I'm guessing here, but maybe it's because the rotations are either applied
in strict order; x,y,z? Normally one would populate a matrix with the
rotations and apply them all in one go but I'd like to think that some
jiggery pokery goes on under the hood to allow one to do this kind of
constraint. Yes? No? maybe?
But then why does the first one work?
> -----Original Message-----
> From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On
> Behalf Of Nicolas Langlois
> Sent: 30 April 2007 14:21
> To: xsi(at)Softimage.COM
> Subject: Refresh Problems
>
> Hello all,
>
> I know many of you have already posted about this problem but a fellow
> animator fell on this this morning and I just though I'd send you repro
> steps this morning as it is too obvious we have refresh issues in
> 6.0.1.
>
> # PYTHON CODE COMING UP
> ##############################
>
> Application.GetPrim("Null", "", "", "")
> Application.GetPrim("Null", "", "", "")
> Application.Translate("", 3, 0, 0, "siRelative", "siLocal", "siObj",
> "siXYZ", "", "", "", "", "", "", "", "", "", 0, "")
> Application.SetExpr("null1.kine.local.roty", "null1.kine.local.rotx *
> 0.1", "")
> Application.SetExpr("null1.kine.local.rotz", "null1.kine.local.rotx *
> 0.1", "")
> # WARNING : 3000 - **** Cycle breaking point
> null1.kine.global.ParentPoseCns_E
> # WARNING : 3000 - Cycle through
> null1.kine.local.ori.euler.roty.Expression
> # WARNING : 3000 - Cycle through null1.kine.global.ParentPoseCns_E
> # WARNING : 3000 - PROBLEMATIC EVALUATION CYCLES ARE IN THE SPECIFIED
> GRAPH
>
> # WHY IS THERE A DEPENDENCY CYCLE CREATED HERE???
> # NOTHING IS WRONG WITH PUTING SOME EXPRESSION ON SOME PARAMETER
> # THAT DRIVES AN OTHER ONE BY EXPRESSION
>
> Application.CopyPaste("null.kine.local.rotx", "",
> "null1.kine.local.rotx", 1)
> Application.SetValue("null.kine.local.rotx", 91.837, "")
>
> # NOW INSPECT THE LOCAL TRANFORM PROP OF null1
> # THE ROT Y AND Z AXIS ARE PROBABLY CORRECT AT 1/10TH THE VALUE OF X
>
> Application.SetValue("null.kine.local.rotx", 0, "")
>
> # NOW INSPECT THE LOCAL TRANFORM PROP OF null1
> # THE ROT Y AND Z AXIS ARE INCORRECT AS IT REMAINED THE SAME AS BEFORE
>
> # CHANGE FRAME AND THE VALUE GETS RE-EVALUATED CORRECTLY
>
> ##############################
>
>
> We also have constraints not seeming to evaluate properly but as soon
> as
> the frame is changed, the refresh puts them back in place properly.
>
> The Application.Refresh command doesn't seem to cut it. neither the
> Application.SceneRefresh.
>
> Any reasons why it is behaving so differently than in 5.11. Why the
> expression now creates a dependency cycle in 6.0.1? and how to fix the
> refresh problems without changing frame?
>
>
>
>
> Thanks for all the help...
>
>
>
> --
>
>
> Nicolas Langlois
> Senior Rigger / TD
> Buzz Image
>
> ---
> Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in
> body:
> unsubscribe xsi
---
Unsubscribe? Mail Majordomo(at)Softimage.COM with the following text in body:
unsubscribe xsi
