Antonio Bleile wrote:
>>> How would I render a tree by just rendering the leaves?
>>> It can't work with travmasks, so I tried with some
>>> reimplementation of the RenderAction
>> I guess that is more or less the only option (you basically need to 
>> replace Action::recurse to omit the enter calls).
> 
> I thought about that as well, but it's not virtual...
> Or has that changed since 1.6? Allright I'll take a look...

No, no need to waste your time, it is not virtual. I meant in your 
reimplementation you need basically the same code with just that change.

>>> and also adding
>>> a custom registerEnter function, but I can't figure out
>>> how to skip the rendering of internal geometry nodes
>>> (without skipping their children!)
>> Sorry, this part remains a mystery to me. What are "internal geometry 
>> nodes" ?  And in what way does that affect children handling (unless a 
> 
> Internal geometry nodes are Geometry nodes with children :)

ah, ok.

>> core's callback calls action->useNodeList() the action defaults to 
>> visiting all children (callback returns Action::Continue) or none 
>> (callback returns Action::Skip or Aktion::Quit) ?
>>
>> Can you give a bit of background of what you are trying to do, maybe 
>> that allows us to come up with some ideas.
> 
> I just add bounding boxes of a geometry nodes as child of that
> node. I know there are 1000 better ways to do so, but I have a
> particular case here....

Oh, now I get it, big misunderstanding on my end. I thought you wanted a 
traversal that only calls the leave callbacks instead of enter and leave 
callbacks, sorry.

You can override an Action's callbacks on a per instance basis 
(registerEnterFunction, registerLeaveFunction), so you can create a 
regular RenderAction and register a different callback for geometry. 
That callback would basically do the same as the regular one, but only 
after testing that the node has no children.

        Hope it helps,
                Carsten


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users

Reply via email to