search

geojson / geojson-ld

GeoJSON in JSON-LD contexts
Other
130 stars 13 forks source link

Support for MultiInstant and maybe MultiInterval #15

Closed jyutzler closed 10 years ago

jyutzler commented 10 years ago

Background: @sgillies stated that "moving features and feature succession/events should be different efforts." In response, I set up this issue to discuss the concept directly.

As I get more familiar with JSON-LD itself, I think it becomes easier, not harder, to support the gamut of temporal types including MultiInstant and MultiInterval. For example, the following works as a MultiInstant:

{
   "@context": {
    "Feature": "http://example.com/vocab#Feature",
    "Point": "http://example.com/vocab#Point",
    "coordinates": "http://example.com/vocab#coordinates",
    "geometry": "http://example.com/vocab#geometry",
    "id": "http://example.com/vocab#id",
    "datetime": "http://www.w3.org/2001/XMLSchema#dateTime",
    "MultiInstant": {
        "@id": "http://example.com/vocab#MultiInstant",
        "@container":"@list",
        "@type":"http://www.w3.org/2001/XMLSchema#dateTime"
    },
    "properties": "http://example.com/vocab#properties",
    "type": "http://example.com/vocab#type"
  },
  "@type": ["Feature", "MultiInstant"],
  "geometry": {
    "type": "Point",
    "coordinates": [
      0.0,
      0.0
    ]
  },
  "MultiInstant": ["2014-04-24", "2014-04-25", "2014-04-26", "2014-04-27"],
  "type": "Feature",
  "properties": {}
}

This parses correctly in the JSON-LD Playground. A similar approach could be used for MultiInterval. What I can't evaluate is the cost of support for these elements compared to the cost of earliest, latest, start, and stop. It seems to me that the thing for a client to do is evaluate the @type entries - if it supports the type then great; if not it is simply ignored. It looks like an easy way to do core/extensions.

This is what I'm looking for personally because I don't care about Interval but I do care about MultiInstant.

sgillies commented 10 years ago

I disagree that there exists a gamut of temporal types. Is there anything to Multi-instants and multi-intervals other than analogy to geometry types? I don't see them in other temporal ontologies (though I confess I haven't made an exhaustive search). Let's leave them out of an event-like features context.