[webmap-discuss] ka-Map and OpenLayers - WMS-C

[Steven M. Ottens]

Cameron Shorter wrote:
> Paul Spencer wrote:
>> The third discussion that also needs to happen on webmap-discuss is 
>> the issue of the proposed WMS-C standard proposal.  I am
>> particularly  interested in this as I can see the tile caching
>> component of ka-Map  forming the basis of a reference implementation
>> of the proposed WMS-C  and living as either a separate project or an
>> optional sub-project of  OpenLayers/ka-Map.
>
> Steven would be best to talk to this as he is implimenting tiling
> within Mapbuilder, but I'm keen to see Mapbuilder's solution to be the
> same as ka-map/OpenLayers and I think that is Steven's approach too.
> If we can use the same code (in the form of a joint library) then that
> would be good too.
Right now I've designed to fit tiling as much in mapbuilder as possible.
This means that after the user finishes panning/zooming the tiles are
requested. I would prefer the more dynamic grid approach done by
OpenLayers, but it requires a rewrite of the event model, which probably
has to happen at some point but now is not the right time I believe.
(the winter with lots of dark nights is much more suited)
Mapbuilder calculates which tiles it requires and request them all as a
standard WMS request. The server has no knowledge of tiling, but does
have caching enabled.
I'm very much aware that there are some issues on WMS-C that have to be
solved server side and as such we need a WMS-C server or preprocessor
(like tile.php) But I'm in favor of a WMS-degradable one if possible,
but I'm willing to be converted.
-Using WMS-degradable standard leaves the option open to use existing
mapservers, albeit with minor tweaks like serverside-caching. The more
advanced features can be added in the process by server developers.
-The ka-map style does make sense from a grid point of view, although
the request can made a bit more clear, like row= and column= instead of
t= and l= If we go onto this path we're leaving all existing WMS servers
behind us and rely on new servers/preprocessors, not sure that's wise.

my 2cents,

Steven
> Is anyone involved in developing the WMS-C spec in the OWS4 testbed?
>
>


-- 
Geodan S&R Amsterdam

-------------------------------------
Geodan S&R
President Kennedylaan 1
1079 MB Amsterdam (NL)
-------------------------------------
Tel: +31 (0)20 - 5711 311
Fax: +31 (0)20 - 5711 333
-------------------------------------
E-mail: steven.ottens at geodan.nl
Website: www.geodan.nl
Disclaimer: www.geodan.nl/disclaimer
-------------------------------------