Spatial Data on the Web Working Group Teleconference

<phila_> trackbot, start meeting

<trackbot> Meeting: Spatial Data on the Web BP Sub Group Teleconference

<trackbot> Date: 15 February 2017

<scribe> scribe:billroberts

<phila> scribe: Bill

<AndreaPerego> chair: jtandy

<phila> scribeNick: billroberts

Preliminaries

<jtandy> minutes of last meeting: https://www.w3.org/2017/02/01-sdwbp-minutes

<Linda_> +1

PROPOSED: approve minutes of last meeting

<jtandy> +1

<ClemensPortele> +1

<LarsG> +0 (wasn't there)

0 - wasn't there

<AndreaPerego> +1

<joshlieberman> +1

<byroncinnz> +0 wasn't there

RESOLUTION: Previous minutes approved

Patent call: no issues raised

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

backlog and sprint plan

<jtandy> sprint plan https://www.w3.org/2015/spatial/wiki/Detailed_planning_BP_document

Jeremy and Linda have prioritised work items for the next sprint: Feb to Mid March

scribe: ready to release a new iteration at the Delft face to face meeting

jtandy: has added some items into the Mid March - end April sprint as well - mainly editorial issues. This sprint is the last opportunity for substantive change
... Coordinate Reference Systems work for the current sprint - Byron is working on it

byroncinnz: I've put in a pull request a couple of days ago

jtandy: thanks! will review and include it in this sprint

byroncinnz: have kept the two BPs separate and tried to make them more distinct. Have cleaned up language around accuracy and precision
... a lot of the issues around datum rather than CRS.
... and took out some of the more general stuff on CRS from BP3 but left it in BP17.
... I suggest moving BP3 later in the document as less important in general than BP17

jtandy: we want to move BP17 into the 'body' of the document. It's currently sitting in the 'Other' section
... and we can recommend WGS84 for most simple/default cases?

byroncinnz: yes

jtandy: other blocks of tasks in the sprint plan: Dataset metadata (BP1)

<phila> Draft new WG

phila: following on from previous DCAT discussions, now preparing a new working group which will probably involve updates to DCAT. The new working group is on content negotiation by profile

jtandy: notes that we shouldn't tie ourselves to a specific version of DCAT because new things are coming

<Zakim> AndreaPerego, you wanted to mention some work planned in OGC about GeoDCAT-AP

joshlieberman: there is work on DCAT in OGC Testbed-12 and will be in Testbed-13. So the recommendation on DCAT should probably say 'keep an eye on it'

AndreaPerego: notes the OGC work on DCAT. Also there was a meeting at JRC on metadata and DCAT and possible collaboration with W3C

phila: let's discuss this in Delft to get input into next working group plans. Also note the intention to continue collaborating between OGC and W3C

jtandy: we need to put the essence of these DCAT discussions into BP1

<AndreaPerego> I can contribute.

jtandy: volunteers to do that for us?
... thanks Andrea. But I know you have work to do on BP8. Can anyone take a lead on BP1, to work with Andrea?

joshlieberman: I can work on BP1 with Andrea

AndreaPerego: is there a risk that the new OGC working group and W3C group will overlap and possibly coming up with competing solutions? To avoid that we should make sure we coordinate

phila: to avoid that risk, we should have a liaison

joshlieberman: it needs to be a collaboration between the metadata and geosemantics groups of OGC and we can finalise that in Delft

<joshlieberman> Yes, as of last night...

<phila> ACTION: phila to talk to OGC, Andrea etc. about liaison wrt Data Exchange WG [recorded in http://www.w3.org/2017/02/15-sdwbp-minutes.html#action01]

<trackbot> Created ACTION-267 - Talk to ogc, andrea etc. about liaison wrt data exchange wg [on Phil Archer - due 2017-02-22].

jtandy: next section is Spatial Data vocabularies and file formats
... Andrea is working on BP8 on different representations of geometries

AndreaPerego: I'm preparing examples to go in BP8. And trying to revise BP8 text to provide more precise guidance. I have a structure based on usage patterns (see mailing list)

<AndreaPerego> http://lists.w3.org/Archives/Public/public-sdw-wg/2017Feb/0376.html

AndreaPerego: waiting for feedback on that
... it might be too simple, but it's a starting point. What are the main approaches we see in practice?
... 3 options listed in the mail linked above
... Option 1 is probably best for web developers. (geometries as literals in the data) - as long as the geometry is not too big
... Option 2 (use HTTP URI for the geometry) is fine for LD people, maybe less so for web developers - allows geometry to be managed separately from other data, so more flexible
... Option 3 is often used by LD people but may be complicated in practice. Geometry is less re-usable
... will review the pros and cons of each of those options

jtandy: please provide feedback on Andrea's idea via the mailing list

<joshlieberman_> Option 2 is best for multiple geometry options, for a choice of geometry roles or scales.

