[solved] Request for verifying OCCT version differences

[mlampert]

I have a unit test that runs just fine on my system, using OCCT 7.3, and it fails in Travis. My assumption is that it depends on the OCC version so I was wondering if ppl could run attached script in FC and let me know if one of them reports "NOT OK" and if the also printed Orientation is the key to success here.

Thanks in advance!

Code: Select all

OS: Debian GNU/Linux testing (buster)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.18.14034 (Git)
Build type: Debug
Branch: feature/chamfer
Hash: 52a00c87f796c4a7a1e90322a1f1c5962547fd9b
Python version: 2.7.15
Qt version: 4.8.7
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: English/Canada (en_CA)

[wandererfan]

Code: Select all

Traceback (most recent call last):
  File "/home/huxster/Documents/CAD/Complaints/mlampert_wire_orientation.py", line 24, in <module>
    if PathGeom.pointsCoincide(-zaxis, ff.Surface.Axis):
<type 'exceptions.AttributeError'>: 'module' object has no attribute 'pointsCoincide'

OS: Linux Mint 18.3 Sylvia
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.17.13522 (Git)
Build type: None
Branch: releases/FreeCAD-0-17
Hash: 3bb5ff4e70c0c526f2d9dd69b1004155b2f527f2
Python version: 2.7.12
Qt version: 4.8.7
Coin version: 4.0.0a
OCC version: 7.2.0
Locale: English/Canada (en_CA)

[mlampert]

RIght - the script requires a 0.18 development build, after commit fb0fc49a3026.
In order to run it on FC 0.17 you need to replace the import of PathGeom with:

Code: Select all

from PathScripts.PathGeom import PathGeom

[chrisb]

I was wondering if ppl could run attached script in FC and let me know if one of them reports "NOT OK" and if the also printed Orientation is the key to success here.

Just in case you are also interested if it works: Yes it does here with output

Code: Select all

CW wire OK (Forward)
CCW wire OK (Forward)

OS: Mac OS X
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.18.13976 (Git)
Build type: Release
Branch: (HEAD detached at fae0de5)
Hash: fae0de58581694157a97d567c151bdce75d387dc
Python version: 2.7.15
Qt version: 5.11.0
Coin version: 4.0.0a
OCC version: 7.2.0
Locale: German/Germany (de_DE)

[mlampert]

Just in case you are also interested if it works: Yes it does here with output

Thanks!

[user1234]

Code: Select all

CW wire OK (Forward)
CCW wire OK (Forward)

tested with:

Code: Select all

OS: Debian GNU/Linux 9.4 (stretch)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.18.14001 (Git)
Build type: Release
Branch: master
Hash: 6014fce390390526f3d74672e79e3d8bc994225e
Python version: 2.7.13
Qt version: 4.8.7
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: English/UnitedStates (en_US)

Greetings
user

[GeneFC]

Works OK for Windows.

OS: Windows 7
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.18.13999 (Git)
Build type: Release
Branch: master
Hash: ab1520b872821414c6ce4a15fb85d471ac2a2b03
Python version: 2.7.14
Qt version: 4.8.7
Coin version: 4.0.0a
OCC version: 7.2.0
Locale: English/UnitedStates (en_US)

Gene

[mlampert]

Thanks guys, that's reassuring.

Did some more tests and the faces created in Travis are the other way around. What's supposed to be up is down and vice versa. If I read the logs correctly Travis uses OCC 6.7.0 ?

Code: Select all

-- Found OCC: /usr/lib/x86_64-linux-gnu/oce-0.15/../../../include/oce (found version "6.7.0") 
-- -- Found OCE/OpenCASCADE version: 6.7.0
-- -- OCE/OpenCASCADE include directory: /usr/lib/x86_64-linux-gnu/oce-0.15/../../../include/oce
-- -- OCE/OpenCASCADE shared libraries directory: 

Can we upgrade that to a reasonable version? Or are we really supporting V6.7 ?

[DeepSOIC]

AFAIR, there was a change in OCCT regarding wire and face orientation, somewhere around version 7.0, and I think it was about older versions doing something to autocorrect errors, and occt7 is more "dumb" and uses all as is.

There was an example on the wiki that had to be patched up because of this change.

[mlampert]

AFAIR, there was a change in OCCT regarding wire and face orientation, somewhere around version 7.0, and I think it was about older versions doing something to autocorrect errors, and occt7 is more "dumb" and uses all as is.

There was an example on the wiki that had to be patched up because of this change.

Great, that explains that. I'll implement the area formula from Gauss and restructure the logic accordingly.
Thanks everyone!