See also: IRC log
<Chris> Scribe: Norm
<Chris> ScribeNick: Norm
Absent: RF
Chair notes a fair amount of administrivia on today's agenda.
5-10 minutes on issues list maintainance
<noah> Regrets for next week
Regrets for next week
<Chris> I can scribe next week
pbc gives regrets as well
<DanC> (I'll be on my way to the airport 17 Jan)
Next meeting: 24 Jan 2005
Meeting of 17 Jan 2005 cancelled.
<Chris> "The Last Call review period ends 28 January 2005, at 23:59 EDT"
<DanC> try all caps ACTION:
<DanC> and ACTION Chris: ... is sometimes better than ACTION: Chris
<scribe> ACTION: CL to Post QA review comments for email discussion.
Minutes of 20 Dec 2004 accepted.
Discussion of http://www.ietf.org/internet-drafts/draft-rsalz-qname-urn-00.txt
DO asks if we want discussion of this ID on www-tag?
This seems related to the QNames as Identifiers issue.
General agreement that it's ok for discussion to go to www-tag
<scribe> ACTION: SW to Respond to DO approving discussion of the ID on www-tag.
<Zakim> DanC, you wanted to comment re minutes/agenda tools
SW opens discussion of TAG contribution to technical plenary day
PC: volunteered to participate in the planning committee but has had to back out. Suggest that we ask for 1+ hours and put together a comprehensive proposal including our issues, perhaps E+V.
... It would have broad interest which is important at the plenary
CL: TP is the first f2f after 1 Feb?
SW: Yes, although we're talking about the Wednesday day.
<Chris> okay
Wednesday is 2 Feb, fwiw.
<noah> Reminder: I believe that David Orchard is planning to join us on this call in part to discuss our plans for versioning discussion at the Plenary. Am I remembering correctly?
Wednesday is 2 MAR, not 2 Feb. Oops.
<DanC> (heh... the dbooth script will change that to "... is 2 MAR, not 2 Mar")
SW: Housekeeping: introduce new members of the TAG.
... WebArch? Substantive issues, E+V or httpRange-14?
Some discussion of how an E+V presentation might work with participation from TAG, Schema and other relevant participants.
SW: We need to introduce the new TAG.
PC: I disagree; Wednesday should be as technical as possible with the minimum amount of administrive overhead
noah: If the reason to have the TAG there as a whole is introductions, that's 3-5 min. What really do we want to have covered though on behalf of the TAG at this session? Versioning is one interesting issue, but it's odd because it spills over beyond the tag. We have others, httpRange-14, looming WS-Addressing issues, perhaps others?
... Did we collect a set of intersting things as we went through the arch document? Should we survey the state of play on unresolved issues; listing them with a 2-3 min introduction to each. Note that we're entering a process of looking at these and tyring to set our agenda for the coming year. Should we share that with the plenary so that they can have some input.
... Perhaps we should do a "state of the TAG" and split E+V off into another slot.
<Zakim> noah, you wanted to suggest maybe we survey unresolved issues?
<Zakim> DanC, you wanted to say oh... no, not httpRange-14 for the big meeting on weds. only a small thing, with position papers prerequisite for attendance and to question the logistics
DanC: I can imagine presenting in a survey style; collecting data will be frustrating if we don't use the WBS survey.
... If we're going to do a survey, we should use the machine to help
... on httpRange-14: please no, not with a large audience.
... I think that should be a small group with a 1 page position paper required to even get in the door
SW: So I've heard noah suggest that we make this a two bite sort of thing: TAG for one and E+V for another.
TBL: For what extent are we concentrating on what we haven't decided yet?
<noah> FWIW, the survey aspect was somewhat secondary in my thinking. I had in mind more of a: "let us remind you what issues look challenging to us and why". Now, we can either discuss a few in the remaining time, and/or solicit your sense of which are important and whether there are other directions we're missing.
TBL: Should we use this time to present the architecture document or a number of things which are not in dispute?
DanC: While I think we should present the webarch document, I wonder if this is the right audience. If there's a new WG, they should get a presentation, but this group participated in review of the document.
