See also: IRC log
NM: F2F minutes review anyone?
DC: Look good
JK: I've looked at 8 and 10
NM: I propose to postpone these until January
<DanC> -1
HST: I would like to publish
NM: Any objections to approving?
JR: Could we say go ahead if no objections in a few days?
NM: RESOLUTION: Minutes will go out unless Chair hears objection by Monday 21 Jan
NM: Obviously I will fix minor bugs, only delay for substantial pblms
<DanC> +1 NM is welcome to edit as he sees fit
NM: Action to Larry to send 3 December minutes
LM: I need help with the originals
HST: I will send you a draft
<DanC> action-215 due 22 Dec
<trackbot> ACTION-215 Send minutes of 3 Dec TAG teleconference to www-tag for review due date now 22 Dec
NM: No telcon 24 or 31 December
... Next telcon 7 January 2010
... Moving next f2f to 24--26 March
... based on change in TAG f2f
LM: IETF is meeting week of 22
March
... Progress on URI work, I really need to be there
<masinter> http://www.ietf.org/meeting/cutoff-dates-2010.html#IETF77
NM: Would you come if we stuck with the old dates?
LM: I am not sure about the 22 week
NM: 17-19 would still work?
LM: Yes
<masinter> yes, 17-19th is better, i just realized the IETF conflict for 21-26
<johnk> I don't mind either way
HST: Do you have firm confirmation from TV that he will come?
NM: I think so. . .
DC: And what about TBL?
... I would rather not make this decision today
HST: I thought TBL had one day
pblm
... in the week of the 22nd
... Doodle poll?
<jar__> it's capability-based
<jar__> :-)
NM: No resolution
<DanC> close ACTION-346
<trackbot> ACTION-346 Collect March 2010 W3C Team day info closed
<scribe> ACTION: Henry to put up Doodle poll [recorded in http://www.w3.org/2001/tag/2009/12/17-minutes.html#action01]
<trackbot> Created ACTION-365 - Put up Doodle poll [on Henry S. Thompson - due 2009-12-24].
<scribe> ACTION: Noah to bring f2f date proposal to group based on poll input [recorded in http://www.w3.org/2001/tag/2009/12/17-minutes.html#action02]
<trackbot> Created ACTION-366 - Bring f2f date proposal to group based on poll input [on Noah Mendelsohn - due 2009-12-24].
action-366 due 6 January
<trackbot> ACTION-366 Bring f2f date proposal to group based on poll input due date now 6 January
<DanC> (the WG doesn't close bugs; the editor does.)
LM: I'm not sure the original TAG submission is actually on record wrt the poll which is now underway
<Zakim> DanC, you wanted to ask the chair to say what he knows about TBL and TVR's availability for the proposed dates and to ask LMM if change proposals in both ways are on the table and
The relevant HTML issue is http://www.w3.org/html/wg/tracker/issues/76
<DanC> I want the poll, not the issue
DC: Change proposals in both directions on the table?
LM: Yes, in both directions
<masinter> http://dev.w3.org/html5/status/issue-status.html
<DanC> http://www.w3.org/2002/09/wbs/40318/issue-76-objection-poll/
<masinter> and http://www.w3.org/2002/09/wbs/40318/issue-76-objection-poll/results
NM: Q2 is objections to proposal to
split; Q3 is objections to proposal to keep
... This looks like an internal exercise on the part of the HTML5
WG
... arising from our concern
... So we could just wait
... Or we could/should make a comment
HST: I think we were being asked to comment
DC: Only members of the HTML WG can comment
HST: OK, I was wrong
DC: What's gotten missed?
LM: The points in the rationale for the bug we raised
DC: Which ones?
LM: [points from http://www.w3.org/Bugs/Public/show_bug.cgi?id=8220]
<DanC> there's a para with "high profile content management system"
DC: I think those are covered
NM: Anyone other than LM like us to do anything here?
<DanC> and "Added Benefit of modularizing Microdata"
HST: Yes, I do -- I want 8220 read into the poll
<DanC> (you want 8220 in the question or in a response?)
