
Thu, 08/11/2011 - 14:54
Forums:
Hi
I am testing performances.
Here is my basic procedure (.tcl file)
box b 0 0 0 1 1 1
compound b cubes
for {set i 1} {$i
copy b b$i
ttranslate b$i $i 0 0
add b$i cubes
}
save cubes c:/temp/10000cubes.brep
It takes hours for duplicating the cube.
Am I missing something?
thanks
Davide
Thu, 08/11/2011 - 16:05
probably. in pythonocc this sort of things is pretty rapid...
Thu, 08/11/2011 - 18:37
thanks
I am not a developer. I am just veryfing the performances.
If you did it, please post here a .brep with 10.000 cubes.
The cube is
box b 0 0 0 1 1 1
the increment, in x only, is 1
thanks in advance
davide
Thu, 08/11/2011 - 18:41
> I am not a developer. I am just veryfing the performances
I'm sorry that's impossible ;)
Thu, 08/11/2011 - 20:06
Sorry jelle, but that's called testing and is often carried out by people other than developers - we call them testers :)
Peye
Thu, 08/11/2011 - 20:07
er, should say Pete
Thu, 08/11/2011 - 20:51
yes / no; a tcl test is not representative of OCC's execution speed, wouldnt you say?
[ but your right, my reply wasnt very constructive / on the verge of rude, sorry... ]
Fri, 08/12/2011 - 13:14
It's a test of creating 10,000 boxes using a scripting language and as such does excecute only a very small amount of the total OCC library, which doesn't give an overall indication of performance (or find performance hotspots). However, its a simple test that can be carried out similarly in the scripting languages of other CAD/CAE apps or libraries and as such could provide a relative measure of performance against other vendors products - as you might carry out in a product selection excercise.
Pete
Thu, 08/11/2011 - 20:41
If you add some echo statement in the loop, you will see that it slows down at each iteration. The most obvious culprit is compound, you should move this command away from the loop:
box b 0 0 0 1 1 1
set boxes "b"
for {set i 1} {$i < 10000 } {incr i 1} {
copy b b$i
ttranslate b$i $i 0 0
lappend boxes "b$i"
}
eval compound $boxes cubes
Thu, 08/11/2011 - 21:45
many thanks Denis
it worked fine
my goal was just to generate a native .brep file with 10.000 cubes
thanks!
davide