Based on IRC log http://www.w3.org/2003/10/01-xmlprotocol-irc
1. Roll Present BEA Systems, Mark Nottingham Canon, Herve Ruellan IBM, David Fallside IBM, Noah Mendelsohn IBM, John Ibbotson IONA Technologies, Seumas Soltysik Microsoft Corporation, Martin Gudgin Oracle, Anish Karmarkar SeeBeyond, Pete Wenzel Sun Microsystems, Tony Graham Sun Microsystems, Marc Hadley Systinet (IDOOX), Jacek Kopecky W3C, Yves Lafon Excused BEA Systems, David Orchard Canon, Jean-Jacques Moreau IONA Technologies, Mike Greenberg Microsoft Corporation, Jeff Schlimmer Oracle, Jeff Mischkinsky Systinet (IDOOX), Miroslav Simek W3C, Carine Bournez Regrets DaimlerChrysler R. & Tech, Mario Jeckle SAP AG, Gerd Hoelzing SAP AG, Volker Wiechers Absent DaimlerChrysler R. & Tech, Andreas Riegg [Yves] http://www.w3.org/2000/xp/Group/3/02/24-soap-attachment-feature.html [scribenm] Chair notes that SOAP Optimized Serialization Use Cases and Requirements has in fact been published as an Ed copy, contrary to what agenda says (URL above) ================ 2. Agenda review no AOB items =============== 3. Approval of last week's telcon minutes [scribenm] Minutes of Sept 24 telcon approved without objection =============== 4. Action items [scribenm] MarkN & DavidF [scribenm] Confirm on the exact name of the content type (ietf vs vnd) [scribenm] Still pending [scribenm] DF tells the scribe that action resolutions need not be logged. ================ 5. Status reports [scribenm] Mark Nottingham summarizes binary XML workshop [scribenm] Wide ranging discussion [scribenm] Result was plan to try and draft a WG. That WG would have as its mission requirements, debating pros/cons, etc. possibly leading to yet another WG that would do an actual binary standard. [JacekK] ACTION: MarkN to go investigate the MTOM-like technology discussed at the binary xml meeting (in time for next week) [scribenm] Mark also reports annecdotally that SVG is considering something like MTOM, but SVG folks seem not to know about MTOM in particular[scribenm] Noah points out that most attendees had not known of MTOM. It was mentioned in passing, so attendees now know it exists, but there was no detailed discussion of it. [scribenm] Anish reports that he was surprised by the number of reported implementations of various approaches. [scribenm] Many of these seemed to be in research/prototype status. [scribenm] Some attendees wanted message or data compaction, others wanted to reduce CPU overhead. [scribenm] ...end of summary of binary workshop [scribenm] DF: OK, we'll keep an eye on it, as well as tracking the SVG-related thing [scribenm] Media type update: Mark Nottingham reports no news [scribenm] Marc Hadley, did you contact WSD regarding who does work on descriptions, and to make sure our work is description-compatible. Marc says, no didn't talk to WSD, but did talk to Philippe LeHegaret [scribenm] Jacek: we had a discussion in WSD regarding methods of marking Media types. WSD plans to draft a note, probably for transmission to the Schema WG, maybe us, etc. [scribenm] No details availble on who will do this, when it will happen, etc. [scribenm] ACTION: Marc Hadley re-contact plh and J. Marsh regarding attachment descriptions in wsd [scribenm] Status of normalization working group notes: publication in progress, should be avail for next telcon. [JacekK] minutes of the WS-Desc f2f at http://lists.w3.org/Archives/Public/www-ws-desc/2003Sep/0218.html, look for 13:00 agenda item [JacekK] that was on WS-Desc discussing describing media types f2f planning [Gudge] +1 for week of Dec 1 [Gudge] +1 for starting Tuesday [scribenm] Regaring f2f: general preference in DF's poll was for a 2 day meeting starting no earlier than tues, week of Dec. 1 [scribenm] Chair decides: we will have 2 day meeting week of Dec. 1 on West Coast. SAP is investigating hosting, presumably in Palo Alto. [scribenm] Mark Nottingham offers BEA as fallback, potentially in San Francisco or San Jose [scribenm] Chair leans toward tues/wed to avoid need for European members to travel on weekend. ================ 6. Attachments [scribenm] Begin discussion of data model proposal [scribenm] Chair's summary: task force recommended adoption. emails leading up to current discussion seem about evenly split, with all sides expressing willingness to live with the others' positions. [scribenm] Poll: your choice is (a) adopt Data Model (b) status quo [scribenm] Poll results: 8 for DM, 1 for status quo [scribenm] Chair has asked if there is strong objection from anyone for either option. No response, so presuming no objection. [scribenm] We have decided to move forward with the data model [Yves] ACTION 4= Editors to merge XQDM draft and current draft. in time for oct 8 [scribenm] Begin discussion of XInclude and reporting to XML Core WG... [scribenm] Chair summarizes email discussion: doesn't think we had concensus or clear trend as to whether to recommend anything to core, and/or whether to use a standardize XInclude [scribenm] Chair solicits summaries of email discussion: [scribenm] Marc Hadley: DM decision needs to be factored in...haven't thought about it in detail. [scribenm] Chair: let's not discuss more details now, and take Marc's suggestion to postpone detailed discussion for a week [scribenm] Mark Nottingham: