[x3d-public] X3D minutes, 17 JAN 2020: X3Dv4 progress, rename hidden as visible, X3DVisibleObject same as X3DBoundedObject?

[Andreas Plesch]

It makes a lot of sense to use X3DBoundedObject for a visible field.
It turns out that x3dom always used X3DBoundedObject for its render
field:

https://github.com/x3dom/x3dom/blob/master/src/nodes/Grouping/X3DBoundedObject.js#L40

x3dom  followed InstantReality:

https://doc.instantreality.org/documentation/nodetype/BoundedNode/

I just did not mention it before to avoid what I thought may be a
distraction from the main goal.

-Andreas

On Sat, Jan 18, 2020 at 5:37 AM Don Brutzman <brutzman at nps.edu> wrote:
>
> Good point about collision.  We discussed this during call, I pursued it further.  More to follow below...
>
> On 1/17/2020 12:15 PM, Andreas Plesch wrote:
> >> 4.a. Mantis 1271: add hidden field to Shape
> >>        https://www.web3d.org/member-only/mantis/view.php?id=1271
> >>
> >> Discussion:  see email thread from last week's minutes and the Mantis issue itself.
> >>
> >> Pending today: rename to "visible"?  Note change to CADAssembly.  Also need example scene.
> >>
> >> TODO: example models for testing implementations.
> >>
> >> Related: whether this has impact on X3DPickableNode.  Our immediate reaction is to confirm that we have formal prose that non-visible geometry is not pickable.
> > Similar prose may be needed for collision detection, for example in walk mode.
> >
> >> We agreed to proceed in this manner.  Once accomplished, this bug is left open that final implementation/evaluation in 2 browsers is needed.
>
> Implementing the object model in schema (and eventually X3DUOM) is very revealing because relationships are forced to work or fail, thus design patterns can be confirmed with high confidence.
>
> The more I work on this issue, the more nodes seemed appropriate.  For example CADFace, NurbsSet have renderable geometry.  Also ReceiverPdu, SignalPdu and TransmitterPdu have whichGeometry field for renderability. So they should have visible field too. That brings us up to 30 nodes (and still may find more).
>
> Finally noticed a pattern:  these all appear to match X3DBoundedObject, which makes sense since that interface includes bounding boxes for renderable content.  So it looks like we should not add an X3DVisibleObject interface and instead just use X3DBoundedObject for the visible field.
>
> So this appears to be a great simplification, both in object model and in how implementations can add the visible field to X3D.  8)
>
> Any counterexamples?  If none, will try to implement visible field as part of X3DBoundedObject and then confirm consistency.
>
> all the best, Don
> --
> Don Brutzman  Naval Postgraduate School, Code USW/Br       brutzman at nps.edu
> Watkins 270,  MOVES Institute, Monterey CA 93943-5000 USA   +1.831.656.2149
> X3D graphics, virtual worlds, navy robotics http://faculty.nps.edu/brutzman



-- 
Andreas Plesch
Waltham, MA 02453