SW: Last year we had theme-oriented panels and that seemed like a good thing
<DanC> yes, please, theme-oriented. E+V
<DanC> yes, content/presentation... CSS, XSL, DI, ... .
CL: I suggest separation of content and presentation because there are several different approaches to the problem (CSS, XSL, vs. Device Independence)
... Might be interesting to see if we have more or less agreement than we expect.
... If we decided to do that, I'm prepared to do some preparation for that.
SW: Is there a third choice, or should we stick with two?
... I can go back to Steve Bratt with those two.
<Chris> I can do some introductory slides on that, to set the scene
<noah> Are we at the point where we should invite DaveO to dial in?
SW: The idea, I think, is to hold one meeting early in the half of the week, a "stakeholders meeting" where we focus on motivating needs and requirements.
Some discussion of the email about this topic
noah: Will the meeting be best served by inviting everybody? It's just a matter of logistics.
Message in question: http://lists.w3.org/Archives/Member/chairs/2005JanMar/0002.html
noah: scratch comment about inviting everybody; was confused, thought we were still talking about the Wednesday panel session
... The Schema WG feels that they have some ownership over the extensibility and versioning issue. But they also see that it's a much larger issue.
... My guess is that we need to do a deep dive to get organized and complement each other going forward
... There are also other communities that are our users and they have different needs as well.
PC: Why would we wait until the plenary to get together with schema on this issue?
... We've got a few meetings in February and at least one more in January
noah: I think it might be more valuable to do it face-to-face.
<scribe> ACTION: SW to To schedule some sort of meeting with Schema between now and the plenary
DanC: any news on getting the Schema work public?
noah: Yes, I did the narrow part of the action to alert the chair.
... Maybe I should take another actoin to follow up?
<scribe> ACTION: NM to Attempt to get the Schema WG to make their extensibility and versioning use cases work public.
Some discussion about administrivia associated with doing this in a WG that has experienced some shrinkage.
SW: wrt E+V, we're talking about a panel session, we might also have a shareholders meeting if we get feedback; and we'll try to have liason with schema before the plenary
Norm: No issues at present, but it's easier to cancel than schedule if we decide we do have issues
<Zakim> DanC, you wanted to ask what of "ACTION NM: to explore means of getting current and future Schema WG work on versioning into public spaces"
They would like to meet. Will work with SW on topics.
They would like to meet. We've had a possible new issue that may be related.
<Zakim> DanC, you wanted to ask didn't SKW take an action to contact ws-addressing WG?
<noah> +1 to idea that we meet with WS-Addressing. Suggested agenda: they walk through their design and issues, preparing to defend non-use of URIs
DanC wonders who has the ball on setting up that meeting
SW: I have the ball.
PC: Which WGs would we meet with if we were going to flatten all the deferred issues
<Chris> Note that I don't have an SVG f2f competing with TAG at the TP, this time
<Chris> However I am meeting with some WG on SVG behalf
<scribe> ACTION: SW to give Steve Bratt a response to how the TAG would like to participate in the Plenary
<Zakim> DanC, you wanted to ask if SKW has all the TP balls, and wonder if spreading the work around would be easier or just introduce more mess
SW: We've got to the piece on E+V. I think you sent out a couple of revised drafts just before Christmas.
... Do you want to tell us what's changed?
<DanC> (hmm... when we left our hero, the ball was with the readers, not the writers)
dorchard: I thought I went through that in Boston.
<noah> Were they just before Christmas, or just before Boston?
SW: Ok, unfortunatly I wasn't present.
<DanC> "ACTION PC: to review parts 1 and 2 of extensibility and versioning editorial draft finding prior to discussion for 10 Jan. ACTION DC: paulc to review parts 1 and 2 of extensibility and versioning editorial draft finding prior to discussion for 10 Jan"
<DanC> oops!
Correction: Before Boston not before Christmas.
<DanC> http://www.w3.org/2001/tag/2004/11/29-30-tag#xmlv41
DanC, PC have not yet completed review; action pending.
