[GRASS-dev] Re: [GRASS-user] how to clean vector file after reprojection

[Maciej Sieczka]

Moritz Lennert wrote:

> http://moritz.homelinux.org/grass/source_v_digit2.jpg
> 
> And zooming even further (I'm in the centimeter range there):
> 
> http://moritz.homelinux.org/grass/source_v_digit3.jpg
> 
> The lines are not closed, but they are still recognized as an area.
> 
> I assume I must be getting something wrong here, but I don't really
> understand what...

Maritz, All,

This *is* weird! There should be node between the boundaries on your
last picture, which would connect them, and let the area be closed.

Although not exactly the same issue, I sometimes too have problems with
areas topology when editing data in v.digit or GRASS Edit /QGIS: after
I split a boundary somewhere near a vertex (eg. on the corner of a
rectangular area), to connect another boundary at this location,
suddenly areas "get bad" after I snap a new boundary to that node. It
it's almost 100% reproducible; please watch this screenkast to see what
I mean:
http://kufaya.googlepages.com/vdigit_bnd.html

As you can see I'm not doing anything wrong. And all seems clean until
I digitize the the 4th boundary. For absolutely no reason the topology
of areas gets bad. And moving vertices around fixes that, after few
blind guesses. Anybody can reproduce that? I can, on 4 machines with
Ubuntu Dapper. The problem was here since I rememeber - on Mandrake
when I was using it, then on Ubuntu Breezy, now Ubuntu Dapper. It is
propably not a bug in v.digit, as the same problem crops out in QGIS.

This reminds me of
http://www.nabble.com/v.buffer-creates-artefact---bug--tf3141568.html#a8733836,
where the area's topology seems OK in v.digit, but v.buffer cannot
handle the input vector correctly until it's "touched by magic".

Maybe Moritze's problem is another manifestation of the same bug? His
boundary is green, which means all-snapped, but there is no node. What
the heck?

Maciek