[ka-Map-dev] future of ka-Map

[Lorenzo Becchi]

Paul, you're comments arrive just in time.
I've planned to spend my participation at the OSGeo Hacking Event [1] 
over ka-Map code and OL integration.


Paul Spencer wrote:

> 
> * a php-based configuration file similar to the existing config.php for 
> specifying various options and identifying maps and tile caches

if I'll be alone at the the Hacking Event I will not touch config.php

> * a php-based tile engine similar to the existing tile.php for 
> dynamically serving tiles from a cache and rendering them as necessary

what changes would you like to do to tile.php?

> * a php-based tool that returns javascript or json to the client with an 
> appropriately configured layer(s) (all the settings for the layer are 
> taken from the configuration of the cache and other settings)


I was thinking to use OWS Context with an extension for the ka-Map 
layer. Another idea is the OLON object but has never come to be included 
in OL trunk. Both XML and JSON output can be easily offered.

> * a client-side way of using the javascript or json from the server-side 
> - could be as simple as including a script tag after OpenLayers.js or an 
> entirely new Control or add-in for OpenLayers.

I still don't know what's the direction I want to take but I've thought 
to both solution.
The Control could help at least to manage the multiple-maps 
configuration. Then the Legend should be adapted in a separate Control 
to let the people choose which layerSwitcher to use. I've already wrote 
a new legend control for OL and it should not take too much time to 
clean it and adapt it to ka-map.
I was even thinking to offer the integration in ka-Explorer, depending 
on the time I'll have available.
If we'll refine the architectural part together before the Hacking Event 
I'll have more time to code.

> 
> This would focus efforts on improving tile.php rendering, tile 
> management tools such as precache.php, and cleaning up that code while 
> leaving the javascript stuff to OpenLayers.
> 

I do agree.

ciao
Lorenzo

[1] http://wiki.osgeo.org/wiki/OSGeo_Hacking_event