 |
|
|
|
|
|
| |
| |
|
|
From: Florian Brucker
Subject: Strange behaviour with looped function calls in POV 3.6
Date: 28 Jun 2004 14:39:49
Message: <40e065f5$1@news.povray.org>
|
|
|
| |
| |
|
|
I get strange error messages running the following code in POV-Ray 3.6
(It works fine in 3.5):
<POV>
#declare MyF = function { 1 }
#macro Test(F)
(1)
#end
#local a = 0;
#while (a<20000)
#debug "Works\n"
#local Dummy = Test(MyF);
#local a=a+1;
#end
</POV>
Here's the output:
<output>
...
Works
Works
Works
Works
Works
Works
File: bugsearch.pov Line: 13
File Context (5 lines):
#while (a<20000)
#debug "Works\n"
#local Dummy = Test(MyF)
Parse Error: Identifier expected, incomplete function call or spline
call found
instead.
</output>
As you can see, it outputs "Works" more than 50 times (did not copy all
of it :) before raising the error. I don't understand this, as each time
absolutely the same code is executed (content of identifiers are
unchanged etc.).
If I use
#local Dummy = Test(function { MyF(x,y,z) });
(i.e. wrapping the function in another function block), it works again.
I'm experiencing this on Gentoo 2004.0, with 1 GB Ram and POV 3.6
official binaries. As I said before, it works without problems under POV
3.5.
Can anyone confirm this behaviour?
Florian
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I can confirm that it runs for a couple hundred calls, then crashes POV 3.6
(official) on a PentiumIII ~730mHz w/ 128Megs of RAM running Windows98.
Strange, indeed...
....Chambers
Post a reply to this message
|
|
| |
| |
|
|
From: Hughes, B
Subject: Re: Strange behaviour with looped function calls in POV 3.6
Date: 29 Jun 2004 02:31:59
Message: <40e10cdf$1@news.povray.org>
|
|
|
| |
| |
|
|
"Chambers" <bdc### [at] yahoo com> wrote in message
news:web.40e090e23e8f0090ca11893c0@news.povray.org...
> I can confirm that it runs for a couple hundred calls, then crashes POV
3.6
> (official) on a PentiumIII ~730mHz w/ 128Megs of RAM running Windows98.
961 times here, then the error. ;-)
Seems to me that there had been talk of this kind of thing before; that is,
repeated macro calls of a function. Not too sure. I don't know why it would
be limited, if so. I'm just thinking #undef could be the only solution then,
as per my unclear memory on the subject. Checking that, with #undef Dummy
added to the end of the loop, lets it continue without a problem.
Bob H.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|
