gdalgorithms-list

[Algorithms] Bezier Triangle Rendering

[Dave S. <Dav...@sd...>]

I can compute points on a Bezier Triangle,
but I'm having a tough time coming up with
a good tessellator for it.  Are there any
references that has a demonstration of this?

I'm currently looking at fixed subdivision for 
the quadratic case, but because its not bi-quadratic 
in the square sense, like bezier patches, it's harder 
to come up with an easy algo for producing the parameter
space points.

Any ideas or pointers would be much appreciated,
Thanks,
-DaveS

ps.

Conor, that sure was a "wacky" patent. :-P

Re: [Algorithms] Bezier Triangle Rendering

[Angel P. <ju...@bi...>]

> I can compute points on a Bezier Triangle,
> but I'm having a tough time coming up with
> a good tessellator for it.  Are there any
> references that has a demonstration of this?

Here's a paper about bezier triangles:
http://www.dgp.toronto.edu/people/jcr/papers/thesis.ps.gz

You can find a program (with source ) that renders
triangular bezier patches at
http://www.witchlord.com/projects.asp

Re: [Algorithms] Bezier Triangle Rendering

[Dave S. <Dav...@sd...>]

Angel Popov wrote:
> 
> > I can compute points on a Bezier Triangle,
> > but I'm having a tough time coming up with
> > a good tessellator for it.  Are there any
> > references that has a demonstration of this?
> 
> Here's a paper about bezier triangles:
> http://www.dgp.toronto.edu/people/jcr/papers/thesis.ps.gz

This kinda stuff is great! Thanks Angel.
(I knew there was a deCastelJau derivation of Triangular
patches but I wasn't going to derive it. :-)

Dave Eberly wrote:
> 
> I have source code for tessellation by recursive subdivision
> of quadratic or cubic Bezier triangles, rectangles, or
> cylinders at http://www.magic-software.com/MgcSurface.html 
> 

Thanks anyway Dave.  Just FYI, I liked your old site
better since it included papers and references on it
as opposed to source with book purchase.  But I can't
dog you for trying to make a living.

As to Tom Forsyth' quick solution, do you have any
textured screen shots?  Just curious as to what 
that parameterization does to texture mapping.

Thanks all,
-DaveS

Re: [Algorithms] Bezier Triangle Rendering

[Dave E. <eb...@ma...>]

> Dave Eberly wrote:
> > 
> > I have source code for tessellation by recursive subdivision
> > of quadratic or cubic Bezier triangles, rectangles, or
> > cylinders at http://www.magic-software.com/MgcSurface.html 
> > 
> 
> Thanks anyway Dave.  Just FYI, I liked your old site
> better since it included papers and references on it
> as opposed to source with book purchase.  But I can't
> dog you for trying to make a living.

I just found out about this list group and my post to
your request was my first.  What a great introduction!
(sarcasm intended)

FYI, the only difference between my 'old' and 'new' site
is that there is more free stuff now.  The papers and
references are still available.  The link I mentioned is to
*free* source code.  Did you check the file preambles?

The book is a separate issue.  The source code that
comes with the book includes a lot of the free stuff at
my web site, but it also has code that may only be
used for non-commercial purposes.  That 'restricted'
code will not be available at my web site.

I am quite saddened that folks would think that
$60 for a meaty book with source code is too
expensive.  The money I make from the book pays
my bills so that I can continue to post *free* stuff at
my web site.  Let's hope enough people have a
different attitude than yours :)

--
Dave Eberly
eb...@ma...
http://www.magic-software.com

[Algorithms] $60 for code that's useful?

[Michael S. H. <mic...@ud...>]

Dave,

  While I can't yet say that I've made use of your free software (although I have checked out the various papers and such) I feel confident that *your* $60 book is likely to be well worth the price, give the number of times your site's been referred to on this list.

  As soon as it's available, I plan on picking it up.  Can't have enough good references on the shelf.

PS.  Ron, if you're reading this, get to work!  I want your book as well  :-)


At 01:45 PM 8/18/00, you wrote:
>> Dave Eberly wrote:
>> > 
>> > I have source code for tessellation by recursive subdivision
>> > of quadratic or cubic Bezier triangles, rectangles, or
>> > cylinders at http://www.magic-software.com/MgcSurface.html 
>> > 
>> 
>> Thanks anyway Dave.  Just FYI, I liked your old site
>> better since it included papers and references on it
>> as opposed to source with book purchase.  But I can't
>> dog you for trying to make a living.
>
>I just found out about this list group and my post to
>your request was my first.  What a great introduction!
>(sarcasm intended)
>
>FYI, the only difference between my 'old' and 'new' site
>is that there is more free stuff now.  The papers and
>references are still available.  The link I mentioned is to
>*free* source code.  Did you check the file preambles?
>
>The book is a separate issue.  The source code that
>comes with the book includes a lot of the free stuff at
>my web site, but it also has code that may only be
>used for non-commercial purposes.  That 'restricted'
>code will not be available at my web site.
>
>I am quite saddened that folks would think that
>$60 for a meaty book with source code is too
>expensive.  The money I make from the book pays
>my bills so that I can continue to post *free* stuff at
>my web site.  Let's hope enough people have a
>different attitude than yours :)
>
>--
>Dave Eberly
>eb...@ma...
>http://www.magic-software.com
>
>
>
>_______________________________________________
>GDAlgorithms-list mailing list
>GDA...@li...
>http://lists.sourceforge.net/mailman/listinfo/gdalgorithms-list 

Re: [Algorithms] Bezier Triangle Rendering

[Dave S. <Dav...@sd...>]

> FYI, the only difference between my 'old' and 'new' site
> is that there is more free stuff now.  The papers and
> references are still available.  The link I mentioned is to
> *free* source code.  Did you check the file preambles?
> 

All I know is that before, when I went to your site to 
look up a topic, it had a paper and some source.
Now, it seems that it's just source.  I saw your source
for this post.  I was looking for references.  I didn't 
see the references which prompted my, I'll admit
hasty, response.

