Optimising Particles - Pooling etc.

The forum is locked. These are the original blitzbasic.com forums, kept so they can still be read. No new topics, replies or edits can be made here. To join in, head over to the Blitz3D+ forums.

dmaz

ok, I originally said in my msg 2 above that looping through a simple embeded list will be "faster" than an array. Since I hadn't tested it I decided to replace that claim with "as fast as". I though it would be faster since the loop would not have to do any pointer arithmetic. the following example seems to show that looping through the list is faster.

SuperStrict

Framework BRL.GLMax2D
Import BRL.Random
Import BRL.StandardIO

Type TPart
	Global livePartList:TPart
	Global livePartArray:TPart[]
	Global total:Int

	Field nextPart:TPart
	Field one:Int
	Field two:Int


	Function Create( total:Int=100 )
		livePartArray = New TPart[total]
		For Local i:Int = 0 Until total
			Local p:TPart = New TPart
			p.one = Rnd(10000)

			livePartArray[i] = p

			p.nextPart = livePartList
			livePartList = p
		Next
		TPart.total = total
	End Function
	
	Function LoopList()
		Local p:Tpart = livePartList
		While p
			p.one = 99
			p = p.nextPart
		Wend
	End Function
	
	Function LoopArray()
		For Local p:TPart = EachIn livePartArray
			p.one = 99
		Next
'		For Local i:Int = 0 Until total
'			livePartArray[i].one = 99
'		Next

	End Function

End Type

TPart.Create(1000000)
Local times:Int[3]

GCSetMode 2
Delay(100)



Local time:Int=MilliSecs()
TPart.LoopList
times[0] = MilliSecs()-time

time:Int=MilliSecs()
TPart.LoopArray
times[1] = MilliSecs()-time

time:Int=MilliSecs()
TPart.LoopList
times[2] = MilliSecs()-time

Print "~nLivePartList: "+times[0]+"ms"
Print "~nLivePartArray: "+times[1]+"ms"
Print "~nLivePartList: "+times[2]+"ms"

anybody see any problems with the above code?

when running, click run from the ide a bunch to get a feel for the timing.

[edit] changed code to put prints at bottom... didn't see a change though on my machine.