Re: [udig-devel] Re: JGrass and uDig

[Antonello Andrea <andrea.antonello@xxxxxxxxx>]

* [A.D. 23/01/07 19:16], Matthias Basler <matthiasbasler@xxxxxxxxxxxxxxx> probably wrote:
> Antonello Andrea wrote:
> [...]
> > Lots of changes are going on at the moment in the JGrass team and
> > architecture. One thought and wish of mine would be that to join the
> > uDig development and merge all the raster algorythms into it, as well as
> > everything suitable from JGrass. Following the last coverage changes in
> > geotools, I guess this should be possible and would bring positive
> > additions to both groups. Please take into consideration that this would
> > anyway mean a hughe amount of work from our side.
> > To understand if this could be in fact feasible, I would need to talk 
> > to someone (or the IRC channel) about advantages and if and how this
> > could be achieved. 
> > 
> > Please let me first of all know if the uDig team is interested in this
> > and to who I may talk to about this issue.
> 
> Hi Antonello,
> 
> I am Matthias Basler, my minor part in uDig is the German translation and bug finding from time to time.

Hi Matthias, minor part? Tell it to me, I noticed right now that the
Italian translation of udig is missing!! That means that we will have to do it
and that is a big work, even if we have the same properties file
approach in JGrass.
 
> I think the idea of collaboration between JGrass and uDig is wonderful. On the university of Jena, Germany, where I studied Geography there Java and hydrology were two big topics. (You may know Prof. Flügel and/or Peter Krause.) Anyway, that's how I have gotten knowledge of JGrass - it was then at pre 1.0. I admit I havn't actally worked with it, for one reason because at that time there was no Englisch user manual.
> Still I liked the idea as much as I like uDig for just making Open Source GIS finally "user friendly".

At Jena you guys did a lot of work on visad's datamodel to use it for
GIS analises. I've seen a lot of nice things.

> Now You will surely have to talk with the more involved people in uDig when it comes to making a plan. May I, however, ask you to provide some technical details about JGrass so that we/I can judge on what areas collaboration would be easy or more difficult. 
> (Just as an example: I can already see from the manual that you use Swing whereas uDig uses SWT/JFace.)

Yes, there are some other things that we used to have and will need in
future, like the R (the statistical package) bindings through Rserv or
the console and scripting engine based on Beanshell (in fact Jython
should be the future?). What would be dismissed (a tear drops from my eye) 
is the opengl pipeline through jogl, that wasn't at a final state, but
it was coming slowly and was extremely fast in rendering, even if we were 
slow in reading the data :) (I noticed however that uDig is a bit heavy on this).

I need to understand what kind of raster rendering is already available
in udig, I wasn't able to display geotiffs, I seem to have the only non
working tiffs on earth :) However I'm checking out the SDK to se how the
eclipse plugin creation is organised. Guess this is the first step...
 
> Imho the plugin approach that you suggested sounds good. You'll find that you will, from time to time, 
> run into problems or gaps trying to get advanced raster operations into uDig in its current state, 
> but if this helps to get the "breakthrough" for advanced raster support and analysis in uDig it is definitely worth the work!

Yes, I hope this will be accepted by the udig community and lead somewhere.

Andrea