<Zakim> DanC, you wanted to say sorry, no, didn't read it, see earlier agend request to discuss issue list maintenance
noah: Is it appropriate to go over at least my recollection of our tentative decisions
<DanC> "ACTION NM: to work with DO to come up with improved principles and background assumptions that motivate versioning finding"
<noah> Right, thanks for finding that.
SW: At the moment we are a little lacking in feedback. Apologies for dragging dorchard here before we were well prepared.
<DanC> ACTION: DC to review blog entry on RDF versioning [pointer?]. CONTINUES.
dorchard: The only thing that's happened since then is that I did write up a paper that examines RELAX NG in this context.
... I tried to come up with a small set of scenarios that I've been working with.
<dorchard> http://www.pacificspirit.com/Authoring/Compatibility/OWLRDFExtensibility.html
<Zakim> Chris, you wanted to suggest an example
CL: One example that might be interesting for RELAX NG would be taking an empty element and adding an attribute co-constraint that says whether or not it should be empty.
<noah> Dave specifically contrasts his scenarios, which are sort of mechanistic (add an attribute to a type) to the more general user level scenarios the schema wg has been using as motivation (I.e. phrased as business scenarios)
SW: Should we continue or cancel some of these actions? If they're continued, can you set expectations?
DanC reviews them:
<DanC> ACTION: NM to to work with DO to come up with improved principles and background assumptions that motivate versioning finding. CONTINUES. bigger than a 1-week thing
NM: to work with DO to come up with improved principles and background assumptions that motivate versioning finding
<DanC> PC asks to withdraw: ACTION PC: to review parts 1 and 2 of extensibility and versioning editorial draft finding prior to discussion
<DanC> http://www.w3.org/2001/tag/2004/11/29-30-tag#xmlv41
<DanC> SKW offers to review part 1
<timbl> To review ... http://lists.w3.org/Archives/Public/www-tag/2003Oct/att-0137/vers-adoc.html ?
dorchard observes that the TAG is in flux until the elections finish
<DanC> ACTION: DC to paulc to review parts 1 and 2 of extensibility and versioning editorial draft finding prior to discussion.
<DanC> (I think I can do it this week)
<DanC> . ACTION PC: paulc to inform QA and Schema WGs of the new version of the e&v draft
<Stuart> http://lists.w3.org/Archives/Public/www-tag/2003Oct/att-0137/vers-adoc.html
<DanC> ACTION: SW to to inform QA and Schema WGs of the new version of the e&v draft
<Stuart> http://lists.w3.org/Archives/Public/www-tag/2004Nov/att-0071/versioning-part2.html
<dorchard> http://lists.w3.org/Archives/Public/www-tag/2004Nov/att-0071/
<dorchard> http://lists.w3.org/Archives/Public/www-tag/2004Nov/att-0071/versioning-part1.html
SW notes that he's put the wrong link in the agenda
dorchard: between now and February 12, is very bad for scheduling additional meetings
noah: but you will be at the Tech Plenary
dorchard: yes
SW: Should we shoot for the 14th?
<DanC> yes, 14 Feb looks like an interesting sync-point
noah: You're hoping Schema WG will be available then?
SW: yes
dorchard agrees that falling back to 21 Feb would be ok
SW: Plan is to have our reviews finished for discussion on 14 Feb
<scribe> ACTION: NM to Coordinate with chair of Schema for meeting on 14 Feb
noah: Schema chair agrees informally
<Chris> NW: posted a draft a while back, went through www-tag discussion
<Chris> nw: mostly discussion was about a single issue
<DanC> (agenda cites http://www.w3.org/2001/tag/doc/xmlChunkEquality.html )
<Chris> nw: xml:lang and case folding and non-ascii characters and stuff
<DanC> ([Editorâ\u20ac\u2122s Draft] TAG Finding 07 September 2004)
<Chris> nw: apart from that, little feedback
<DanC> action 11 = NM: Coordinate with chair of Schema for meeting on 14 Feb
<Chris> sw: how was that recieved by xml core?