HST: No, I want it entered as an objection to the 'keep' option
NM: But we already filed the bug -- we can use that to come back in
<DanC> (ah. as a response. I can get that into the team input to the poll)
DC: So now I understand HST wants
this in the response
... I can get it in the W3C Team response
HST: I would prefer not to dilute the team response with the TAG's points
NM: So we need someone in the WG
LM: I'm willing to let this go -- NM has convinced me
NM: Are we comfortable with that?
HST: What I would like is for NM to ask the HTML5 chairs to treat our 8220 bug as input to the poll, specifically as "An objection to keeping Microdata in"
<masinter> 1+ to chairs & archive
<DanC> in particular, www-archive@w3.org ; it predates the public- convention
DC: Copy to www-archive?
HST: sure
<scribe> ACTION: NM to ask the HTML5 chairs to treat our 8220 bug as input to the poll, specifically as "An objection to keeping Microdata in", cc to www-archive@w3.org [recorded in http://www.w3.org/2001/tag/2009/12/17-minutes.html#action03]
<trackbot> Created ACTION-367 - Ask the HTML5 chairs to treat our 8220 bug as input to the poll, specifically as "An objection to keeping Microdata in", cc to www-archive@w3.org [on Noah Mendelsohn - due 2009-12-24].
NM: we got email from Maciej http://lists.w3.org/Archives/Public/www-tag/2009Dec/0087.html
and replied http://lists.w3.org/Archives/Public/www-tag/2009Dec/0088.html
... I thought Maciej's response was weak wrt Mike Smith's draft, so
pushed back on that front
... LM also pushed back a bit http://lists.w3.org/Archives/Public/www-tag/2009Dec/0095.html
... Any action needed now?
DC: I think we've gone too far
<Zakim> masinter, you wanted to say that i'm ok with letting this go
DC: We should just look at what they now do
LM: I'd like to see it on a schedule or list of deliverables
DC: What list? Their charter?
LM: Not clear
<Zakim> ht, you wanted to back NM's response wrt Mike Smith's doc
NM: I'm sympathetic to LM's point --
if a WG takes on a commitment to produce a WD and take it to Last
Call, there is some implied commitment to put effort into
that
... Not a Charter change, I think it's already covered
... But I would like to see a public commitment to do the work,
with success criteria
<DanC> (basically, I now where maciej lives; if he doesn't give us what we ask for, I know how to start the discussion about why not.)
HT: My reading of what Maceij said made a firm and binding commitment to the author view, which was one of the things we asked for. There was no commitment on the Mike Smith draft, and I think it was appropriate to push back.
HST: We didn't get what we asked for
NM: I think the requests crossed
<DanC> (what we asked for is last call on the html 5 reference; how could we have gotten that already? sigh.)
HST: I don't think we should micromanage the WG
NM: What about the first request, wrt the authoring view -- are you happy?
HST: Yes, I think that commitment was adequate
NM: Anything further we should do?
<masinter> "plus the fact that it was actively maintained and reviewed by itself for quality"
LM: If you agree with my concern
about the authoring view, . . .
... I agree that the editor has agreed to move stuff if it's
misclassified
... but what's missing is any commitment from the HTML5 WG to
review the result for quality
... But I am willing to drop this
NM: Me too, reluctantly
<DanC> (we just dealt with ACTION-359 so I'll close it.)
<DanC> close ACTION-359
<trackbot> ACTION-359 Communicate TAG resolution to HTML WG closed
LM: New input from JR on version
identifiers
... My inclination is not to update the existing doc. http://larry.masinter.net/tag-versioning.html, but write a new one
that's shorter
... I'm also waiting for input from HST on XML-compatibility
guidelines
... As long as you're careful to distinguish implementations from
specifications
... and implementations evolve faster than specs do
... then the utility of VIs is limited
<DanC> (tracker, note we're touching on ISSUE-41)
<noah> I'm not sure I buy the "version of implementations" point, at least as a typical idiom
LM: because it doesn't map to
implementations reliably
... But 'limited' doesn't mean 'none'
NM: So you're suggesting closing 283 and opening a new one?
<DanC> -1 "schedule" actions. please just actions to do technical work; they'll naturally get scheduled.
