[Geojson] Request for comments (Re: Fixing problems in section 4)
sgillies at frii.com
Fri Mar 14 09:48:33 PDT 2008
Martin Daly wrote:
> I think that this is indeed an improvement. Is there any way to
> (slightly) modify it to use the "type" and "properties" sort-of model in
> other parts of the spec?
> All you implementers out there,
> Were this to become the CRS part of 1.0, would it adversely affect you?
> Or, in other words, any -1's?
Thanks, Martin. It even has a touch of theory in it :)
Rename "value" to "properties"? That doesn't feel right to me. I think
the larger issue (than changing "properties" to "value") is the impact
on early adopters of discarding the "EPSG", "OGC", and "URL" types in
draft version 5. We've been instructing GeoJSON processors to switch on
the value of "type", not on the value of "properties".
More information about the GeoJSON