Points of Interest Working Group Teleconference -- 04 Aug 2011

<trackbot> Date: 04 August 2011

<robman> sorry...rebooting my phone

<karls> hi good morning, silence on the phone?

<karls> i'll dail in again

<karls> matt - have to step away for a minute or 2

<cperey> perhaps we should begin with the end of the agenda?

<cperey> F2F meeting stuff?

<scribe> scribe: Matt

Next F2F

<cperey> the registration page works :-)

matt: Raj, do you have any additional information on the space?

<scribe> ACTION: matt to create registration form [recorded in http://www.w3.org/2011/08/04-poiwg-minutes.html#action01]

<trackbot> Created ACTION-94 - Create registration form [on Matt Womer - due 2011-08-11].

<scribe> ACTION: Andy to create F2F details page [recorded in http://www.w3.org/2011/08/04-poiwg-minutes.html#action02]

<trackbot> Created ACTION-95 - Create F2F details page [on Andrew Braun - due 2011-08-11].

<rsingh2> registration page: v

<rsingh2> https://portal.opengeospatial.org/public_ogc/register/110919poi.php

matt: The dates are 9/19-20-21

<robman> sorry...redialling

matt: I'll copy this data into our system.
... Then pass it along to you Raj.

raj: This won't register you for the OGC meeting, if you want to attend, I can get you in.

cperey: The WG meetings there, some are closed, some are open, right? How will we know?

rsingh2: The link at the bottom of the reg page saying "more information", you can see a tentative agenda posted.
... Anything that is marked a "DWG", you can likely attend.

-> http://www.opengeospatial.org/event/1109tcagenda Agenda page

matt: Any recommendations?

rsingh2: Mass market is the most relevant to web technologies.

cperey: ARML meeting? Is that within Mass market group?

rsingh2: There's an ad hoc for ARML. Monday at 1600.

cperey: I think everyone in POI would be interested.

matt: I think we should as part of our agenda go to that.

cperey: My opinion is that the mass market DWG on Wednesday from 1-3 we should go to as well.

rsingh2: Geosemantics is relevant, it's RDF stuff.

karls: We'll have critical mass, that's a great thing.

cperey: The 3d summit might be of interest.

rsingh2: The 3d summit has a separate registration.

-> http://www.opengeospatial.org/event/11093dim 3D Summit page

matt: We should probably follow their agenda building and see if we should attend.

rsingh2: There's the Context SWG that is similar to what we're doing, building on Atom.

karls: Can we get someone from that group to present to us?

rsingh2: That's a better idea.

<scribe> ACTION: Raj to find someone from Context SWG to present to POI [recorded in http://www.w3.org/2011/08/04-poiwg-minutes.html#action03]

<trackbot> Created ACTION-96 - Find someone from Context SWG to present to POI [on Raj Singh - due 2011-08-11].

<cperey> thank you for showing us the OGC meeting agenda, Raj, looks like a busy week!

<cperey> for those who arrive early, I will be there

<rsingh2> http://2011.foss4g.org/

<cperey> is that the Open Street Map?

matt: The week before is the FOSS4G meeting, open source geo stuff.

rsingh2: There will be about 800 geo fanatics around Denver the week before.

matt: Any other missing details for the F2F?

rsingh2: There's also the State of the Map, OSM conference.

<rsingh2> http://stateofthemap.org/

rsingh2: It is September 9-11 in Denver.

Next Draft

-> http://www.w3.org/2010/POI/documents/Core/core-20110804.html latest draft

-> http://www.w3.org/2010/POI/documents/Core/core-20110804.html#CRS CRS stuff

matt: I tried to introduce the CRS stuff a bit easier, and then refer back to that section within each primitive.

karls: We dropped center, but have navigation point and point.

matt: What's the difference between CRS and SRS?

rsingh2: I can't remember exactly.

