> 1. A raster algebra with  a natural syntax (i.e. like r.mapcalc), and,
> as a second step, a GUI. The current implementation has such an odd syntax 
> that most
> users using it early in their qgis learning would conclude that qgis is too 
> complex and odd for them.
> As far as I know, Barry is not thinking on further developing this tool.

> 3. C++ implementation of the (vector) Tools, as a dissolve operation on a 
> complex layer
> takes hours to complete in qgis while seconds with other software (i.e. 
> TNTmips) on
> the same machine.

are you referring to the performances of the operations made thorugh
the python ftools plugin? obviously they're times slower then any
other implementation in C/C++, mostly because of the slowness of the
python loop routines...
the 'analysis' in trunk should go that way, isn't it? (both for raster
and vector)

> In my opinion, these 6 improvements/additions are critical for the normal use 
> of
> QGIS as an operational GIS. Without them, I think we are still below the 
> minimum
> set of requirements for a GIS, even if in other aspects QGIS is very advanced.

+1
_______________________________________________
Qgis-user mailing list
[email protected]
http://lists.osgeo.org/mailman/listinfo/qgis-user

Reply via email to