On Wed, Mar 23, 2016 at 10:15 AM, Moritz Lennert < [email protected]> wrote:
> > But I still think that the layer concept is nothing unique or special > > in this GRASS data model. If you "rename" the layers (that is: set > > layer1=column1 and layer2=column2) and put the columns in a shapefile > > (for example), then you have exactly the same chances to relate shape > > data to different external files via the different columns, while the > > geometric feature is digitized only one time. Or am I wrong? I know > > that shapes are different because they hold only one type at a time, > > but that don't matter here. > > Just to make sure that this is clear: different layers do not > duplicate the same geometric feature. Categories in layers are like > attributes of features, but they are directly linked to the geometry, > thus avoiding the overheads of attribute table handling, and not all > features have to have categories in all layers. You can have features > with no categories in no layer, features with one category in one > layer, features with several categories in the same layer and features > with several categories in different layers. You can also say that saying "in layer" is a bit misleading. The geometry (feature) is not in a layer, rather there is one or more layers of additional information on top of the geometry. There are different application of the layer+category system. In some special (!) cases, layer+category can indeed replace attribute table. If it is a good practice or not is a matter of ongoing discussion: https://lists.osgeo.org/pipermail/grass-dev/2015-September/076552.html Uwe, once you get things sorted out, feel free to edit these wiki pages: https://grasswiki.osgeo.org/wiki/GRASS_GIS_for_ArcGIS_users https://grasswiki.osgeo.org/wiki/Terminology_comparison_between_ArcGIS_and_GRASS_GIS Vaclav
_______________________________________________ grass-user mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/grass-user