LM: Suggesting closing 283 and scheduling short discussion for guidance before I decide whether to take a new document forward
<DanC> a pending review action is implicitly an action for the chair to schedule discussion.
NM: Is your existing email sufficient?
<masinter> http://lists.w3.org/Archives/Public/www-tag/2009Dec/0094.html
DC: Could we review whether HT minuted LM's analysis correctly above?
<johnk> P&C = http://www.w3.org/TR/2009/CR-widgets-20091201/ (from Larry's email)
<masinter> i think the minutes are OK so far, the email says more
LM: So given that what users want VIs for is to identify implementations, they don't work in the case cited
<Zakim> noah, you wanted to ask for clarification on the implementations versioning
NM: I don't quite understand where
you're trying to go with implementation versioning
... I think we need to be clear about the role of senders and
receivers
<DanC> (examples are great; let's go slow enough to get them recorded)
NM: Suppose we had stability for a
while wrt version 3
... and now innovation has set in
NM: We had code written which was generating instances of that version of the language
NM: And there were agents which implemented that version of the language to process documents
<DanC> I can go with "name implementations" as a paraphrase for "name receiver understandings"
<jar__> it's an ontology ... hierarchy of classes whose members are implementations
<DanC> I wrote it as an ontology.
<DanC> my ontology for this stuff: http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0136 'formally defining W3C's namespace change policy options w.r.t. recent TAG versioning terminology'
<Zakim> masinter, you wanted to note that, in the cases where implementations evolve faster than specfications, version indicators are not useful in producer/consumer communication.
LM: Even in that lossy situation, there is still value for VIs in the production context
<noah> I think the problem is that we're not being clear on what is being identified. In many cases, people grab pieces of different specs, knowing full well which they're grabbing from
<DanC> tx for writing it down, LMM. I see the key point: "in the cases where implementations evolve faster than specfications, version indicators are not useful in producer/consumer communication."
LM: And sometimes, when there's an
incompatible change, there can be utility even in the
producer-consumer communication
... context.
LM: So that could all be an update to what we say in AWWW about VIs
<noah> I tend to be anti-VI in general.
<masinter> Danc, are not *as* useful
<DanC> I just copied/pasted what you wrote, masinter . but ok, *as* useful.
<masinter> DanC, I was correcting myself, the utility in producer/consumer communication is limited to the unfortunate situation where it's necessary to introduce incompatible changes
<DanC> gotcha, lmm
DC: I am somewhat positive about this, but not sure updating AWWW is the most effective route -- I like blog articles these days
NM: Languages evolve in multi-faceted
ways -- sometimes it's chunky, when there's agreement to move from
V3 to V4
... When you talk about implementations evolving, that's masking
that fact
... because some impls are innovating wrt the video tag, and some
other impls are innovating somewhere else
NM: In that kind of situation, you onlyneed VIs if there are conflicting interpretations somewhere
NM: as long as it's all "this tag is
supported, or it isn't" then there's no need
... Differentiating the production pipeline from the
producer-consumer connection doesn't really get at that issue
NM: I'm not sure it helps
<DanC> (trying to get consensus on this doesn't seem like a good use of TAG discussion time; I'd much rather see LMM write it up as he sees it and let NM either comment on that or do a separate piece.)
<Zakim> masinter, you wanted to disagree with NM about 'needing multi-dimensional version indicators'
LM: Multi-dimensional VIs? Where I'm
going is that you only need VIs for specifications
... If you need a VI for "version 3 plus the video tag from Apple
impl. of xxx and the other tag from foobar impl of ..."
LM: then you need to write a spec.
that says that, and you can have a VI
... I'm only trying to find a minimal utility claim: there is value
for VIs for published specs.
... Maybe there are other use cases, but that's the one I care
about right now
<masinter> My belief is that you only need version indicators in the language to indicate versions of specifications. And if there is a need for a version indicator for something, you would have a spec for it
<DanC> (I guess I should turn http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0136 into a blog article; wish I'd done it a while ago.)
