[Bf-committers] physics concerns

Matt Ebb matt at mke3.net
Thu Jan 31 06:44:30 CET 2008


On Thu, Jan 31, 2008 at 6:55 AM, Ton Roosendaal <ton at blender.org> wrote:
> Hi,
>
>  Thanks for the review; I also wonder how this was approved for trunk...

Michael's not just talking about cloth, which seems to be performing
pretty well to me, he's talking about physics in general. While I
definitely wouldn't agree with his assessment of things being worse
then they ever were, there are a few things that really need more work
- it seems that the particles side of the new particle system has been
neglected so far (understandably) compared to hair/fur.

The worst part right now is the point cache system and especially
using it with deflector/force fields/multiple particle systems. Nearly
every time you change some setting, it requires several round trips to
and from the particle buttons, forever clicking 'clear cache' to get
the system to update correctly. There's very little realtime feedback
when you do change settings, and things can often get messed up when
you change something in the middle of the timeline, rather than right
from frame 1.

All of these things together gives the particle system a really flakey
feeling to use, since you're never quite sure if what you're looking
at is 'real' or the results of an out of date cache, or weirdness
caused by updating the cache midway.

Actually at least in the very short time I've played with the new
cloth stuff, it hasn't suffered from these problems quite so much as
softbody and particles currently do...

Anyway, I hope these things can be worked out better before people
start thinking of releases etc. Just because this stuff may not have
been used as much in Peach as hair etc doesn't mean it's ready too :)

cheers

Matt


More information about the Bf-committers mailing list