<josh> sigh+ remind of webex password, please

<josh> sdw

<ByronCinNZ> +present

<BartvanLeeuwen> http://www.slideshare.net/semanticfire/action-85-64205032

<scribe> scribe: Linda

<eparsons> Topic : Approve last week's minutes

eparsons: lets start

<eparsons> PROPOSED : Approve last week's minutes

<eparsons> http://www.w3.org/2016/07/06-sdw-minutes.html

<BartvanLeeuwen> +1

+1

<billroberts> +1

<kerry> +1

<Nicky> +1

<RaulGarciaCastro> +1

<aharth> +1

<jtandy> +1

<eparsons> RESOLUTION : Approve last week's minutes

<ChrisLittle> +1 minutes

<eparsons> Topic : Patent Call

eparsons: checks attendees - ok

<Maik> +present Maik

<eparsons> https://www.w3.org/2015/spatial/wiki/Patent_Call

<eparsons> Topic : Demo

[silence]

<josh> +1 minutes

<BartvanLeeuwen> http://www.slideshare.net/semanticfire/action-85-64205032

eparsons: bart and nicky have put together a demo. Mentioned at BP call last week but better to show it in this bigger group.

BartvanLeeuwen: link to the slides is here.

<billroberts> yes, got it

<ByronCinNZ> +1

<kerry> +1 works

BartvanLeeuwen: we do a lot of work for emergency response orgs who need to work together in crisis situations
... for proper information sharing between them we created the Firebrary, a SKOS vocabulary
... people started putting things on maps so then sharing geo-information became a topic
... for example here a map of the Netherlands with companies at risk for high water
... represented as purple dot but user has no idea what it means

<eparsons> jtandy just the one...

BartvanLeeuwen: in the demo we have geo data in a postgis data, served in a standard WFS but also R2R mapping to RDF
... used a rdf:about attribute on the data, linking to Firebrary
... and can visualise the data based on the semantics
... based on DWBP work
... SKOS can also be used to map icons from different communities or countries.
... could this be a BP?
... Is this in scope?
... is rdf:about the right way of doing a backlink?

<BartvanLeeuwen> http://geo.resc.info/geoserver/netage/wms?service=WMS&version=1.1.0&request=GetMap&layers=netage:fb_demo&styles=&bbox=7.00324749210733,52.2615836944433,7.00881387704046,52.2640494501209&width=768&height=340&srs=EPSG:4326&format=application/openlayers

BartvanLeeuwen: this map shows red dots, each has a link to RDF information about the item

<aharth> +1

jtandy: use of rdf:about to include identifier which allows users to lookup information, right?

BartvanLeeuwen: yes, this separates use of WFS only for markers on maps, the information is elsewhere

jtandy: why rdf:about?

BartvanLeeuwen: had to pick something, any suggestions?

<aharth> rdfs:seeAlso

jtandy: anyone on the call has suggestions?

aharth: rdf:about is a remnant of rdf/xml
... but not worrying to use this. Cool demo.
... is there live lat/lon information in the maps?

BartvanLeeuwen: no, it's a scenario drawn on a map, but live info is also possible

aharth: do you have datasets with polygons as well?

BartvanLeeuwen: for me representing geometries in RDF was out of scope for this demo

josh: there is an RML extension that encorporates the geosparql spatial functions
... would be interesting to standardize that

<jonblower_> +1 to what Josh is saying about ontop-spatial. It's good but documentation is lacking

josh: Interesting way of linking information to spatial data, valuable.
... but not easy

eparsons: if this isn't a BP, what would it need to be to become one?

josh: should be general; in linking spatial data the links should be available

BartvanLeeuwen: the links should be from the SDI to the data published on the web

eparsons: would be valuable to give a contribution on this in the BP

jtandy: so the BP is if you provide data through an SDI, publish the information in a complementary manner and link them

BartvanLeeuwen: yes, and somehow do the backlinking

eparsons: any comments on the backlinking?

we hear you rob

<josh> It seems to me to be a mismatch to develop links "back" to non-linked data. It would make more sense to me to make sure that SDI data is (alternately) accessible as linked data.

+1

roba: see redundancy in the identifiers in the example
... you could generate some automatically during rendering
... Also, in LD world there are often multiple views, one of them can be backlinks.
... e.g. links from data back to services where you can get the data

eparsons: Thanks Bart for this interesting example.

