[Geomoose-users] using WFS-T

[Brent Fraser]

Hi All,

   Now that I've configured my Geomoose demo to do WFS-T, I need to 
think about how to use it in my old Geomoose v2.4  Ice Digitizing 
application.

   That application grouped digitized polygons by "image_id" 
(representing the Landsat image used as a backdrop).  There are hundreds 
of images and thousands of polygons in the system and when the user 
selected an image, Geomoose passed the image_id to mapserver to do a 
substitution in the mapfile SQL so only those polygons related to that 
image were displayed (and available for edits etc).  That worked great.

How do I do the same thing with Geomoose v2.8 for my app?  Some possible 
"solutions" (I like #3):

1.  Convert the old 2.4 vector editing system to a user extension and 
use it instead of WFS-T
         - a lot of work, uses deprecated code, and no benefit to the 
Geomoose community
or

2. Create a layer per image_id in the TingyOWS config.xml file and have 
Geomoose switch layer name when the user selects the image.
     - as images get added to the system the config.xml file must be 
updated (could be automated, but clunky)
     - each layer would point to a table in the database which would 
make the database ugly
(hundreds of tables).  I might be able to use views instead of tables to 
make it slightly less ugly, but TinyOWS might not do inserts into a 
view.  Dunno about that.
     - could be some limitations on the number of layers in the 
config.xml file (or performance issues?).

or

3.  Add functionality to Geomoose to accept attribute filters in the 
mapbook and pass the filters to the OpenLayers definition.
     - adds complexity to the mapbook syntax (so more documentation, 
testing, maint., etc).
     - the logic may be similar to the  mapbook's search/query mechanism 
of specifying input types of layer,template,fieldname,comparitor,value 
so maybe some of the code could be used.

or

4.  Other?

Thanks for any input!

-- 
Best Regards,
Brent Fraser