Sun, 16 Oct 2011 21:51:17 +0100 http://flintparticles.org/forum/ Lussumo Vanilla 1.1.10 & Feed Publisher http://flintparticles.org/forum/comments.php?DiscussionID=526&Focus=1741#Comment_1741 http://flintparticles.org/forum/comments.php?DiscussionID=526&Focus=1741#Comment_1741 Thu, 06 Oct 2011 20:09:57 +0100 evo
I guess I'll start by showing a picture explaining the issue:



At the top if my memory usage profiler (Flashdevelop). Middle are the "Live Objects" Max Count / Current Count / Memory. Last is my code.

In the memory profiler pic you can see a seesaw during the first third, this is before the emitter is created. After that I create the emitter and you can see the memory steadily increase. While watching the Live Object viewer I could see that PortalBlueDot Count (2nd column) was pretty much increasing/decreasing in sync with what the profiler shows. It does this until reaching a Max Count (which I assume is the maximum objects of this type alive at any one time) of 1500 than does the complete drop off you see at the end of the profiler. From then on the current count stayed around 300, expected for 100 particles/sec with a lifetime of 2-3, right?

What I'm trying to figure out is what caused the large buildup of extra particles when they weren't required, what triggered the "evening out" of memory usage, and finally how can I prevent the buildup and keep the particle creation "steady" as the counter name implies?

Additional info:

PortalDotBlue is a simple 5px circular sprite created in Flash CS4.
This entire process was done with the application running idle with the exception of me clicking to start the emitter.

Thanks in advance! ]]> http://flintparticles.org/forum/comments.php?DiscussionID=526&Focus=1742#Comment_1742 http://flintparticles.org/forum/comments.php?DiscussionID=526&Focus=1742#Comment_1742 Sat, 08 Oct 2011 08:28:58 +0100 evo