[Geojson] Closure?

Tim Schaub tschaub at openplans.org
Wed Sep 12 09:19:29 PDT 2007


I like the idea of finalizing this.

>    4.1.3. The use of a "crs" member for EPSG:4326 is discouraged since the
> default of not having a "crs" member would result in the same coordinate
> values. 

Does anybody feel strongly about keeping this?  We're already pretty 
weak in our handling of crs, and I think this point is a bit silly.  Are 
we hoping that EPSG:4326 will go away if we discourage its use?

>    (4.1.2 is new.)
>    4.1.2. To use a URL as a unique identifier to a coordinate reference 
>           system, the "crs" member should have the following structure:
>           "crs": {"type":"URL", "properties":{"url":
>           "http://example.com/mysrs/"}. When used as a value,
>           "URL" must be uppercased.         
> The specification does not offer any information on how
>                to convert this URL into a spatial reference system: its
>                use is inteded to provide users the ability to define
>                their references outside the EPSG namespace *only*.

I like the idea of proposing a convention for crs definitions via url, 
but I wonder if this should be a bit different.

At some point, someone may want to know if they are getting WKT or proj4 
text via a url.  The above recommendation doesn't really leave room for 

What about

"crs": {
     "type": "WKT",
     "properties": {
         "url": "http://example.com/mysrs/ogcwkt"

or (I don't really like this one)

"crs": {
     "type": "URL",
     "properties": {
         "wkt": "http://example.com/mysrs/ogcwkt"

If neither of those sounds good, I'm ok with including Chris' proposal 
and seeing how things evolve (no shame in having a version 2 in my mind).


More information about the GeoJSON mailing list