Page 3 of 7

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Sat Sep 17, 2022 3:30 pm
by mfro
wsteffe wrote: Sat Sep 17, 2022 8:22 am Regarding the L-Profile constraint, I think that now, after merging of toponaming3 which provides support for sketch, the problem is solved also in branch development/toponaming.
Tried that as well yesterday and can confirm it's fine now.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Sun Sep 18, 2022 8:50 am
by wsteffe
adrianinsaval wrote: Thu Sep 15, 2022 8:55 pm part extrude with taper angle works incorrectly (nested profiles are filled)
I have just checked that this bug persists after merging of toponaming3. I have also seen that it affects also the current RT branch LinkDaily. So I do not expect it can be fixed in the next PR (toponaming4).

It seems that this bug was actually introduced by toponaming related PRs. The only other possibility is that it was preexisting but fixed in a recent Master Commit not yet included in currrent development/toponaming and in RT/LinkDaily.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Sun Sep 18, 2022 3:26 pm
by adrianinsaval
I doubt that, it was working fine in 0.20 and later IIRC

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Sun Sep 18, 2022 4:18 pm
by wsteffe
adrianinsaval wrote: Sun Sep 18, 2022 3:26 pm I doubt that
But it is also quite strange that this kind of bug can be caused by the topo-naming stuff: It works fine with angle=0 while it fails with angle>0. The topology is the same in both cases.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Sun Sep 18, 2022 4:28 pm
by adrianinsaval
the extrusion algorithm is not the same for both cases, for angle=0 it directly uses a extrusion function from occt, when there is a taper angle it has to use different occt functions, I suspect there was some conflict there, realthunder had his own extrusion taper implementation, perhaps some of that got mixed in and causes trouble.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Thu Sep 22, 2022 2:00 pm
by realthunder
PR submitted for the taper angle problem. The topo naming enabled algorithm is what originally implemented before the upstream version. I choose to retain it because it problem extra function like independent outer and inner tapper angles (for holes). The PR fixed the backward compatibility by adding a new property AutoTaperInnerAngle defaults to True, which will auto fill the inner taper angle as the negative of outer one which is the same behavior as upstream master.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Fri Sep 23, 2022 10:58 am
by lamikr
I did a git pull but did not see the fix in development/totonaming branch.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Fri Sep 23, 2022 11:10 am
by Kunda1
@lamikr hasn't been merged yet. ;)

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Fri Sep 30, 2022 12:36 pm
by obelisk79
Has the 4th PR, implementing TNP into Part Design, been submitted yet? I was checking github and don't see it anywhere, however admittedly I'm far from an expert at navigating github.

Is there any ETA on when this may be getting merged?

Thanks to all involved, I realize that incorporating this code is a major undertaking and I am not attempting to rush the process.

Re: [Ongoing] Testing Toponaming in upstream FreeCAD

Posted: Fri Sep 30, 2022 5:29 pm
by Kunda1
These are the toponaming PRs that have been merged:
Screenshot_20220930_132730.png
Screenshot_20220930_132730.png (80.54 KiB) Viewed 22067 times
The pending toponaming PR in queue is https://github.com/FreeCAD/FreeCAD/pull/7534


Edit: Now all toponaming related PRs use the 'Toponaming' label.