[OSGeo-Standards] The OGC requests comments on the OWS Context Conceptual Model and ATOM Extension candidate standards

Wed Feb 20 15:07:44 PST 2013

We are getting much closer Raj. I read over your response a few time make sure I understand it. It would help if you could provide an example.

I was aiming to pull the information up to the <ows:entry> level, and yes I am starting from the base assumption that a client has to perform version negotiation and issue a new GetMap request to match their abilities (supported protocol) and environment (screen size and DPI).

I will think about it (entry or operation). It may be advantageous to list multiple <ows:operations> incase the information was available from several sources (an inline geotiff overview, a WMTS and a WMS for example).

Although my feedback was harsh, I am really keen on several improvements that have been made.

Having access to the GetCapabilities endpoint fills in a major gap. It allows me as a client to go to the original server and see what is available - perform version negotiation if required and determine what abilities of the server I can make use of as a client.  Previously this was an implicit assumption that a GetCapabilities would be available at the same end-point as the GetMap operation.

Idea: Only thing that could be better here is to allows (a) GetCapabilities to respond to a request for the capabilities of a single layer, and (b) inline that cut down Capabilities document. This would allow the context document to (in the best case) communicate the information required to issue a GetMap request.

Still explicitly listing the GetCapabilities endpoint is the big win here.

The final thing I need to figure out is how vendor extensions work, so that a reference to a local shapefile can be made. Still the main use case for sharing a map between applications.  By going with a specific GML entry, rather than feature, I do not have the option of just changing mime type and referencing a different file.

--  
Jody Garnett

On Thursday, 21 February 2013 at 2:33 AM, Raj Singh wrote:

