scribenick, webr3

<scribe> scribe: nathan

<dbooth> scribenick: webr3

<jar_> this one: http://www.w3.org/2001/tag/awwsw/ir-axioms/

<scribe> ACTION: nathan to add links to wiki [recorded in http://www.w3.org/2011/03/01-awwsw-minutes.html#action01]

<trackbot> Created ACTION-41 - Add links to wiki [on Nathan Rixham - due 2011-03-08].

Requirements doc

nathan please add links to http://www.w3.org/wiki/HttpRange14Webography

<scribe> Scribe: webr3

TAG issue tracking

http://lists.w3.org/Archives/Public/www-tag/2011Feb/0150.html

<jar_> http://www.w3.org/2001/tag/group/track/actions/534

<jar_> here's the relevant email re 57: http://lists.w3.org/Archives/Public/www-tag/2011Mar/0000.html

<jar_> Change title to: "Mechanisms for obtaining information about the intended

<jar_> meaning of a given URI"

jar: does anybody object to the title

all: no objections

<jar_> (no particular input from group)

<jar_> nathan thinks that's a good title

next steps on www-tag (doc review, or a call of some kind)

jar: i had two questions
... 1: next step on the broader continuation of uri-meaning work
... if you guys want to give input, you can

dbooth: this is tag work yes, seems like we need to get a draft for tag review

jar: there are two things going on here, consensus doc for issue-57 review, and awwsw tf report
... two different docs, w/ intertwining paths

dbooth: it seems ambitous to do two docs

jar: i think we need 2 or 3 docs

dbooth: so a report of where we're at or?

jar: produce some kind of ontology, or vocabs or

<jar_> series #1: awwsw reports/notes (consensus within awwsw, reporting to tag & community)

<jar_> series #2: tag reports/notes (consensus in TAG and/or in community)

<jar_> this TF is responsible for #1

<jar_> #2 would be tag finding and/or rec track

<jar_> jar thinks intent to do #2 should be announced sooner rather than later, since otherwise situation will continue to fray

<jar_> dbooth: wants #1 note in hand before going to www-tag. awwsw consensus

<jar_> db: need to get over barriers of terminology and confusion

<jar_> jar: framing the tag issue?

<jar_> oar background?

<jar_> nathan: 1) summary of space and views 2) consensus in awwsw re interoperability 3) axioms or ontology

<jar_> dbooth: don't try to be too ambitious re #1... too comprehensive gets out of control... but need to say something

<jar_> ... #3 could be separate

<jar_> (agreement that #3 stands on its own)

<jar_> #1 & #2 are a second document

nathan: can we do 3 until 2 is done?

<jar_> dbooth #3 can be helpful in getting clarity... work on them in parallel

all: general agreement - sounds like a plan

<jar_> jar: how quickly can we do this?

<jar_> ... what do we need to do to get there?

<jar_> dbooth: start with ir-axioms doc & owl

<jar_> ... doesn't solve issue 57 of course

<dbooth> http://dbooth.org/2009/lifecycle/

<dbooth> http://dbooth.org/2010/ambiguity/

jar: we want documents as short as possible

<dbooth> Those documents both address the social obligations involved

<dbooth> but we could start with those documents in addressing issue-57

<jar_> port 80 means http, without a marker...

<jar_> 200 is supposed to mean IR without a marker... but the obligation is being resisted

<jar_> implied obligation

<jar_> Plan A = refer to IR using its URI, refer to thing-described-by-doc using 303 URI, #, .well-known, tdb: etc

<jar_> Plan B = refer to tdb using 200 URI, refer to IR *** UNKNOWN *** (maybe IRW or ir-axioms)

<jar_> 1) summary of space and views

<jar_> we need a vocab that doesn't assume either plan...

<jar_> two classes of things, OVERLAPPING. IR, and thing-described-by-document.

<jar_> maybe instead IR, it's IR-with-reps-available-at-URI

Axioms doc

<dbooth> http://www.w3.org/2001/tag/awwsw/ir-axioms/

<jar_> a simple IR is like a representation, except that it's an IR

<jar_> simple IR is like content-location:

<jar_> dbooth: unseasy with simple IR

and REST: ''Some resources are static in the sense that, when examined at any time after their creation, they always correspond to the same value set.''

<jar_> a simple IR is one that has only one representation (fixed resource)

<jar_> dbooth: instead of trinity, we have 4 things, URI, IR, simple IR, representation

<jar_> jar tbd: A simple IR has only one representation... by definition

<jar_> the writeup says this, but not clearly enough. will fix

both returning the same int - /mary/age vs /bob/age

<jar_> TimBL says these are different IRs with the same representation

<jar_> "have different meaning"

<jar_> dbooth: How to make this more palatable?

<jar_> dbooth: model an IR as a function... a simple IR would be a constant function...

<jar_> doc needs to say very plainly that a simple IR is one that has a single representation

<jar_> nathan: Is the IRI / URI part of the simple IR? Does a simple IR 'know' its own URI?

<jar_> source URI is like provenance

<jar_> simple IR could be modeled as a pair (rep, prov) where rep is a represenation and prov is provenance

<jar_> provenance might or might not involve some URI

<jar_> provenance is history and/or material context

<jar_> IR -> {simple IR = (rep, bits)} -> rep

<jar_> rep may be shared, under different provenances, but

<jar_> sorry scratch

<jar_> IR -> {simple IR = (rep, provenance)} -> rep

<jar_> rep (mathematical; bits) can be shared among multiple simple IRs (puns, coincidences)

<dbooth> I'm trying to frame this in terms of IR as a function from (Time x Request) to Representation, and Simple-IR is a constant function from (Time x Request) to Representation.

<jar_> can't model simple IR as a constant function AND have 2 simple IRs with same rep

<dbooth> "Does Mary authorize this?" vs "Does Bob authorize this?"

<jar_> SIR1, SIR2 both have same rep, but Mary authorizes one, Bob the other.

<jar_> nathan: like having the question + the answer.

<dbooth> GET on SIR1 yields: "yes"

<dbooth> GET on SIR2 yields: "yes"

<jar_> 'speak for' is a better idea... does Mary authorize the resource to speak for her?

<dbooth> I would say that SIR1 is bound to one URI U1, and SIR2 is bound to another URI U2, and that's how you know who authorized.

[[ Even given an enumeration of syntactic parts, a simple IR's identity is not determined - two simple IRs might have all the same parts yet have distinct origins (provenance). ]]

<jar_> Maybe convince TimBL and others to agree that representations are information resources?... no way

<jar_> TimBL says, that there can be two fixedresources (simple IRs) that have the saame representation, yet are different

<jar_> (intentional identity, not extensional)

<jar_> two blank sheets of paper...

<jar_> suppose Bob's doc and Mary's doc come from the same server, different URIs...

<jar_> nathan: That seems fair

<jar_> Suppose one IP address, multiple domain names resolving to same IP...

<jar_> these are different resources...

<jar_> If no host: then the requests would be identical

<jar_> URIs have distinct meaning

This is scribe.perl Revision: 1.135  of Date: 2009/03/02 03:52:20  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Succeeded: s/would/would be/
Found Scribe: nathan
Found ScribeNick: webr3
Found Scribe: webr3
Inferring ScribeNick: webr3
Scribes: nathan, webr3
Default Present: DBooth, jar, [IPcaller]
Present: Jonathan_Rees Nathan David_Booth
Got date from IRC log name: 01 Mar 2011
Guessing minutes URL: http://www.w3.org/2011/03/01-awwsw-minutes.html
People with action items: nathan