[Behaviour suggestion] Sketcher_ConstrainSymmetric

About the development of the Part Design module/workbench. PLEASE DO NOT POST HELP REQUESTS HERE!
Post Reply
User avatar
Shalmeneser
Veteran
Posts: 2556
Joined: Wed Dec 23, 2020 12:04 am

[Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by Shalmeneser »

Code: Select all

* Selection before tool only
* '+' = 'then'
* '&' = 'and' (no order)

TODAY  :) 
* 2 points + 1 (center) point --> 2 symmetric points around the center point
* 2 points & 1 line /axis     --> 2 symmetric points around the line / axis
* 1 line & 1 (center) point   --> symmetric line around the center point

TOMORROW  8-) 
* 1 line & axis               --> symmetric line around the axis
* 1 line + 1 (center) line    --> symmetric line around the center line
* 1 line                      --> symmetric line around the 'best' axis
* 2 points                    --> 2 symmetric points around the 'best' axis

(best) : each side of an axis (2 axis = origin) OR smaller angle change
chrisb
Veteran
Posts: 40150
Joined: Tue Mar 17, 2015 9:14 am

Re: [Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by chrisb »

I like the first very much, the second pretty much (not sure if it is intuitive). Don't understand the concept of best axis.
A Sketcher Lecture with in-depth information is available in English, auf Deutsch, en français, en español.
GeneFC
Veteran
Posts: 2952
Joined: Sat Mar 19, 2016 3:36 pm
Location: Punta Gorda, FL

Re: [Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by GeneFC »

I think I understand "best", but I do not think it is a good idea.

If the program chooses the "best" axis then the operation tends to be unstable. A location of the line that is only one micron different could completely change the final orientation after the symmetry operation.

That sort of rapid output change for a tiny input change should almost always be avoided in programming or mechanics.

Gene
User avatar
Shalmeneser
Veteran
Posts: 2556
Joined: Wed Dec 23, 2020 12:04 am

Re: [Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by Shalmeneser »

GeneFC wrote: Sat Jan 15, 2022 3:20 pm If the program chooses the "best" axis then the operation tends to be unstable. A location of the line that is only one micron different could completely change the final orientation after the symmetry operation.
This is just for sketcher symmetry selection mode. At the end, you have the same symmetry constraint with 2 points around a point/line than today.
If the 'best' choice seems an error, just ctrl-Z it.
User avatar
Shalmeneser
Veteran
Posts: 2556
Joined: Wed Dec 23, 2020 12:04 am

Re: [Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by Shalmeneser »

Examples of selection on just one line. Just finding the 'best' axis.
Attachments
BEFORE
BEFORE
Capture du 2022-01-15 23-35-56.png (6.63 KiB) Viewed 333 times
AFTER
AFTER
Capture du 2022-01-15 23-36-36.png (7.29 KiB) Viewed 333 times
GeneFC
Veteran
Posts: 2952
Joined: Sat Mar 19, 2016 3:36 pm
Location: Punta Gorda, FL

Re: [Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by GeneFC »

Shalmeneser wrote: Sat Jan 15, 2022 10:34 pm If the 'best' choice seems an error, just ctrl-Z it.
I was thinking of the problem of flipping that was brought up again just in the past day. Someone was using a spreadsheet to control the sketch.

I understand and agree that anyone working "live" could easily reverse any unwanted choice.

Gene
TheMarkster
Veteran
Posts: 3654
Joined: Thu Apr 05, 2018 1:53 am

Re: [Behaviour suggestion] Sketcher_ConstrainSymmetric

Post by TheMarkster »

Shalmeneser wrote: Fri Jan 14, 2022 2:46 pm

Code: Select all

* Selection before tool only
* '+' = 'then'
* '&' = 'and' (no order)

TODAY  :) 
* 2 points + 1 (center) point --> 2 symmetric points around the center point
* 2 points & 1 line /axis     --> 2 symmetric points around the line / axis
* 1 line & 1 (center) point   --> symmetric line around the center point

TOMORROW  8-) 
* 1 line & axis               --> symmetric line around the axis
* 1 line + 1 (center) line    --> symmetric line around the center line
* 1 line                      --> symmetric line around the 'best' axis
* 2 points                    --> 2 symmetric points around the 'best' axis

(best) : each side of an axis (2 axis = origin) OR smaller angle change
Seems to me the first 2 would be fairly simple where when line1 and line2 are selected or where line1 and axis1 are selected we treat it as if the selection was line1-point1, line2, line1-point2 (or line1-point1, then axis1, then line1-point2).

For the "best" options it could be sensible to presume the origin when only a single line is selected. This would be consistent with other operations, such as setting a horizontal length constraint and only a single point has been selected, in which case the origin is presumed.
Post Reply