A room is available for our f2f in SF for Tues/Wed Dec. 2& 3, can switch to 3rd and 4th or cancel [scribenm] Chair thanks Mark and BEA [scribenm] Begin discussion of Use Cases. [davidF] UC & Rec doc URL is http://www.w3.org/2000/xp/Group/3/02/24-soap-attachment-feature.html [anish] http://lists.w3.org/Archives/Public/xml-dist-app/2003Sep/0080.html [scribenm] Discussion of UC4 (see link from Anish above) [herve] Rxx: It must be possible to encrypt the message or its portion(s), [herve] including non-XML data (e.g., binary data and XML fragments). [herve] Rxx: It must be possible to sign the message or its portion(s), [herve] including non-XML data (e.g., binary data and XML fragments). [scribenm] Quoting from that proposal: "I had taken an action to formulate UC-4 [1] as a requirement. This email [scribenm] fulfills that action. UC-4 is split into two requirements: [scribenm] Rxx: It must be possible to encrypt the message or its portion(s), [scribenm] including non-XML data (e.g., binary data and XML fragments). [scribenm] Rxx: It must be possible to sign the message or its portion(s), [scribenm] including non-XML data (e.g., binary data and XML fragments). [scribenm] The above two requirements are accepted without objection. [JacekK] ACTION: editors of requirement document to incorporate the two above requirements [scribenm] Discussion of UC-6: streaming requirements. [scribenm] Seamus: is it an issue of change of philosphy from SOAP+Attachments to using a more REST formulation? Client can pull stuff later? [scribenm] Noah: The REST proposal really doesn't deal with the same use cases as MTOM, e.g. store and forward, occasionally-connected systems, etc. [scribenm] Chair asks Noah to expressed in earlier emails: [scribenm] Noah: Basically, I think there are lots of tricky use cases including a single large object, multiple large objects following each other, multiple large objects progressing in parallel, situations in which you want to make things easier for sender, others where you want it to be easier for receiver etc. [scribenm] Noah's position is: if we want to do something for streaming, we should consider all of these, and see if there is a set we believe to be high value and tractable from an implementation point of view. [scribenm] Jacek: maybe we do nothing, and be quietly happy that what we're building anyway will meet some needs. [scribenm] Chair: should we drop the streaming use case [scribenm] Gudge: as Noah has said, there are multiple things people call streaming. It's not clear that we should just pick one before we know which it is. [scribenm] Chair's clarification: we would dive in and figure out which one(s) they are [scribenm] Gudge: some are indeed useful, we should dive in and find out [scribenm] Yves: +1 [scribenm] Noah +1 [scribenm] Chair: we'll take this back to email and bring it up again next week, hoping to nail down one or more use cases. [scribenm] Jacek: task force? [scribenm] Chair: no, not now. [scribenm] Our team is a bit smaller [scribenm] Streaming discussion returns to email...end discussion of UC-6 [scribenm] Also; request WG to look over use cases document so we'll be prepared to discuss other questions. [scribenm] Discussion of issues 432 and 435... [Gudge] http://lists.w3.org/Archives/Public/xml-dist-app/2003Sep/0010.html [scribenm] Noah points out that we need to think of this in light of Data Model decision. The DM gives you (as did the infoset) the information to determine which elements look like base64 binary. [scribenm] Proposal from Jacek and others: there is no property explicitly suggesting what to optimize. The rule is that a binding MAY optimize any, all or none of the items that are (a) of dm:type xsd:base64binary and (b) in the canonical form of that type. [scribenm] Noting that we have yet to nail down a canonical form for our purposes. [Gudge] That looks fine to me ( and I *think* the schema errata canonical form is OK ) [scribenm] +1 to both of Gudge's positions. [JacekK] JacekK has joined #xmlprotocol [anish] isn't there an issue of CRLFs for canonical base64? [scribenm] I believe the schema erratum makes clear that there needs to be (and now is) a canonical form of base64 [scribenm] Marc Hadley; do we really need to refer to canonical [Gudge] yes, and it defines line length and line ending sequences [scribenm] Some discussion between Noah and Marc, where Noah tries (and fails) to convince Marc that it's not OK for whitespace to change [scribenm] Proposal from Jacek and others: there is no property explicitly suggesting what to optimize. The rule is that a binding MAY optimize any, all or none of the items that are of dm:type xsd:base64binary. We agree to discuss later whether data to be optimized must be in canonical form, and if so what canonical. [scribenm] Proposal passes without objection. Meeting adjourned [RRSAgent] I see 4 open action items: [RRSAgent] ACTION: MarkN to go investigate MTOM-like technologies discessed at Binary workshop [2] [RRSAgent] ACTION: Marc Hadley re-contact plh and J. Marsh regarding attachment descriptions in wsd [3] [RRSAgent] ACTION: Editors to merge XQDM draft and current draft. in time for oct 8 [4] [RRSAgent] ACTION: editors of requirement document to incorporate the two above requirements [5]