On Wed, 21 Aug 2002, Norbert Bollow wrote: > The newest version of the draft is at > http://maillist.info/rfc-draft.txt
A couple of comments (some very minor, others slightly more substantive). (1) Could you put some kind of version number or last modify time on that so that we know whether we are commenting on the same draft. (2) I'm wondering if we want finer distinctions than just the four listed. And whether a multidigit code would be better ala SMTP response codes. Furthermore allowing several comma separated codes per line. So 1xx (Forbidden) 2xx (Ask) 3xx (munge) 4xx (allowed) while x0x (all publically accessable modes) x1x (http accessible) x2x (nntp accessible) x3x (some mail access protocol (IMAP, POP) x4x (mail query mechanism, archive sent in response to mail) For munge we could have finer distinctions. xx1 (all email addresses) xx2 (header addresses only) xx3 (only header addresses picked up by NNTP XOVER) Thus we could have a header like: List-Policy-Archive: 210, 423, 431 which would say. Web archives must ask. NNTP archives are must munge relevant header addresses, and IMAP-like archives must munge everything. I agree that this gives up the human readable virtue of what you have now. Maybe a mixed protocol which would allow something like List-Policy-Archive: Ask: 210, 423, 431 (Comment goes here) Making each part optional (though at least must be present). -j -- Jeffrey Goldberg http://www.goldmark.org/jeff/ Relativism is the triumph of authority over truth, convention over justice
