POV-Ray : Newsgroups : povray.programming : Vector arithmetic using .t problems Server Time
23 Dec 2024 06:56:10 EST (-0500)
  Vector arithmetic using .t problems (Message 1 to 3 of 3)  
From: Massimo Valentini
Subject: Vector arithmetic using .t problems
Date: 20 Dec 2002 14:01:02
Message: <3e0368ee$1@news.povray.org>
Bugreport taken from:

http://www.povray.org/download/3.5-bugs.php

in particularly

http://news.povray.org/3c4bb31d$1@news.povray.org

This bugreport is not marked as a job, I thought it was because 
already fixed for the next release, but I see now it's in the 
bottom half of the list. So here's a note:

The problem is that an expression like:

#declare vvv = <1, 0, 0> +1*t;

results in a 4D vector vvv = <1,0,0,0>

this is because 1 is firstly promoted to a 3D vector (<1,1,1>)
as a right hand side of the binary +, and then it is promoted
to a 4D vector (<1,1,1,0>) as a left hand side of the * operator.
Multiplying it by t (<0,0,0,1>) results in a 4D <0,0,0,0> that 
added to <1,0,0> only promotes it to <1,0,0,0>.

Here's a patch that shows a possible solution, but I'm not
certain it's fully correct. (The patch is against express.cpp
of the linux sources version 3.50a)

HTH Massimo

--- src/express.cpp.ex 2002-12-20 19:27:58.000000000 +0000
+++ src/express.cpp 2002-12-20 19:28:15.000000000 +0000
@@ -1582,7 +1582,7 @@
 
    Parse_Num_Term(Express,Terms);
 
-   Local_Terms=*Terms;
+   Local_Terms=1;
 
    EXPECT
      CASE (PLUS_TOKEN)


Post a reply to this message

From: Thorsten Froehlich
Subject: Re: Vector arithmetic using .t problems
Date: 20 Dec 2002 17:06:14
Message: <3e039456@news.povray.org>
In article <3e0368ee$1@news.povray.org> , "Massimo Valentini" 
<six### [at] inwindit> wrote:

> Bugreport taken from:
>
> http://www.povray.org/download/3.5-bugs.php
>
> in particularly
>
> http://news.povray.org/3c4bb31d$1@news.povray.org
>
> This bugreport is not marked as a job, I thought it was because
> already fixed for the next release, but I see now it's in the
> bottom half of the list.

No, it doesn't have a job number iirc.  Here is the current state of most
bugs listed on that page:


job000038

 Poly Object Bug
 
 Ron Parker says:
 This one is unlikely to be fixed soon, unless somebody sees something I
 didn't.  It looks like a root-finding inaccuracy of some kind.
 
 Suspended by Thorsten
 
 -----
 
  from       povray.bugreports
  date       Wed, 10 Nov 1999 05:50:04 -0800
  posted by  Ken <tyl### [at] pacbellnet>
  message    <382a3063@news.povray.org>
 
       This bug concerns the poly object using POV-Ray Windows v3.1g.watcom
   on a Pentium 200 Win98 platform. The error manifests itself as a black
   pixelated line on the objects surface on the same plane and level as the
   camera sits. The strum keyword does not appear to have any influence on
   the problem.
 
      It should be noted however that the same object does render correctly
   using POV-Ray v2.2 for DOS and in POV-Ray for Windows v3.01.watcom.win32
   either with or without the sturm keyword. This suggests a strong
possibility
   that it is indeed a version related bug.
 
      This same problem can be seen with the Twin_Glob 6th order poly object
   from the shapesq.inc file that is included with the distribution of
POV-Ray.
 
 
      This bug has been verified by Ron Parker.
 
 
   // Example of problem
 
     camera { location <0,0,-10> look_at 0 angle 25 }
 
     light_source { <0,1,-100> color rgb 1 }
 
     #declare Heart =
     poly { 6,<
     8.0,0,0,0,12,0,0,12,0,-12,0,0,0,0,0,0,0,0,0,0,6,0,0,12,0,-12,0,0,0,0,6,
    -0.1,-12,0,6,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,3,0,-3,0,0,
     0.0,0,3,-1,-6,0,3,0,0,0,0,0,0,1,0,-3,0,3,0,-1>
     }
     // sturm }
 
     object { Heart rotate <-90,90,0> pigment { red 1 } }
 
 
  RLP, 10 Jan 2001:  The bug does indeed exist, but it is not in the camera
plane. It is in
  the y=0 plane, as can be seen by moving the camera upward while focusing
on the same point.
  Whatever it is, it sure isn't obvious.

job000093 on 2001/03/11 by thorsten *suspended*

 Implement 16 bit image map reading for PNG and/or Targa image files.

job000137 on 2002/03/29 by thorsten *suspended*

 Highlight bug?
 See article <3b2fdb5b@news.povray.org> by "Rune"
<run### [at] mobilixnetdk>
 Highlights, when used together with normals, are visible in places where
 they should be in shadow. The same problem once existed for reflection, but
 it was fixed so reflected rays couldn't go inside the object. The same fix
 need to be applied to highlights as well.
 
 In the code below if you change Sign to be 1, the light_source is behind
the
 plane, and yet highlights are still visible. Note that because the
 reflection is fixed, there are no incorrect reflections of the sphere even
 though it is located at the same location as the light_source.
 
 #declare Sign = -1;
 light_source {Sign*100*z, color red 1}
 sphere {Sign*100*z, 20 pigment {color green 10} no_shadow}
 camera {location -5*z}
 background {color rgb 0.1}
 plane {
    -z, 0
    pigment {color <0,0,0>}
    normal {bumps 3}
    finish {phong 2 reflection 1}
 }
 
 The bug exists in POV-Ray 3.1, MegaPov 0.7 and POV-Ray 3.5 pre-beta 10.
 
 Also reported in http://news.povray.org/povray.beta-test/18760/

job000177

 (feature request) radiosity sampling, with fix method
 Report:
 http://news.povray.org/povray.beta-test/19143/
 Note: According to Nathan the effect is "at least somewhat desired". There
might
 also be problems with the proposed solution, so I marked it as feature
request
 and suspended the job because radiosity is experimental.

job000207

 radiosity doesn't take Double-illuminate into account
 
 Reported:
 http://news.povray.org/povray.beta-test/20319/ and
 http://news.povray.org/povray.beta-test/20102/

job000226

 "Camera inside non-hollow object" reported in some light_groups
 
 Reported:
 http://news.povray.org/povray.beta-test/20803/137765/

job000230

  Using media and fog has strange effects (incorrect colors)
 
  Reported in:
  media & fog bug?
  http://news.povray.org/3c1379e0$1@news.povray.org
  Another Media+fog problem report with unclear bug status:
  Funky fog vs. media
  http://news.povray.org/3c642f94$1@news.povray.org





____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org


Post a reply to this message

From: Apache
Subject: Re: Vector arithmetic using .t problems
Date: 20 Dec 2002 19:05:05
Message: <3e03b031@news.povray.org>
I'd love to see radiosity take double illuminate into account! Especially
with cloth-tents with radiosity it would be too cool to be true!  :-)


Post a reply to this message

Copyright 2003-2023 Persistence of Vision Raytracer Pty. Ltd.