[Feature request] Automatically apply symmetry constraints to geometry created with symmetry tool in Sketcher workbench

[Yosyp]

Title is self explanatory. Newly created geometry with the symmetry tool appear green, while, in fact, does not apply any symmetry constraint. Upon moving, one can realize no constraints have been applied, requiring manual action.

[chrisb]

Title is self explanatory.

A few more words would not hurt. I guess you are talking about Sketcher workbench? There it would be in general possible to apply symmetry constraints, however, there are cases where this doesn't work.

Here is an example with Y-axis as symmetry line:

SnipScreenshot-863156.png (4.82 KiB) Viewed 1938 times

While for the upper line you need two symmetry constraints, you need for the lower only one.

[wsteffe]

you need for the lower only one.

I think that also in this case you still need two new constraints.
The constraints associated with the new (symmetrical) endpoint which stays on the symmetry line may be a symmetry constraints or a coincidence constraints.

[adrianinsaval]

While for the upper line you need two symmetry constraints, you need for the lower only one.

will the solver fail if the lower part has two symmetry constraints? In my opinion applying symmetry constraints is a good default, when someone has a need for different constraint in some specific points he can change it manually. Maybe have two versions of the tool, one that applies symmetry constraints and one that doesn't. My guess is that most of the time you would want the symmetry constraints.

[GeneFC]

I suspect this may be a matter of slightly different green colors for "selected" and "fully constrained".

Gene

[chennes]

See also:
Sketch symmetry and constraints
Mirror sketch does not mirror constraints
Sketch symmetry doesn't copy constraints

[drmacro]

...My guess is that most of the time you would want the symmetry constraints.

My workflow does symmetries with pattern tools. This keeps my sketches easy to read, easy to make one change effect all symmetries (less error prone), minimizes the constraint requirements (thus minimizing compute time in the solver and subsequent model recomputes).

But, that's just me.

[chrisb]

The constraints associated with the new (symmetrical) endpoint which stays on the symmetry line may be a symmetry constraints or a coincidence constraints.

In theory it could be a symmetry. But the way it is implemented it is not possible to have a symmetry on coincident points - which I think makes sense.

[wsteffe]

it is not possible to have a symmetry on coincident points

Ok, then the requested feature (which I agree would be nice to have) could be implemented with following actions:
- Enforce a symmetry constraint on every edge endpoint (or arc center) with the exception of endpoints (or arc centers) that are placed on the symmetry axis. For the latter the coincidence constraint should be used instead of the symmetry constraint.

[chrisb]

- Enforce a symmetry constraint on every edge endpoint (or arc center) with the exception of endpoints (or arc centers) that are placed on the symmetry axis. For the latter the coincidence constraint should be used instead of the symmetry constraint.

That would be even closer to how it would be constrained manually.