[GRASS5] problems with v.digit or dbmi?

[Michael Barton]

Radim,

I am copying you on this as you seem the primary expert. I copy my 
notes and error messages below. Essentially, while I managed to create 
a closed area, I can't yet get any data associated with it in GRASS 
5.7. I can export it to GRASS 5.0.2 and do this without a problem. 
Either, I am simply missing some essential concept, this is a bug in 
the code, or there is something missing from the version of 5.7 that I 
have. After several hours of messing with it, I can't tell which of 
these it is.
____________________
C. Michael Barton, Professor
Department of Anthropology
PO Box 872402
Arizona State University
Tempe, AZ  85287-2402
USA

Phone: 480-965-6262
Fax: 480-965-7671

- - - more GRASS 5.7 test notes - - -

Preface:
I tried a lot of things to associate some kind of table with a closed 
polygon I have created in GRASS 5.7--including working in v.digit and 
experimenting with the v.db  and db. commands. My current goal is to 1) 
create a polygon in v.digit, 2) associate with a single category of 
information, and 3) output it to a raster file. (For the curious among 
you all, I am using this as a mask to create a 'hole' of null values 
for a network of barrancos/ravines in a DEM. I am then using 
r.fillnulls--still only in 5.0.2--to create a model of the terrain 
before the barrancos were eroded. Since I needed to do this anyway, I 
thought it would be a good excercise to try out several capabilities of 
version 5.7.)

Please note that I am not a DBMS novice, but have developed 
applications of at least passable sophistication in x-base (dBase and 
FoxPro) and more recently in Filemaker. I have PostgreSQL up and 
running and have used it with GRASS on a test basis. However, for most 
of my needs, the internal database in the new GRASS 5.7 will serve much 
better currently than a PostgreSQL/PostGIS or even ODBC connection 
(though one to Filemaker would be nice eventually). So this is why I am 
trying to get it to work smoothly. I understand conceptually (or at 
least **think** I do) how the new vector format interfaces with various 
DBMS, including the internal one--and like it. Also, I am comfortable 
with DBF and it will allow me to rather seamlessly transition data from 
MapInfo and ArcView--both of which maintain attribute data in dbf 
formats.

This background, of course, does not prevent me from completely and 
stupidly missing some important point here which may be patently 
obvious to everyone else ;-)

11/2/03
*****************************
OK, I exported my area to an ascii file (v.out.ascii) and reimported  
it. Ran v.build. Then ran v.digit and created an associated table. It 
seems to have an attached database and table now (checked with 
v.db.connect.)

It will let me output to a raster file, but nothing in the raster file 
(i.e., NO DATA and nothing visible in d.rast). Tried to run v.category 
to add something to the category field in the hopes that this might 
solve the problem. Did the default using the same input and output 
files (i.e., overwrite original). Here is the result.

v.category input=polop_barranc2 output=polop_barranc2 type=area 
option=add cat=1 field=1 step=1
WARNING: Vector 'polop_barranc2' already exists and will be overwritten.
0 new centroids placed in output map
DBMI-DBF driver error: (null)
WARNING: Cannot open select cursor: 'select * from 
polop_barranc2'ARNING: Cannot copy table
Registering lines:
2 primitives registered
Building areas:   50%100%1 areas built
1 isles built
Attaching islands:  100%
Attaching centroids:   50%100%

Number of nodes     :   2
Number of primitives:   2
Number of points    :   0
Number of lines     :   0
Number of boundaries:   1
Number of centroids :   1
Number of areas     :   1
Number of isles     :   1

 >>>>>>>>>>>>>>>>>>>>>>>
Do I have a problem with the version I have in the dbmi drivers or am I 
just doing something wrong. I've reread the tutorial sections that seem 
appropriate and I can't find the answer there.
********************************