That’s right. And in fact, I think this would be a GREAT opportunity to apply what has been learned from Blender’s shortcomings, and what’s necessary if the project is ever intended to generate revenue. Even if revenue isn’t in the picture, there are certain design principles that will make the software a friendlier, more useful application.
why not put your effort on dynamica or nurbana? at least dynamica programmers were (dont know if still) were asking people to contribute to their work
I’d love do that 
but we talking about blender, if blender becomes into a frozen proyect then this other projects will need another base-software to go on expanding posibilities to the people, in other way the users of this parts will be limited by a software without development and then them will be limited too 
Mhhhh!
One of the open source strenghts is the idea of not reinvent the wheel, there are many open source programs that can model, animate and render, then why you want to do this again?
One of the open source strenghts is the idea of not reinvent the wheel
This has never been true, not even in the early days. Most interesting projects have several implementations active at the same time. For example you note: “there are many open source programs that can model, animate and render,”
I don’t think that what is being discussed is to re-invent the wheel, but to re-implement our favorite wheel. Maybe polishing in some of the corners along the way. (If we have to.)
enjoy
scott
would be cool…
I agree with vidigiani that there should be
a core team and plugin-developers.
Of course it should have nearly the same interface like blender. And the developers shouldnt do the same mistakes like most
opensource programs: look on interface design and usability!
To the renderer: we first should use a
renderman interface (many renderers available) and then maybe ask the developers of one
renderer to join the new-blender group…
would be cool!
cya henrik
i’d like it so the user interface is familiar, yet gives access to newer features, but the guts needs to be modernised to take advantage of available hardware better and allow extension more quickly and easily. having a kernel and plug in model for this would certainly be a good way forward. (i’d also like Ruby scripting in addition to Python . . . hehe)
in the manual i think ton mentions that Blender is a struct visualiser (ho hum). if this project gets anywhere near going forward i’d start thinking about prototyping some of it in Ruby (proper OO) like the whole materials, mesh etc. maybe as far as resolving a scene graph. before that could happen though i think some discussion about the good and the bad of blenders data structure should occur. personally i feel that recent things like the armatures (whilst great) don’t integrate into the current system that well. maybe i’m missing the point, the kernel should provide a place for the structures defined by plugins . . . .
anyone?
ok i’ve got one
what i’d really like is to be able to attach procedural modifiers to meshes
imagine being able to say take the mesh of that, intersected with the mesh of that and then make a catmull clark surface of it.
I’d like to have a good modeler for VirtuaLight. Somebody should start to develop it … 
I say we wait til Ton posts something, and then we decide what to do. If what we get back isn’t good news, does someone here want to set up a website at sourceforge.com? (it’s exactly what sourceforge is for - lots of webspace, lots of stuff which is great for net projects eg. revision control, bug report posting - the works).
Someone wanna volunteer?
LethalSideParting
One of the open source strenghts is the idea of not reinvent the wheel, there are many open source programs that can model, animate and render, then why you want to do this again?
Because no one want’s a square wheel. I have yet to see an opensource app take an even halfway decent stab at animation, especially character animation for instance. It wouldn’t be much help to yank code from those opensource apps and bunch them all togather to make one frankenstein app either since the only thing that would be relevant to your finished design are fundamental concepts that if you understood them properly, would want to code yourself.
But thats just my opinion. Take it or leave it.
This is not strictly relevent but may be useful non the less - I have been looking at options to create my own script-able player for 3D models. It’s for my own use and, if I ever get around to doing it for proper, not intended for anyone else to use. I mention this only to save people firing critism. Anyway, my latest selection includes XBasic as the development language, it’s completely free (even for commercial use), requires less brain work than C/C++ and you can dl the source code (allowing to streamline the language for a particular use if need be). It has a flexible array/tree implementation but no direct way of storing lists of varying structures and also no OOP features (you can easily allocate, read and write to memory so it’s not totally impossible). This and other reasons led me to search about for scene-graph managers and I came across OpenRM, again free and with source code. I have yet to try the combination. For scripting I found LUA (also free) which already seems to be used by a lot of projects very successfully. As far as graphics go, XBasic has no direct support for 3D but it’s an easy language to extend and someone has already wrote a wrapper to OpenGL. I have already taken this wrapper library, modified it slightly and converted around 10/11 of NeHe’s OpenGL tutorials. The results of the last part are comparable with C/C++ programs and in some cases smother and/or faster, impressive for a compiled BASIC. Hope this provides some pointers and inspiration if anyone decides to go ahead with the project - I’m all for it.