POV-Ray : Newsgroups : povray.beta-test : #persistent : Re: #persistent Server Time
15 Jun 2021 09:00:33 EDT (-0400)
  Re: #persistent  
From: green
Date: 29 Dec 2018 21:20:00
Message: <web.5c28296b874d1a8ed0c6faf90@news.povray.org>
ingo <ing### [at] tagpovrayorg> wrote:
> in news:web.5c27863e874d1a8e765e06870@news.povray.org Bald Eagle wrote:
>
> > For short-term
>
> Meh. With #write a lot can be done already, below's the resulting file
> of ~170 lines of macro. I'd opt for doing #persistent right. That doesnt
> take a way that it is an important feature that I've missed often and
> for me would be high on the list.
>

i can't say i can connect the code with the topic.
  #write might be more useful in this context if it wrote to a ram disk.

i believe my question to lipka was off-target.  linux generally spawns povray
for each batch of frames, so persistent objects should be lost after a job.  it
is hard to see how name conflicts would arise without contrived examples.
  for windows it is different.  i've already bumped my nose hard against it.  i
use a construct such as this to work around it;

#declare klock=clock; // or whatever time you are interested in
#if( klock=0)         // or whatever time you are interested in
    #ifdef( thatthing)
        #undef thatthing
    #end
#end
#ifndef(thatthing)
    #persistent thatthing=
    ...
#end

i think this type of thing should be sufficient.

i have not exercised due diligence in determining how much time uberpov saves
for long-parsing frames, but the small amount of experimentation i have done
suggests ten or twenty percent.  which is an enormous amount when making frames
for days or weeks or months.  so i will keep on using uberpov for animations
unless 3.8 has something i /have/ to have.


Post a reply to this message

Copyright 2003-2021 Persistence of Vision Raytracer Pty. Ltd.