EAI November 20, 2008 Harald Alvestrand, chair The group first reviewed the status of the documents; the core group has been published as rfcs. A few others have expired or were refreshed very late. Only one in queue with the ADs. Downgraded is the one in the AD queue, with issues raised. The group reviewed the AD issues: ABNF for envelope addresses, 2047 encoding instructions, ORCPT handling, reservation of Downgraded-*, Less critical issues The ABNF for envelope addresses is inconsistent; the AD recommended that a single form be selected. This avoids the need for a different parsers. The AD recommended using the From form (nested angle brackets versus adjacent angle brackets) consistently throughout. Consensus to use the > throughout. 2047 encoding instructions: some text implies encoding of structured fields; not explicit about breaking fields into multiple encoded words. The resolution agreed was to add advice to use multiple encoded words when needed. ORCPT encoding was the third issue. It has text that says it MUST be converted to one of two forms; this seems unworkable. The group concluded that it would accept the ADs suggested change: "The transformation that needs to occur on the content of ORCPT is to 1. remove xtext encoding 2. convert the result of step one to utf-8-addr-unitext form where all non-ASCII characters are represented as EmbeddedUnicodeChar, 3. re-apply xtext encoding to the result of step 2." The group then discussed the reservation of downgraded-*. The document references procedures which are not present. Discussion of the problem noted that there might be a way to bend the rules rather than changing them (making it more a suggestion with a note, rather than changing the rules: asking them not to include new downgrade- headers during the EAI experiment). AD suggests that this document contain instructions to the expert reviewer, rather than IANA. Re-wording to fall under the "conflict with existing IETF activity" provision of expert review instead. The working group chair will ask the editor to produce a new document incorporating these changes as soon as possible; the other issues raised will be taken to the list. The group then discussed how comprehensive to make the list of headers to be downgraded; Auto-submitted was one example. Chris was concerned that converting auto-submitted to downgraded-auto-submitted might allow for the creation of mail loops. A total of six will be added to the list: Original-recipient, Auto-submitted, Disposition-to, Content-transfer-encoding, Content-language, Accept-language (the comments on these may be where the UTF-8 occurs). Further discussion will be on the linst. The group then discussed the downgrade-display draft. Only two present had read the draft recently. The group discussion focused on the security considerations. John Klensin felt that it out to be re-couched as a set of experiments to be performed. No consensus on disposition, but the chair stressed that further review is needed here to make progress. The group then discussed the EAI deployment practices draft. John's comments here mirrored the previous draft; it is a useful place to record experience, but it needs to be updated with real deployment experience. The chair took on the task of pinging the editors of Pete Resnick then discussed the IMAP draft,first asking for implementation and deployment experience, then reviewing the changes. During the discusion of the APPEND syntax problem, the chair noted that the room's participants were glazed over. Recommendation that there be a working group last call of the draft revision, knowing that experimental results might cause it to be changed. John Klensin volunteered to push the working group toward producing an update of the charter. He believes we basically know what has to be done. There are documents evaluating the experiments, then ship documents to the IESG for standards track. As a strawman, a ship date of late 2009 for the proposed standards was discussed, but with an intent of reviewing the question in March 2009. There being no other business, the chair closed the meeting.