 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3CF540FE.6B43B2F5@gmx.de> , Tim Nikias <tim### [at] gmx de> wrote:
> I'm not sure of the status of those reports, if
> they were confirmed officially, or if something
> else might be causing the memory increase.
Those have been explained a very long time ago in p-beta-test: Some people
observed memory use increasing by a few KB with each render using the system
monitor, which is exactly the size needed to store the statistics and other
message output. So there is no problem and never was.
I should point out that the system monitor just like any other external tools
are completely unsuitable to measure any memory usage in a correct manner.
There is no way to get the correct information without access to the source
code and linking it with libraries suitable for that task or enabling the
POV-Ray internal memory tracking. All other measurements gained with external
methods could just be replaced by random data which would say as "much"!
> I'm aware that you are actually a little "afraid"
> of rumors, so I'll be less POV-offensive
> the next time, okay? ;)
Well, you already responded to a rumor so I suppose you now know why they are
a problem :-) Even if they were explained long ago as being false and
baseless someone will have missed that explanation...
Thorsten
____________________________________________________
Thorsten Froehlich
e-mail: mac### [at] povrayorg
I am a member of the POV-Ray Team.
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>
> Well, you already responded to a rumor so I suppose you now know why they are
> a problem :-) Even if they were explained long ago as being false and
> baseless someone will have missed that explanation...
>
Yeah, like me... ;)
--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3cf612bb$1@news.povray.org> , "Gilles Tran" <tra### [at] inapginrafr>
wrote:
>>So there is no problem and never was.
>
> Memory problems that became apparent in animations have been reported before
> :
>
> Change 1512 on 2002/04/14 by chrisc@lilliput
> Fix pigment_pattern memory leak reported in
> http://news.povray.org/3cb3217f$1@news.povray.org
>
> Change 1300 on 2001/12/30 by thorsten@host27
> Should fix the JPEG image reading memory leak
> http://news.povray.org/3be2ef0d@news.povray.org
> (note: this one was fixed, apparently...)
These are not those reports I have been referring to or which this thread is
about. There was a long thread in p.beta-test (I don't feel like searching
for it now) that was about someone insisting there was a problem while there
was not.
> A last one was reported recently but not confirmed.
> http://news.povray.org/3CDFFDB3.F24905A9@gmx.de
As explained, looking at the system monitor is absolutely unsuitable to check
memory usage...
Thorsten
____________________________________________________
Thorsten Froehlich
e-mail: mac### [at] povrayorg
I am a member of the POV-Ray Team.
Visit POV-Ray on the web: http://mac.povray.org
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Thorsten Froehlich wrote:
>
> [...]
>
> > A last one was reported recently but not confirmed.
> > http://news.povray.org/3CDFFDB3.F24905A9@gmx.de
>
> As explained, looking at the system monitor is absolutely unsuitable to check
> memory usage...
>
I have not been able to supply a test scene for this until now so no need
to have much fuzz about it, but if the harddisk starts trashing after 250
frames despite 256 Mb RAM this IMO is a suitable indication for memory
consumption.
Christoph
--
POV-Ray tutorials, IsoWood include,
TransSkin and more: http://www.tu-bs.de/~y0013390/
Last updated 05 May. 2002 _____./\/^>_*_<^\/\.______
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3CF622D4.238E3ABA@gmx.de> , Christoph Hormann
<chr### [at] gmxde> wrote:
>> As explained, looking at the system monitor is absolutely unsuitable to check
>> memory usage...
>>
>
> I have not been able to supply a test scene for this until now so no need
> to have much fuzz about it, but if the harddisk starts trashing after 250
> frames despite 256 Mb RAM this IMO is a suitable indication for memory
> consumption.
Well, how many messages are in the message window by then? I suppose many.
And they are likely all over the application memory space which suggests a lot
of memory fragmentation. Consequently after each frame there can a little
memory that is no use as it is a leftover fragment when rendering the next
frame which uses a similar memory allocation pattern.
The effect many not be as visible when rendering different scenes, but it is
simply in the nature of memory allocation. Just like your harddisk memory
gets fragmented, and even worse, with memory there is no way to "defragment"
it other than freeing all memory allocated, which is essentially restarting
the program ... essentially you are chasing ghosts here.
Thorsten
____________________________________________________
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Well, how many messages are in the message window by then? I suppose many.
> ...
This note got me thinking. Is there a way to have POV-Ray
(in Windows) flush the message-stream? Perhaps that's whats
causing the most trouble for Windows users...
And I mean something different than restarting POV manually.
Something automatic would be helpful, I guess...
Suggestions?
--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
You could just stop POV-Ray from printing messages with -GA
If you are rendering a long animation, you usually don't need to see
the scene statistics for anything.
--
#macro M(A,N,D,L)plane{-z,-9pigment{mandel L*9translate N color_map{[0rgb x]
[1rgb 9]}scale<D,D*3D>*1e3}rotate y*A*8}#end M(-3<1.206434.28623>70,7)M(
-1<.7438.1795>1,20)M(1<.77595.13699>30,20)M(3<.75923.07145>80,99)// - Warp -
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Aha, yup. I'll try that. Would it switch debug messages
(that my particle-system outputs) off too?
(That was a rherotical question, I'll see for myself...
Its not too hard to test, if you know what I mean)
Warp wrote:
> You could just stop POV-Ray from printing messages with -GA
> If you are rendering a long animation, you usually don't need to see
> the scene statistics for anything.
--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Had a look at it. Thanks!
"Kevin R." wrote:
> Try Start -> Programs -> Accessories -> System Tools -> System Monitor, and
> add the charts you want. It's a pretty powerful tool.
--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Another idea I had...
If this message window stream in Windows is the cause
of some errors (memory problems, to be precise),
how about putting a note in this in the
docs? It would be best to make this behaviour very
obvious, unless someone wants to hear someone really
upset talking about how he tried to raytrace a 5000
(for 200 secs) animation and ran into some problems...
Thorsten Froehlich wrote:
> Those have been explained a very long time ago in p-beta-test: Some people
> observed memory use increasing by a few KB with each render using the system
> monitor, which is exactly the size needed to store the statistics and other
> message output. So there is no problem and never was.
--
Tim Nikias
Homepage: http://www.digitaltwilight.de/no_lights/index.html
Email: Tim### [at] gmxde
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
