Re: [brlcad-users] Parallel g-stl conversion?

[Joshua S. <jos...@gm...>]

On Tue, Apr 9, 2013 at 2:26 PM, Christopher Sean Morrison <br...@ma...>wrote:

>
>   Of course, you still need to convert to a BoT mesh first via the
> facetize command, and I suspect that's more likely where you're currently
> spending the majority of your time.
>
> Another option would be to perform the conversion in parallel at a
> scripting level, running multiple g-stl at the same time (on different
> objects).
>
>
I think you are right; this scripting approach is probably what I'll end up
doing.  Right now I'm unioning a bunch of right circular cylinder
primatives into a region, and then converting to stl with g-stl (I've found
the time scaling for that to be quadratic in the number of cylinders [1]).

An alternative approach would be to facetize the cylinders (maybe at really
low resolution) and perform Boolean ops on those BoTs (perhaps
incrementally rather than all n cylinders at once?).  Would you expect
unioning BoTs to be faster than unioning rcc primitives?

I'm happy to report back with timing results for any other approaches folks
on the list recommend.

Thanks!

[1]
http://www.variousconsequences.com/2013/01/octet-truss-memory-time-and-dollar-costs.html

-- 
Joshua Stults
Website: variousconsequences.com
Hackerspace: daytondiode.org