Archived issue #0023575
Bounding box for a planar edge has non-zero dimension in direction orthogonal to an edge plane
Description
The problem is caused by fix on #566: bounding box computed for a curve is computed by evaluating fixed set of points on a curve and then extended by some heuristic value, which leads e.g. to bounding box of planar curve having non-zero dimension in direction orthogonal to the plane.
Steps to reproduce
The commands:
bsplinecurve c 3 2 0 4 1 4 0 0 0 1 0 1 0 1 1 1 0 1 1 0 0 1
mkedge e c
bounding e
produce this result:
-0.15023140314433286 -0.15023140314433286 -0.15023140314433286 1.150231403144332
8 0.90023140314433281 0.15023140314433286
the expected dimensions are:
0. 0. 0. 1. 1. 0.
bsplinecurve c 3 2 0 4 1 4 0 0 0 1 0 1 0 1 1 1 0 1 1 0 0 1
mkedge e c
bounding e
produce this result:
-0.15023140314433286 -0.15023140314433286 -0.15023140314433286 1.150231403144332
8 0.90023140314433281 0.15023140314433286
the expected dimensions are:
0. 0. 0. 1. 1. 0.
Public activity
6 archived notes
Participants are labeled by their role within this record.
Tests executed on current master with reverted fix for #566 show only one regression (checkshape faulty: InvalidSameParameterFlag) on
sewing tol_100 J3
There are also some differences in images, notably on sewing tol_100 C5, but none of them seems real problem. CPU and memory usage seem to remained the same
sewing tol_100 J3
There are also some differences in images, notably on sewing tol_100 C5, but none of them seems real problem. CPU and memory usage seem to remained the same
On current master, the computed bounding box looks like this:
The added value 1e-7 in both +Z and -Z directions is expected, as the method of box computation adds a gap equal to the shape tolerance.
So, I think this bug can be closed.
Draw[8]> bounding e -9.[revision removed]-008 -9.[revision removed]-008 -9.[revision removed]-008 1.0000001000000001 0.90023140314433281 9.[revision removed]-008
The added value 1e-7 in both +Z and -Z directions is expected, as the method of box computation adds a gap equal to the shape tolerance.
So, I think this bug can be closed.
I agree, this seems to be somehow fixed since OCCT 7.0.0. Please add a test case and close the issue.
Branch [archived branch] has been created by Participant.
[revision removed]
Detailed log of new commits:
Author: apv
Date: Wed Jun 14 16:42:49 2017 +0300
0023575: Bounding box for a planar edge has non-zero dimension in direction orthogonal to an edge plane
# Test case bugs/moddata_3/bug23575 creation
[revision removed]
Detailed log of new commits:
Author: apv
Date: Wed Jun 14 16:42:49 2017 +0300
0023575: Bounding box for a planar edge has non-zero dimension in direction orthogonal to an edge plane
# Test case bugs/moddata_3/bug23575 creation
Branch [archived branch] has been deleted by Participant.
[revision removed]
[revision removed]
Branch [archived branch] has been deleted by Participant.
[revision removed]
[revision removed]