quite done yet.) I'm not clear on the suggestion about MAY. We already present the choice; a MAY here would mean you can opt not to do any of these, which isn't redaction at all, and this whole memo doesn't apply. _______________________________________________ marf mailing list [email protected] https://www.ietf.org/mailman/listinfo/marf
