[GRASS-user] overlapping areas seem valid to v.build: why?

[Benjamin Ducke]

Would it be a solution to add tolerance
for data with partial or bad topology to those vector
modules that deal with polygons and must not strictly
rely on error free topology? v.rast.stats is one 
them. I imagine the group is not too large.

I don't think it would be very much work to change
the check for topology level 2 that those modules
currently use into one that just issues a warning
if only level 1 is present, but then lets the
user proceed at his/her own risk(?)

The critical task would be to decide which
modules strictly require level 2 for usable 
result.

Ben

-- 
Benjamin Ducke
{*} Geospatial Consultant
{*} GIS Developer
  
  benducke at fastmail.fm


On Saturday, December 03, 2011 4:39 PM, "Hamish" <hamish_b at yahoo.com>
wrote:
> Ben wrote:
> > So from a user point of view, getting rid of "-c" in
> > GRASS 7 would remove another source of uncertainty.
> 
> if that means that tasks like the ones I described
> in the third bullet point of this email:
>   http://lists.osgeo.org/pipermail/grass-user/2011-December/062818.html
> are no longer possible, then I can not support that.
> IMO relying on v.external as work-around to be able to
> deal with that class of data is a bit of a cop-out.
> 
> 
> Hamish
>