Direction of travel for BP7 and BP14

Action 248<https://www.w3.org/2015/spatial/track/actions/248>

Hi- I'm tasked with updating BPs 7 and 14 (see Action 248) in line with
discussions at the f2f meeting in London.

To give a heads-up of the changes that need to be included, I'm sharing my
notes from the f2f discussion about BP7 here. If you look closely, updates
to BP14 are mentioned. See below.

Minutes of the meeting are here: https://www.w3.org/2016/12/15-sdw-minutes

Regards, Jeremy

---

*BP7: Use globally persistent HTTP URIs for spatial things*

   - [note that Bart wants to add stuff here too; e.g. R2RML]
   - @clemens sees no reason for forcing people to [re-]use an
   authoritative identifier assigned from a third-party domain
   - @bill: the use case to consider is when publishing _statistical
   information_ about a known spatial thing - use the persistent identifier of
   the target spatial thing …
      - [in hindsight, the subject of interest that @bill is publishing is
      a _statistical record_ which references the spatial thing]
   - @clemens suggests we turn things around - mint your own identifiers as
   a default approach, _and_ in specific cases (like @bill’s) re-use existing
   identifiers … the issue is resolving the URL to return your information
   - the key question here is “what’s the subject?” … in @bill’s case, it’s
   a census record, therefore we need the links to the well-known
   (authoritative) identifier for the topic of the census record; e.g. the RDF
   object of the triple(s) that comprise the census record
      - this is BP14 “Publish links from spatial things to related
      resources” - or more accurately, the “census record” may not be a spatial
      thing, but the topic of the record is …
      - *we need to clearly distinguish between BP7 (identifying) and BP14
      (linking) … in the introductory material to §13.4 indicate that the
      “re-use” statement is dealt with in BP14*
      - these updates should be sufficient to resolve ISSUE 440 …
      - [ACTION] @jeremy to close ISSUE 440 after updating the best
      practice text
   - note that it’s still useful for users to be able to search for census
   data for a particular spatial thing - this is a query API: “search for
   census data by location” or “statistics for _this_ place”
   - [ACTION] @jeremy to rewrite BP7 ‘possible approach’ using a simpler
   example than “operational catchment” … @linda has already provided a new
   example; see Example 12.
   - [ACTION] @jeremy to close ISSUE 358 - we covered enough of the concern
   about structuring URIs for authority and uniqueness in the preceding
   paragraph and we recognise that although it would be nice to include a
   spatial hierarchy in the URL structure, this may result in URLs not being
   durable - the URI would have to change if the “containment” changed.
   - leave ISSUE 208 for review