[GeoJSON] spec clarification of CRS

Martin Daly Martin.Daly at cadcorp.com
Wed Dec 11 14:52:34 PST 2013


> To show one data point that confusion exists, I'll relate my own experience:
> ...
> On the other hand, what I'd really like is a similar specification and implementations that allow arbitrary CRSs
> without qualification so that I could implement the examples I described.

No spec as simple as GeoJSON (deliberately) is could hope to cover all possible use cases, exceptions, or edge conditions. So it is no surprise to me that something out of the ordinary is not easily, or well, supported. That, perhaps, is the price worth paying for simplicity.

Of course, if you want to supply data within a relatively small domain of use, then go ahead and use GeoJSON, but document how you do it, and how it differs from the core spec, and don't expect users outside of your domain to be able to use your data.

However, if you really want an extensible framework on which to base highly controlled, domain-specific data structures, then GML is what you are after. That is what you get with a several hundred page standard, not a two page one.

Martin




More information about the GeoJSON mailing list