Unable to scale

thomthom · 22 Apr 2010, 14:16

Why is it that I can't scale the component in the attached model?

TIG · 22 Apr 2010, 14:23

It must be some bug relating to the proportions of the xyz dimensions.
If you make a similar component 1m long or even just half the original length it scales perfectly.
Recommendation: make a shorter component and scale if longer...
Ironically after making it half the length and scaling it x2 it will still scale later on ???

remus · 22 Apr 2010, 14:26

Very strange. My findings so far: something to do with the long faces, none of which will scale individually. The end faces do scale, however.

thomthom · 22 Apr 2010, 16:36

So I'm not going mad then... good to know.

pmolson · 22 Apr 2010, 19:16

@thomthom said:

So I'm not going mad then... good to know.

No one said that.

Chris Fullmer · 22 Apr 2010, 21:15

It seems if one length is 1000 times longer than the other lengths, it does not let you scale. If you make it so it is 50mm x 50000mm then it scales. But 50mm x 50001mm does not scale. Very interesting find.

Chris

TIG · 22 Apr 2010, 21:48

But strangely... after you've scaled a scalable one beyond those 'limits' it still scales OK - even when its new proportions are worse than the non-scaling version's ???

Chris Fullmer · 23 Apr 2010, 03:03

Perhaps because the definition is still under the ratio? That is untested, just a thought.

TIG · 23 Apr 2010, 08:49

So to recap - if you want to scale something it must have xyz / rgb values less than 1000 different - not usually a problem ? If you want something that is going to be worse than that make it less than this proportion initially them you can scale it to beyond the limits...
So if something is made 1x1x1000 it is unlikely to be scalable... but make it 1x1x1 and scale it x1000 in the z/b axis and it's OK and becomes 1x1x1000, then scale it again x10 it becomes 1x1x10000 which would not have been possible in the initial form

thomthom · 23 Apr 2010, 09:19

This is ridiculous bug!

Is it somewhat related to SU's unit precision 1/1000"?
Just seem so odd to occur at exactly 1000 the scale... it can't possibly be by design, can it?

TIG · 23 Apr 2010, 09:28

Question for Google....

Gaieus · 23 Apr 2010, 10:48

You cannot scale anything by 0.0001 either (you get an "invalid scale" warning)

thomthom · 23 Apr 2010, 11:14

@gaieus said:

You cannot scale anything by 0.0001 either (you get an "invalid scale" warning)

True - another annoying issue.

brookefox · 28 Apr 2010, 22:26

Irrelevant, but you might consider adding a little post-modern detail to that otherwise pretty interesting model.

"Just ideas." (No poke.)

Forget it; I hate post-modern.

xrok1 · 28 Apr 2010, 22:43

works with fredoscale!