Yet Another Mail Working Group Minutes Meeting : IETF78, Thursday 29 July, 2010 Location: MECC, Maastricht, Athens, 16:10 to 16.10 Chairs : Tony Hansen Minutes : S. Moonesamy Version 0.1 The filename for audio feed is ietf78-ch5-thur-noon.mp3. The YAM session starts around the 133 minute mark and cuts off before the end of the session. AGENDA A. Meeting Administrivia Note well Agenda bashing B. Mime type registration issue Alexey Melnikov C. Redux of plenary, discussion of two-track and their affect on YAM's work and charter D. Any Other Business Meeting Report A. Meeting Administrivia Tony Hansen chaired the the IETF78 YAM Working Group meeting in Maastricht, Netherlands. S. Moonesamy was the scribe. Barry Leiba was the Jabber scribe. The Chair brought the Note Well to the attention of the meeting participants and asked them to sign the blue sheets. B. Mime type registration issue Alexey Melnikov, Applications Area Director, asked for feedback about the image/svg+sml MIME type registration. The definition of the MIME type specifies two file types extensions. http://www.w3.org/TR/SVG11/mimereg.html The explanation as to why this was not two different mime types was because one is used for the XML version and the other for the gzip version of the same thing. The initial reaction of the Applications Area Director was why it was not two different MIME types. it was two different transfer encodings used in HTTP of the same thing, one is identify and the other is gzip version. Mark Nottingham commented on the two different kinds of encoding used in HTTP where transfer encoding is hop by hop is invisible to the payload whereas content encoding is a property of the content encoding itself. Mark mentioned that HTTP does not have Content Transfer Encoding. Chris Newman pointed out that HTTP Content-Encoding acts like a MIME Content-Transfer-Encoding but it's an HTTP-only MIME feature. gzip is a Content-Encoding in HTTP. There is no gzip Content-Transfer-Encoding defined. Pete Resnick commented on the decoded "thing" getting the .gz on it and that is wrong, i.e. the filename having the compress type attached to it is wrong. John Klensin agreed that it was wrong as there are some "things" that pay attention to the content type, some "things" that pay attention to the file extension and some "things" that pay attention to both. The result is that it can cause confusion and lead to interoperability problems. It is also a fundamental break with the core of the specification which specifically says that content type is the content type and the filename is indicative of something the receiver might find useful. Chris Newman said that the filename for gzipped-SVG needs to be in a different spec, not the spec for image/svg+xml. Use of the gzipped-SVG file name with a different MIME type is fine but that MIME type needs to describe all the bad gzip security considerations and that content type can't be +xml. According to John Klensin, drawing clues from filenames breaks the MIME specs. Julian Reschke asked whether filename registrations are purely advisory. C. Redux of plenary, discussion of two-track and their affect on YAM's work and charter The Chair provided a quick update on what has happened since the last meeting. Three pre-evaluation documents have been published. One of them has gone through IESG review and they have very little to same about it. The Chair summarized the Two-track discussion that happened during the plenary the last night. With draft-housley-two-maturity-levels there would have be two levels and a six month wait is no longer required to move to Full Standard. As a side effect, all existing documents at Draft Standard would be moved immediately to Internet Standard. Barry Leiba exclaimed that the working group has been successful. Alexey Melnikov comments that there could be a different interpretation where all Full Standards are downgraded to Draft Standards. Randy Gellens agreed with Alexey's interpretation. There was some discussion between Randy and Barry about the documents being put on the top tier whatever the semantics of the tiers are. Dave Crocker pointed out that the working group is a test of the ability to move documents to the top tier and the goal was to exercise the mechanisms. The Chair mentioned that if the documents the working group has adopted are already at Full Standard, the working group might consider doing updates to the documents. Pete Resnick proposed that the working group could consider suspending its activities until it knows whether draft-housley-two-maturity-levels will go through. Barry responded that if the working group believes that the working on the documents is important, then it does not have any bearing on the other decisions. According to Pete Resnick, the changes to RFC5322 are not extensive but quite necessary. He stated that he would continue working on the document. Barry was of the opinion that the IESG is not interested in dealing with more pre-evaluation documents. Alexey Melnikov, Area Director, said that he needs to ask the IESG about that. Dave said that "we have a clear agressive tenacious statement from the IESG that they do not want to do this stuff". In the face of institutional resistance, there isn't any reason to continuing the work as there is other important work to do. John Klensin agreed with Dave. Alfred Hones was less optimistic about the success of the Two-track proposal. He was of the view that it would not be completed by the current IESG and suggested that the working group continues the work. John Klensin disagreed with Alfred and pointed out that some participants in the working group have other work with a higher priority. Although he does not know what the right answer is stop or it is pause but he is pretty sure that the right answer now is not move forward. Barry ask the Area Directory about the time schedule for processing Two-track and Alexey replied that he has no idea. As a personal opinion, Alexey would like people to update documents to fix bugs. There was a discussion about how the Two-track proposal was produced during the IESG retreat and Alexey mentioned that there was a view that moving documents to Full Standard was pointless. Tony Hansen asked how much of a change is acceptable for a document at Full Standard not to be recycled to Proposed Standard. John Klensin mentioned that it is subject to the unpredictability of the IESG. Alexey said the IESG finds the pre-evaluation documents too painful. Randy proposed one approach which considers pre-evaluation as shepherd write-ups. Jacob Palme stated that the working group's task is to do the technical work. The task is whether we want to modify the document. There are two alternatives, publicize a new revised document or publicize an errata document. If there is an errata document, there is a risk that implementors will not find it. It is much better to change the entire document so that there is a full document with all the changes. Chris Newman asked whether there was any energy to continue. John Klensin mentioned that sooner or later a revised version of RFC 5321 would appear. The odds of doing it in the next few months is very low. Pete mentioned that he could probably get RFC5322bis done. A participant commented that RFC 5321 is just finished and there is already a new version in the works. Barry took the comment as a vote about advancing documents as it creates too much churn. The Chair wanted to get a sense of the room between three choices: (a) continue YAM (b) pause YAM (c) close YAM Of the twenty-six people in the room, six people were in favor of continuing, six person were in favor of pausing and nobody were in favor of closing the working group. The Chair asked the questions again. Seven people were in favor of continuing, nine people were in favor of pausing. There was no consensus. D. Any Other Business There wasn't any other item for discussion. The Chair ended the meeting. Regards, S. Moonesamy YAM WG Secretary