Results 1 to 5 of 5

Thread: Theories invited about puzzling new behaviour of this model post-patch

  1. #1

    Theories invited about puzzling new behaviour of this model post-patch

    Here is an .obj - [url]https://www.dropbox.com/s/jsgrby637czw4gi/Crystal%20Ball%20base1.obj?dl=1[/url] - I made it in Blender a few months ago for a Workshop upload (it's a sort of stand for a crystal-ball to sit on). Here is a collider for it - [url]https://www.dropbox.com/s/zh11oxl7pn4lixy/postcompoundneedballholder3.obj?dl=1[/url] - though whether you use the collider or not, the same behaviour that puzzles me occurs.

    Spawn an object from the chest, like a marble or a pawn - whatever. Place it on top of the ball-stand. Create a Joint between the two. Pick up the stand with your mouse - picks up normally. Now pick up the marble (or whatever) you put on top. Instead of lifting UPward, the two Jointed objects will move DOWNward, some small way into the table-surface (and will then launch itself out when you release, generally). Repeating with the other object up in the air above the ball-stand rather than touching it will have the same result - it pushes down. Jointing not from on top but from the side, however, and you get the correct, upward-lift.

    This wasn't happening previously, though I'm not certain which patch has introduced it (if that is indeed the cause) - I know for sure it wasn't doing it 6 weeks ago. I've played around with Joint settings, Physics settings, and Toggles of both objects, without improvement.

    Anyone have any theories about what is going on? Why this one specific object?

  2. #2
    Hey cluny

    That's a nice model and a good use of the new compound collider, Kudos for that for a start


    This seems to be tied to an objects lift height. for some reason the object on the stand is not applying its current Y displacement to its lift height.

    Effectively it thinks its sat flat on the table so when you pick it up by this top object it has to go down to reach the set lift height for an object that being picked up off the table surface. If that makes sense.

    Raise the global lift height and once you pass the point were the lift height surpasses the top objects current origin point. it will start to lift up rather than push down.

    Seems like a bug to me as the only way to mitigate it is to boost the global lift height.. Which is not always practical, or reset the top objects origin point to match that of the object its sitting on... Which is not always possible like the marble in your example.

    and interesting and unusual find there Cluny

  3. #3
    Interesting - that's a little more light shed on this peculiarity. The natural next question would be what property there is about this one model that invokes this lift-height bug - so it could either be changed as a workaround, or at least avoided when scratch-making other objects.

    Perhaps a developer might have thoughts?

  4. #4
    I think you might find it isn't just related to that model, I checked a few other things while testing yours out

    If you take a simple block from the chest, scale it so its about the same size as your stand then place the same marble on top of it, joint them together then lift the assembly by the marble it does exactly the same thing. the block it sits on is forced into the table as it try's to set the marble to the default lift height.

    A more interesting question is which of your creations with a similar set up does this Not happen

  5. #5
    You're right - it seems to simply be something about certain sizes of the objects and/or relative-size between them that's capturing the bug - the ball-stand is just one example. Or, at least, they need to be those sizes for the effect to be at its most obvious, when the reverse-lift is strong enough to submerge partially below the table surface.

    I'll add this as a Bug Report to the other sub-forum.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •