[OpenLayers-Users] OpenLayers WFS compliance

[Tim Schaub]

There is an ongoing effort to improve how vector layers behave.  The WFS 
layer is a prime candidate for better behavior.

When this effort is done, we will (if the current design holds) have a 
WFS protocol, a BBOX strategy, and a smarter Vector layer.

The BBOX strategy and the WFS protocol will largely do with the current 
WFS layer (and tile and format and feature maybe) stuff does now.  This 
protocol will also know how to deal with filters, for which we currently 
have limited support in the library.

If you care to follow or participate, see the vector-behavior sandbox.
http://trac.openlayers.org/log/sandbox/vector-behavior/lib/OpenLayers

Tim

Some detail on the idea:
http://trac.openlayers.org/wiki/Proposal/VectorBehavior

Also, for what it's worth, I've rewritten the GML parser again.  We're 
bringing over the WFS CITE tests (those that are appropriate).  In the 
end, we'll have WFS compliance (for the limited set of things we support).


Christopher Schmidt wrote:
> On Sat, May 17, 2008 at 09:53:24PM +0200, Arnd Wippermann wrote:
>> It's possible to add a filter with mergeNewParams.
>>
>> http://gis.ibbeck.de/ginfo/apps/OLExamples/OL26examples/wfs_filter_without_e
>> ncoding_26.html
>>
>> But in this case the bbox will not be used and the request returns a
>> response for features outside the view.
> 
> Accepting both of these parameters and not throwing an error is a
> violation of the spec. Additionally, there is no reason to do this,
> because it still reloads with the dragging o the map, but it doens't
> need to, because the data is always the same. Instead, one should just
> use a GML layer, since that doesn't refresh unneccesarily when you drag.
> 
> Regards,