[ka-Map-users] ka-Map and OpenLayers

Schuyler Erle sderle at metacarta.com
Wed Jun 21 12:18:18 EDT 2006

Hello! I'd just like to follow up on John's email with a few
additional thoughts.

* On 20-Jun-2006 at  3:18PM EDT, Steve Lime said:
> What does OpenLayers offer that ka-map doesn't ( besides a better
> name ;-) )? The list you reel off of limitations is pretty
> substantial.

OpenLayers was designed from the ground up to be totally
object-oriented JavaScript, using (as Paul pointed out) the nicer
features of Prototype.js. The code base is intended to be very clean
and very readable, and has over 450 unit tests written:


OpenLayers currently supports WMS and WFS:


The project is also quite close to supporting Google, Yahoo!, and MSN
Virtual Earth base layers as well.


More importantly, OpenLayers is designed as a library API for building
map applications, much like the Google Maps API. OpenLayers is pure
JavaScript, and does not have any particular server-side dependency.
Consequently I would imagine that though OpenLayers might subsume the
client-side portions of ka-Map, ka-Map as a project would continue to
focus on the server-side portions of its codebase.

> * core tiling engine is very similar, but it is not quite as optimal
> (doesn't reuse images for instance)

Based on what Paul showed us at Where 2.0, this will be quite easy to


> * lacking tools (layer controls, scale bar etc) and the windowing
> stuff

We have some of these tools -- pan, zoom in/out, zoom box, ruler,
layer switcher. I have the impression that other things like the scale
bar will be pretty easy to port as well.

> * lacking query capability

I'm not totally sure what this means. OpenLayers supports an event
model that makes this kind of thing easy to add on a per-application
basis. If you're thinking of the sort of query mode supported natively
in MapServer, this should be no problem at all.

> * tile caching

If you're referring to server-side caching, we're excited about
adopting ka-Map's existing tile caching method.


> At this point, I am proposing to join the two projects into a single
> project.  My desired approach would be to start with the OpenLayers
> code and start to bring over the parts of ka-Map that I(we) feel are
> missing.

Perhaps a bit of history is in order. About this time last year,
MetaCarta (a geographic technology company based in Cambridge, MA)
decided it was high time that an Open Source equivalent to
Google/Yahoo/MSN maps was available. (Note that this was before ka-Map
existed as a public software project.) After a couple of meandering
false starts - during which we learned an enormous amount about how
mapping needs to work on the web - we put together the code base
currently available from http://www.openlayers.org/.

For the last several months, MetaCarta has been supporting four
developers (including me) to work part or full time on OpenLayers,
with the intention of giving away the code, in the hopes that the Open
Source community would eventually adopt the project and help carry it
forward. We'd be thrilled to join forces with the ka-Map community,
and would be glad to begin working on the features needed to make that


More information about the ka-Map-users mailing list