Points of Interest Working Group Teleconference

05 May 2011


See also: IRC log


Matt, Vinod, cperey, Raj, Andy, Ronald


<trackbot> Date: 05 May 2011

<Ronald> Hi all, the meeting room I reserved is still occupied, so I will call in in a couple of minutes

cperey: I need to take regrets for the next two weeks of calls

<scribe> scribe: matt

Some links: http://www.w3.org/2010/POI/documents/Core/POI%20Core%20Draft.html http://www.w3.org/blog/POI/2011/05/03/poi-core-editors-draft-released/

Next F2F

Andy: Still working on logistics for Budapest f2f, which will hopefully be resolved by Monday

cperey: The poll was before the Jonathan said the OMA people would be interested in hosting.

Andy: I think what we really want to get done today is general consensus on publishing as FPWD.


Andy: FPWD doesn't mean the spec is done, unable to change, etc. It means: "Here's something we've put work into and want to show the public." It also kicks off some official W3C-ish actions.
... I think we should try to get it out next week, because of a publishing moratorium it would mean late this month otherwise.
... Anyone have comments on that idea?

<cperey> +q

rsingh2: Are we putting it out before integrating the discussions we've had this week?

Andy: I think we can integrate and then release on the 12th.
... So, let's use this call to have the discussions we need.
... Talk about them, get them integrated and move ahead with fpwd. I think it would be a good step for this call.

rsingh2: Sounds good to me. The one thing missing was a real overall "this is how you do a POI".

matt: agrees

rsingh2: There's the model and primitives then the XML and the example.
... I wanted to see a little conceptual graphic.

<cperey> +1 on needing introduction

rsingh2: So maybe spending time on getting an introduction.

matt: I am all for more introductory text, especially if someone wants to help write it!

cperey: The examples are good, do they take the place of the use cases? Or are they complimentary to the use cases?

matt: How we want to integrate the use cases into the document I left undecided.

cperey: I'd like to make sure the first draft have reference to them.

rsingh2: I'd say leave the use cases out or put them into an Appendix.

-> http://www.w3.org/2010/POI/documents/Core/POI%20Core%20Draft.html#poi-data-model Data Model

rsingh2: There should be some sort of list of what the data model is, or a graphic.

<rsingh2> http://www.w3.org/2005/Incubator/geo/XGR-geo-20071023/

matt: I think I also need to explain the data model vs syntax a bit better, some responses seem to think the XML syntax elements would map one to one in the data model, when the intent is that all bits of the data model are covered.

rsingh2: We struggled with this too in the Geo Vocab
... We used a very simple UML model. We put brackets around words that looked like XML, but not really needed. Then in the 3rd section we talk about serializations and encodings.

matt: I should steal from this doc for the draft.

Andy: We publish to get the comments like we're getting thus far. They'll fall out from here.

cperey: The use cases in the geo vocab are in the appendix.

rsingh2: I think I could put together something like section 2's graphic.

matt: I'll outline it a bit and share it with rsingh2

<scribe> ACTION: matt to outline datamodel, pseudo-DTD and send to rsingh2 [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action01]

<trackbot> Created ACTION-65 - Outline datamodel, pseudo-DTD and send to rsingh2 [on Matt Womer - due 2011-05-12].

cperey: What does Member mean here?

matt: I was just trying to come up with a word to express the relationship, I stole member from C++

Ronald: How about contains?

cperey: Yep!

<scribe> ACTION: Matt to change "Members" to "Contains" [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action02]

<trackbot> Created ACTION-66 - Change "Members" to "Contains" [on Matt Womer - due 2011-05-12].

cperey: And 3.1.6 object has no members.

matt: It hasn't been written yet.
... I don't really know what we mean by Object, there's the Karl thing in the wiki, cperey's thread on the mailing list

cperey: And danbri did a bunch of work too.

matt: Is that okay to leave it TBD for now?

cperey: I think AR folks will shoot it down.
... The fact that it's not filled-in is going to confirm their prejudices.
... They already think we're looking at this from a geo-reference/spatial point of view, and that non-GPS based is being left out.

matt: Maybe it will attract them to it?

cperey: Not so far.
... We've looked at this a number of times, why they're not participating, etc.

Andy: Who should we talk to about it?
... Who will be upset about this section not being completed on FPWD?

cperey: The companies that do AR using objects or things that are not part of this activity.

