[Bf-committers] Meeting minutes May 1 + todo for release

[Ton Roosendaal]

Hi,

>> - Bump map fixes
>
> What's the situation with the 'temporary' 'make my render work  
> correctly' button?

Uhm... you mean the "Correct normal map" option?
It actually belongs in the "Map to" panel, which is completely stuffed.  
That will be solved of course later... waiting for the Big Panel  
Shuffle. :)

I agree it's not the most beautiful option... but I also don't know how  
to code bumpmappping in a way that satisfies all user demands. Applying  
bumpmaps in a "2D" way, as was in Blender since beginning, just is  
something that "looks" better in many situations. Correcting it to have  
it "rotation-invariant" (or for normal-maps) is a new feature... that's  
why I made it a choice with button.


>> --------> Softbody
>>
>> Todo:
>> - Baking (done)
>> - docs & demos!
>
> Are the softbodies still considered WIP and open to proposals and  
> changes, or has it basically been locked off now? There are a number  
> of usability problems with it, and from what I can see on the forums  
> most people who aren't physicists and try to use it (including myself)  
> are getting very confused about what all the different buttons mean,  
> what units they are in, what effects they have and how they  
> interrelate with each other. At least in my testing, it's been one big  
> session of 'tweak unknown settings, test, repeat' but with all the  
> different permutations of variables affecting each other in mysterious  
> ways, it's very hard to predict, and hard to get good results for  
> things that are more complex than square sheets falling on spheres.  
> I've seen good stuff in Jens' test files, so I know the core system  
> can do a lot - it's just very hard  for people like me, who don't have  
> a deep understanding of the code behind it, to get it working nicely.  
> So far the buttons seem to have been tacked on incrementally,  
> reflecting changes in the code along the way (and the baking popup..  
> ouch..), without a general plan/design for interaction and workflow.  
> It would be good if the interaction here could be properly analysed,  
> designed, considered and re-considered.

I've really tried to improve on that as well. The issue really is that  
the implementation choosen actually *is* scientific, with no obvious  
hooks to present it in a way that makes it predictable for users who  
don't understand the values. The only idea I have now is replacing  
"Spring constant" with "Stiffness".
(Apart from, later, creating presets like "Cloth" or "Yello" or "Skin"  
of course).

The main reference we had was the way Maya presents physics to users.  
There you can also find a menu with "Solver methods" providing "Runge  
Kutta", or settings like "Collision tolerance" or the concept of  
"Goal". Their manual states about this that "the best way to experiment  
is to create a particle grid and play with the settings".

By providing in Blender simulation code based on the state-of-the art  
(Jens Ole used as reference the *standard* papers on this topic from  
Mr. PhysicsGod Baraff himself, now employed by Pixar) we at least offer  
in Blender advanced animation options any "professional 3d application"  
would go for.

My suggestion would be that in time we might find ideas how to present  
such a system better for users. The fact our competitors don't really  
do that at least tells me it's not an easy task... if you want to play  
with power, you'll have to become powerful yourself too. :)


-Ton-



>
>> --------> Post-production effect: ZBlur
>>
>> New:
>> - it's the zblur sequence plugin, tied to render output itself
>> - functionality highly limited... antialias & blur errors by   
>> implementation
>
> I have significant misgivings about this and will write another email.
>
>> --------> New .B.blend
>>
>> Todo:
>> - general re-check of standard color conventions (Matt?)
>
> I would say if there are larger UI changes afoot (2.4) then it would  
> be much better to leave any default theme changes until then and do it  
> all in the one go, to isolate disruption and also for the purposes of  
> the 2.3 manual being valid for the 2.3 series.
>
> Cheers,
>
> Matt
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at projects.blender.org
> http://projects.blender.org/mailman/listinfo/bf-committers
>
>
------------------------------------------------------------------------ 
--
Ton Roosendaal  Blender Foundation ton at blender.org  
http://www.blender.org