plib-devel

[Plib-devel] Releasing PLIB 1.2 *soon*.

[Steve B. <sjb...@ai...>]

Well, the tornado(s) that took out downtown FortWorth earlier this
evening passed about four miles to the North of my home - we got
away without any damage at all - pretty awesome stuff though.

I wanted to mention that I'd like to release PLIB 1.2 (a stable
release) sometime soon.  I have a couple of mods of my own to
put in - if anyone else needs to commit anything or knows of
any significant problems that need fixing - please let me know
so that we can arrange to catch all of these prior to release.

Pretty much as soon as PLIB 1.2 is out, I'll make a 1.3
(unstable/development) release for ongoing changes.

The only changes to 1.2 after release will be bug fixes.

The things I *know* I have to do for 1.2 are all pretty
minor:

* Add Dave McClurgs' Windoze/MSVC project files.
* If I have time, I'll try to do something about the
  GL_COLOR_MATERIAL problem that Curt has been concerned
  about...but I don't have a lot of time - so we may be
  out of luck right now for 1.2.0 - that doesn't preclude
  a fix in 1.2.1 if we find that problem after 1.2.0 the
  release.
* Loring Holden's fix to enable things to compile on
  Solaris and AIX.
* Darrell Walisser's fixes for CodeWarrior on MacOS.

Ah - the Joys of Portability!   :-)

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
sj...@ht...      (work)

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[Per L. <li...@ho...>]

Steve Baker wrote:
> I wanted to mention that I'd like to release PLIB 1.2 (a stable
> release) sometime soon.  I have a couple of mods of my own to
> put in - if anyone else needs to commit anything or knows of
> any significant problems that need fixing - please let me know
> so that we can arrange to catch all of these prior to release.

Well, the 3ds-stuff is still broken when it comes to textures, but
currently I can't work on it since my system refuses to display textures
in *any* OpenGL program - I'm really starting to hate this nvidia/glx
stuff. I will try to fix it as soon as I get my system working, but if
anyone else want to look at it before the release, go ahead.

(If somebody has any clue to why I get a GLXBadRenderRequest in
GLXRenderLarge, even after reinstalling my old XF86_SVGA, glx.so and
libGL.so.1.0, which seemed to work fine before, would you please mail
me? I should have known better than installing new untested drivers...)

Best regards,
Per Liedman

-- 
===========================================================================
Per Liedman                                       
md...@md...
Nilssonsberg 7                       
http://www.mdstud.chalmers.se/~md6pl/
411 43 Göteborg                               tel: 031-825659 /
0705-520455
===========================================================================

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[Steve B. <sjb...@ai...>]

Per Liedman wrote:

> Well, the 3ds-stuff is still broken when it comes to textures, but
> currently I can't work on it since my system refuses to display textures
> in *any* OpenGL program - I'm really starting to hate this nvidia/glx
> stuff. I will try to fix it as soon as I get my system working, but if
> anyone else want to look at it before the release, go ahead.
> 
> (If somebody has any clue to why I get a GLXBadRenderRequest in
> GLXRenderLarge, even after reinstalling my old XF86_SVGA, glx.so and
> libGL.so.1.0, which seemed to work fine before, would you please mail
> me? I should have known better than installing new untested drivers...)

I have no clue about *that* - if I were you, I'd restore the X-win version
of Mesa rather than messing with the SVGA version - which isn't exactly
well maintained these days.

I did get this encouraging message about nVidia drivers a couple of
weeks ago.  Dunno if it's of any help:

> ---------- Forwarded message ----------
> Date: Tue, 7 Mar 2000 21:58:18 -0000
> From: Allen <al...@cl...>
> Reply-To: me...@iq...
> To: me...@iq...
> Subject: [mesa] GeForce256 X GLX
> 
> For all those who have tried to get this to work (both the Nvidia binaries
> and building from source), you can try the version I built, it does work,
> but as they said it ain't about to set any new speed records. It was built
> on a RH6.1 system. Give it a shot. I would like to know if it works with the
> Riva128,TNT and TNT2. I think it should.
> 
> www.clarke-aj.freeserve.co.uk
> 
> Allen.