<robman> that's spatial vs coordinate no?

rsingh2: Always use CRS, but for some reason the parameter is always called srsName.
... SRS doesn't include some important concept, I can't remember which.

matt: What's the S stand for? spatial? spherical?

rsingh2: Spatial.

matt: Would it be okay to use CRS consistently?

<scribe> ACTION: matt to make srsName into crs [recorded in http://www.w3.org/2011/08/04-poiwg-minutes.html#action04]

<trackbot> Created ACTION-97 - Make srsName into crs [on Matt Womer - due 2011-08-11].

matt: Will URNs always be used or should it be http URIs?

rsingh2: That's OGC policy now

-> http://www.ietf.org/mail-archive/web/geopriv/current/msg00941.html URN thread in Geopriv

-> http://geoserver.org/display/GEOS/Unambiguous+Communication+of+CRS

matt: And then I was confused about the CRS sub-element stuff.
... I suspected that the "geometry element" mention was meant to mean within a location primitive?

rsingh2: I think it means everything.

matt: Does that mean you can't do <pois> and have it default, and then you specify a location with a crs that you can't do that?

rsingh2: Interesting. I don't know.

matt: I could see people mixing within a POI, e.g. to have an envelop in 2d but a point in 3d.

rsingh2: A clean way to do this would be to not allow the specification at the top level.
... Only specifying it at the top level isn't really clear either
... If you're not just doing latitude and longitude, it's an edge case, and you should specify it at the POI level.
... So, if you're just doing lat/lng you don't have to do anything, but each POI would have to specify the CRS if not lat/lng.
... The only case I can see that being annoying for is for 3d coordinates.

matt: I can see people frequently wanting a 3d point.

rsingh2: How about no default at POIs level.

<jens> that would be very verbose

matt: So it's only at the POI level?

rsingh2: For collections in 3d they'd have to write it for every POI.

<robman> +1 to that being too verbose

<Ronald> a bit too verbose indeed

matt: I can see a use case of a 3d point, but maybe a 2d poly.

rsingh2: So, make it so that within a location object having only one CRS.

karls: Relationship is a primitive?

matt: No, it's a bug. Moved back to being a locationship type.

karls: Labels for terms, link for membership, category for type of description and relationships for spatial.
... So, hours of operation would be a link relationship?

<robman> links sound good to me 8)

<scribe> ACTION: matt to work out example on representing hours using link and Good Relations. [recorded in http://www.w3.org/2011/08/04-poiwg-minutes.html#action05]

<trackbot> Created ACTION-98 - Work out example on representing hours using link and Good Relations. [on Matt Womer - due 2011-08-11].

<karls> be right back, boss calling

<scribe> ACTION: matt to remove whitespace/serialization from data model before publication [recorded in http://www.w3.org/2011/08/04-poiwg-minutes.html#action06]

<trackbot> Created ACTION-99 - Remove whitespace/serialization from data model before publication [on Matt Womer - due 2011-08-11].

ISSUE-43?

<trackbot> ISSUE-43 -- Do we want polygons to have complex topology with holes, etc -- raised

<trackbot> http://www.w3.org/2010/POI/track/issues/43

<cperey> sorry, had to drop off

<cperey> bye bye!

-> http://lists.w3.org/Archives/Public/public-poiwg/2011Aug/0002 Rob's proposal

matt: I'd say, let's raise an issue and talk about it next week.

robman: I think it makes profiles easier, but doesn't have to go now that we're close.

<karls> am back

matt: So, I'd like to update the XML syntax over the next two days, then publish on Tuesday. Sound good?

[[yes]

matt: I'll also update label and link.

karls: I've got some folks geared up to write examples. I told them to pause for a bit, and talked to them last night. They're re-engineering their spec, and I'd love for them to comply.
... I'm trying to contemplate on when to pull the trigger on getting them to do this.

<robman> +1 to test implementations/usage

matt: I would say have them do it Tuesday.

- DRAFT -

Points of Interest Working Group Teleconference

04 Aug 2011

Attendees

Contents

Next F2F

Next Draft

Summary of Action Items

Scribe.perl diagnostic output