Drop a cube.

Ctrl click the cube

Set z orientation to 0.2 or 0.1 (component properties).

Unselect the cube

Ctrl click the cube

z orientation is back to 0.0

October 1, 2010 at 1:06 pm /

Drop a cube.

Ctrl click the cube

Set z orientation to 0.2 or 0.1 (component properties).

Unselect the cube

Ctrl click the cube

z orientation is back to 0.0

February 28, 2011 at 1:20 pm /

I don’t see that I can do anything about this one. I had a look, and It’s the accuracy of the math. Eventually, .2 degrees probably seems like a fairly large distance from 0 degrees, but not with the accuracy of numbers I work with. It is something like .99993. Very close to "1". Eventually the rounding results in it being "1". This is especially true because it isn’t just the angle of this one group, but the math involved in relation to the world, it’s parent etc.

So, I’ll move this to "outstanding issues", and I’ll keep it in mind for a possible solution.

March 7, 2011 at 9:10 am /

"DOM107":3lj9nbo2 wrote:

Drop a cube.

Ctrl click the cube

Set z orientation to 0.2 or 0.1 (component properties).

[/quote:3lj9nbo2]

At this point, the cube is rotated as expected…

Ctrl click the cube

Set z orientation to 0.2 or 0.1 (component properties).

[/quote:3lj9nbo2]

At this point, the cube is rotated as expected…

"DOM107":3lj9nbo2 wrote:

Unselect the cube

Ctrl click the cube

z orientation is back to 0.0[/quote:3lj9nbo2]

Ctrl click the cube

z orientation is back to 0.0[/quote:3lj9nbo2]

… then it’s back to original orientation.

March 9, 2011 at 4:13 pm /

Sorry, I don’t understand. .1 and .2 degrees round down to 0 degrees due to the vector math involved. These (euler) angles are not stored internally, they are converted to a matrix and it is the conversion to these matrix that results in the rounding error.

You must be logged in to reply in this thread.