<noah> Larry says: you only need VIs for specs. My (noah's) position is: you need VIs whenever the same content might mean different things, or when tools or consumers want early warning
NM: I'm pointing to another important case, which is when there's a conflict between two interpretations, i.e. if the documents are ambiguous
[scribe is not keeping up]
<masinter> Noah, there may be other uses for which you might also want version indicators to help with, but they don't, or can't, in the situation where languages evolve independently outside of standards
LM: So I hear you asking for VIs in cases where they are needed, but it's hard to see how to get them
<DanC> +1 we've made some progress and are reaching diminishing returns; please sketch actions
<Zakim> ht, you wanted to ask NM to write up what he said
NM: The IRC log will have to do
DC: I will write up my ontology in this space
<DanC> ACTION Dan write up version change ontology as blog item http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0136
<trackbot> Created ACTION-368 - Write up version change ontology as blog item http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0136 [on Dan Connolly - due 2009-12-24].
<jar__> +1 to LM's new story, too.
LM: So no value in the email?
HST: No, I like it, please write it up
<masinter> ok, i wanted enough discussion to get feedback
<DanC> action-368 due 1 Mar 2010
<trackbot> ACTION-368 Write up version change ontology as blog item http://lists.w3.org/Archives/Public/public-swbp-wg/2005Sep/0136 due date now 1 Mar 2010
<masinter> action-283?
<trackbot> ACTION-283 -- Larry Masinter to update document on version identifiers w.r.t. Cambridge June discussion -- due 2009-12-10 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/283
<DanC> close action-283
<trackbot> ACTION-283 Update document on version identifiers w.r.t. Cambridge June discussion closed
<DanC> . ACTION Larry write a shorter document on version indicators
<DanC> ACTION Larry write a shorter document on version indicators
<trackbot> Created ACTION-369 - Write a shorter document on version indicators [on Larry Masinter - due 2009-12-24].
<masinter> action-309?
<trackbot> ACTION-309 -- Henry S. Thompson to draft input to HTTP bis draft re sniffing based on 8 Dec discussion -- due 2009-12-09 -- PENDINGREVIEW
<trackbot> http://www.w3.org/2001/tag/group/track/actions/309
<johnk> I'll note that I (this morning) bumped the dates on all of my agenda-linked actions until the 7th
<johnk> (of Jan)
HST: My new proposed input is at http://lists.w3.org/Archives/Public/www-tag/2009Dec/0068.html
<masinter> +1 like Henry's text
<DanC> +1 "don't always meet their obligation to provide correct headers
<noah> Propose: "don't always meet their obligation to provide correct Content-type headers"
<DanC> and +1 "...configuration issues..."
DC: Not happy with the "servers don't supply correct ..."
NM: Agree with DC
<masinter> I'd be happy to have the TAG send it with just a note saying that we'd like them to review it
<jar__> I don't like "obligation"
<masinter> i don't like Noah's rewording
<johnk> Propose: "provide Content-type headers which do not correctly identify the content sent"
JAR: I prefer HST's wording
<masinter> johnk's is good
<DanC> +1 "...headers which do not correctly..."
<masinter> provide a content-type header, to be accurate
<noah> suggest s/alter/increase/ the security exposure
<jar__> "correct" is a weasel word, and I like it that it is
<jar__> +1 Johnk
<DanC> +1 send on behalf of the TAG as ammended
NM: Any objections to adopting JK's proposal
<jar__> +1 TAG
[none]
NM: replace 'alter' with 'increase'
HST: Agreed
<DanC> +1 send on behalf of the TAG as ammended x2
HST: Propose to send this from me on behalf of the TAG
<jar__> +1
<johnk> +1
NM: RESOLUTION: HST to send a revised-as-amended version of http://lists.w3.org/Archives/Public/www-tag/2009Dec/0068.html to the HTTP bis list on behalf of the TAG
<scribe> ACTION: Henry HST to send a revised-as-amended version of http://lists.w3.org/Archives/Public/www-tag/2009Dec/0068.html to the HTTP bis list on behalf of the TAG [recorded in http://www.w3.org/2001/tag/2009/12/17-minutes.html#action04]
<trackbot> Created ACTION-370 - HST to send a revised-as-amended version of http://lists.w3.org/Archives/Public/www-tag/2009Dec/0068.html to the HTTP bis list on behalf of the TAG [on Henry S. Thompson - due 2009-12-24].