> Ah I think I get it now. So you're saying that now all the parameters of the GetMap request have to be pulled out of the href property of <ows:operation> instead of each parameter being separate properties of <ows:operation>. Is that correct?  
>  
> If so, then that's a decision we went around and around on for a long time. The only concrete reason to do it this way (that I can remember) was that even a client app that wasn't very smart about Geo "stuff" could still have a good chance of reproducing the original map because all they have to do is resolve the hrefs and overlay them on top of each other. In a more parameterized format, the client would need to know how to put together a WMS request to get anything.  
>  
> However, I don't think anyone was completely against the old way of doing things, and getting feedback on issues like this is what public comment periods are for. Anyone else have a preference/comment?
>  
> ---
> Raj
> The OGC: Making location count.
> http://www.opengeospatial.org/ogc/organization/staff/rsingh
>  
>  
> On Feb 19, at 7:20 PM, Jody Garnett <jody.garnett at gmail.com (mailto:jody.garnett at gmail.com)> wrote:
>  
> > So what we have on the table is: to "read back" a WMS layer I need to gather this information from either a) a WMS GET Request b) a WMS PUT Request or c) a WMS PUT Request with optional inlined Image
> >  
> > The previous copies of OWS Context offered this information explicitly (Name, Title, SRS element, etc...), allowing me to get buy with just an XML parser (or in the case of javascript clients attacking a DOM).
> >  
> > Now I notice the current standard has some of this information pulled out (as noted in the previous email). I could hijack the standard and use the <id> element to hold the layer name, and go through the remaining standards and perform a similar "mapping".
> >  
> > What we need to facilitate discussion is a working example:
> >  
> > <feed xmlns=“http://www.w3.org/2005/Atom”
> > xmlns:owc=“http://www.opengis.net/owc/1.0”
> > xml:lang=“en”>
> > ...
> > <entry>
> > <id>http://someserver.com/#base</id>
> > <title>Base World Map</title>
> > ....
> > <owc:offering code=“http://www.opengis.net/spec/owc/1.0/req/atom/wms”>
> > <owc:operation
> > method=“GET”
> > code=“GetCapabilities”
> > href=“http://www.someserver.com/wrs.cgi?
> > REQUEST=GetCapabilities&
> > SERVICE=WMS &
> > VERSION=1.1.1”/>
> > <ows:operation
> > method="GET"
> > code="GetMap"
> > href="http://www.someserver.com/wrs.cgi?
> > REQUEST=GetMap&
> > SERVICE=WMS&
> > VERSION=1.1.1&
> > layers=topp%3Aabse&
> > styles=&
> > srs=EPSG%3A4326&
> > bbox=-145.15104058007,21.731919794922,
> > -57.154894212888,58.961058642578&
> > width=780&
> > height=330&
> > format=image%2Fpng"/>
> > </owc:offering>
> > <entry>
> >  
> > --  
> > Jody Garnett
> >  
> > On Wednesday, 20 February 2013 at 9:01 AM, Raj Singh wrote:
> >  
> > > You have to explain your point better for me. I understand that there are two distinct tasks here -- being a WMS client (an encoder) and being able to parse any WMS request into its parameters (a parser). And I understand that most clients are not also parsers. But if you want the additional capability of reading in the state of someone else's WMS map, how do you avoid needing to build a parser? Doesn't the additional code needed match up with the desired functionality? I thought we have asked for exactly the right amount of work ;)
> > >  
> > > ---
> > > Raj
> > > The OGC: Making location count.
> > > http://www.opengeospatial.org/ogc/organization/staff/rsingh
> > >  
> > >  
> > > On Feb 18, at 8:59 PM, Jody Garnett <jody.garnett at gmail.com (mailto:jody.garnett at gmail.com)> wrote:
> > >  
> > > > So back to my origional feedback to Jim.
> > > >  
> > > > The conceptual model makes great re-use of existing standards, you can watch it call out to the other fish in the pond and it very carefully does not repeat anything making it easier to write as a standard.
> > > >  
> > > > If you swap perspective to that of an implementor you are left with … a gap of understanding.
> > > >  
> > > > The document produced by the ATOM Encoding requires client authors to implement both an encoder (so they can make request) and a parser so that they can understand the information that has been cleverly reused by the standards body.
> > > >  
> > > > This standard is lazy on the part of the standards body, and hostile to client authors.
> > > >  
> > > > As an example: Just because GeoTools has the ability to encode WMS requests, DOES NOT imply there is a ready-at-hand WMS request parser available.
> > > >  
> > > > As such the payload delivered by ATOM to describe a WMS request may as well be on the moon. Perhaps a bit further out of the payload is the content of a PUT request, or needs to call out for more schema information as with GML and WFS content.
> > > >  
> > > > So kudos on the reuse of definitions, missed the target audience when considering implementors (you have asked for too much work).
> > > > --
> > > > Jody Garnett
> > > >  
> > > > On Tuesday, 19 February 2013 at 10:41 AM, Jody Garnett wrote:
> > > >  
> > > > > Last time I looked at this one with Jim … I hated it (even though he was keen on how ATOM had been used).
> > > > > For me It was not providing enough context for a client to make its own request.
> > > > >  
> > > > > I would really like to see a context document take shape, it would be very valuable between systems (which we tend to like on these osgeo lists).
> > > > >  
> > > > > Tracing through your links I arrive at some word docs, don't suppose you can format shift those to PDF or something more portable.
> > > > >  
> > > > > --
> > > > > Jody Garnett
> > > > >  
> > > > > On Tuesday, 19 February 2013 at 3:39 AM, Carl Reed wrote:
> > > > >  
> > > > > > I am not sure if this group saw this announcement. This is a preliminary comment period. There will be another comment period later in the process.
> > > > > > The announcement is here:
> > > > > > http://www.opengeospatial.org/node/1765
> > > > > > Please ignore the Comment Due date. The OGC OWS Context Standards Working Group is happy to receive comments on this candidate standard at any time.
> > > > > > Regards
> > > > > > Carl Reed, PhD
> > > > > > CTO and Executive Director Standards Program
> > > > > > Open Geospatial Consortium
> > > > > > www.opengeospatial.org (http://www.opengeospatial.org)
> > > > > >  
> > > > > > The OGC: Making Location Count!
> > > > > >  
> > > > > > ---------------------
> > > > > >  
> > > > > > This communication, including attachments, is for the exclusive use of addressee and may contain proprietary, confidential or privileged information. If you are not the intended recipient, any use, copying, disclosure, dissemination or distribution is strictly prohibited. If you are not the intended recipient, please notify the sender immediately by return email and delete this communication and destroy all copies.
> > > > > >  
> > > > > > "The important thing is not to stop questioning." -- Albert Einstein
> > > > > > "Security is mostly a superstition. It does not exist in nature. Life is either a daring adventure or nothing." -- Helen Keller
> > > > > > _______________________________________________
> > > > > > Standards mailing list
> > > > > > Standards at lists.osgeo.org (mailto:Standards at lists.osgeo.org)
> > > > > > http://lists.osgeo.org/mailman/listinfo/standards
> > > > > >  
> > > > >  
> > > > >  
> > > >  
> > > >  
> > > > _______________________________________________
> > > > Standards mailing list
> > > > Standards at lists.osgeo.org (mailto:Standards at lists.osgeo.org)
> > > > http://lists.osgeo.org/mailman/listinfo/standards
> > > >  
> > >  
> > >  
> >  
> >  
>  
>  
>  

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/standards/attachments/20130221/1144cc69/attachment.html>