Post Number: 106
|Posted on Monday, February 05, 2007 - 04:34 pm: ||
I've got some unespected results with recatngle in draft.
1- draw a rectangle in the draft
2- draw a segment from a corner of the previous recantgle with snap point on.
3- zoom in very closely (if neccessary increase the accuracy to max n16)
If you explode the rectangle it will change of size.
You should see that the segment is NOT exactly on the corner.
Here ie a sample file.
This affect only rectangle.
Polylign seem to work fine
This problem has caused booleans fail with imported entities from draft to another project.
We have not diagnosed it immediatly...
Post Number: 113
|Posted on Wednesday, February 07, 2007 - 09:59 am: ||
Nobody can reproduce this?
Post Number: 1529
|Posted on Wednesday, February 07, 2007 - 10:07 am: ||
Yes, we are reproducing this, but this scale is at the size of individual atoms. Therefore I am not sure if this is to be expected, so I have forwarded this to our developers for clarification. We should have an answer for this soon.
Post Number: 115
|Posted on Wednesday, February 07, 2007 - 03:43 pm: ||
I've seen this litle unaccuracy is atomic...
In fact it's not really a problem, except it caused some problem with booleans in modeling.
I think without this, I may never realize this.
Here is a new file with a correct accuracy (0.001 cm) with a rectangle in the draft. i've draw an enclosure with point snap in a corner.
After I've copied it in the moedling than extruded the two shape.
You'll find the result in the modeling window.
If you try to make an boolean union, it will fail.
I suppose you can reproduce this.
In fact from this, we avoid to use any rectangle in our draft destined to modeling.
Of course there is a simple work arround by using an extand face to face before the boolean operation.
Thank you very much for your professionalism.
Even the small problems are studied which is very appreciable.
This is the origin of a stable software.
Post Number: 1535
|Posted on Wednesday, February 07, 2007 - 05:00 pm: ||
Thanks for the additional information. We have now verified that there can be a small (but greater than sub-atomic) roundoff error with rectangles in drafting. Thanks again for your report, we will get this corrected.