<eparsons> Topic : Simple Spatial Ontology,

<josh> https://www.w3.org/2015/spatial/wiki/Further_development_of_GeoSPARQL

josh: put some info in the wiki
... there needs to be a common way to express spatial data so that you can have a common way to query it using geosparql.
... the spatial ontology in geosparql 1.0 has good things but is not well-used
... if we create version 1.1 it would be an OGC update of an existing standard. Not sure what it would be in SDWWG terms
... geosparql 1.0 is considered too complicated, but just using lat/long is too simple and unexpressive
... update could be to add simple geometry properties
... another update is to have a way to infer relationships between features from relationships between geometries
... Features and Geometries are disjoint
... but are both SpatialObjects and share the same possible relationships
... There was an issue with three different types of spatial relations which many found confusing.
... explains the new proposed spatial ontology
... Another possible improvement is the addition of serialization types, e.g. GeoJSON
... theoretical foundation in ISO191xx
... simple geometric properties could be added
... not done, some things are not completely clear yet.
... maybe define a practice for CRS reference and definition, perhaps machine readable?
... or for UOM
... there's a version on Webprotege, but this is hard to work with. Will work with ttl file on Github

<Zakim> aharth, you wanted to ask about conneg for geometry serialisations and to ask about possible compatibility with W3C wgs84_pos vocab

aharth: maybe possible to make the new voc compatible with the simple geo vocabulary by Brickley.
... would be good for existing data on the web
... Also, using conneg for serialization: can that be considered?

josh: compatibility with W3C vocabulary: looked at it, but in existing data the use of it is ambiguous.
... so worthwile doing, but existing data might have meant different things.
... About conneg: you could offer different serializations of the data. Have this in separate properties?

aharth: depends on which community you do it for. Using conneg on URI is fav of some communities

josh: how do you negotiate against a property within a geometry? Useful, but not clear how to.
... will think on it

eparsons: continue this on email list

<eparsons> Topic : Coverages update

<eparsons> https://lists.w3.org/Archives/Public/public-sdw-wg/2016Jul/0172.html

billroberts: we made progress, in this email there's a proposal for what we should do in the coverages subgroup
... a fair bit of work but achievable
... need to move on to really doing stuff
... take the CoverageJSON work, though incomplete, and see if that could become a W3C rec / OGC standard
... conceptually it's compatible with existing work, but addressing a user group that is not happy with existing approaches. Web developer type audience.
... also looking at how to use data cubes, different audience.
... For discussion: pure RDF datacube is useful for representing coverages, and a note could be writtne on how to do that
... Also, outstanding tasks for BP narrative.
... Also, plans to write a Primer alongside the spec.
... we welcome opinions through maillist and our call next week.

ChrisLit: extensions to datacube model are needed.
... e.g. datatiles, webtiles. Could be interesting followup work. Will get in touch with bill about this.

<josh> Any new coverage encodings or representation do need to share the conceptual model of existing coverages.

eparsons: out of time

<billroberts> thanks Chris - will follow up wiht you on that

eparsons: will send email about meetings during holiday

<josh> One issue to explore is that the coverage conceptual model is really a mapping between domain and range spaces. Expression in grids or cubes is convenient, but not fundamental.

eparsons: closes meeting

bye!

<eparsons> linda thank you !!!

<BartvanLeeuwen> thx bye

<RaulGarciaCastro> bye

<kerry> bye1

<BartvanLeeuwen> thx Linda

<jtandy> bye ... thanks to Ed for Chairing and Linda for scribing

<roba> bye

This is scribe.perl Revision: 1.144  of Date: 2015/11/17 08:39:34  
Check for newer version at http://dev.w3.org/cvsweb/~checkout~/2002/scribe/

Guessing input format: RRSAgent_Text_Format (score 1.00)

Found Scribe: Linda
Inferring ScribeNick: Linda
Present: eparsons NickyVanOorschot BartvanLeeuwen kerry ScottSimmons billroberts aharth ByronCinNZ Linda josh jonblower RaulGarciaCastro jtandy MattPerry roba ChrisLit

WARNING: Replacing previous Regrets list. (Old list: Rachel, Lars, SimonCox)
Use 'Regrets+ ... ' if you meant to add people without replacing the list,
such as: <dbooth> Regrets+ Andrea, Danh

Regrets: Andrea Danh
Agenda: https://www.w3.org/2015/spatial/wiki/Meetings:Telecon20160720
Found Date: 20 Jul 2016
Guessing minutes URL: http://www.w3.org/2016/07/20-sdw-minutes.html
People with action items: 

WARNING: No "Topic: ..." lines found!  
Resulting HTML may have an empty (invalid) <ol>...</ol>.

Explanation: "Topic: ..." lines are used to indicate the start of 
new discussion topics or agenda items, such as:
<dbooth> Topic: Review of Amy's report