[georss] [Opensearch-discuss] Spatio-temporal query extension

[Ron Lake]

It is easy to make a WFS discoverable by Google Search.


R

 

________________________________

From: georss-bounces at lists.eogeo.org
[mailto:georss-bounces at lists.eogeo.org] On Behalf Of Stefan F. Keller
Sent: March 15, 2007 1:40 PM
To: georss at lists.eogeo.org
Subject: Re: [georss] [Opensearch-discuss] Spatio-temporal query
extension

 

Sorry, got trapped by reply-to user instead reply-to georss... 

2007/3/15, Pat Cappelaere <pat at cappelaere.com>:

	Stephan,
	
	I do not follow your train of thoughts at all.

 

Sorry, my fault. That's why I asked about the use case before.


 

	The intent here is to provide an improved and more widely
accepted interface to WFS-Simple using OpenSearch (which is already
integrated in the browser and can be discovered).  We all agreed that
WFS-Simple does not have much in terms of filtering capabilities
(spatio-temporal-keyword search). 

 

WFS Simple has bbox, time capabilities and we actually proposed regular
expressions for content filtering. Raj let regex stay there explicitely
for discussion. It even has a defined exception handling.

 

The main thing WFS Simple 'lacks' compared to OpenSearch is an explicit
relation to autodiscovery. This could be where OpenSearch, Google
Sitemaps or KML come in (Google was'nt part of the OpenSearch spec. team
but now it seems that Google GData API supports OpenSearch responses). 


 

	WFS-Simple targets the mass-market GIS users and outputs
Atom/GeoRSS format (among others).
	 

	Using an OpenSearch interface would allow distributed queries
across GIS and non GIS repositories.
	The RESTful interface is great.
	Passing a bbox attribute is becoming widely accepted. The
expected response is Atom (or Atom/GeoRSS for GIS Servers) and not XML
in our case.
	
	http://www.loc.gov/standards/sru/  does not seem to have
spatio-temporal query capability that I could find.  CQL seems
interesting though but might be complex for developers to implement.
Query languages become problematic very quickly. 
	
	Pat.

 

With ISO 23950/SRU you can make spatio-temporal distributed online
queries. See e.g. http://www.search.gov/geospatial/  which seems to be
also a attempt comining GIS and non GIS repositories. But I agree that
implementing a query interpreter can add complexity. Thats why we
suggested regex in WFS Simple. 

 

Now without knowing your use case better I can't judge which path to
follow. Just my few cents. -- Stefan 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.eogeo.org/pipermail/georss/attachments/20070315/4747a6ab/attachment.html