- From: Jeremy Tandy <jeremy.tandy@gmail.com>
- Date: Tue, 30 Aug 2016 13:00:51 +0000
- To: SDW WG Public List <public-sdw-wg@w3.org>
- Message-ID: <CADtUq_3r=bc-G2xTZmKyDssuM1wB_j-9H0DCw_DhwDM2jvUJFA@mail.gmail.com>
BP 16 "Provide a minimum set of information for your intended application" [1] is one of the few 'orphaned' best practices left in the document. ... In the BP sub-group call on 24-Aug (minutes [2]), we were discussing what to do with another best practice "Provide context required to interpret observation data values" [3] (now removed from the Editor's Draft following that discussion). We thought that that was too specific for the general best practice doc (it was purely about sensors) - but was an example of a more general concern about how to interpret "data instance level metadata" (as Rob Atkinson referred to it); how we find out what a piece of data actually means. We need to be careful not to stray beyond our self-imposed scope of "spatial". It's tempting to provide guidance on publishing data with well-described semantics for all properties, attributes and link relations - but these general concerns are covered in by DWBP's best practices 15 "Reuse vocabularies, preferably standard ones" [3] and 16 "Choose the right formalisation level" [4]. SDW best practice 16 "Provide a minimum set of information for your intended application" [1] seems to extend DWBP's best practice 16 (and probably best practice 15 too) by looking at the properties that are needed to differentiate between spatial things. This is dependent on the intended application ... As a bear minimum (e.g. for _all_ applications) we need a human-readable label (provided in at least one language that is familiar to the intended user community) and a Type declaration (e.g. what kind of spatial thing is this?) that can be interpreted using on-line documentation (published somewhere). However, I think that these elements are required (or at least strongly recommended) when you use schema.org to provide structured mark-up ... ref. SDW BP 4 "Make your data indexable by search engines". I propose that we merge the points about including Type and Label into BP 4. Concerns about spatial vagueness (e.g. definition of things without precise geometries), Req 5.43 "Spatial Vagueness" [6], could be covered off using a best practice about the use of spatial relations [7] within the section on Describing Location [8]. Thoughts please. Jeremy [1]: http://w3c.github.io/sdw/bp/#minimum [2]: http://www.w3.org/2016/08/24-sdwbp-minutes.html [3]: http://w3c.github.io/dwbp/bp.html#ReuseVocabularies [4]: http://w3c.github.io/dwbp/bp.html#ChooseRightFormalizationLevel [5]: http://w3c.github.io/sdw/bp/#indexable-by-search-engines [6]: https://www.w3.org/TR/sdw-ucr/#SpatialVagueness [7]: http://w3c.github.io/sdw/bp/#spatial-relations [8]: http://w3c.github.io/sdw/bp/#bp-expr-geo
Received on Tuesday, 30 August 2016 13:01:36 UTC