<DanC> close action-309
<trackbot> ACTION-309 draft input to HTTP bis draft re sniffing based on 8 Dec discussion closed
DC: Been closed
<DanC> ACTION-356 due 12 Jan 2010
<trackbot> ACTION-356 Work to schedule followup meeting on xmlnames next week due date now 12 Jan 2010
<DanC> http://lists.w3.org/Archives/Public/www-tag/2009Dec/0070.html
<DanC> "proposals are due January 16, 2010"
<DanC> thread continues http://lists.w3.org/Archives/Public/www-archive/2009Dec/0025.html
<Zakim> masinter, you wanted to note I also submitted a bug/change proposal
DC: [asks people to read the email at
http://lists.w3.org/Archives/Public/www-tag/2009Dec/0070.html]
... Polling the group
NM: I'm not spun up
LM: I've already pushed back on this issue
<noah> Could someone paste a link to the HTML 5 usage that's causing concern?
JAR: I'm sympathetic to the proposition "there is no such thing as what you [Julian Reschke] call a resource" -- Ian Hickson, in http://lists.w3.org/Archives/Public/public-html/2009Sep/1133.html
JAR: We've used this word for a long time, it confuses things to use it in a contradictory way
DC: I expect that if TBL were here he'd say the Hypertext web doesn't need this distinction
AM: There's been alot of stuff
written about this
... It would be useful if we could agree and write something
small
<noah> +1 to being very careful, agree with HT
<masinter> http://www.w3.org/Bugs/Public/show_bug.cgi?id=8264
<masinter> "The result is also at odds with reality (since some resources have no representation and content-negotiated resources may have many)."
<masinter>
<jar__> Ian "There is no such thing as what you [Julian] call a resource" - Dan & Jonathan sympathetic
<noah> That's the bug report, but where's the HTML 5 text that's causing concern. Quick search of the HTML 5 draft doesn't reveal it to me.
<noah> Which HTML 5 section are we discussing?
<DanC> lots of them, noah; one of hixie's msgs to www-archive says which
HST: Trying for a "yes and" response -- I liked what Rhys Lewis was trying to do before we lost him
<masinter> Noah, it's spread throughout the document, there was someone from Oracle who did an analysis
<Zakim> masinter, you wanted to note that HTML should normatively reference IRI spec which also uses 'resource' and 'representation'
LM: I made some arguments in this bug report http://www.w3.org/Bugs/Public/show_bug.cgi?id=8264
<DanC> (oh. we have like *1* minute left. didn't hear that somehow.)
LM: One important point is that you are going to have to
read other documents to implement anything on the web
... and understand them
... and they use these words differently
... The bad news is that fixing this is a big and messy job --
these terms are spread out throughout the HTML5 draft
<DanC> noah, when is our next meeting?
<jar__> I think the harder problem is that no one likes "representation", really... it forces you to buy into REST/AWWW
<noah> Section 2.1.1 has title "resources", but seems to define the terms "supported" and "Mime Type". I'm a bit at sea. Is that implicitly the definition of resource as advertised?
<DanC> ooh... good point... "resources which have multiple resources" is goofy.
LM: A concrete problem is that there are resources with no representation, and resources with many: if you have only one word, you can't address this at all
NM: We will have to come back to this
JK: We need both concepts, we have to do whatever we have to do to fix that
NM: Adjourned until 2010-01-07
<jar__> I like "document" for Ian:resource and awww:Representation ...
+1 to JAR
<jar__> ... but timbl likes "document" = awww:Resource ...
OK, I am about to start quoting FRBR, watch out :-)
<DanC> FRBR is good stuff.
<DanC> ACTION-363?
<trackbot> ACTION-363 -- Jonathan Rees to inform SemWeb CG about market developments around webfinger and metadata access, and investigate relationship to RDFa and linked data -- due 2010-01-31 -- OPEN
<trackbot> http://www.w3.org/2001/tag/group/track/actions/363
<DanC> jar's TAG actions http://www.w3.org/2001/tag/group/track/users/38732