> I am quite saddened that folks would think that
> $60 for a meaty book with source code is too
> expensive.  The money I make from the book pays
> my bills so that I can continue to post *free* stuff at
> my web site.  Let's hope enough people have a
> different attitude than yours :)
> 
>

I wasn't criticizing your $60 book.  I was critizing
the web site change.

That said, I didn't mean to offend you.  I just wanted
you to know my honest opinion.  I respect you as an
authority in this area and having another "master"
on this list is greatly needed.  

If it'll make you happy, I'll buy your book. :-)

Thanks,
-DaveS

Re: [Algorithms] Bezier Triangle Rendering

[Tom H. <to...@3d...>]

At 10:45 AM 8/18/2000, you wrote:
>I just found out about this list group and my post to
>your request was my first.  What a great introduction!
>(sarcasm intended)

Welcome to the list.

>FYI, the only difference between my 'old' and 'new' site
>is that there is more free stuff now.  The papers and
>references are still available.  The link I mentioned is to
>*free* source code.  Did you check the file preambles?
>
>The book is a separate issue.  The source code that
>comes with the book includes a lot of the free stuff at
>my web site, but it also has code that may only be
>used for non-commercial purposes.  That 'restricted'
>code will not be available at my web site.
>
>I am quite saddened that folks would think that
>$60 for a meaty book with source code is too
>expensive.  The money I make from the book pays
>my bills so that I can continue to post *free* stuff at
>my web site.  Let's hope enough people have a
>different attitude than yours :)

We just got spoiled :)

As for the book, I'll probably buy it when it ships, although I've never 
heard of anyone making enough money off a tech book to actually pay the 
bills :)

Tom

Re: [Algorithms] Bezier Triangle Rendering

[Dave E. <eb...@ma...>]

From: "Tom Hubina" <to...@3d...>
> Welcome to the list.

Thanks.

> As for the book, I'll probably buy it when it ships, although I've never 
> heard of anyone making enough money off a tech book to actually pay the 
> bills :)

The presales indicate that this might very well be
possible.  But I am conservative.  So there are a
couple more books in the works, one on geometry
tools for computer graphics (co-authored with Philip
Schneider at Disney Feature Interactive) and one on
game physics.  And I gave up my job at NDL to
concentrate on Magic Software and do some contract
work.  I believe all these cover the bills...

From: "Dave Smith" <Dav...@sd...>
> All I know is that before, when I went to your site to 
> look up a topic, it had a paper and some source.
> Now, it seems that it's just source.  I saw your source
> for this post.  I was looking for references.  

Apparently I need to work on ease of navigation at
my site.  All the documentation is easy to get to by
going directly to http://www.magic-software.com/Documentation

> I wasn't criticizing your $60 book.  I was critizing
> the web site change.

Sorry, I must have misread your previous comments,
in particular the part about 'dog you...'.

> Thanks anyway Dave.  Just FYI, I liked your old site
> better since it included papers and references on it
> as opposed to source with book purchase.  But I can't
> dog you for trying to make a living.

From: "Akbar A." <sye...@ea...>
> just a little curious, for some reason i can't wait :)
> but; did you write most of the api code with opengl or d3d?

The book focuses on scene graph management.
There is an abstract renderer class.  The code
includes a software renderer for MS Windows, an
OpenGL renderer for MS Windows, and a GLUT-based
renderer for MS Windows and Linux.  I'll admit that
the renderers are simple and could be greatly improved.
For example, there is no display list or compiled
vertex array code in it.  I avoided D3D because (1) I
wanted to minimize time spent on implementing a
renderer and (2) I wanted code that ran on Linux as
well as MS Windows.

If I get the time to write a D3D renderer, I'll include it
as an update at my web site.  As it is, the first update
(to be posted when the book ships in mid-September)
will have a lot of stuff in it that I did not have ready when
the CD-ROM was due at the end of July.

--
Dave Eberly
eb...@ma...
http://www.magic-software.com

Re: [Algorithms] Bezier Triangle Rendering

[Pierre T. <p.t...@wa...>]

> This kinda stuff is great! Thanks Angel.
> (I knew there was a deCastelJau derivation of Triangular
> patches but I wasn't going to derive it. :-)

DX7 exe + source:
http://www.witchlord.com/projects.asp


Pierre

Re: [Algorithms] Bezier Triangle Rendering

[Dave E. <eb...@ma...>]

----- Original Message ----- 
From: "Dave Smith" <Dav...@sd...>
To: <gda...@li...>
Sent: Wednesday, August 16, 2000 11:06 AM
Subject: [Algorithms] Bezier Triangle Rendering


> I can compute points on a Bezier Triangle,
> but I'm having a tough time coming up with
> a good tessellator for it.  Are there any
> references that has a demonstration of this?
> 
> I'm currently looking at fixed subdivision for 
> the quadratic case, but because its not bi-quadratic 
> in the square sense, like bezier patches, it's harder 
> to come up with an easy algo for producing the parameter
> space points.
> 
> Any ideas or pointers would be much appreciated,

I have source code for tessellation by recursive subdivision
of quadratic or cubic Bezier triangles, rectangles, or
cylinders at http://www.magic-software.com/MgcSurface.html 

--
Dave Eberly
eb...@ma...
http://www.magic-software.com