Doctor John wrote:
> Rafal 'Raf256' Maj wrote:
> 
>>"Apache" <apa### [at] yahoocom> wrote in
>>news:3dec0a34[at]news.povray.org
>>
>>
>>>Time For Trace:   14 hours 12 minutes  27.0 seconds (50902 seconds)
>>>    Total Time:   14 hours  8 minutes  22.0 seconds (50902 seconds)
>>
>>How fast was Yours computer moving while rendering ;) ?
>>
> 
> I'm glad I'm not the only one getting weird time estimates. Until now I
> assumed it was because there was something broken in my compile.
> Now I know the true reason - I'm moving at 96.657463% of c!
> For my next trick I will send a copy of a few other weird stats files
> yesterday. ;-)
> 

I see the Time For Trace hours/minutes/seconds part offbase all the time.
The seconds listed in the parentheses will be correct. Here's an example
from rendering biscuit.pov (320x240 no AA):

Time For Trace:    0 hours 30 minutes  60.0 seconds (3 seconds)
     Total Time:    0 hours  0 minutes   3.0 seconds (3 seconds)

What I think it's been doing is calculating the trace time for the
hours/minutes/seconds part based on how long the POV-Ray render window
stays open. I know that sounds weird but testing seems to bear that
out. For the biscuit.pov times above, I started the render and left
the display window open, made a futile attempt to catch up on newgroups,
then returned to the display window and clicked on it to close it. So
you can see I spent about 30 minutes trying to comprehend the many
POV-Ray newsgroup postings.

This is with the Linux version btw, I have not tested the behavior on
any other platform.

I like Tek's suggestion though and we need to look into implementing
some sort of backwards timewarp so that my 15 hr renders only take
3 minutes ;)

-Roz

Post a reply to this message