[vos-d] thoughts for 0.24

[Karsten Otto]

Am 28.04.2006 um 17:17 schrieb Reed Hedges:

>>> But that's just a second idea I had, not sure yet if it's good or  
>>> not.
>>>
>> [...explanation...]
>
> OK that makes sense, though I think it might be a rare vobject that  
> denies
> parenting requests? Anyway, we should come up with a test  
> application, like a
> game or something, and try things out.
>
 From the top of my head, I would say a vobject is justified in  
rejecting new parents when it is supposed to be unique. For example,  
imagine you just spent a fortnight designing your own individual  
avatar model. Now some *@#& comes by and says, "Cool, I want to look  
like this too". That person just links its own avatar as a parent to  
your "individual" model property, and nobody can distinguish you from  
this *@#& !

> [snip]
>
> CS can handle hierarchical scene graphs, but the VOS plugin for CS  
> doesn't yet.
> I started working on that (it's pretty much required for VRML to  
> work right),
> but haven't finished yet.
>
> One way to do it, once hierarchies are finished, will be to have  
> the blocks
> objcets and the pyramids objects be under two root "transform  
> groups" that can
> be included in one sector.  (Currently in my half done work, any  
> object3d can be a
> "transform group" but that can easily be changed to a special  
> "transform group"
> object.)  And nothing stops those transform group objects from  
> *also* being
> new sectors themselves :)   (So there would be three sectors, two  
> of which
> happen to just group contents for the third.)
>
Ok, that is an alternative - just creating a meta-world that  
*contains* the other two, then displaying this instead of one of the  
individual worlds. The meta-world can also be constructed locally in  
Ter'angreal when necessary, based on link information as I described.
Ok... I'll just wait for you to finish the hierachical csvosa3dl :-)

Regards,
Karsten Otto