![](/i/fill.gif) |
![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Hi,
I previously [2000-06-26] reported the same bug for version 0.5a
that was crashing MegaPOV on some architectures under UNIX, but not others.
The same thing is still here in version 0.6a - actually I'm using the linux
Try rendering the simple scene below, alternatively commenting one
of the two boxes. The problem arrises from the 'transform cube_t' which
seems to be still buggy due to the Compose_Transforms() function under
CASE(TRANSFORM_TOKEN) in parse.c; may use an unassigned pointer.
//-----------------------------------------------------------------
#version unofficial MegaPOV 0.5;
camera{ right x*4/3 up y angle 40 location -z*50 look_at 0 }
light_source{ -z*50 rgb 0.66 }
light_source{ -z*50 rgb 1.0 rotate<45,-25,0> }
plane{ y, -5 pigment{checker rgb 1,rgb 0.7 scale 5} }
fog{ rgb 1 distance 200 fog_type 2 fog_offset -2 fog_alt 10 }
#declare cube_t = transform{
translate<-5,-5,-5>
rotate x*20
rotate y*20
}
box{
<0,0,0><10,10,10> pigment{ rgb<1,0,0> }
transform cube_t
}
box{
<0,0,0><10,10,10> pigment{ rgb<0,1,0> }
transform cube_t
translate x*5
}
//------------------------------------------------------------------
*** Nicolas Calimet
*** http://pov4grasp.free.fr
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> #version unofficial MegaPOV 0.5;
Well it should be 'MegaPOV 0.6' but there's no difference...
Can anyone tells if this also append under Win/MacOS/others ?
*** Nicolas Calimet
*** http://pov4grasp.free.fr
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Nicolas Calimet wrote:
>
> Hi,
>
> I previously [2000-06-26] reported the same bug for version 0.5a
> that was crashing MegaPOV on some architectures under UNIX, but not others.
> The same thing is still here in version 0.6a - actually I'm using the linux
mpov-0.6.1 is based on megapov 0.6. I just tried with mpov-0.6.2 which
is based on 0.6a and didn't see any problem (linux mandrake 7.1 using
the binary that's available from my site)
--
* Abandon the search for truth, * mailto:ber### [at] iname com
* Settle for a good fantasy. * http://www.enst.fr/~jberger
*********************************
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
> mpov-0.6.1 is based on megapov 0.6. I just tried with mpov-0.6.2 which
> is based on 0.6a
Gosh ! The hell with me !! ;-)
I don't know why I didn't downloaded your mpov update just after Nathan
announced version 0.6a... I unfortunately assumed that mpov-0.6.1 was already
based on the 0.6a sources ('m completely blind man).
Okay, so there's actually no bug anymore...
Sorry for bothering with good-old-damn-buggy code, and forget my posts !
*** Nicolas Calimet
*** http://pov4grasp.free.fr
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Sorry, but when again trying on a Digital Alpha architecture,
my cube example above crashes with floating exception. This is for
mpov-0.6.2 which is based on MegaPOV 0.6a.
Looking back into the source, the infamous Compose_Transform()
with an unassigned &Local_Trans is still there. It does not crash
Linux and SGI, but does for DEC. I cannot test other architectures,
including, in particular, Win or MacOS (maybe could with SunOS).
The weird image I obtained previously with the mpov-0.6.1 code
has been solved by another portion of code in parse.c; using a 'diff'
on files of the two distro, mpov-0.6.2/parse.c has some change from
the lines 9004 to 9007 in Parse_Transform() function. The problem I'm
talking about is at line 1081 in Parse_Object_Mods().
I hope this will be fixed in the next MegaPOV release, or I
will have to report it again ;-)
*** Nicolas Calimet
*** http://pov4grasp.free.fr
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
Nicolas Calimet <pov### [at] free fr> wrote:
> Sorry, but when again trying on a Digital Alpha architecture,
> my cube example above crashes with floating exception. This is for
> mpov-0.6.2 which is based on MegaPOV 0.6a.
>
> Looking back into the source, the infamous Compose_Transform()
> with an unassigned &Local_Trans is still there. It does not crash
> Linux and SGI, but does for DEC. I cannot test other architectures,
> including, in particular, Win or MacOS (maybe could with SunOS).
>
> The weird image I obtained previously with the mpov-0.6.1 code
> has been solved by another portion of code in parse.c; using a 'diff'
> on files of the two distro, mpov-0.6.2/parse.c has some change from
> the lines 9004 to 9007 in Parse_Transform() function. The problem I'm
> talking about is at line 1081 in Parse_Object_Mods().
This is indeed wrong and a bug. However, it doesn't crash the Mac
either.
Anyway, looking at the source, it seems to me that it has been a bug
even before the TransformPatch.
> I hope this will be fixed in the next MegaPOV release, or I
> will have to report it again ;-)
Well, I have fixed that now so if there is ever going to be a next MP it
will be fixed.
I guess I overlooked that specific problem, sorry for that!
Smellenbergh
--
e-mail:sme### [at] skynet be
http://users.skynet.be/smellenbergh
Post a reply to this message
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |
|
![](/i/fill.gif) |
| ![](/i/fill.gif) |
|
![](/i/fill.gif) |