[CITE-Forum] [wfs-dev] WFS 1.1.0 CITE tests:multiplegeometrytypes ona single layer

Satish Sankaran ssankaran at esri.com
Tue Apr 29 11:49:00 EDT 2008


The OAB during its last meeting took some decisions regarding this
issue. At a high level, one of the decisions was to allow the tests to
accommodate testing wfs implementations that work against a storage
model that supports only a single geometry. This would be something that
the OAB would recommend to CITE.

As Arctur suggests in his e-mail, the OAB would be recommending CITE to
create wfs compliance tests that will test for different levels
(profiles) of compliancy - testing for filter support, different
payloads, no geometries, xlink support etc.

Carl has the minutes for the OAB meeting and would probably address this
with a more official statement.

In summary, the recommendation (from OAB) for immediate action was to
create a wfs test that would work against storage models that support
just single geometries. The resulting compliancy certificate would
clearly mention the "compliancy level" of such a test.


On a related note about geometryless features: In the current wfs
capabilities document, the bounding box is mandatory but this prevents
feature types without geometry from being served by a WFS. This will be
addressed (making bounding box optional) in the next version of the wfs
spec.

Thanks.

Satish



-----Original Message-----
From: cite-forum-bounces+ssankaran=esri.com at lists.opengeospatial.org
[mailto:cite-forum-bounces+ssankaran=esri.com at lists.opengeospatial.org]
On Behalf Of David Arctur
Sent: Sunday, April 27, 2008 8:06 PM
To: 'Raj Singh'; 'Andrea Aime'
Cc: Tom.Kralidis at ec.gc.ca; wfs-dev at lists.opengeospatial.org;
Ian.Painter at snowflakesoftware.com; Simon.Cox at csiro.au;
cite-forum at lists.opengeospatial.org; assefa at dmsolutions.ca
Subject: Re: [CITE-Forum] [wfs-dev] WFS 1.1.0 CITE
tests:multiplegeometrytypes ona single layer

Well, but I think what started this was that the current tests require
servers to support two geometry properties on a feature, do I remember
that
right? That seems unnecessarily onerous to require of all
implementations.
I'd like to ask again: could we modify the tests so that (1) compliance
does
not require supporting multiple geometry properties (zero or one
geometry
property should be required); and (2) for implementations that do
support
multiple geometry properties, that represents a higher level of
compliance
or capability..? Can we have this level of nuance in our compliance test
for
WFS with SF-GML? 

dka

-----Original Message-----
From: Raj Singh [mailto:rsingh at opengeospatial.org] 
Sent: Sunday, April 27, 2008 1:52 PM
To: Andrea Aime
Cc: Simon.Cox at csiro.au; Ian.Painter at snowflakesoftware.com;
wfs-dev at lists.opengeospatial.org; Tom.Kralidis at ec.gc.ca;
darctur at ogcii.org;
cite-forum at lists.opengeospatial.org; assefa at dmsolutions.ca
Subject: Re: [wfs-dev] [CITE-Forum] WFS 1.1.0 CITE tests:
multiplegeometrytypes ona single layer

OK. Simon and Andrea are convincing me of the need for "geometryless"  
features. End of discussion!
---
Raj


On Apr 27, 2008, at 4:41 AM, Andrea Aime wrote:
> Simon.Cox at csiro.au ha scritto:
>> Raj Singh wrote:
>>> (a document with 0 geometric properties should also pass, but I  
>>> think that is "a bad thing")
>> Why is it a bad thing? And would you count "time" as a geometry?
>> The OGC/ISO General Feature Model certainly does not require that  
>> every
>> feature must have a geometric property. We should be careful about  
>> introducing constraints that makes SF deviate
>> strongly from the expectations of the general model.
>
> Chiming in to add a user perspective. At GeoServer we try to keep
> "geometryless" features working, and we have users definitely using
> WFS for that kind of data. Proof is that every time we fiddled with
> reprojections and the like and introduced a code path expecting
> a geometry to be there, we promptly had users complain about
> their geometryless data not working anymore.
>
> So despite the WFS being targeted at data with a geometric component,
> there seem to be significant usage of it with geometryless data
> as well. Couple that with the consolidation of javascript based UI  
> and a
> WFS that can generate (geo)JSON and you'll see geometryless data
> being used more and more.
>
> Cheers
> Andrea

_______________________________________________
CITE-Forum mailing list
CITE-Forum at lists.opengeospatial.org
https://lists.opengeospatial.org/mailman/listinfo/cite-forum




More information about the CITE-Forum mailing list