22 Dec 2009


See also: IRC log


DBooth, jar, DanC, mhausenblas, TimBL




<jar_> gotta reboot my computer. very weird, and weird timing

<jar_> oops seems to be ok now.

<jar_> also weird timing - it fixed itself just when i threatened it


<mhausenblas> http://www.w3.org/2001/tag/dj9/story.html

<timbl> JAR in principle

debrief from TAG meeting

<dbooth> ScribeNick: jar_

<DanC> jar's msg http://lists.w3.org/Archives/Public/public-awwsw/2009Dec/0017.html

<dbooth> Scribe: jar

<DanC> http://www.w3.org/2001/tag/2009/12/08-agenda

<mhausenblas> http://www.w3.org/2001/tag/2009/12/08-agenda#HttpRedire

danc finding relevant part of tag meeting for us to review

<DanC> jar presented on the 9th http://www.w3.org/2001/tag/2009/12/09-minutes.html#item01


<DanC> slides as presented http://lists.w3.org/Archives/Public/www-archive/2009Dec/0024.html

jar: The slides are in two parts, the awwsw part and the jar experimental part

danc: orresponds_to goes backwards

it gets indexed backwards

in ordinary speech, corresponds to is symmetric, which this isn't

<DanC> RESOURCE has information ENTITY

<DanC> (where in N3, 'has' is a noise word)

david: i like the speaks for stuff

jar: michael, do you see connection between conneg slide and your feb email to www-tag?

mh: yes

david: Looks like corresponds to is broken down into speaks for and is about?
... httprange-14 slide - ??

danc: speaks for is its own world, different from the http world
... the slide asks: what if you connect the two?

david: what if you treat sf as a subproperty of corr_to?

danc: walk through examples - that will show answer

mh: What precisely is missing - for final report?
... The deadline in dec was very good

danc: An audience is its own deadline. In Dec audience was the TAG.

<DanC> (pointer to list of outstanding issues?)

mh: Need something more complete by march

<mhausenblas> pickup agenda item 2

speaks_for stuff http://www.w3.org/2001/tag/dj9/story.html

<DanC> Revision: 1.26

slight fixes from 25 to 26.

<timbl> wonder which of 30 windows and tabs is running a talk on language acquisition

danc: Different editorial style compared to jar

david: It's going very smoothly, very nice

mh: What's the motivation? You could have used FOL?

danc: The later use is supposed to motivate it

we do use FOL with quoting

danc: but you need some way to have statements with term
... i have a reasoner that uses URIs as terms

david: Why does WSJ have different relationship?

danc: When bob reads the WSJ web page he doesn't read any further

<DanC> * perhaps elaborate WSJ c:controls { Fed rate ?R }. using quoting

david: Why different from [other case]?

danc: ... it introduces says, speaks for a la ABLP axioms
... the 'quoting' operator

david: Why do you need 'quoting' - not just 'says'?

danc: sugar
... cwm has a lot of trouble with dequoting ('controls')
... That onto is broken. borrows from cyc.

<timbl> danc: The owl time ontology is horribly broken

mh: Overlapping intervals and so on, seems good?

<DanC> http://www.w3.org/2001/tag/dj9/httpspeech.html

danc: Don't expect people to want to read n3 rules in this document

<mhausenblas> Michael: I was referring to http://www.w3.org/TR/owl-time/

danc: Borrowed http in rdf -- need to fix it if it turns out to be broken
... (seems related to powder?)
... Status code numbers - ontologically unfortunate

<timbl> ("status codes are alreday numbers")

danc: Referring to 'get' method for usage of 'information ... with'
... Connection - found a bit of ABLP in HTTP spec even without the controversial assumption

after 2nd note in httpspeech

<timbl> Maybe s/corresponds to/expresses/ ?

about what the *connection* says.

danc: jar wanted to account for buggy proxy servers

<mhausenblas> I'm looking http://rdfs.org/ns/void-guide#sec_2_1_Classic_vs_3rd_party

mh: looking at a special case of linked data where there is a third party linking data sets
... hugh glasser from southampton... does this still hold?

danc: You need an independent policy that determines whether the 3rd party controls the information
... 3r party = non authoritative. Trusted means you start believing them

david: Only origin servers are authoritative ... trying to understand
... The statement in question is res information E, yes?

danc: yes

<DanC> http://www.w3.org/2001/tag/dj9/httpsocial.html

jar: Be very careful about how and when the corresponds / speaksfor axiom is used

david: Re misconfigured server issue
... couple of ways to think of the resource. We could say there's no such thing as a misconfigured server, only bogus resources

danc: So when does the resource speak for the domain owner? Let's see....

david: Sort of like a misconfigured telephone answering system

danc: Inadequate level of mistrust - people vs. computers - we trust computers too much

david: Melissa on the phone is different from resource on internet, but similar

danc: How do we figure out that conneg of is/is about is a lose?
... limits of conneg? picture and description of picture?

<mhausenblas> DanC, you actually answered to my question (indirectly ;) see http://lists.w3.org/Archives/Public/www-tag/2009Nov/0083.html

'indirection needs redirection' - danc reviewing

david: stuck on n3 - replaced the manifesto file with some data

"The organization replaced the manifesto file on the server with some metadata about the manifesto,"

david: comes from which server?

danc: This is just a statement about the world

<Zakim> mhausenblas, you wanted to ask on http://www.w3.org/2001/tag/dj9/badmeta-pf.html

'contraction' axiom in logic

<DanC> (I can live with light scribing)

<mhausenblas> scribenick: mhausenblas

<DanC> * perhaps picture in inches, description in centimeters

DanC: maybe an example with a picture (in vs cm)

dbooth: depends on the coding process (lossy/gainy)
... for example JPEG

DanC: Mashup Speech Acts (with XSRF example)

<DanC> typo: "credentials to for all requests that the user is authorized to make:"

<DanC> perhaps change conclusion to: bobAccount debit "1000000". malloryAccount credit "1000000"

<DanC> * clarify: "he has a current cookie"

<DanC> * fix "We know that Q1 speaks for mallory, not bob, "

<DanC> * perhaps "Lets look at the bank's (faulty) reasoning."

<dbooth> DBooth: Need to say "Mallory causes Bob's browser to issue this request ..."

DanC makes dbooth read 'Mashups and the Same Origin Policy' out loud

<DanC> * perhaps another layer between hs:messageEntity and information for parsing

Michael: what's the plan with this document?

DanC: got some todos in there
... review is useful

<dbooth> Here is ABLP: http://www.hpl.hp.com/techreports/Compaq-DEC/SRC-RR-70.html

<dbooth> DBooth: This is nice, Dan! The more you can do on this, the better.

<dbooth> ADJOURNED

thanks a lot DanC - KUTGW! ;)

<DanC> SRC-RR-70 has the same content, but for copy/paste purposes, the other citation works better http://homepages.inf.ed.ac.uk/gdp/publications/Calculus_for_Access_Control.pdf <- http://www.w3.org/2001/tag/dj9/story.html#toplas93

cya all in 2010

<dbooth> :)


<DanC> AwwswHome topic

<DanC> s///

Summary of Action Items

[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2009/12/22 15:37:31 $