ClemensPortele: need to clarify the context. eg GeoJSON often falls in category 3.
... and note that we agreed we would not only take an RDF perspective. Will provide more detail via mailing list

joshlieberman_: there is a distinction between having a geometry as a literal and having coordinate positions as a literal
... often people represent the position as a literal but provide other information about the geometry in a non-literal way
... so we should make a distinction between geometry and coordinate positions

jtandy: BP10 is another high priority activity
... choice of vocabulary for describing spatial data
... We should provide advice people on how to make a choice of vocab

<joshlieberman_> could we use vocabulary, encoding, serialization? file format is a blunt instrument.

jtandy: this could be a big piece of work
... Bill, can you take ownership of BP10?

yes

jtandy: note that Bill will need assistance

<AndreaPerego> I can help - considering the relationship with BP8

jtandy: BP9
... Josh had an action to talk to Christine about the augmented reality community
... we should start a discussion on whether to merge BP9 and BP10

joshlieberman_: relative positioning is often about a reference system, so a bit different to spatial relationships

jtandy: so maybe closer to the CRS work that Byron has been doing?

joshlieberman_: it's about describing a position relative to another object or to a human perspective

jtandy: I know you are working on BP1 too, but could I ask you to lead BP9 as well?

joshlieberman_: yes

jtandy: I'll start a discussion thread for whether we need a 'samePlaceAs' property
... BP14 depends on vocabulary choices. Depends to some extent on BP10. Lars said he'd check with German colleagues about the Beacon format

LarsG: yes, it looks like it should work fine

jtandy: I'll take the lead on BP14
... BP16 is vague and partly covered by DWBP so I suggest deleting and moving anything useful to BP10
... any objections to that?

<Zakim> AndreaPerego, you wanted to say that this is partially included also in BP8 - geometry complexity

AndreaPerego: I see a relation between BP16 and BP8
... so good idea to delete

jtandy: Section 11 on how to use the BPs will be easier once the individual BPs are more advanced
... next block is Spatial data access and APIs
...BP11: 'convenience APIs'
... since we started the WG, a lot of what we wanted to say has been added to DWBP
... so we need to think about how we tie things back to DWBP and what is special about spatial data with respect to APIs
... Can anyone on the call take the lead on moving BP11 along?

ClemensPortele volunteers to work on BP11

jtandy: Bart's work is relevant to BP11 - coordinate with Clemens on this

AndreaPerego: I can add another example for BP11 around the CSW API
... many of our examples are more about Spatial Things than geometries

joshlieberman: OGC APIs serve Features rather than Geometries
... there is an aspiration to extend that but not yet an established practice

jtandy: can we add a placeholder for this before the Delft meeting (as no substantive changes after that)
... minor work to do on BP12 and BP13. Could move those to next sprint if necessary

<joshlieberman> That said, OGC services are used to reference features with geometries, that are then linked to by objects representing other feature properties (e.g. demographic info linked to population units).

jtandy: properties that change over time. BP6 nearly complete. Linda asking Geonovum colleagues for examples
... Editorial questions. BP2 - long discussion on mailing list about units of measure
... suggest removing this BP unless we can identify something specifically spatial

AndreaPerego: there is an example of units of measures in the examples, copied from DWBP, about spatial resolution
... example 15 is taken from Data Quality Vocabulary
... and we need to make sure we don't provide conflicting recommendations

jtandy: will start new discussion thread around deleting BP2
... we need to review public comments

<Zakim> phila, you wanted to ask for clarity on a couple of points in the new publiction

<phila> https://www.w3.org/TR/2017/NOTE-sdw-bp-20170216/#sotd

phila: I need help on the Status of this Document section.

jtandy: item 3 still pending. Item 4 will be done this time. Still want feedback on BP7

<joshlieberman> bye

jtandy: thanks to folks who agreed to own secions of the document

<AndreaPerego> Thanks, and bye!

thanks and bye

<ClemensPortele> thanks, bye

<BartvanLeeuwen> thx