IRC log of tagmem on 2003-08-18

Timestamps are in UTC.

[TBray]

TBray scribing

... discussion of agenda...

PCotton had made a point of coming to this meeting because provisional agenda said we were going to be discussing extensibility

Letting it slide has a very significant impact on our plan

PC: what is impact on downstream agendas of moving extensibility off this week?

RESOLVED: accept July 28 and Aug 4 minutes

[DanC_]

(we decided in Vancouver to cancel 1Sep)

[TBray]

Next two meetings have large numbers of regrets

SW: suggest next meeting 8 September

RESOLVED: next meeting 8 September

ACTION: SW to review work plan from Vancouver F2F to help with schedule

TBL: Possible conflict with other F2F meetings Sep 8

TBL: Might affect IJ too

[Norm]

FYI: 8 Sep was to focus on namespaceDocument-8

[DanC_]

(I haven't starting making travel arrangements for Bristol and still have conflicts and don't know how they'll be resolved.)

[TBray]

SW: Should he arrange a single hotel for Bristol?

DC: Consider net access

ACTION: SW to make a suggestion re hotel on email

----------------------

XML Binary Workshop: anything further

CL: those who are going should go, but not as TAG reps

SW: OK to answer questions on where TAG is at

PC: +1

Proposed new issue, see http://lists.w3.org/Archives/Public/www-tag/2003Jul/0022.html

[DanC_]

aye.

[TBray]

Discuss

[Norm]

discuss

[Stuart]

discuss

[Chris]

discuss

[TBray]

DOrchard: raises same issue re extensibility as PC did above

TBL: Deep issue

TBL: Cuts across many SW, RDF, CG, TAG issues

TBL: Risk of philosophical ratholes

TBL: BOF at Budapest conference, need to get this written down

TBL: IMHO need to write down how to interpret

TBL: Easy in RDF

[Stuart]

http://lists.w3.org/Archives/Public/www-tag/2003Jul/0022.html

[TBray]

TBL: confusion re 'denote'/'mean' etc

TBL: how are RDF/HTTP/OWL tied together

TBL: when you deref a predicate's URI, you can use the URI to get more info about it

TBL: discussions need to make sure they use URIs the way the rest of the Web does

SW: need TAG input?

[Chris]

If P means that given binary relation is *asserted* then thats ok, but that it *holds* is a different level of social meaning

[TBray]

TBL: Yes, TAG is 50% implicated

DC: history accurate, but missed technical issue

[Zakim]

TBray, you wanted to say that I feel poorly qualified to help RDF people & OWL people sort out their disagreements

[TBray]

TB: having trouble understanding the issue

SW: task force?

DC: Possible outcome: text in webarch saying how URIs are shared

DC: If you connect protocols to logic, you have answered the question of meaning

[Chris]

the architecture is that a single meaning is given to each URI (such

as P), that the URI ownership system makes statements by owners

authoritative weight, despite what other documents may say.

[TBray]

TBL: confusion between mean/denote

TBL: we agree that it's good for URIs to produce information

CL: you are claiming that if someone makes an assertion that a URI means X, that can never be changed

TBL: if I'm sesnding an RDF assertion to order a coat, and I have a URI for Pantone #1003, then when I dereference that I should get a colour chart

[Chris]

and who is 'someone' and what exacty is that meaning and where is it written and to what degree of precision

[TBray]

TBL: so to avoid disputes in future, that URI is authoritative

TB: OK, but what's RDF-specific?

TBL: meaning of statement is determined by predicate

TBL: which you can find out more about by dereferencing

TB: so do you want to say that when you identify things in RDF, you should make the URIs yield useful information

[Chris]

what TimB says is true, but is not reallt the issue at hand it seems to me.

[Zakim]

Chris, you wanted to point out edge flaws of TimBLs example

DanC_, you wanted to point out what folks might think is in the RDF spec that is, in fact, not there any more.

[TBray]

DC: current RDF spec makes no linkage between use of URIs in RDf and their use in HTTP

[Chris]

some of the statements in section 3 of TimBLs email are, to me, self evidently false

[TBray]

DC: It used to, but people objected

[Zakim]

TBray, you wanted to say that I don't understand potential outcomes well enough yes to say "yes" to issue

[TBray]

TB: seems that situation described by Dan is indeed bogus

[Roy]

There are hundreds of different uses of URI in HTTP

[DanC_]

straw poll pls

[TBray]

SW: straw poll

[Chris]

yes there is clearly an issue, and we should take it up

[DanC_]

yes, I think there's an issue that's worth the TAG's time.

[Norm]

yes

[Roy]

abstain

[TBray]

Abstain for now, but would like to ask a couple quewstions

PC: abstain

[DanC_]

the chair will please read the IRC responses

[TBray]

TBL: yes

[Stuart]

yes because needs more discussion

[TBray]

DO: leaning to "no" because there's probably an issue here, but in past when the issue-raiser hasn't been clear enough, we will say "we don't get it, more info please"

[DanC_]

Zakim, aaaa is TimBL

[TBray]

TBL: reprises last para of his email referenced above

oops, actually reprises whole email

[Chris]

that seems like an open-ended list to me

[Zakim]

TBray, you wanted to ask politics question

[Norm]

I'm content to accept the issue, but I have real concerns that TimBL is suggesting an attempt to "legislate morality". If I own a URI for my car and I assert my car is Blue, that doesn't make it true. And if eleven other people assert that it's Green, the fact that they're other people doesn't make their assertions false.

