You're a champ Halfy.
Only you could boast about new bugs and get away with it. ;)
-----Original Message-----
From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On Behalf Of Halfdan Ingvarsson
Sent: January 9, 2007 2:01 PM
To: XSI(at)Softimage.COM
Subject: RE: Weird Realflow problem in XSI 6
I'm guessing that this particular case best be solved by passing context data to the operator indicating a base evaluation time or something like that, correct?
The original shifting of the timeline was a weird-ass hack that was added in before there was an operator evaluation code in place (Twister) and had some significant costs in terms of added complexity and odd side-effects.
Like I've said before, the rendering core has been rewritten from scratch and so most of the old bugs and side-effects are gone and you simply get new ones instead :)
- ½
-----Original Message-----
From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On Behalf Of Guy Rabiller
Sent: 09-Jan-07 12:38
To: XSI(at)Softimage.COM
Subject: Re: Weird Realflow problem in XSI 6
Except that, when the topology change during a frame time ( for motion
blur ), you have to refer to the scene time, or else you endup by
interpolating between different files, wich is not gonna work.
So you actualy *have* to refer to one frame only, the frame scene time,
to get one file ( topology ) only, then use the time passed to the
oprator to compute delta from wich deformation can be handled according
to the given velocities.
If you only refer to the the time passed to operator, then it will
failed if topology change from frame to frame.
--
guy rabiller | 3d technical director (at) LaMaison
Halfdan Ingvarsson a écrit :
> We don't think there's any particular need to inform them that what
> they're doing is wrong when the correct method has been spelt out for
> them for a good while now in the SDK documentation. Operators
> don't evaluate using scene time, they evaluate using the time given,
> it's as simple as that.
>
> - ½
>
> ------------------------------------------------------------------------
> *From:* owner-xsi(at)Softimage.COM on behalf of Chris Marshall
> *Sent:* Tue 09-Jan-07 08:36
> *To:* XSI(at)Softimage.COM
> *Subject:* Re: Weird Realflow problem in XSI 6
>
> Do Soft inform plugin manufacturers of changes such as this, so they're
> aware their software doesn't work any more?
>
>
> Halfdan Ingvarsson wrote:
> > The difference now is that we don't shift the current time on the
> timeline when rendering. Instead we insist that operators actually
> evaluate based on the time passed to them, rather than the current
> time. Y'know. As they should've been doing in the first place.
> >
> > - ½
> >
> > ________________________________
> >
> > From: owner-xsi(at)Softimage.COM on behalf of Kim Aldis
> > Sent: Tue 09-Jan-07 06:48
> > To: XSI(at)Softimage.COM
> > Subject: RE: Weird Realflow problem in XSI 6
> >
> >
> >
> >
> >
> >
> >
> > From: owner-xsi(at)Softimage.COM [mailto:owner-xsi(at)Softimage.COM] On
> Behalf Of Chris Marshall
> > Sent: 09 January 2007 11:44
> > To: XSI(at)Softimage.COM
> > Subject: Re: Weird Realflow problem in XSI 6
> >
> >
> >
> > Way Hey! That works!
> >
> > [kim aldis]
> >
> >
> >
> > Excelent!
> >
> >
> > So where do you think the problems lies?
> >
> > [kim aldis]
> >
> >
> >
> > Either in Spain or Montreal.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
> >
>
>
>
> ---
> 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
---
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
