POV 3.5 b 13 icl on WinNT Sp 6 PII 233 with 128 MB

Below script:

  #local g=function(_){ceil(_)-_};
  plane{-z,-5 pigment{function{ g(atan2(x,y))}} finish{ambient 1}}

returns:

  Rendering Error: Stack full. Possible infinite recursive function call.

but when I inline function and write script this way:

  plane{-z,-5 pigment{function{ ceil(atan2(x,y))-atan2(x,y) }} finish{ambient 1}}

then it is rendered fine. Any confirmation ?
As extended testing example use my signature :-)
Btw: Am I only active beta-tester or just lucky one ?

ABX
--
disc{z,-z 5#macro O()asc(substr("-+((1*(,1,/.-,*/(,&.323/'1"e,1))*.1-4#declare
e=e-1;#end#local e=26;pigment{#local g=function(_){ceil(_)-_}function#local//X
k=function{pattern{object{sphere_sweep{linear_spline 13#while(e>0)<O(),O()//AB
>.01#end}}}}{k(g(atan2(x,y)),g(ln((y+x)^2+1e-5)),0)}}finish{ambient 1}}//POV35

Post a reply to this message

"W?odzimierz ABX Skiba" <abx### [at] babilonorg> wrote in message
news:at849uk68rnqhu92mhbkng6ho01tl23178@4ax.com...
> POV 3.5 b 13 icl on WinNT Sp 6 PII 233 with 128 MB
>   #local g=function(_){ceil(_)-_};
>   plane{-z,-5 pigment{function{ g(atan2(x,y))}} finish{ambient 1}}
> returns:
>   Rendering Error: Stack full. Possible infinite recursive function call.
> but when I inline function and write script this way:
>   plane{-z,-5 pigment{function{ ceil(atan2(x,y))-atan2(x,y) }}
finish{ambient 1}}
> then it is rendered fine. Any confirmation ?
> As extended testing example use my signature :-)

Confirm (PIII-1000 512 Mb Win2000)
Your signature fails with the same message.

Gleb

Post a reply to this message

> Btw: Am I only active beta-tester or just lucky one ?
>

Maybe you are just one of the few who try these weird function/macro things
I have no clue about :)

--
-Jide

Post a reply to this message

>...
> Btw: Am I only active beta-tester or just lucky one ?

No, you're not alone out in the
dark. I too torture functions.

Right now I am experiencing
severe crash problems when my
macros builds _VERY_ large
functions: I have to reboot my
PC after each crash  =(

So maybe you'll consider me
"lucky" as well.


Tor Olav

Post a reply to this message

On Sun, 17 Mar 2002 13:11:19 +0100, Tor Olav Kristensen
<tor### [at] hotmailcom> wrote:
> No, you're not alone out in the
> dark. I too torture functions.

Sounds very promising for functions advocate :-)

> Right now I am experiencing
> severe crash problems when my
> macros builds _VERY_ large
> functions: I have to reboot my
> PC after each crash  =(

Did you reported this problem ? If your time is limited or something I can
help you with bug investigation if you can share your source.

> So maybe you'll consider me
> "lucky" as well.

We are really lucky with those functions in 3.5 :-)

ABX

Post a reply to this message

> 
> On Sun, 17 Mar 2002 13:11:19 +0100, Tor Olav Kristensen
> <tor### [at] hotmailcom> wrote:
> > No, you're not alone out in the
> > dark. I too torture functions.
> 
> Sounds very promising for functions advocate :-)

=)


> > Right now I am experiencing
> > severe crash problems when my
> > macros builds _VERY_ large
> > functions: I have to reboot my
> > PC after each crash  =(
> 
> Did you reported this problem ?

No, because I'm not sure if I have
read reports about similar problems
earlier. I'll have to check that.


> If your time is limited or something I can
> help you with bug investigation if you can share your source.

Thank you.

I'll soon post a zip file to povray.
beta-test.binaries, containing two
test files so that you can check if
it crashes POV on your PC.

I wish that I knew the limitations
for the sizes of functions.


