[Geojson] Inclusion in OData

[Carl Smyth]

Arlo,

I think you get a usable result (on your schedule) with your outline here. The key idea is that you make a copy of the spec to which you want to make a normative reference (but can't) and make that copy normative. At any later time when a version of the spec exists that can support a normative reference, you replace the copy with the reference.

My outline had the extra step of wrapping the copy in a standard in a standards body (OGC) so it could be referenced but skipping that step is actually a better way to meet your needs.

...steve

From: geojson-bounces at lists.geojson.org [mailto:geojson-bounces at lists.geojson.org] On Behalf Of Arlo Belshee
Sent: Sunday, November 06, 2011 9:43 PM
To: Sean Gillies; geojson at lists.geojson.org
Subject: Re: [Geojson] Inclusion in OData

Makes sense.

So, we're going to publish OData V3 early next year (under OSP and/or OWF). In that spec, we are required to follow the same rules for normative references as the standards bodies do (due to lawyer action).

Since it sound like GeoJson may well go RFC, but not in that timeframe, then does an adaptation of Carl's approach make sense?

Something like:


1.       In the OData V3 spec, published under OSP and in final edits now, include:

a.       An informative reference to GeoJson.

b.      In a normative section, state exactly the subset of GeoJson that we want to use (an instantiation of the normative reference).

c.       In the examples, include GeoJson examples, copied from published GeoJson examples.

2.       When and if we take this to a standards body, they'll be producing the next version of OData. At that time, recommend that the body replace the GeoJson copy with a normative reference to the GeoJson RFC (assuming that it is released by then).

Does this approach work? It feels, to me, like OData is temporarily forking GeoJson, but we're not intending to update our fork. We're just freezing some version of GeoJson. Our intent is to, in the next release, merge back to GeoJson's main spec (once we can us a normative reference to it).

Does that meet the spirit of use that this community would like?

Arlo Belshee
Sr. Program Manager, OData, Microsoft

From: geojson-bounces at lists.geojson.org<mailto:geojson-bounces at lists.geojson.org> [mailto:geojson-bounces at lists.geojson.org]<mailto:[mailto:geojson-bounces at lists.geojson.org]> On Behalf Of Sean Gillies
Sent: Friday, November 04, 2011 8:51 AM
To: geojson at lists.geojson.org<mailto:geojson at lists.geojson.org>
Subject: Re: [Geojson] Inclusion in OData

Yeah, that was me and my mouth. Until now we've never heard from anybody that said GeoJSON had to be a "real standard" before they could use it, so I've made myself put RFC writing aside. Even now, I'd like to hold off until we finish revising (or not revising) the spec.
On Fri, Nov 4, 2011 at 2:55 AM, Martin Daly <Martin.Daly at cadcorp.com<mailto:Martin.Daly at cadcorp.com>> wrote:
There was some talk about submitting an IETF RFC, but it never happened. I believe that there would be more traction with the authors in that approach than the OGC one.

I see no reason that we couldn't start that process, either with 1.0 as-is, or with whatever 1.1 becomes.

Martin

From: geojson-bounces at lists.geojson.org<mailto:geojson-bounces at lists.geojson.org> [mailto:geojson-bounces at lists.geojson.org<mailto:geojson-bounces at lists.geojson.org>] On Behalf Of Arlo Belshee
Sent: 04 November 2011 01:19

To: geojson at lists.geojson.org<mailto:geojson at lists.geojson.org>
Subject: [Geojson] Inclusion in OData

As I mentioned a while back, we're using GeoJson to format the geospatial primitives in OData V3. A new wrinkle has arisen; I'm looking for input.

We are strongly considering taking OData to a standards body. If OData goes through standardization, then it will only be able to take normative references on other standardized formats. So we can't include GeoJson by reference, as it currently is defined.

I'm trying to figure out how to reference GeoJson. In the end, one of the most important considerations is the preference of this community.

I see the following options.


*         GeoJson goes through official standardization, and then OData references it.

*         OData copies in the parts of GeoJson that we use (the representations for the spatial types, but not the features or rest of the doc), then modifies as necessary.

o   We state this to be GeoJson.

o   We do not state this to be GeoJson, and have explicit permission from you all to do so.

Does anyone see any more? Does anyone have any preferences between options?

I've been happy with our choice of GeoJson so far. I appreciate any guidance you can give on how best to link to it from an official standard. Thanks.

Arlo Belshee
Sr. Program Manager, OData, Microsoft




--
Sean Gillies
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.geojson.org/pipermail/geojson-geojson.org/attachments/20111107/0c005644/attachment-0005.htm>