CURRENT_MEETING_REPORT_ Reported by Terry Gray/University of Washington Minutes of the Internet Message Access Protocol Working Group (IMAP) Internet Message Access Protocol - Version 4 Mark Crispin, the document author, described a few terminology clarifications he made since the last version of the draft. References to ``message numbers'' and ``sequence numbers'' are now uniformly changed to ``message sequence numbers,'' and the term ``sets'' is now used for lists or ranges of message sequence numbers. The principal agenda item was to discuss issues raised by John Klensin (our Area Advisor) during his review of the IMAP4 specification in preparation for submission as a Proposed Standard. John raised the following issues: o The document is hard to follow, because you must continually refer back and forth between the commands, responses, and syntax sections of the document. - Result: the rationale for the current organization was discussed (it is oriented toward the way an implementation would be written). The need for an IMAP tutorial was acknowledged. There will not be a change to the document organization at this time. o The document needs a few more words at the beginning under a heading titled ``How to Read this Document,'' especially since you must consult the grammar to do an implementation. - Result: the text will be added. o Another sentence or two is needed to clarify cases where the optional keyword MAILBOX is included in a command (for backward compatibility); e.g. is there a difference between ``subscribe mailbox'' and ``subscribe mailbox mailbox''? - Result: the text will be added. o The 30 minute minimum value for an autologout timer on a server was thought to be too long in certain dialup contexts. - Result: After extensive discussion, four options were identified: 1. Remove reference to any timeout value in the specification. (This is bad because a value is needed to set an upper bound for client keep-alive processing.) 2. Change value to something other than 30 minutes. (There was no consensus on what a better minimum would be.) 3. Have the server probe the client. (This is dangerous because of the potential for flow control deadlock!) 4. Create a command to allow the client to tell the server what the timeout value should be. (This was agreed to, but not in the base document. Rob Austein will propose---the first!---new CAPABILITY extension.) o There needs to be a small amount of meta-wording added to explain that a subsequent document describing a protocol modification or extension (new CAPABILITY) can override or modify the behavior described in the base specification, but that new extensions (CAPABILITIES) only modify the base specification behavior by mutual consent of the client and server. This is similar to the SMTP extensions model. - Result: the text will be added. Also, an agreement was reached on modified wording concerning handling of zero-length literals. Next Steps Version ``05'' of the Internet-Draft will be drafted, incorporating the above changes. It is expected to appear within a few days. John Klensin will then request a Last Call for Proposed Standard notification for the following three documents: o ``Internet Message Access Protocol - Version 4'' (draft-ietf-imap-imap4) standards track o ``IMAP4 Authentication Mechanisms'' (draft-ietf-imap-auth) standards track o ``IMAP4 Compatibility with IMAP2 and IMAP2BIS'' (draft-ietf-imap-compat) Informational Everyone is requested to look at the last two as soon as possible for any final suggestions. After the Last Call interval, assuming no major problems surface, John will initiate balloting of the IESG on the above three documents. Also, Rob Austein requested more feedback on the Informational ``How to use IMAP4 for Disconnected Operation'' document, which is not yet ready for publishing.