> > So maybe you'll consider me
> > "lucky" as well.
> 
> We are really lucky with those functions in 3.5 :-)

They are wonderful new tools, but if
we don't treat them with great care
they often break.


Tor Olav

Post a reply to this message

On Tue, 19 Mar 2002 01:37:14 +0100, Tor Olav Kristensen
<tor### [at] hotmailcom> wrote:
> I'll soon post a zip file to povray.
> beta-test.binaries, containing two
> test files so that you can check if
> it crashes POV on your PC.

Yes, It makes illegal operation. But fast look into your zip gives me the
answer - there are 4022 selects not even talking about different tokens.
Thorsten explained me some day that number of tokens is limited but handling
of length checking was breaken. I though it is fixed now but look like your
source contain another exception. Check
http://news.povray.org/3c63f967%241%40news.povray.org

> I wish that I knew the limitations
> for the sizes of functions.

Check: http://news.povray.org/3c83a056%40news.povray.org

ABX

Post a reply to this message

> POV 3.5 b 13 icl on WinNT Sp 6 PII 233 with 128 MB
> then it is rendered fine. Any confirmation ?

Confirmed. This bug is breaking most of my newest scenes.
Here's where I met it:


#macro Interpolate (p1,p2,TT)
#local p1x=p1.x;
#local p1y=p1.y;
#local p1z=p1.z;
#local p2x=p2.x;
#local p2y=p2.y;
#local p2z=p2.z;

#local fx = function (T){p1x + T*(p2x-p1x)}
#local fy = function (T){p1y + T*(p2y-p1y)}
#local fz = function (T){p1z + T*(p2z-p1z)}

#local vec = <fx(TT),fy(TT),fz(TT)>;

vec
#end

#local my_vector =
Interpolate (<0,0,0>,<1,1,1>,0.5) ;

--
Jonathan.

Post a reply to this message

In article <3c9786bf@news.povray.org>, jrg### [at] hotmailcom says...

> > POV 3.5 b 13 icl on WinNT Sp 6 PII 233 with 128 MB
> > then it is rendered fine. Any confirmation ?
> 
> Confirmed. This bug is breaking most of my newest scenes.
> Here's where I met it:
> 

It seems that the bug shows up when the function has fewer than the 
default 3 parameters.  Try your code like this instead:

> 
> #macro Interpolate (p1,p2,TT)
> #local p1x=p1.x;
> #local p1y=p1.y;
> #local p1z=p1.z;
> #local p2x=p2.x;
> #local p2y=p2.y;
> #local p2z=p2.z;
> 
> #local fx = function (T,a,b){p1x + T*(p2x-p1x)}
> #local fy = function (T,a,b){p1y + T*(p2y-p1y)}
> #local fz = function (T,a,b){p1z + T*(p2z-p1z)}
> 
> #local vec = <fx(TT,0,0),fy(TT,0,0),fz(TT,0,0)>;
> 
> vec
> #end
> 
> #local my_vector =
> Interpolate (<0,0,0>,<1,1,1>,0.5) ;
> 

This workaround fixes the original reporter's signature, and also fixes 
at least f_sqr in math.inc.  Now it is up to the developers to make a 
real fix for the bug.

--
Geoff Reedy                                                gre### [at] umredu

Post a reply to this message

> 
> On Tue, 19 Mar 2002 01:37:14 +0100, Tor Olav Kristensen
> <tor### [at] hotmailcom> wrote:
> > I'll soon post a zip file to povray.
> > beta-test.binaries, containing two
> > test files so that you can check if
> > it crashes POV on your PC.
> 
> Yes, It makes illegal operation. But fast look into your zip gives me the
> answer - there are 4022 selects not even talking about different tokens.
> Thorsten explained me some day that number of tokens is limited but handling
> of length checking was breaken. I though it is fixed now but look like your
> source contain another exception. Check
> http://news.povray.org/3c63f967%241%40news.povray.org
> 
> > I wish that I knew the limitations
> > for the sizes of functions.
> 
> Check: http://news.povray.org/3c83a056%40news.povray.org

Thank you for the info.

I'll look further into this later this week.


Tor Olav

Post a reply to this message