On 28.06.2007 03:58, Mario Lapin wrote: > Hi, > >>> * Adapt current map loader (including physics loader) to communicate >>> to PhysX API - then we continue to use blender2crystal and its great >>> interface (thanks Pablo) to create worlds with physics data; >>> * Use a different map loader cs plugin (for instance Collada) and >>> then we could inject physics data directly into PhysX, through its own >>> import feature. We would be free to use to tool of choice (at least among >>> the ones supporting Collada). > >> The current physics loader is an add-on - a plugin loaded at runtime. >> Via addons maps can be augmented with pretty much arbitrary data. Thus >> you could also use CS maps together with a custom loader for the >> physics-related data (be it Collada or some other format). > > Sorry, not sure I understand. Do you mean merging Collada physics data with > CS maps, and then having each loader to handle its own part? Like using > <addon><plugin>..., for inserting e.g. Collada data into CS maps?
Yes. (Although in practice it's probably more sensible to have the physics data in a separate file. But for the purpose of this discussion that's not that relevant.) > > One aspect that really concerns me is tooling for creating/exporting > contents to the engines (ie. CS and PhysX). > > Merging Collada physics data into CS maps doesn't sound at first a > straightforward solution, because there wouldn't be a ready tool for > exporting everything together, and the resulting format would be a mix of > standards. To accomplish that, modifying b2cs might be an option (not sure > if allowed, or even worth doing). Then we would be able to export everything >>from the same creation environment (Blender + b2cs) at once. CS map data > would be handled by current map loader and Collada physics data, exported by > a custom/modded exporter, would be handled by a PhysX custom loader. Though, if b2cs is extended to export some kind of physics, it could also be extended to write in CS' own format for physics data (ie the one used by the existing dynamics loader). > Another option would be to load world geometry, uvmapping, etc. and physics > data from Collada format through a more comprehensive custom loader, leaving > to CS map format only basic engine initialization or specific features (if > needed). I'm not sure if the work in progress on the Collada plugin by Scott > Johnson would accomplish to or simplify this task (the on-the-fly mode...). > What sounds interesting on this approach is to stick on a standard for 3D > data. From what I gather that should be doable (Scott, am I not mistaken?). However, from a performance point of view it would probably be better to use CS' native format for world data. > Last would be to use world files exactly as exported by b2cs (including > physics) and implement a custom physics loader, providing an abstraction > layer for instancing data compatible to PhysX. A drawback might be the > difficulty to explore and adjust exclusive PhysX features, but this can be > worked around by custom tags... I use to prefer 'standards' than 'customs' > but this might be the less painful way to obtain results. The current CS physics loader loads things through the iDynamics interface - so if you'd implement a PhysX iDynamics plugins, you'd get a loader "for free". But as you pointed out, the support for PhysX specifica would be probably not as good as some loader entirely tailored to that. -f.r.
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------- This SF.net email is sponsored by DB2 Express Download DB2 Express C - the FREE version of DB2 express and take control of your XML. No limits. Just data. Click to get it now. http://sourceforge.net/powerbar/db2/
_______________________________________________ Crystal-main mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/crystal-main Unsubscribe: mailto:[EMAIL PROTECTED]
