Hi all,
I have almost finished to develop a web client for GeoServer based on MapBuilder.
The advantages of this client are:
- The filter forms and contexts are fully customizable because they are based on MapBuilder framework. This client is a kind of framework itself.
- The client prints out the features using the WMS service instead of the WFS. It does different requests to the WFS ONLY when the user wants the feature details. In other words first it builds the filters for the representation of the features usnig the WMS-SLD and then it builds the standard WFS filters to retreive the selected features. By this way the traffic is highly reduced because the client doesn’t need to load all the features.
My question is: how is the state of the work on the SLD filter improvements? Do you guys need some help? I would like to be involved in this topic.
Cheers,
Alessio.
Alessio Fabiani wrote:
Hi all,
I have almost finished to develop a web client for GeoServer based on MapBuilder.
The advantages of this client are:
1) The filter forms and contexts are fully customizable because they are based on MapBuilder framework. This client is a kind of framework itself.
2) The client prints out the features using the WMS service instead of the WFS. It does different requests to the WFS ONLY when the user wants the feature details. In other words first it builds the filters for the representation of the features usnig the WMS-SLD and then it builds the standard WFS filters to retreive the selected features. By this way the traffic is highly reduced because the client doesn't need to load all the features.
My question is: how is the state of the work on the SLD filter improvements? Do you guys need some help? I would like to be involved in this topic.
What Filter improvements are you thinking of?
I can think of four?
1. Some parser code for Filter 1.0 (and Filter 1.1) that Justin wants
2. Filter 1.1 API (for WFS 1.1 and Catalog)
3. Construction of Filter 1.0 and Filter 1.1 factories that subset the available range of GeoTools constructs...
4. Inclusion of SLD 1.1 ideas like enviornment variables
Cheers,
Alessio.
Quoting Alessio Fabiani <alessio.fabiani@anonymised.com>:
Hi all,
I have almost finished to develop a web client for GeoServer based on
MapBuilder.
The advantages of this client are:
1) The filter forms and contexts are fully customizable because they
are
based on MapBuilder framework. This client is a kind of framework
itself.
2) The client prints out the features using the WMS service instead
of the
WFS. It does different requests to the WFS ONLY when the user wants
the
feature details. In other words first it builds the filters for the
representation of the features usnig the WMS-SLD and then it builds
the
standard WFS filters to retreive the selected features. By this way
the
traffic is highly reduced because the client doesn't need to load all
the
features.
Not quite sure I completely understand this, though it sounds cool. Is
it still pure javascript? Are you planning on rolling the improvements
back into mapbuilder? Or is this something different? Did you write a
human text filter parser (like length < 5, translate to an ogc filter),
or do you have to write out the full xml filters? And/or are you making
use of some of the geoserver framework? I would really like a web
client framework, as one of our future plug-ins to geoserver, something
like the arcims authoring tools - I think that's the only thing we
really lack vis a vis arcims.
My question is: how is the state of the work on the SLD filter
improvements?
Do you guys need some help? I would like to be involved in this
topic.
Are you talking about the performance stuff? Passing the filters to the
backend datastore? I would very much like to see more, and remember
David commented on how it's not actually passing as much as possible
back. I wonder if there's ways to further optimize that? I think that
would be the best - I think Dave's concern was the if/else stuff, but
if the sld has just one rule, then it should be fully turned to the
backend filtering...
Chris
Cheers,
Alessio.
----------------------------------------------------------
This mail sent through IMP: https://webmail.limegroup.com/