[DanC_]

interesting point, norm.

[TBray]

TB: suppose we do nothing, toss it back, what happens?

[Norm]

My car is, in fact, green. A pretty ugly green, in fact. :-)

[TBray]

TBL: they might build a consistent logic system with nothing to do with the web

CL: want to take it up although SW people might not like the answer

[Chris]

well said, norm

[DanC_]

yes, issue for the TAG

[TBray]

DC: if we do nothing, discussions will go on diffusely, do we want to be at center

[Roy]

abstain

[TBray]

Yes

[Norm]

yes

[Stuart]

yes

[Chris]

yes it is an issue but I request a clearer problem statement or statements

[TBray]

PC: abstain

TBL: Yes

DO: abstain

RESOLVED: issue accepted

[DanC_]

rdfURIMeaning

[Chris]

issue 42?

[Norm]

lol

[TBray]

RESOLVED: RDF-URI-Meaning-[++Ian]

[Chris]

** please ***

[TBray]

SW: approach SWCG for joint meeting?

ACTION: DC to take this back to SWCG

TBL: not clear that SWCG people are the right people

DC: which time slot?

various: their slot

[DanC_]

where "their slot" means "not the TAG slot"

[Norm]

FYI: it appears to be issue 39

[TBray]

PC: need to discuss how this issue affects progress to last call

PC: does this go to top of list?

PC: one reason I abstained is that I'm concerned about adding items to worklist

[Zakim]

DanC_, you wanted to set expectations to resolve this in Q1 2004

[TBray]

TB: don't see this one on path to last call

DC: spring 2004

[Norm]

The advantage of meeting on their time is that it doesn't have to step on our time as we progress towards last call

[TBray]

------------------------------------------------------------

Webarch

SW: RF's action item on sect3 re-write?

RF: if not done by Aug 18, won't get done for a while

[Chris]

Create an illustration of two resources, one designated by URI without fragment, and one designated by same URI with fragment...

please someone point me to a whiteboard photo, then i can draw it

[DanC_]

I felt so good after the planning session in Vancouver; ah well, "life is what happens when you're making other plans"

[TBray]

They'll be in the photosummary I iposted, Chris

ACTION TB: bring meeting photos to Ian's attention

[DanC_]

I can edit the ftf record, as can chris

[TBray]

modify action item, point mailing list at photos

[DanC_]

i.e. I am technically capable; my question is: May I? ah... yes, stuart answered. thx.

[TBray]

Action item to CL: in re bullleted-list re SVG reference

... discussion of various action items ...

TB/CL action item on "text-based" done

TB "xml-based" not done

NW: WIll get to his actions this week

redraft of Moby Dick section?

leave it pending

http://www.w3.org/2001/tag/2003/webarch-20030801

DC: .../tag/webarch/tim

"Integrate findings"?

TB will make himself available to IJ to work on this

-----------------------------------------------------------

Rewrite of intro

RF: Need to be able to change sections of the document

RF: not worth working on if it's not open to change

TB: Roy was trying to make a technical point about def'n of Web. That aside, I thought he prior language was a bit clearer

TB: obviously OK to change doc, but we need to have better feeling as for what parts of the doc are cooked

RF: still need to address problem of def'n of web.

RF: currently starts out by defining things as an information system, & follows on to resources from there

RF: but that leaves out SW, need to start from further back and then work forward to the description of the current browser-centric hypertextual web

RF: the web isn't an "information system" , it's the space of resources that are interconneected

[Chris]

in what way is a space of resources not a system?

[TBray]

RF: depending how you define the web constrains hwo you define what resource means

TB: def'n excludes software components?

RF: yes, because components change & are used depending on what you're doing

DO: <missed question>

[Stuart]

acl TimBL

[DanC_]

(re-starting discussion of what the web is doesn't speak well for our hopes for last call)

[TBray]

RF: My dissertation explicitly limits itself to the information system

TBL: not productive to go back and argue about what web really, really is

TBL: one subset is what you can get at with HTTP GET

TBL: email is part of the information space, but HTTP is very different from SMTP

TBL: could say info space (includes|doesn't include) things like email and HTTP

TBL: so don't need to spend time on Web "for purposes of this document"

TBL: ... fuzzy edges of what the Web is ...

TBL: the only way to get a handle is to write an ontology

TB: happier with a definition that includes the software as part of the web, but acknowledges that you might be able to start with a definition based purely in information

DO: webservices people have wrangled over what a web service is at length, settled on a definition explicitly limited to the doc they're writing, admit their may be things outside that are considered web services but that's not what we're talking about

[Zakim]

DanC_, you wanted to say yes, let's focus on the interaction between the terms we've using/defining in our doc... I hadn't appreciated the connection between 'Web' and 'Resource'

... that Roy points out, but I dunno what to do about it off the top of my head

[TBray]

TB: Aug 1st text is getting pretty close in quality to July 16 text, modulo my specific suggestions (in particular see notes on "effect of following web architecture")

DC: neither July nor August version is acceptable to all the TAG as of now