[wfs-dev] WFS-Basic naming

[Rob Atkinson]

If you do not have a concept of a Feature Type, then its not a WFS,

here's why I think it matters:

* encoding isnt the issue - its whether you can
  1)  look at a feature definition, and
  2) predict how to generate a query and interpret the response 
(structurally)

The proposed solution might apply to to #1 (notwithstanding that you 
will have unlimited number of possible schema patterns to interpret, 
whereas a ISO feature type can be understood)
but it wont apply to #2

Perhaps you really have something more akin to a vocabulary service - 
give me objects _named_  using pattern X in which case

The overloading of WFS semantics with the special cases proposed may 
look temptingly simple, but in fact requires a whole raft of external 
conventions to be defined and maintained, and semantic deviations to be 
documented, that in practice its anything but simple.

An ATOM/GeoRSS discovery service specification might be useful, but 
better to constrain an external spec (if there are OGC governed 
semantics that matter)  that generalise an OGC spec (whose complexity 
and generality is already apparently a hurdle)




Simon.Cox at csiro.au wrote:
> Raj 
>
> - I fear this would undercut two of the key OGC messages 
> - i.e. that 
> (1) normative encodings of community data-models are a key to deep
> interoperability, and 
> (2) normative use of common components is a key to any inter-community
> interoperability. 
>
> I recognise the difficulties of heavyweight GML. 
> But we should not throw the baby out with the bathwater. 
> GML provides some really really useful things, particularly in encoding
> patterns (the Object-property rule; the by-rep or by-ref pattern). 
> And a standard encoding for point geometry including CRS references,
> which is being adopted in other communities now.  
> Without these, XML encoding is chaos. 
>
> Perhaps another approach would be an alternative UML-GML encoding rule,
> that allows mapping from UML to "flattened" (denormalised) GML. 
> i.e. something like UML -> GML-SF level 1
>
> Simon
>
> -----Original Message-----
> From: wfs-dev-bounces+simon.cox=csiro.au at opengeospatial.org
> [mailto:wfs-dev-bounces+simon.cox=csiro.au at opengeospatial.org] On Behalf
> Of Raj Singh
> Sent: Wednesday, 1 November 2006 4:09 AM
> To: wfs-dev at opengeospatial.org
> Subject: [wfs-dev] WFS-Basic naming
>
> There is an emerging data service specification targeted at mainstream,
> generic data platforms being developed at: http://
> www.ogcnetwork.net/wfsbasic.
>
> I wanted to query this group on whether anyone would 'offended' by a
> spec with WFS in the name that didn't mandate GML support. Comments?
>
> thanks,
> Raj
> _______________________________________________
> wfs-dev mailing list
> wfs-dev at opengeospatial.org
> https://mail.opengeospatial.org/mailman/listinfo/wfs-dev
>
> _______________________________________________
> wfs-dev mailing list
> wfs-dev at opengeospatial.org
> https://mail.opengeospatial.org/mailman/listinfo/wfs-dev
>