[GRASS5] options order in GUI

regarding bug http://intevation.de/rt/webrt?serial_num=3017

Maciek Sieczka wrote:

I encounter a segfault with one particular dataset with the following
options:
GRASS 6.0.cvs:~ > v.clean input=i output=j type=area err=j_err
tool=break,rmdupl,rmdac,bpol,rmarea

Radim Blazek wrote:

I fixed that, however:
1) bpol should be used as first cleaning tool, it does the same as
break, but it is faster on clean (identical boundaries) polygons

I can't change the order of cleaning tools using v.clean in GUI. The order
is hardcoded. Could that be fixed Michael? Options order issue also refers
to v.type in GUI http://intevation.de/rt/webrt?serial_num=2969.

2) bpol works only with type=boundary, the same break,rmdupl etc.

Radim, do you mean that "rmdupl" and "break" should not be used with
type=line?

Maciek

P.S.
This is the 3rd time I'm sending this to the list in last 2 days. I've been
also forwarding to Michael and Radim but nobody answered thus I assume
nobody got my email by now. Maybe now, sending only to the list, will deliver my message somewhere.

Maciek Sieczka wrote:

regarding bug http://intevation.de/rt/webrt?serial_num=3017

Maciek Sieczka wrote:

I encounter a segfault with one particular dataset with the following
options:
GRASS 6.0.cvs:~ > v.clean input=i output=j type=area err=j_err
tool=break,rmdupl,rmdac,bpol,rmarea

Radim Blazek wrote:

I fixed that, however:
1) bpol should be used as first cleaning tool, it does the same as
break, but it is faster on clean (identical boundaries) polygons

I can't change the order of cleaning tools using v.clean in GUI. The order
is hardcoded. Could that be fixed Michael? Options order issue also refers
to v.type in GUI http://intevation.de/rt/webrt?serial_num=2969.

2) bpol works only with type=boundary, the same break,rmdupl etc.

Radim, do you mean that "rmdupl" and "break" should not be used with
type=line?

Maciek

P.S.
This is the 3rd time I'm sending this to the list in last 2 days. I've been
also forwarding to Michael and Radim but nobody answered thus I assume
nobody got my email by now. Maybe now, sending only to the list, will deliver my message somewhere.

I think that I answered the 2)
http://grass.itc.it/pipermail/grass5/2005-March/017673.html

The order could be changed in C code, but it is not systematic solutin, because tools can also repeat and there is no correct order for all situations.

Radim