<Chris> nw: reluctant as there was not a single correct response for all of xml
<Chris> nw: however, this was juts 'a' notion of equality not 'the' one, so they declined
<Chris> nw: thus, the ball is in TAG court again
<Chris> nw: Core is fine with TAG doing 'a' way
<Chris> dc: looks like an interesting note
<Chris> nw: can tag publish a wg note?
<Chris> dc, cl, sw: yes (probably)
<Chris> pc: why bother changing the finding to a note?
<Chris> nw: one psychological step less normative. not in all caps.
<Zakim> DanC, you wanted to review actions from http://www.w3.org/2001/tag/2004/11/29-30-tag#xmlv41 and to and to say that my instinct about xml:lang is to write a test case and then see
<Zakim> noah, you wanted to ask about precendents of findings as notes
<Chris> nm: don't want to get into duplicat work, republish findings and notes. rather train people how to read findings
<Zakim> Chris, you wanted to expound about xml:lang
<Zakim> DanC, you wanted to say that my instinct about xml:lang is to write a test case and then see what implementations do, but hey, with a note, just stick a "xml:lang looks hairy.
<noah> slight clarification: I'm not necessarily against publishing as notes if we can convince ourselves there's a good reason that justifies the duplicate investment in publication and ongoing maintenance in the face of possible bugs. I'm suggesting we decide on the criteria in general. I think Dan is now suggesting such a criterion.
<pbc> as an example of another algorithm.
<Chris> cl: bogus language codes do not affect well formedness, but anything not conforming to the prose of xml still does not conform to prose
<pbc> should we also point to http://www.w3.org/TR/xquery-operators/#func-deep-equal
<Chris> cl: its well defined, except for this theoretical corner case
<Zakim> timbl, you wanted to ask whhy we started this.
<DanC> yes, noah
pbc: yes, we should
<Chris> tbl: can we take norms finding and see to what exent dsig breaks on things that are the same, if RDF would be happy with it as a definition of RDF litteral, etc
<Chris> tbl: xml has a deep equality issue (from scribbled notes)
<Chris> pc: xquery and xpath, rather than xml??
<timbl> s/scribbled notes/http://lists.w3.org/Archives/Public/www-tag/2004Jan/0013.html
<Chris> pc: so, this is one possible way to compare two chunks
<timbl> Woudl they have prefered it with ':'?
<Chris> pc: it does not really demonstrate that other algorithms exist, when you might use them etc
<DanC> (RDF uses the c14n one)
<Chris> pc: deep=1 from F&O, plus???
<noah> Tim's scribbled notes say: XML schema has the "Deep equality" issue as to when any two chunks
<noah> are "equal".
<noah> FWIW: I'm not immediately calling to mind any reason that XML Schema would care.
<Chris> pc: dsig have various views on canonical representations
<Chris> pc: so if we added some other alternatives, its fine as a finding
<DanC> (hmm... it's now starting to smell like a survey of the literature on XML chunk comparison, more in the finding genre)
<Chris> tbl: a list of things to avoid or known potholes, is valuable
<noah> Schema does have equality rules for typed values of particular fields, e.g. when typed as an integer attribute AT="123" is equal (as a key for example) to AT="00123".
<Chris> yes, more like a finding now
<noah> I don't think there are open issues in this area.
<Chris> sw: not overriding support for making it a note
<noah> I note that DSIG achieves something close to chunk equality via its canonicalization rules (which I believe are user-pluggable).
<Chris> pc: ok as a finding if we had time to discuss it, so suggesting improvements
<Chris> sw: ok so very much as a finding
<Chris> ACTION: NDW to make editorial improvements, point to other different schemes, why use them, things to avoid in XML Chunk Equality.
<Chris> tbl: is there vagueness in RDF literal?
<Chris> pc: good to see when F&O deep= works and when it does not
<Chris> pc: looking for a way to select an algorithm to get least astonishment
<Chris> tbl: for URI comparison, if they are trivially equal they are always equal with more complicated meythods
<Chris> tbl: is the same thing true here? if two chunks are norm=, are they always equal??
<Chris> sw: oops, running out of time, this is interesting
<Chris> pc: put at front of next agenda
<Stuart> Adjourned