WFS GetCapabilities 1.0.0 not valid

[Steve Lime]

Nice frigging standard, real flexible. Tom, we worked around this by
creating our own
version of the WFS schema with the extra elements defined (at Peter's
suggestion). I
guess I don't have a better alternative at the moment. Easy enough to
supress those
lines and I have done so just now (will be fixed in beta 3).

You can still ask for GML 3 and related schema as before...

Steve

>>> "Kralidis,Tom [Burlington]" <Tom.Kralidis at EC.GC.CA> 9/5/2006
2:14:22 PM >>>
I would have to agree.  Though I love the new GML support, I think it
is
more important to validate against the OGC WFS 1.0.0 schemas.

Does anyone object to this (looking at the other OGC developers and
Steve who did the GML3L0 implementation :)) ?  If not, I will submit a
bug.

..Tom


> -----Original Message-----
> From: Bart van den Eijnden (OSGIS) [mailto:bartvde at xs4all.nl] 
> Sent: 05 September, 2006 3:12 PM
> To: Kralidis,Tom [Burlington]
> Cc: MAPSERVER-USERS at LISTS.UMN.EDU 
> Subject: Re: [UMN_MAPSERVER-USERS] WFS GetCapabilities 1.0.0 not
valid
> 
> Hi Tom,
> 
> I would be in favour of obeying the WFS 1.0.0 schema.
> 
> So leave the elements out, but support them " in the background".
> 
> Best regards,
> Bart
> 
> Kralidis,Tom [Burlington] schreef:
> >> using version 4.10 beta2, the WFS GetCapabilities (1.0.0
> >> version) response is not valid.
> >>
> >> Problems:
> >>
> >> 1) <SFE_XMLSCHEMA/> is not known in the 1.0.0 schema so should be

> >> omitted
> >>       <SchemaDescriptionLanguage>
> >>         <XMLSCHEMA/>
> >>         <SFE_XMLSCHEMA/>
> >>       </SchemaDescriptionLanguage>
> >>
> >> 2) GML3 in not allowed as result format
> >>
> >>       <ResultFormat>
> >>         <GML2/>
> >>         <GML3/>
> >>       </ResultFormat>
> >>     
> >
> > Because we added GML3L0 as a format, we added these formats for 
> > Capabilities to differentiate btw GML2 and GML3.  However, OGC:WFS

> > 1.0.0 is hardcoded.  The WFS 1.0.0 specs only define XMLSCHEMA as a

> > description language for DescribeFeatureType, as well as 
> only GML2 for 
> > GetFeature.
> >
> > So we're in a bit of a quandry, because we support WFS 
> 1.0.0, GML2 AND 
> > GML3L0.  But GML3L0 is only legal in WFS 1.0.0 (which uses 
> OWS Common 
> > 1.0.0, which abstracts constructs like outputformats [soft 
> typing] as 
> > opposed to defining hardcoded element names for same).
> >
> > If we want to be compliant to OGC:WFS 1.0.0, which must remove the

> > GML3 refs in Capabilities.  They can still be around in the 
> interface, 
> > just not advertise as such in Capabilities XML so as not to 
> break the schema.
> >
> > Comments?
> >
> > ..Tom
> >
> >
> >
> >
> >   
> 
> 
> --
> Bart van den Eijnden
> OSGIS, Open Source GIS
> http://www.osgis.nl 
> 
>