[mapserver-dev] GeoServer superseeding MapServer in Europe?

[Daniel Morissette]

Bart van den Eijnden wrote:
> Hi Daniel,
> 
> I think there are a few reasons:
> 
> 1) Java fits in better with existing IT infrastructures

Not sure I agree with this one... if you are already a Java shop then
GeoServer fits better, but if you are not then I would have thought that
MapServer would have fit better (as lightweight CGI under any HTTP
server) and saved you from installing and managing a complete Java
server environment.

Is it a matter of lack of a click-through installer for Windows, or
really an IT infrastructure issue?

> 2) Geoserver's strong and ongoing support for WFS, especially the
> application schemas, which are needed for INSPIRE

Good point.

> 3) The admin tool
> 

Yeah, this one keep coming up...


> I think you can probably also read some of the reasons of why the Dutch
> government chose Geoserver, around page 46 in the following document (it
> also contains a section on Mapserver):
> http://www.geonovum.nl/sites/default/files/algemeen/MOT20100618TechnicalDesignv1.0.pdf
> 

Sounds to me like the integrated stack is what made GeoServer win in
this case, not the superiority of the software itself.

I have to admit that the alternative to "the stack" in the MapServer
case is a bit messy, a set of independent tools each with their own
paradigms, quirks and config files, and in different languages:

- mapserv.exe (native binary)
- MapScript (PHP, Python, or one of the others if you dare)
- TinyOWS (for WFS-T, which works great BTW), also a native binary
- TileCache (Python)
- FeatureServer (Python)

One of the things we've been talking about here (at Mapgears) is coming
up with an integrated set of tools for all of the above with a
consistent config file and possibly all native binaries... but that
would require some work.

Daniel
-- 
Daniel Morissette
http://www.mapgears.com/