Assembly 3 - [Solved] Issue on solving when count element are used

[nemesis]

Hello,
I got a new issue on A3.
see the video below :

phpBB [video]

file available here (plateauAssy) :
https://github.com/Nemesis81/testPrinterASM3

the recompute "tick" never disappear.
and solver fail when I add new constraint, which normally shouldn't be problematic.
No message in the report view

[nemesis]

I'm Idiot
one of my sketch was not fully constraints.
that's why.
Nope adding a new constraint make the problem coming back

[nemesis]

So I achieve to reproduce my initial issue with a basic assembly
see attached file.

it looks like it is not possible to solve the assembly constraints when constraints are apply to parts of the "elemenetcount" array after the use of multiply function.
elements are "moved"
not sur my explanation is clear... (sorry for my english) it's clearer in the file

strangething.png (182.84 KiB) Viewed 945 times

still 0.8.6

Code: Select all

OS: Debian GNU/Linux testing (buster)
Word size of OS: 64-bit
Word size of FreeCAD: 64-bit
Version: 0.18.14778 (Git) AppImage
Build type: None
Branch: (HEAD detached at FETCH_HEAD)
Hash: 6c5d33ad61142eb849bbc736ae4f0e02f61c2c18
Python version: 2.7.6
Qt version: 4.8.6
Coin version: 4.0.0a
OCC version: 7.3.0
Locale: French/France (fr_FR)

[realthunder]

So I achieve to reproduce my initial issue with a basic assembly

Sorry for the late reply. I am stuck with something else. And also because this constraint multiplication feature is much trickier to implement than I originally thought. I have fixed quite a few problems, thanks for you report. The tricky part is to make the order of array instance stable. When new circular edges are added, the user has no control of the edge order, which causes unexpected and hard to notice constraining element pair reordering. The reordering is usually harmless, but will cause serious problem when the reordered array instance is involved in other constraints. I have now stabilized the order by sorting the geometry element distances, so new instances always go to the nearest unpaired element, while old instances stays the same. There are probably many other corner cases, and need more test. And due to other pending feature, I am afraid the new version I promised is going to be delayed.

[nemesis]

Sorry for the late reply......., I am afraid the new version I promised is going to be delayed.

As long as It helps you to fix bug, there is no need for excuse

I didn't know if you prefer to have the feedback in the forum or on github, so I created an issue in github.
let me know how you prefer to work.
I also ask for feature improvements, as I have some ideas as I am using ASM3 to design a 3D printer from scratch this time, I see some room for improvement (mainly UI ideas).

thanks again for the workbench!

[realthunder]

I didn't know if you prefer to have the feedback in the forum or on github, so I created an issue in github.
let me know how you prefer to work.

Using github issue for bug report is easier to track. Feature request in forum is better, because it can get more feedbacks from other users.

BTW, I have also fixed the problem you mentioned in the other post.

[nemesis]

Using github issue for bug report is easier to track. Feature request in forum is better, because it can get more feedbacks from other users.

BTW, I have also fixed the problem you mentioned in the other post.

Awesome
Thanks a lot.
can't wait for the new version now