henesua wrote...
Funky, did you find at what point GNObT outperforms GObT? How many objects (roughly speaking) do you need to have for this to matter?
We found that GNoBT outperformed GObT radically even when our mod was much smaller - there was no evidence of any kind of optimization on GObT. It could take a second or two, even, which is outrageous as compared to GNObT.
Mavrixio wrote...
Yes, i have tested it.
It certainly depend on the number of object the number of object in the area, i dont have stat about that, but for if there's 100+ object int the area, then GetObjectByTag is 2-3 times faster than GetNearestObjectByTag, I dont know you, but personally we have areas with 1000+ placeables, so GetNearestObjectByTag is a no-no.
You say you've tested it, but you don't have any statistics on performance by object count? How exactly did you test it? I'd be very interested to see your profiling on it.
Also, if you have areas with 1k+ placeables, I'm a lilttle less inclined to take your claims about performance seriously. That's an unbelievable number, even if you're really decking an area out. Have you actually counted, or are you guestimating? The most obscene placeable count on HG is in the 500s, where I went seriously overboard decking out the Forest of Poisoned Dreams in Shedaklah with extra mushrooms in order to get the right effect.
For both of you, here's our mod set out in placeable and tile count, for purposes of comparison:
HG placeable/tile countsFWIW, we stopped using GObT when the mod was less than half its current size.
Funky