On Thu, Jan 26, 2012 at 5:15 PM, Tim Sutton <[email protected]> wrote: > > Martin I have been doing quite a bit of QGIS python programming the > last few weeks and I must say, it's is a really awesome development > environment and I could see a lot of merit in replacing what we have > in app with pure python and then using C++ for core, analysis and gui > libs. I can see that on new platforms (e.g. android mobile) new > interfaces are going to be needed and in my opinion it isn't something > we should discount over the long term. However the massive work > involved in doing that is probably a show stopper. Just mixing in > python functionality in the way of gdaltools and ftools seems to me a > viable way to continue for the medium term. What exactly would you > vision of a more python centric QGIS amount to?
My vision is roughly the same as you describe: first of all, leave the libraries (core, gui, analysis) strictly in c++. These are our engines and they should be robust and fast. However the application code (in src/app) taking care of GUI usually does not run tasks that need to be super-fast. This is the area where the possibility to switch from c++ to Python might be useful. The python code could be more compact, not needing to compile and in case of fatal errors it would just show an error message instead of crashing whole QGIS. I certainly do not support rewriting stuff from c++ to python. The whole point is that by having the possibility to put python code directly in qgis app we may be more efficient implementing new features and/or reusing existing code from plugins. Martin _______________________________________________ Qgis-developer mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/qgis-developer