Andy: One or two thought leaders?

cperey: There's Total Immersion where place is completely irrelevant, they always use visual recognition.
... Mark Billingsworth from HIT Labs NZ.
... Timo from Fraunhoffer who works on X3D DOM.
... And Alex.

matt: How about you and I work together and try to figure out what the latest/best is?

cperey: I'd say put what you think is best even if it is controversial. An object is non-stationary POI that can be 2 or 3d, it is a thing with a physical manifestation.

<vinod> Sorry, I have to go.

cperey: You can give stationary examples, but you can also give non-stationary examples.

matt: I'll put something in there to signal we're not ignoring it.

cperey: You can send me your text if you want review. I don't think we've got any new thoughts about it.

Andy: Any other gaps that if it was released today would be detrimental? Other than references, introduction and object?

cperey: And use cases
... Time? Categorization?

matt: Time had an action item on Martin and Thomas, I was hoping they'd get to me. Categorization I didn't really have an idea of what we said.

cperey: Somewhere in the last few weeks we decided we shouldn't reinvent it.

<Ronald> +1 to cperey

matt: I tried to establish a category that points to the US Census, and then have an association to a particular label within the the system.

<scribe> ACTION: matt to flesh out category section (see example) [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action03]

<trackbot> Created ACTION-67 - Flesh out category section (see example) [on Matt Womer - due 2011-05-12].

matt: I'll ping Martin and Thomas on time primitive.

-> http://www.w3.org/2010/POI/documents/Core/POI%20Core%20Draft.html#object Object

rsingh2: On things that move, you have to have multiple geographies. Or time stuff for it. Moving, I'm not necessarily advocating it for the fpwd.

cperey: The people who care about moving are not part of the discussion, they'll probably ignore it.

rsingh2: I don't know if you want to try to get it in there in five days.

cperey: I'd rather have something in there that is wrong.

matt: I'd lean towards something in there that is wrong than empty.

rsingh2: Put multiple georeferences in there with different start/end times.
... I think the map georeference is also best left to client-side.


matt: I should have noted in the doc that I don't get the map geo-reference either.
... Drop it or big notes?

cperey: No preference.

rsingh2: I'd say drop it.


Ronald: I'd say drop it too.

<scribe> ACTION: matt to drop map geo-reference [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action04]

<trackbot> Created ACTION-68 - Drop map geo-reference [on Matt Womer - due 2011-05-12].

rsingh2: map might be a good candidate for an extension

<Ronald> +1

matt: So, I'll churn out a new doc based on these actions and minutes and send out a tentative publication request for next Friday.

Andy: Do we have enough for this call?

rsingh2: I've got more comments: Carl wanted addresses to be handled using the IETF spec and for spatial relationships we have a canonical nine relationships that should always be included. He can write those up pretty quickly.
... So, touch base with Carl.

<danbri> sorry couldn't make it, will read up the logs here

<danbri> am v pleased to see a candidate draft draft doc in the mix :) really helps firm up discussion

<danbri> congrats!

Summary of Action Items

[NEW] ACTION: Matt to change "Members" to "Contains" [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action02]
[NEW] ACTION: matt to drop map geo-reference [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action04]
[NEW] ACTION: matt to flesh out category section (see example) [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action03]
[NEW] ACTION: matt to outline datamodel, pseudo-DTD and send to rsingh2 [recorded in http://www.w3.org/2011/05/05-poiwg-minutes.html#action01]
[End of minutes]

Minutes formatted by David Booth's scribe.perl version 1.135 (CVS log)
$Date: 2011/05/11 17:59:55 $

Scribe.perl diagnostic output

[Delete this section before finalizing the minutes.]
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)

Found Scribe: matt
Inferring ScribeNick: matt
Default Present: Matt, +3539149aaaa, Vinod, +1.617.848.aabb, cperey, +1.617.764.aacc, Raj, +1.919.439.aadd, Andy, Ronald
Present: Matt +3539149aaaa Vinod +1.617.848.aabb cperey +1.617.764.aacc Raj +1.919.439.aadd Andy Ronald
Agenda: http://lists.w3.org/Archives/Member/member-poiwg/2011May/0003

WARNING: No meeting chair found!
You should specify the meeting chair like this:
<dbooth> Chair: dbooth

Found Date: 05 May 2011
Guessing minutes URL: http://www.w3.org/2011/05/05-poiwg-minutes.html
People with action items: matt

WARNING: Input appears to use implicit continuation lines.
You may need the "-implicitContinuations" option.

[End of scribe.perl diagnostic output]