Ok, here's an update..

roughly what's changed :

Added the Fir tree.. (Remco's)
Worked on the algorithm's a bit. No good results.
removed some spaces in outputfile..
Changed the interface.
Added option for precision.

Will look a bit more about breaking the unions... I'm not sure wether I'll make it..


Here are some stats on the fir tree...
Only as a comparsion. this is on a p166mmx 32Mb, overclocked to 200

included FirTree : 
        objects : 10873
        parse   : 0m 51s 
        trace   : 0m 38s
        total   : 1m 29s
        Mem     : 23 641 189 b

#macro FirTree : 
usage : object { xtree(30, .25, .4, T_Branch1 ,O_Needle , seed(666)) } 
        objects : 10880
        parse : 0m 19s
        trace : 0m 36s
        total : 0m 55s
        Mem   : 12 072 040

#macro FirTree : 
usage :#declare Fir1 =  object { xtree(30, .25, .4, T_Branch1 ,O_Needle , seed(666)) }
        objects : 10880
        parse : 1m 2s
        trace : 0m 37s
        total : 1m 39s
        Mem   : 23 656 943

This shows that It is not a big gain between including a FIR, and using the #macro.
But
when you have more, I think It will become better.
I may remake the program to take a texture for the needle, and then break unions.
perhaps.


//Spider

Post a reply to this message

Attachments:
Download 'wintree.zip' (140 KB)