[Bf-committers] Re: proposal for new feature: DupliMats

Ed Halley ed at halley.cc
Thu Dec 8 03:43:57 CET 2005

[Sorry for the crosspost-- I don't know how much overlap there is 
between these list groups.]

Ed Halley wrote:
 > My draft proposal for DupliMats in detail:
 >   http://mediawiki.blender.org/index.php/Requests/DupliMats

After I proposed this feature on the bf-funboard list, I decided to 
learn my way around the Blender codebase.  I've actually gotten 80% of 
the solution coded, and need about 10% more to be "feature complete" as 
I envisioned it.  The rest is testing and refinement and performance 
tuning, though it seems quite usable now (except for long animations).

A test render is also posted onto the wiki now:

Because this is a moderately-sized core feature with a lot of potential 
uses, I would like to have a discussion with some of the core coders 
about the methodology and the remaining concerns before handing it over. 
  My biggest fear is that I'm somehow making unseen breakage in the 
struct ID chains or something equally bone-headed.

Also worth discussing is whether anyone objects to the general concept, 
the feature's limitations according to this implementation, whether it 
will conflict with other planned concepts, or even the term 'DupliMats.' 
  I really want this kind of feature in Blender, but I want others to 
see the value in it also, so my patches are ultimately adopted mainstream.

This was quite a tour of the codebase, to reassure myself that makesdna 
works as advertised, to try three or four times to come up with a 
hashing method that works in animations and isn't a complete hack job, 
and to develop the user interface to something even simpler than 
proposed.  I look forward to further refining the feature with the team.

(I know this isn't meant for 2.40 which is nearing a roll-out.  I had 
always considered this for 2.41 or 2.42 or whatever the next release 
would be called.)

[ e d @ h a l l e y . c c ]

More information about the Bf-committers mailing list