RE: [K3d-development] Re: Question

[Gregory J. <gj...@sh...>]

Agreed, and LOL yes I do remember. :)

However, it's not so much a matter of simply adding an object, that I
can tell. This would be as much an inherent part of the system as is the
Hierarchy view, in terms of the user experience, and I see the
procedural support existing at the same level as the Hierarchy, rather
than as a Module. This would involve adding a new tab to the Document
view in which the node network lives.  (Preferably I would have the
Document view window be dockable in the Camera view because I hate
desktop clutter, but that's just me and I can live with it until I get
around to making that change as a GUI option. ;)

Greg

> -----Original Message-----
> From: Timothy M. Shead [mailto:ts...@k-...]
> Sent: Wednesday, January 29, 2003 7:03 PM
> To: Gregory Junker
> Cc: k3d...@li...
> Subject: Re: [K3d-development] Re: Question
>=20
>=20
> Gregory Junker wrote:
> > From what I can tell support for procedural methods would=20
> fix this. You
> > would have a Lissa node that feeds another node that=20
> creates the mesh,
> > or does some form of manipulation on it, and still be able=20
> to edit the
> > two independently, without the destructive creation process=20
> that a lot
> > of operations in K-3D exhibit. The same thing happens, for=20
> example, with
> > the basic primitives; once the object is created it's a=20
> mesh, and there
> > is no way to, say, go back and change the radius.
> >=20
> > The more I think about it and the more I look into it, Tim,=20
> the more I
> > think that it's not such a radical departure to add support for
> > procedural modeling. K-3D seems awfully insistent on getting objects
> > into mesh structure as soon as possible, but that need can=20
> be met while
> > still giving the user a more abstract experience by keeping=20
> around the
> > notion of a sphere or cube or what have you. I have spent a=20
> lot of time
> > playing around with the Houdini Apprentice download and=20
> looking at the
> > Houdini videos, and it's becoming a lot clearer how a=20
> similar approach
> > might work in K-3D. It's a radical departure for the user,=20
> but from an
> > internal viewpoint it doesn't really affect even the=20
> current codebase.
>=20
> You're preaching to the choir, here - I was the optimistic one in=20
> response to Paul's original "Radical design thoughts" post, remember?
>=20
> I would further add that this kind of effort doesn't even=20
> have to wait=20
> for me or my blessing - just create a new type of object, call it=20
> "Procedural Mesh" or what have you, and go to town.  I would=20
> highly-recommend a look at "Andrew Glassner's Other Notebook" for=20
> inspiration.
>=20
> I plan to develop laths and update the existing tools without=20
> trying to=20
> accomplish the above, but if this turns-out to be easy and overtakes=20
> that development, great!  The way I see it, laths are going=20
> to provide a=20
> clean framework for moving geometry around whether we stick with the=20
> current way of doing things or not.
>=20
> Cheers,
> Tim
>=20
>=20
>=20