[Geojson] FW: Features without geometry

[Stephen Battey]

> Would this alleviate the concerns of the original poster, if 
> we said "must have a property 'geometry', which must be a 
> GeoJSON Geometry or null" ?  
Yes, that would certainly help.
Although semantically a null geometry means 'this feature has no
geometry' I think I can get away with using it to mean 'geometry not
specified'. The rationale being: if a client specifically asks not to
receive any geometry then the client should know to interpret
"geometry": null  as 'geometry not specified'.
I think that's a reasonable condition to ask of client applications, and
it means clients can use a GeoJSON parser to parse the JSON regardless
of whether they query for geometry or centroids, bounding boxes,
metadata, etc.


>From a purely aesthetic point of view, would it be better to use a
consistent syntax for empty/null geometry and properties?
We have empty geometry defined as ..
	"geometry": null
.. and empty properties defined as ..
	"properties": {}

Actually, now that I'm thinking out loud (well, on paper), maybe the
difference is a positive thing. They are, after all, two different types
of information.
Also, using 'null' clearly indicates "an empty geometry" whereas an
empty object might give the impression there is geometry but the data
("type" and "coordinates") was inadvertently omitted when the JSON was
constructed.

The more I think about it the more I like  "geometry": null
I think it is a useful addition to the spec.


Steve


> -----Original Message-----
> From: geojson-bounces at lists.geojson.org 
> [mailto:geojson-bounces at lists.geojson.org] On Behalf Of 
> Christopher Schmidt
> Sent: 02 October 2007 21:24
> To: Sean Gillies
> Cc: geojson at lists.geojson.org
> Subject: Re: [Geojson] FW: Features without geometry
> 
> 
> On Tue, Oct 02, 2007 at 01:45:57PM -0600, Sean Gillies wrote:
> > Thank you for the correction. I much prefer
> > 
> >    { "geometry": null }
> > 
> > over
> > 
> >    { "geometry": {"type": "Empty", ...} }
> 
> I can get down with that. My primary concern is that if it 
> looks like a duck, it has the same properties as a duck. 
> Since 'null' doesn't look like a duck, then it doesn't need 
> to have any particular properties.
> 
> Would this alleviate the concerns of the original poster, if 
> we said "must have a property 'geometry', which must be a 
> GeoJSON Geometry or null" ?  
> 
> Is there anyone against this? 
> 
> http://wiki.geojson.org/index.php?title=GeoJSON_draft_version_
4&diff=153&oldid=151

Regards,
-- 
Christopher Schmidt
MetaCarta
_______________________________________________
Geojson mailing list
Geojson at lists.geojson.org
http://lists.geojson.org/listinfo.cgi/geojson-geojson.org

Incoming e-mail scanned by Altman Technologies
Email has been scanned for viruses and spam by Altman Technologies