[GRASS-user] grass vector model, cats and layers concept

[Moritz Lennert]

On 06/06/13 14:52, Nikos Alexandris wrote:
> Nikos A wrote:
>
>>>>> Still, I find it counter-intuitive that it's required to create
> a new
>>>>> vector map to achieve that.
>>>>> --%<---
>>>>> v.category in=wrs2_tiles_of_interest_testing at PERMANENT layer=2
>>>>> option=add
>>>>> ERROR: Output vector wasn't entered
>>>>> --->%--
>
>
> Markus Metz wrote:
>
>>>> Because you need to modify vector geometries in order to add a
> new
>>>> layer. Categories and layers are first and foremost stored
> together
>>>> with the geometries. Whenever geometries are modified, the result
> will
>>>> be a new vector map (with the exception of v.edit and the
> digitizers).
>
>
> Nikos A:
>
>>> Just for (even more) completeness and understanding:  adding a new
> layer
>>> does not really change the geometries per-se (that is the
> topology, the
>>> shape).
>
>
> Markus Metz:
>
>> No, it does change the geometry directly.
>
> O-K.  This is where I am confused.
>
>
>>> It's the inner design of the vector data model, i.e. the fact that
>>> categories and layers are bound together. Is it wrong to describe
> it as an
>>> internal "indexing" issue or similar?
>
>> Yes, it is wrong. Categories and layers are first and foremost
> stored
>> together with the geometries.
>
> I'll try to consult the programmer's manual or else about the vector
> data model.  I need to clarify this to myself :-).
>
>
>> Therefore adding a new layer will modify the geometries directly.
> IOW,
>> categories and layers are part of the (GRASS) vector geometry
> objects.
>
> Rigth, I need to (re-)define to myself the "geometries" term.
>
>
>> Also, layers and categories have nothing to do with topology.
>
> A-ha...
>
>
>> A new layer really is only added by modifying the geometries, only
> after
>> that you might want to attach a table to that new layer.
>
> Will check the manuals...
>
>
>> I guess you confuse layer with database connection. The existence of
> a
>> layer is a condition for an optional database connection. A database
>> connection itself does not mean the layer of this connection
> actually
>> exists.
>
> Danke Markus.
>
> I am not really confused over the layer/db-connection stuff.  Rather,
> it is the topology - geometry - categories and layers stuff that I am
> confused about.

I think the confusion comes from the fact that what Nikos calls 
"geometry" is the part "drawn" on the screen, i.e. the geometrical 
features visible to the user when displaying a map. Markus speaks about 
"geometry" in the programming sense, i.e. the programmed structure that 
represents the geometrical features, but which contains more than just 
the "visible" geometric parts (i.e. the coordinates), e.g. it also 
contains the information about the type of geometry and the possible 
layers (aka fields) and category values. See:

http://grass.osgeo.org/programming7/vlibFormat.html#vlibCoorFileFormat

Moritz