RE: UCR ISSUE-23: phrasing of the spatial operators requirement

Hi all,

First of all, is it very important that the requirements can be understood by everybody? This is very important for the Best Practice, but for the UCR document it’s most important that we ourselves can understand it.

But adding something about the purpose of spatial operators, as you suggest, does make the text clearer. So no harm in adding it and we can reuse this in the BP.

Linda

Van: Frans Knibbe [mailto:frans.knibbe@geodan.nl]
Verzonden: woensdag 7 oktober 2015 19:14
Aan: SDW WG Public List
Onderwerp: UCR ISSUE-23: phrasing of the spatial operators requirement

Hello,

This thread can be used for discussion of UCR issue 23<https://www.w3.org/2015/spatial/track/issues/23>: phrasing of the spatial operators requirement<http://w3c.github.io/sdw/UseCases/SDWUseCasesAndRequirements.html#SpatialOperators>. Currently the requirment reads: "There should be a recommended way for the definition and use of spatial operators. Spatial things can have spatial relations: topological relations, directional or distance relations. Operators based on these relations (e.g. 'Contains'. 'Intersects', 'Nearest') should be well defined and easy to use.". This issue is about clarity of the requirement, keeping in mind that people with different backgrounds should understand the requirement.

Questions that can be asked are:

1) Is it clear what is meant by 'operators'?  Is the difference between operators and relationships clear (compare GeoSPARQL fucntions and relationships)? For spatial relationships there is a separate requirement<http://w3c.github.io/sdw/UseCases/SDWUseCasesAndRequirements.html#SpatialRelationships>. Perhaps we should add something about the purpose of spatial operators (to filter, select or derive data)?

2) Should we specify that we understand operators to work on numerical (coordinate) data only? Could there be a need to support a query like 'give me all oases in the Sahara desert' when there is no way to compute the answer based on geometry (because the Sahara desert has no clear boundary)? Or put in another way: Would it be a problem if we leave the way operators should be implemented open, for phrasing the requirement?

Next to these questions, if you think there are other ways in which the requirement is unclear or could be improved, please share your thoughts!

Regards,
Frans