Sketcher solver redundancy error

[openBrain]

ForceRecompute macro shows several Errors upstream,

Where do you see those errors? Treeview shows error only on Sketch019, Report view shows only blue messages.

Recomputing the whole design shows error in BottomRightCorner, BottomLeftCorner, TopLeftCorner & TopRightCorner.
I tried to find links with Sketch019, but the Dependency Graph is ... unreadable.
Didn't investigate more.

[chrisb]

I reopened the ticket.

[kisolre]

Running the ForceRecompute macro shows several Errors upstream, and I have to wonder if those Errors are contributing to the Solver problems.

Recomputing the whole design shows error in BottomRightCorner, BottomLeftCorner, TopLeftCorner & TopRightCorner.

I am very curious why do you get errors? Those are simple sweeps and recompute works ok here (except for Sketch019 of cource). See that video

phpBB [video]

[abdullah]

I reopened the ticket.

I confirm I can reproduce.

I do not think it is exactly the same case as the "fixed" ticket, as here there is no reference constraint. But it appears related.

I will take a look.

[abdullah]

I reopened the ticket.

I am trying to start a new trend regarding (at least) solver bugs.

I want to have a Freecad macro that generates new solver bugs. I want to make unit tests out of them, so that when I commit a fix, we put a new unit test to better cover solver issues and prevent fixes from reintroducing old problems.

Maybe somebody can create a small macro for the minimal reproducible situationt of this bug, which is attached to this post while I try to fix the bug...

minimal_reproduction.FCStd

(941.38 KiB) Downloaded 20 times

[bejant]

Using the file from the opening post, and in Sketch019, deleting either the Coincident Constraint or the Tangency Constraint (Constraints 49 and 50) at the blue arrow I drew in the image below fixes the Solver error.

20190420a.png (21.58 KiB) Viewed 835 times

Deleting the Tangnecy Constraint, then adding a Tangency Constraint there using Box Selection to select the Coincident Endpoints creates the Solver error for me. Applying a Tangency Constraint by selecting the edges instead of the Endpoints works as expected and displays the Constraint Substitution message box.

I can't drag the segments apart after I first delete the Coincident Constraint 49 instead of the Tangency Constraint, although I expected to be able to do so.

OS: Ubuntu 18.04.2 LTS
Word size of OS: 32-bit
Word size of FreeCAD: 32-bit
Version: 0.18.16093 (Git)
Build type: Release
Branch: releases/FreeCAD-0-18
Hash: 690774c0effe4fd7b8d2b5e2fb2b8c8d145e21ce
Python version: 2.7.15rc1
Qt version: 5.9.5
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: English/UnitedStates (en_US)

[kisolre]

adding a Tangency Constraint there using Box Selection to select the Coincident Endpoints creates the Solver error for me.

I remember that I might have done the same.

[bejant]

I am very curious why do you get errors?

Sorry, I forgot to mention in my previous reply that I don't know why those other errors don't appear for you.

If you include the link to your YouTube video inside video tags:

Code: Select all

[video]Your video link here[/video]

the video will appear directly in the forum post:

phpBB [video]

[kisolre]

Thanks for the video tag reminder

why those other errors don't appear for you.

I think that the quesion should be the other way around - why are there errors at your end. Those are relatively simple sketches. They are all independent copies of one another and depend only on the spreadsheet. Those lofts are between the same geometry sketches. Sweeps have the same number of edges and are along single arc which does not protrude after the second sketch. Where do that errors come from?

[openBrain]

Unfortunately FC isn't very verbose about the error.
The only clue in the tooltip is "TopoDS::Face".
Notice that despite this error, Part/Check Geometry doesn't complain at all.