View Full Version : snap to curve tool

07-21-2003, 10:03 AM
would it not be handy with an "snap to curve" tool?

07-23-2003, 06:15 AM
I have under 90's used the Aladdin 4D from ADSPEC Programming in USA, on the Amiga.
In Aladdin 4D there was lopt of usefull stuff that LW lacks.
e Time Envelope.

There was even the amazing tool in Aladdin 4D "Align":

You could create a circle or a curve, then create 3D text, or any othe object for that matter, and finaly use the tool Align Objects with Curve.
much like in the DTP software with all functiones spacing, stright or inverted, etc.
Still today there is no such things in LW.

Etc, etc. All that existed in Aladdin 4D already in 1993, more or less 10 years a go.
Today year 2003 we still not have so simple yet powerfull tools/functions in LW.
Above that the Aladdin 4D costed about 1/10th of the price of LW.

When I started work with LW, and still today, I am a bit disapointed of lack of such small things making work more efficient and pleasent.

I hope that NT progammers/developers team will take all our reuquests into concideration and make the reality of them.

What about that.

07-23-2003, 06:20 AM
After reading all of your post, and while I agree that plenty of things could and should be improved with LW:

Why do you use it? After all, Aladdin seems to have all the features that you miss in LW, why use a tool that's not appropriate to your work?

Mike :confused:

07-23-2003, 07:29 AM
1. Aladdin is an Amiga based software, Amiga is not developed since 1996.

2. Aladdin has been sold to Nova Design, since then it has not been improved and updated. Therefore it out of date.
Even thought thoise fanatstic tools.

3. mainly I work with video production and since the Amiga was not longer developed I need to change to PC to get my job done.

4. I've got a good offer when buying the Aura and LW as a package. Both are compatible with each other.
And I need bothe 3D and 4D software.

5. I knew NT and LW since the LW was made first for Amiga back in 90's.
But at that time Aladdin was beter devloped han LW.

6. I am not such a jerk who buy stuff and is silently suffering.
When I think something is wrong I say it and demand improvement.
I am iproving my products to my customers on demand, thus I demand improvement from whom I buy stuff. I think that is fair.

7. I do not demand impossible things, it was already made once by someone else. Then it cant be so hard to do.
So I just try to get better stuff not only for me but for other user too.
This is how I see the others requests.

8. Wasn't that you requested the Align tool?
I wrote to you in hope that you will be glad that I support your request but you seem to be hostail instead, are you out of your mind or you just kiding me.

pawel c

07-23-2003, 07:41 AM
I didn't ask for the "snap to curve", that was miQlo.
After reading your third post praising Aladdin, I was just wondering why you switched. You could just as well nag Nova about porting to the PC :)
The grass is always greener on the other side, but telling people (or developers) that something is so easy to do because it has been done by others isn't really helpful. NT have theris priorites I guess, most of them from user feedback, but slagging them off for not implementing features that may be essential to you doesn't help anybody.
You know, plenty of stuff existed 10 years ago, or even 15 years ago, and not everything is in Lightwave, you can't really expect it to be either.

07-23-2003, 07:51 AM
"You know, plenty of stuff existed 10 years ago, or even 15 years ago, and not everything is in Lightwave, you can't really expect it to be either. "

maybe not everything but some stuff would be nice....some stuff considered as "industry standard"...snap to curves could be easily obtained with a core Nurbs modeling toolset :)

A Nurbs toolset is an EXTRA feature...so it is basically a matter of $$ in developpement and patents....no need to change anything in the poly/subdiv workflow...but instead of having crappy cardinal splines we could use Nurbs splines...notice that every package out there has a Nurbs implementation...why LW doesnt have it? are NT the only ones that cannot afford it?

by using some specific tools LW loses the chance to integrate a much larger market where standards are used (and in a certain way can be considered as a proprietary soft)...i really think that this will be a benefit for NT to really think about it but i'm afraid that this is more of a investement problem...somebody at NT is not fully convinced of the benefits of investing in a modeler overhaul (too expensive?) ...maybe i'm wrong...but that is the felling i have by the way things are evolving...

07-23-2003, 07:56 AM
Now that is something that I absolutely agree with. I'd love to see decent NURBS (allthough that would take NT at least two years to integrate with a full tool set).
As a start I wouldn't mind real patches in Modeler and Layout though, no freezing, attached to splines, variable densities.
Oh well...

07-23-2003, 08:06 AM

i have no idea of how long it would take to NT...since the time i'm asking for Nurbs (and that lots of reactionary LW users says "what for?") it will be already there believe me ....

i dont get your idea of real patches...what ya mean? real "bezier" patches or real "Nurbs patches" because now we have real "poly" patches....to have real Nurbs patches you will certainly have to implement Nurbs technology anyways and plug it to the Opengl outpout anyways....no? and then deal with poly display tesselation and the like...

now i'm not a techie so i dont know but as far as i know subdivision surfaces are based in an algorithm inspired by Nurbs and the way they are displayed in modeler reminds me of Nurbs patches display...i dont know what is inputing/outpouting in the Opengl but AFAIK when rendering there is a tesselation pass (subpatches to triangles) before image rendering so maybe is not that hard to implement Nurbs ...this said Nurbs display is probably a differnet thing since every point on surface can be picked....while in subdivison surfaces there is no way to pick a point inside the boundaries of a control poly...

07-23-2003, 08:12 AM
Well, we currently have "manual" patches. Create a spline cage, select the splines, hit <ctrl>-f , enter the subdivisions and it creates a polygon patch based on the selected splines.
I would like to see this workflow:
Create spline cage
Select splines, <ctrl>-f, enter subdivisions
Now Modeler creates a patch that is hooked up to the splines. You can select a spline segment and change the number of divisions (which should propagate through the mesh to avoid cracks). Or you could move vertices and splines around and the patches follow.
You could also take these (let's be stupid and call them "Patchigons" :D ) over to Layout and tesselate them there (allowing for a multiplier, a bit like the current SDS).
It won't be a full NURBS toolkit, since the basic splines are different, and there is no filleting etc. But it would at least evolve the current workflow with fairly little programming work involved (..compared to a full NURB toolkit).

07-23-2003, 08:19 AM
ok i see...well this could be easily doable if LW could use a dynamic geometry system with history where all the input/output realtions are kept like in errr Maya :D

if i make a square in Maya (Nurbs or poly output is the same) i can post edit the splines and update the surface and also change mesh density in U or V (i suppose that is what u mean by setting the subdivison U and V bein parallel or perpendicular in LW)


07-23-2003, 08:22 AM
Hi Sailor,
what would be needed within the SDK would be:
* Edges
* Dynamic Geometry (as you said)
A patch would then just be generated by looking at the edges (parts of splines, defined by a start and ending point) that form the patch, and reading out the UV values stored within the edges.


07-23-2003, 08:30 AM
Agree with you Lightwolf...but hey then we are soooo close to what Nurbs does....why dont put an extra efforst and have Nurbs and history ? :p

07-23-2003, 08:34 AM
Hi sailor,
:) Well, I'd prefer that too. But I think what I sketched in post would be much easier to implement on the current core of LW. A first step (they can always improve that later on, can't they :) ).