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
sj...@ht...      (work)

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[Wolfram K. <w_...@rz...>]

Steve Baker <sjb...@ai...> wrote:

>- we got
>away without any damage at all - 

Good!

>The things I *know* I have to do for 1.2 are all pretty
>minor:
>
>* Add Dave McClurgs' Windoze/MSVC project files.

I am a bit confused about this..
The MSVC project files in CVS work well.
You will base 1.2 on what is in CVS, wont you?
Daves newest additions to the CVS repository regarding DXF-files are
cool.


>* If I have time, I'll try to do something about the
>  GL_COLOR_MATERIAL problem that Curt has been concerned
>  about...

This is the same problem that PPE has sometimes with the grid looking
shaded/lighted, is it not?
On my home system with new Voodoo-3-drivers this problem is
reproducable. It depends on the mode you are in, wireframe or filled.

Bye bye,
Wolfram.

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[Steve B. <sjb...@ai...>]

Wolfram Kuss wrote:
> 
> Steve Baker <sjb...@ai...> wrote:
> 
> >- we got
> >away without any damage at all -
> 
> Good!
> 
> >The things I *know* I have to do for 1.2 are all pretty
> >minor:
> >
> >* Add Dave McClurgs' Windoze/MSVC project files.
> 
> I am a bit confused about this..
> The MSVC project files in CVS work well.

Oh?  Now I'm puzzled.  Maybe Dave checked in those files himself.

I'll check *carefully* before I mess with it.

> You will base 1.2 on what is in CVS, wont you?

Absolutely...but since it's a CVS and can potentially be changed
right up to the last 10 seconds before I make an official 1.2
release out of it - I need to be sure that everyone can agree on
a short-term "feature" freeze while the last few wrinkles are
ironed out.

> Daves newest additions to the CVS repository regarding DXF-files are
> cool.

Yes - I was using it to view the plans of my house in PPE!
(Our architect kindly provided the CAD files)  I need to get
PPE's extrude code working so I can turn them into a 3D model :-)

> >* If I have time, I'll try to do something about the
> >  GL_COLOR_MATERIAL problem that Curt has been concerned
> >  about...
> 
> This is the same problem that PPE has sometimes with the grid looking
> shaded/lighted, is it not?
> On my home system with new Voodoo-3-drivers this problem is
> reproducable. It depends on the mode you are in, wireframe or filled.

Yes - it shows up in a few places in my Tux game too.

I think these are all symptoms of the same problem.

The problem is that all these cases are too complex to track down
what's going on.  I really need to write some stand-alone OpenGL
programs that do essentially what I think SSG is doing and see
if those work.  That'll help me to see if I have a misunderstanding
of the OpenGL Spec, or if one or more implementations are broken,
or whether there is a simple bug in SSG such that it isn't doing
what I think it is.

The glColorMaterial API isn't all that well described and quite a
few of the OpenGL Guru's out there cannot agree on what it's *supposed*
to do under some of the situations that can arise in SSG.  If the
guys who wrote the OpenGL spec can't agree then there is little chance
of all the implementations out there getting it right.

As always, if Quake doesn't use it - all bets are off.

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
sj...@ht...      (work)

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[<Va...@t-...>]

Steve Baker wrote:
> 
> I wanted to mention that I'd like to release PLIB 1.2 (a stable
> release) sometime soon.  I have a couple of mods of my own to
> put in - if anyone else needs to commit anything or knows of
> any significant problems that need fixing - please let me know
> so that we can arrange to catch all of these prior to release.

I'd like to add a bit of functionality to SG. I think I write it up in
no-time (but not compile it as my box is in major reordering at the
moment). Is it still possible that it makes it to 1.2?

It'd be 

void sgProjectOnVector3           ( sgVec3 dst, sgVec3 src1, sgVec3 src3
);
void sgProjectOnNormalizedVector3 ( sgVec3 dst, sgVec3 src1, sgVec3 src3
);

that projects one vector on the other and

void sgProjectOnNormal3           ( sgVec3 dst, sgVec3 src1, sgVec3 src3
);
void sgProjectOnNormalizedNormal3 ( sgVec3 dst, sgVec3 src1, sgVec3 src3
);

that maps the vector onto a plane.

