Block instances and quad osnap

insert a block instance, rotate it and try to osnap to it’s quad… the quad option is rotated as the block instance.
I expect that the quad object snap is Cplane related…

I agree with you
Ciao Vittorio

I always thought it was seam related. I’ve never had the end point anywhere but on a quad point.

Well, imho the quad is similar to the entity’s bounding box…
We have more than one quad snap on an object so don’t understand how you can have the seam point on each of them (in a closed curve there is only one seam and at least 4 quads).

Yea, the seam can be anywhere. The quads should be along the axis in the current cPlane. An oversight… (aka bug)

Sorry if I was unclear @lucio_zadra. I mean one qaud on seam, one on mid, and the other two at the 1/4 and 3/4 marks. I thought the quad marks were always oriented with the seam, but my seams do generally run with the cplane.

For blocks, they retain the orientation of the original block definition. When you rotate that block with a curve, you are rotating the block’s position, not the original geometry. The object is still aligned to the world coords of the block definition.

Setting a block’s gumball to align with object for example, always aligns to the world coords of the original block definition.

Hi Lucio- I see that, thanks. All snaps and everything on a block instance have to be transformed back to the underlying block definition, and that original circle is probably oriented to the cplane. So we need to do some more work to account for the transform there, obviously… thanks for pointing this out.


(these are not visible to the public yet.)


Ok Pascal,

Hi Lucio - this is fixed in V6, it could not be done in V5 code, from what I understand.

thanks, -Pascal

doh! So I have to make a workaround for the next 4 years… :grin:
Thanks for the info Pascal!