New particle issue from patch

It seems that the changes applied from the irrisor patch leave odd leftover particles at the end of a particles lifecycle. You can see this by running the editor, pulling up smoke, and turning down the life expectancy. You can also see this by running dynamicsmoker and spinning around… Just letting you all know in case you see odd particle artifacts… looking into it today.

Update: rolling back the changes to the two particle classes and fixing the tests that ref them fixes the issue… I’ll see if I can slowly remove the camera dependency and see if that fixes things.

Ok, fixed and in CVS. The issue was that the rerouting of the updates to the draw call made it so the nio buffers were updated AFTER the draw call, causing incorrect positioning to be displayed.



I also did away with the comparison to the camera direction. needsUpdate was set to true on every active update which made the check and the constant set of the orientation vector superfluous except when the particles are dead. I don’t see a real point, but feel free to tell me I’m wrong :slight_smile:



Anyhow, particle bug gone.



PS: I’m not knockin’ irrisor… I think his change gave us a bump to particle update speed… It updates once per cycle instead of twice. :slight_smile:

8-O, I never noticed that problem - sorry for causing it ://

You changed quite a lot in that checkin - hope it did not take you too long!

actually I didn’t, I rolled back the massive format change that was also applied at the same time as your patch because it was easier to start from the working version and remove the camera dependency. So most of those lines are formatting differences