I can also supply the functions for 'sgd' and perhaps other dimensions
(if that makes sense...). If the names could be differenet if that makes
more sense...

CU,
Christian

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[Steve B. <sjb...@ai...>]

Christian Mayer wrote:
> 
> Steve Baker wrote:
> >
> > I wanted to mention that I'd like to release PLIB 1.2 (a stable
> > release) sometime soon.  I have a couple of mods of my own to
> > put in - if anyone else needs to commit anything or knows of
> > any significant problems that need fixing - please let me know
> > so that we can arrange to catch all of these prior to release.
> 
> I'd like to add a bit of functionality to SG. I think I write it up in
> no-time (but not compile it as my box is in major reordering at the
> moment). Is it still possible that it makes it to 1.2?
> 
> It'd be
> 
> void sgProjectOnVector3           ( sgVec3 dst, sgVec3 src1, sgVec3 src3
> );
> void sgProjectOnNormalizedVector3 ( sgVec3 dst, sgVec3 src1, sgVec3 src3
> );
> 
> that projects one vector on the other and
> 
> void sgProjectOnNormal3           ( sgVec3 dst, sgVec3 src1, sgVec3 src3
> );
> void sgProjectOnNormalizedNormal3 ( sgVec3 dst, sgVec3 src1, sgVec3 src3
> );
> 
> that maps the vector onto a plane.
> 
> I can also supply the functions for 'sgd' and perhaps other dimensions
> (if that makes sense...). If the names could be differenet if that makes
> more sense...

That sounds pretty non-contentious - I don't see a problem with
popping that stuff into the SG lib.

A not-normalized normal is a vector and a normalized vector is
a normal - so I find the names a bit confusing.

I'm not quite sure what you intend these to do but how about

sgProjectVec3OntoPlane
sgProjectNormal3OntoPlane

etc.

-- 
Steve Baker                  http://web2.airmail.net/sjbaker1
sjb...@ai... (home)  http://www.woodsoup.org/~sbaker
sj...@ht...      (work)

Re: [Plib-devel] Releasing PLIB 1.2 *soon*.

[<Va...@t-...>]

Steve Baker wrote:
> 
> A not-normalized normal is a vector and a normalized vector is
> a normal - so I find the names a bit confusing.

A normal is a vector that decribes the 'direction' of a plane. So if
  you have got a 4th number you could write the plane equation 
  (A x1 + B x2 + C x3 + D = 0 with A, B and C defining the normal)

A normalized vector is a unit vector (i.e. legth == 1.0).

This was that I thought of. But you are right it's confusing, it
should be e.g. sgProjectOnUnitVector3

> I'm not quite sure what you intend these to do

With this misunderstanding it's obvious that you don't know it...

> but how about
> 
> sgProjectVec3OntoPlane

this would be the sgProjectOnNormal3 as the normal is defining the plane

> sgProjectNormal3OntoPlane

this wouldn't make sense as the resuld would be a zero vector...

Yesterdy I've sent you the code (+ a patch a short time after I've sent
the code) privately. It comes with descriptions of the functions. But
please change the names to something less confusing (and update the
CVS as I don't have SSH). For the interested here are the descriptions:

// sgProjectOn[Normalized]VecX ( sgVecX dst, const sgVecX src1, const
sgVecX src2 )
// 
// calculate the projection of src2 onto src1:
//         _ 
//         /|
//  src2 /  :
//     /    :
//   /      :
//  -------->-------------->
//     dst      src1
//
//        src1 o src2
//  dst = ----------- * src1
//        ( src1 )^2


// sgProjectOnNormal3 ( sgVec3 dst, const sgVec3 src1, const sgVec3 src2
)
//
// calculate the projection of src2 onto a plane that is specified by
src1,
// its normal vector
//
// This works by calculating temp, a vector that lies on the plane and
has the
// direction of the projected vector and then do the same stuff as above
//
//                                      temp o src2
// temp = ( src1 x src2 ) x src1; dst = ----------- * temp
//                                      ( temp )^2

CU,
Christian