[GeoJSON] Greetings and schemas

[Simser, Bil]

Hi Erik,

Totally understand and I'm the last person to want JSON to even begin to resemble XML. I like simplicity. I was just looking for a good way to validate GeoJSON files (without going to an online service).

FWIW, http://geojsonlint.com/ is good for this and can be invoked with a POST.

Thanks.

-----Original Message-----
From: Erik Wilde [mailto:dret at berkeley.edu] 
Sent: Tuesday, May 27, 2014 9:24 AM
To: Simser, Bil; geojson at lists.geojson.org
Subject: Re: [GeoJSON] Greetings and schemas

hello bill.

On 2014-05-27, 7:24 , Simser, Bil wrote:
> First time on the list so be gentle. I’m currently building a .NET 
> library that converts KML to GeoJSON (basically a .NET version of a 
> JavaScript project but I needed it for client/server operations). As 
> I’m working through the project I found an initiative around creating 
> schemas for JSON (http://json-schema.org/) and thought it might be 
> useful to have a GeoJSON schema for validation (currently I’m just 
> validating through code checks in unit tests).

the issue with JSON schemas is that there is no stable language to write them in. there were and are a couple of proposals, but there isn't the one language do rule them all such as DTD or XSD for XML.

the reason is that parts of the JSON community don't like the idea of a schema language at all. there is a lingering perma-thread on the IETF JSON list, and because of the heated debates, i think nobody ever had the energy to really follow through with a proposal and define and standardize a schema language.

that doesn't mean that there aren't scenarios where schemas might be useful. just don't be surprised to get passionate reactions when you bring up the idea of schemas and a schema language in JSON-land.

cheers,

dret.

--
erik wilde | mailto:dret at berkeley.edu  -  tel:+1-510-2061079 |
            | UC Berkeley  -  School of Information (ISchool) |
            | http://dret.net/netdret http://twitter.com/dret |