Document: draft-ietf-sieve-notify-mailto-07 Reviewer: Ben Campbell Review Date: 2008-03-28 IETF LC End Date: 2008-03-31 IESG Telechat date: (if known) NOTE: This is a review of version 07 of this draft. The gen-ART assignment was for version 06, but the draft was revised prior to my actual review. Summary: This draft is almost (really close) ready for publication as a draft standard. I have a small number of comments that should be considered prior to publication. Comments: Section 2.7, third to last bullet on Page 6: "URI headers with hname "received" are considered unsafe, and will be ignored." Was that intended to be MUST be ignored? I noticed that similar statements are made about other header fields that do say MUST. It is not clear to me if this particular sentence was meant as normative language for the notifier, or descriptive language of what the recipient of the notification would do. References: IDNITS reports that draft-ietf-sieve-variables has been published as RFC 5229. Also, the [Notify] reference describes a very outdated version of that draft--I suggest checking all of the references to make sure they are current. Editorial nits (opinions only--none of these need to block publication) Section 1.1: I found the first sentence a little hard to digest. I think it would read better if expanded into 2 or 3 sentences. I also found the style of references disruptive to the flow of reading. The paragraph uses a rather clever overloading of references to serve as nouns in sentences. Although one can get away with this when the noun in question is the document being referenced, as in "...notations are as in [Sieve] section 1.1 ...", I find it visually disruptive to read "The [Notify] extension to the [Sieve] mail filtering language...". My brain keeps stopping to wonder what is with all the square brackets--it is not immediately clear from context that the brackets are there to denote references.