[GRASS-user] Report from ongoing GRASS GIS Community Sprint in Prague

[Benjamin Ducke]

OK, I think I can see my mistake now.
I always assumed that Victor had developed the QGIS bindings
on top of the existing Java libraries, using one of the
available Java to Python bridges or the Java Native Interface
for C++.

However, this does not seem to be the case. The code in:

http://code.google.com/p/sextante/source/browse/trunk/soft/bindings/qgis-plugin/src

does not contain just another binding for SEXTANTE to QGIS (even
though the SVN folder structure suggests this) -- Victor, please
correct me if I am wrong here.

Rather, it looks like it contains a complete re-write of SEXTANTE in
Python (excluding the original SEXTANTE processing tools, which seem
to be only available in their Java versions).

I am entirely unsure how to further maintain a code base in two
different programming languages and how to keep my own work
on the GRASS/SAGA/R interfaces in Java synchronized with this.
Any ideas welcome (but should probably be discussed on the SEXTANTE
user mailing list).

As far as the GRASS code sprint is concerned, please consider
that many people use SEXTANTE under gvSIG, OpenJUMP, GeoTools and
other Java-based GIS at this point, and that they are just as
interested in seeing the GRASS interface improved, as QGIS
users are. For now, I will monitor the QGIS bug tracker and
synchronize any relevant tickets manually with bugs.gvsigce.org.

Best,

Ben

On 05/28/2012 03:08 PM, Markus Metz wrote:
> On Mon, May 28, 2012 at 2:47 PM, Paolo Cavallini<cavallini at faunalia.it>  wrote:
>> Il 28/05/2012 14:33, Benjamin Ducke ha scritto:
>>
>>> SEXTANTE and gvSIG CE share a bug tracker
>>> at http://bugs.gvsigce.org, a decision that
>>> was made together with Victor Olaya a while
>>
>>
>> Hi all.
>> Sorry for the misunderstanding.
>> I am not sure Victor (main SEXTANTE dev, AFAIK) is listening here - if so, he can
>> explain much better.
>> Anyway, I (and I suspect also Marckus) was referring to the QGIS Python plugin called
>> sextante. No shared code with the Java version, AFAIK.
>
> I have just compared the QGIS SEXTANTE plugin, the gvSIG SEXTANTE
> plugin, and the generic SEXTANTE. For example, the contents of the
> grass/description folder of QGIS SEXTANTE are completely different
> from gvSIG and generic SEXTANTE. The former has hand-crafted .txt
> files, the latter have automatically generated (<grass command>
> --interface-description) XML files. QGIS SEXTANTE is pure python
> without any java code, the others are pure java without any python
> code.
>
> There is a dedicated bug tracker for QGIS SEXTANTE, and from a user
> perspective this is the logical place to report issues for QGIS
> SEXTANTE, in particular when keeping in mind that the GRASS command
> interfaces in QGIS SEXTANTE are hand-crafted, see v.buffer.angle.txt,
> v.buffer.column.txt, v.buffer.distance.txt,
> v.buffer.minordistance.txt, four different interfaces to the same
> GRASS command v.buffer. In gvSIG SEXTANTE, there is only v.buffer.xml.
> How should a user take these differences into account? Report to
> http://bugs.gvsigce.org and explain that this does not apply to gvSIG
> SEXTANTE, only to QGIS SEXTANTE?
>
> Markus M
> _______________________________________________
> grass-user mailing list
> grass-user at lists.osgeo.org
> http://lists.osgeo.org/mailman/listinfo/grass-user



-- 
Benjamin Ducke
{*} Geospatial Consultant
{*} GIS Developer

   benducke at fastmail.fm