[Qgis-developer] Directions needed for GSOC Proposal

[Victor Olaya]

>> I would vote for the Framework. About deciding which of the two, I think sextante
>> should be more mature, therefore a better candidate, so:
>> - please Victor share your feelings about that
>> - what is the license of sextante?

I had thought of a MIT license, but I have included some GPLed code
from existing QGIS algorithms (from ftools and mmqgis plugins), so
those algorithms are released under GPL (I have added the
corresponding notices in the source code)

However, I selected MIT for the Java SEXTANTE code to allow for
integration in any kind of software. Since this Python version is just
developed to work with QGIS, GPL does not sound bad to me, in case the
MIT license is an issue.

Regarding ideas for improving SEXTANTE, here are two of them:

-Adding more algorithms: connecting with external apps or libraries,
or writing them directly; This is the easiest and most obvious way of
improving SEXTANTE. The Orfeo Toobox app is one of my priorities in
this, but I am sure there are many other interesting ones...
-Better data handling: Make it easier for algorithms, no matter if
they are 'native' or use an external app, to use any kind of input.
Now they have no problems with file-based data, but using data for a
DB connection or remote sensing is not that easy when you have to pass
it to an external app.

Any other question you might hAve qbout SEXTANTE, feel free to ask

Regards