[WF-Protocols] Summary of ideas from wallaba's first proposal - meshes only

[Aglanor]

On Sun, 2003-07-06 at 03:54, Eason Choo wrote:

> - I'm thinking screen-centric, i think thats more common..

	This is also what was stated on IRC. If noone disagrees, then the
format will be screen-centric.

> - I think we shoudl split the file up like the cal3d format. I.e. have a 
> mesh file, a material file, a skeleton file and a skeleton animation 
> file.

	Agreed. As a possibility, we could also zip the lot like OpenOffice
does, as Peter Amstutz has suggested.

> I also think we should leave LOD out, unless of course the automaticlly 
> generated LOD models are going to be supported. If its manual LOD where 
> a new mesh is assigned, i think that should have nothing to do with this 
> format, not even linkages in the model to another model representing a 
> lower/higher LOD state. 

> Also, I think that the way manual LOD like this is handled by each 
> engine is different, as in how many levels, when to swap the levels, etc.

	Well, perhaps it would be convenient to have all the meshes for the
different LODs for the model in a single zip file so you could get it
whole. I don't know in which units is LOD measured, distance from
camera? percentile? But perhaps it could be useful to define the
"standard" meshes for the model, then a mesh for when the LOD is 50%,
then another for when it's 25% etc.

	The trick here is, as you say, to know if all engines that use manual
LOD use it the same way (i.e. percentiles all), then it might be useful
including it in the format. If they do it quite differently then it is
best out.

> If we do decide to provide support for manual LOD, I think it should 
> handled as multiple meshes in the same model chunk.

	Agreed.

> 
> We still need to decied whether to allow more than one model/mesh in the 
> same model file.
> 
> I actually don't like the idea of multiple meshes per model anymore. I 
> dont' know why I even proposed it. Unless of course its for different 
> LOD levels.

	Well, theoretically it's for swapping several parts of the model. I.e.
you have a human model with a set of torsos, heads, hats, etc. that can
be grouped in different combinations while sharing a similar look (and
the same skeleton).

	I think multiple meshes for swapping and for LOD can be included at the
same time under the model chunk. In the simplest case a model would have
only 1 mesh (no swapping, no LOD), in the most complex case a model
could have N swappable parts each with M levels of LOD with a total of
N*M mesh chucks. As long as they are well identified, I don't think this
would bear technical problems for storing it. 

	Another issue is whether we desire to include swappable meshes in the
model format or not, I suppose this depends more on what use will the
engines make of them (i.e. if they will support them). Opinions?

	Regards,
	
	Aglanor