it's kinda off-topic but still it pertains to the title of the
conversation....

is it possible to make a hemispherical hole with jiglibflash?
preferably in another jbox... something like booleans on the physical
primitives of box and sphere.

On Sep 3, 2:35 pm, savagelook <[email protected]> wrote:
> Like katopz and I said, this really looks and sounds like a zsorting
> issue that could be handled with lite's .layer or there's a branch of
> lite that has .canvas.  I had the same exact problem in an Away3D demo
> of jiglib and resolved it using ownCanvas and pushback on the "floor"
> plane, which is the equivalent of using layer or canvas in lite.
>
> On Sep 3, 6:23 am, katopz <[email protected]> wrote:
>
> > are you sure "a little inside the box" is not z-sort? if not that case last
> > try would be adjust JConfig for more precise collisson detection
>
> > and just say yes if you think you can hack jiglib, i'm free to accept any
> > patch for any issue you found :)
>
> > hth
>
> > On 3 September 2010 17:07, dsirijus <[email protected]> wrote:
>
> > > yup, did that too, same thing happens...
>
> > > anyways, before you give up ( :D ) please do tell me is it possible to
> > > do this withjiglibflashand away3dlite, that's the main thing i am
> > > exploring now. i need to tell the client today is it possible to do
> > > this anyway without seirous hacking on libs.
>
> > > On Sep 3, 11:32 am, katopz <[email protected]> wrote:
> > > > a little inside the box mean face z-sort? in this case you need to use
> > > > .layer or .canvas to separated each mesh
>
> > > > hth
>
> > > > On 3 September 2010 16:30, dsirijus <[email protected]> wrote:
>
> > > > > tried that already, same as with 100, it gets a little inside the box
> > > > > (very perceptible) and then it floats out.
>
> > > > > On Sep 3, 10:58 am, katopz <[email protected]> wrote:
> > > > > > what about 1000,1000,1000 ?
>
> > > > > > On 3 September 2010 15:41, dsirijus <[email protected]> wrote:
>
> > > > > > > a little better, but it still gets 'in' the box if moved 
> > > > > > > rapidly...
>
> > > > > > > On Sep 3, 10:08 am, katopz <[email protected]> wrote:
> > > > > > > > maybe JBox is too thin, do try increase JBox height e.g. new
> > > > > > > > JBox(...,1000,1000,100
> > > > > > > > and you can hide it later then replace with plane (if you want 
> > > > > > > > to
> > > > > make it
> > > > > > > > look thin)
>
> > > > > > > > hth
>
> > > > > > > > On 3 September 2010 14:57, dsirijus <[email protected]> wrote:
>
> > > > > > > > > tried upping the framerate and tweaking the
> > > > > physics.engine.integrate
> > > > > > > > > parameter, tried separating renders from physics updates, 
> > > > > > > > > tried
> > > > > moving
> > > > > > > > > objects on separate layer
> > > > > > > > > no avail
>
> > > > > > > > > @katopz: the way to reproduce it is to move the ball to any 
> > > > > > > > > end
> > > of
> > > > > > > > > level (doesn't even have to be completely to the end) then
> > > swiftly
> > > > > > > > > move mouse to diagonally opposite direction (as to rotate the
> > > level
> > > > > > > > > fast). Even if the ball doesn'tfallthrough, it goes 'in' the
> > > > > level
> > > > > > > > > slightly.
>
> > > > > > > > > I'm using latestjiglibflashand lite from the lite branch. My
> > > > > player
> > > > > > > > > version is debug lnx 10,1,82,76 but i get this even on 
> > > > > > > > > windows.
> > > > > flex
> > > > > > > > > sdk is 4.1. so pretty much latest everything.
>
> > > > > > > > > On Sep 3, 5:33 am, katopz <[email protected]> wrote:
> > > > > > > > > > well it look fine to me there, maybe you hit the time step
> > > sync i
> > > > > bet
> > > > > > > > > > if in that case do try
>
> > > > > > > > > >  physics.engine.integrate(0.12);
>
> > > > > > > > > > instead of
>
> > > > > > > > > >  physics.step();
>
> > > > > > > > > > hth
>
> > > > > > > > > > On 3 September 2010 05:36, dsirijus <[email protected]>
> > > wrote:
>
> > > > > > > > > > > Sorry for the !BUMP! but i really need answer to this. The
> > > ball
> > > > > > > keeps
> > > > > > > > > > > slipping through even though the speed ain't that fast
> > > > > (certainly
> > > > > > > not
> > > > > > > > > > > bullet time).
>
> > > > > > > > > > > Katopz, please? I guarantee you you'll get premium game 
> > > > > > > > > > > for
> > > > > > > showcasing
> > > > > > > > > > >jiglibflashwithin 3 to 4 weeks. :)
>
> > > > > > > > > > > On Aug 2, 5:11 pm, dsirijus <[email protected]> wrote:
> > > > > > > > > > > > Click to reset ball, mouse movement to rotate
> > > surfacehttp://
> > > > > > > > > > > anteperic.com/dev/iskon/
>
> > > > > > > > > > > > Code ishttp://pastie.org/1071117
>
> > > > > > > > > > > > Issue is that ball falls through if movement of the
> > > surface
> > > > > is
> > > > > > > too
> > > > > > > > > > > > fast. Anyone knows what property to increase/decrease
> > > > > (presumably
> > > > > > > :D
> > > > > > > > > )
> > > > > > > > > > > > to get it working more precise so this doesn't happen?
>
> > > > > > > > > > --
> > > > > > > > > > Regards
> > > > > > > > > > -----------------------------------------------------------
> > > > > > > > > > Todsaporn Banjerdkit, katopz,http://sleepydesign.com
> > > > > > > > > > Away3DLite andJigLibFlashDeveloper Team
>
> > > > > > > > --
> > > > > > > > Regards
> > > > > > > > -----------------------------------------------------------
> > > > > > > > Todsaporn Banjerdkit, katopz,http://sleepydesign.com
> > > > > > > > Away3DLite andJigLibFlashDeveloper Team
>
> > > > > > --
> > > > > > Regards
> > > > > > -----------------------------------------------------------
> > > > > > Todsaporn Banjerdkit, katopz,http://sleepydesign.com
> > > > > > Away3DLite andJigLibFlashDeveloper Team
>
> > > > --
> > > > Regards
> > > > -----------------------------------------------------------
> > > > Todsaporn Banjerdkit, katopz,http://sleepydesign.com
> > > > Away3DLite andJigLibFlashDeveloper Team
>
> > --
> > Regards
> > -----------------------------------------------------------
> > Todsaporn Banjerdkit, katopz,http://sleepydesign.com
> > Away3DLite andJigLibFlashDeveloper Team

Reply via email to