XMLP WG telcon minutes, 27 October 2004

Based on IRC log

1. Roll
Present 6/5 + 2
Canon, Herve Ruellan
IBM, David Fallside
IBM, Noah Mendelsohn
Microsoft Corporation, Martin Gudgin (scribe)
SAP AG, Gerd Hoelzing
Sun Microsystems, Marc Hadley

Addison Philips, I18N WG
Andrea Vine, I18N WG

Canon, Jean-Jacques Moreau
IBM, John Ibbotson
Microsoft Corporation, Jeff Schlimmer
SAP AG, Volker Wiechers
Sun Microsystems, Tony Graham

IONA Technologies, Suresh Kodichath
Nokia, Michael Mahan
Oracle, Anish Karmarkar
SeeBeyond, Pete Wenzel
W3C, Yves Lafon

BEA Systems, David Orchard
BEA Systems, Mark Nottingham
Oracle, Jeff Mischkinsky

2. Agenda review

3. Minutes
[Gudge] no modifications requested for minutes
[Gudge] minutes of 2004-10-20 approved with no objection

4. Action items
[Gudge] First two actions will probably be resolved by our discussion in this call
[Gudge] Yves has now got all the files to post to the impl page

5. Status reports and misc
[Gudge] Anish noted in e-mail Media Type doc has been sent to the team and will
be published as a last call document RSN
No other reports.

6. Candidate Rec
[Gudge] test status; the impl page being updated with IRI examples

[Gudge] doc status:
[Gudge] docs in CVS do NOT have PR boilerplate

issue 502
[Gudge] http://www.w3.org/2000/xp/Group/xmlp-cr-issues.html#x502.
[davidF] http://www.w3.org/mid/"
[Noah] resolution suggested in http://lists.w3.org/Archives/Public/xml-dist-app/2004Oct/0017.html.
[Gudge] davidf link has response from I18N\Martin to Gudge
[Noah] Text is fine with me
[Gudge] no discussion on proposed text for 502
[Gudge] proposal to add the proposed text
[Gudge] no objection to closing issue 502 by adding the new text

[Gudge] Issue 501
[apphillips] http://lists.w3.org/Archives/Public/xmlp-comments/2004Sep/0018.html
[davidF] I18N precedence comment incorp in http://lists.w3.org/Archives/Public/xmlp-comments/2004Sep/0018.html
[Gudge] I18N clarify that their issue is about precendence of information attached
to resource representations
[Gudge] specifically, if there are conflicts, who wins?
[Gudge] Noah says that although we use a variety of specs; Unicode and optionally
multipart MIME
[Gudge] we don't necessarily list all the things they do
[Gudge] we try to use the underlying specs appropriately
[Gudge] where possible, we ensure the information in the INfoset and the information
at the MIME level (if any) match up
[Gudge] But we might not have MIME at all, or we might not have any information
at the Infoset level with which to populate any MIME headers that we might be using
[Gudge] We provide a generic mechanism for shipping around blocks of octets
[Gudge] our spec doesn't care whether they are jpegs, gifs, or whatever
[Gudge] our spec is completely neutral WRT what the bytes mean
[Gudge] addisson: we understand that, but our concern is that people can base64
encode a text file
[Gudge] and how do people figure out what encoding they have
[Gudge] for that text file?
[Gudge] we want to ensure that when possible the content-type includes a charset
[davidF] http://lists.w3.org/Archives/Public/xmlp-comments/2004Oct/0004.html
[Gudge] Gudge proposes that we add text to say 'If the xmime:content-type attribute
is present on a RRSHB and the resource is a text based resource, the value of the
attribute SHOULD include a charset parameter"
[Gudge] david: There is an illustrative example shown in the e-mail link above
[davidF] We also note that the working draft at [3] says of the xmime:content-type
[davidF] attribute:
[davidF] "The [normalized value] of the contentType attribute information item MUST
[davidF] be the name of a IANA media type token, e.g., "image/png", "text/xml;
[davidF] charset=utf-16""
[Noah] +1 to Gudge proposal.  The key is that this is driven by the attribute.
We do NOT IMO want to get into discussions that have to do with precedence of
arbitray combinations of info in the multipart.  The XOP model only deals with
what you can see in the Infoset, and the multiparts follow from that,
[Gudge] yup, that's where I'm trying to go
[Gudge] Addison: This would satisfy my concern
[Gudge] andrea: MIME sometimes doesn't carry the charset info. 
[Gudge] andrea: so you might not have the information required
[Gudge] Gudge: For us, MIME is a serialization detail. We always work at the level
of the XML infoset
[Gudge] david: Do Addison and Andrea feel that adding this statement will
adequately address 501?
[Gudge] addision; I think the proposal reasonably addresses our concerns
[Gudge] andrea: i concur with addisson
[Gudge] david: proposal is to adopt a 'SHOULD' for including the charset parameter
in cases where the media type identifies a text based type
[Gudge] no objection to the proposal
[Gudge] 501 is closed

[Gudge] davidf: this closes our remaining open issue. Are there any other issues?

[Gudge] davidf: If not, are we prepared to seek PR?
no issues are brought up

[Gudge] davidf: Any objection to seeking PR status after incorporating #501 and
#502 resolutions?
[Gudge] no objections
[Gudge] davidf: we will seek proposed rec

[Gudge] questions on process
[Gudge] davidf: PR to Rec is managed by team and voted by AC

[Zakim] leaving.  As of this point the attendees were Addison_Phillips,
David_Fallside, Noah, Yves?, Gudge?, Marc, Canon, Andrea

[RRSAgent] I see 6 open action items:
[RRSAgent] ACTION: Editors and Yves to update editor's copies to contain PR boilerplate [1]
[RRSAgent] ACTION: Gudge to send closing e-mail for #502 to Martin Duesrt and xmlp-comments [2]
[RRSAgent] ACTION: Editors to replace existing text with new from http://lists.w3.org/Archives/Public/xml-dist-app/2004Oct/0017.html [3]
[RRSAgent] ACTION: Editors to incorporate resolution of 501 as described in minutes of 2004-10-27 [4]
[RRSAgent] ACTION: Gudge to send closing e-mail for #501 to I18N and xmlp-comments [5]
[RRSAgent] ACTION: DavidF to sketch out the PR->Rec process and send e-mail to WG list [6]