Radial scale offsets object

[changrilla]

Hi. I'm fairly new to 3d modelling in general so please forgive me if this is in error. I'm doing a radial scale using the y axis of this model http://www.mgcsoft.com/junk/Thread_Base_Unit.wings When I do a 500% LMB y axis radial scale the z axis of the top and bottom faces shows an offset of 0.00627, the scale is correct. If I do the same using the RMB or MMB using the top face there's no offset. This doesn't appear to happen with a simple cylinder without the helix.

The link in the preview here doesn't work but it does when I paste it into a browser. It's definitely there. Sorry if it doesn't work, there's not much more I can do to verify it.

Wings3d 1.5.4
Windows 7 Pro 32 bit
openGL 3.3.10750

Kind regards

Gordon

[micheus]

Gordon, I couldn't verify the problem. Sorry.
I think that can happens when you are scaling for a high values. In computer math this small difference can appear because of the way floating-point works:

Code:

So you’ve written some absurdly simple code, say for example:
    0.1 + 0.2

and got a really unexpected result:
    0.30000000000000004

* from this link

You can try to use the Tools->Bounding Box->Bounding Box... and create a bounding box with the exact dimensions you want. Then you select your entire object and use Tools->Bounding Box->Scale To Saved Bounding Box Proportionally->All.
It probably will give you the same result, but you can try to see if it doesn't.

[changrilla]

I hear what you're saying but it *is* scaling properly, according to the readout. If it was a scaling problem wouldn't that imply that 0,0 (zx) was out in the first place by something like .00125. I was hoping my PC maths processor was better than that!, and it does work with the radial MMB and RMB options, so I have a work around.

But, if you can't reproduce it there's not a lot you can do I suppose.

Cheers

Gordon

[micheus]

I don't know what I looked before, but I can notice that offset. It's interesting it happens for the body mode, but not for face. If you select the top and bottom faces and use the LMB Scale option we don't get the offset.
Something to be studied some day.
Thanks for report it.

[changrilla]

If you rotate it 90° before scaling the error moves to the X axis. It would appear to be something to do with the start and finish of the helix.

Cheers

Gordon