BaseSplPwr component issue?

[Colen]

If you're changing 9 spells all at once, I would actually build that into one foreach if you can? For example:

foreach pick in hero from BaseSpell where "thingid.spell1 | thingid.spell2 | thingid.spell3 | etc etc"
eachpick.field[sDC].value += 1
nexteach

It'll make your code look a lot nicer, and it'll be comparable performance to your 9 findchilds I think.

You can actually find out the performance of a script by going to Develop -> Floating Info Windows -> Show Task Performance. That info window shows you the evaluation time, in microseconds, of all scripts that ran last evaluation. When I'm doing profiling stuff, I typically do 3-5 runs of each to get a good average, then compare the averages against each other.

[AndrewD2]

Originally Posted by Colen

You can actually find out the performance of a script by going to Develop -> Floating Info Windows -> Show Task Performance. That info window shows you the evaluation time, in microseconds, of all scripts that ran last evaluation. When I'm doing profiling stuff, I typically do 3-5 runs of each to get a good average, then compare the averages against each other.

OK filing that away for future use. I had no idea.

[ShadowChemosh]

Originally Posted by Colen

If you're changing 9 spells all at once, I would actually build that into one foreach if you can? For example:

foreach pick in hero from BaseSpell where "thingid.spell1 | thingid.spell2 | thingid.spell3 | etc etc"
eachpick.field[sDC].value += 1
nexteach

That is pretty much what I have. Was just wondering if 9 findchilds would be faster as I was trying to prevent the need of two foreach loops at two different timings. I currently found a more "middle" ground timing and it all appears to be happy again.

Originally Posted by Colen

You can actually find out the performance of a script by going to Develop -> Floating Info Windows -> Show Task Performance. That info window shows you the evaluation time, in microseconds, of all scripts that ran last evaluation. When I'm doing profiling stuff, I typically do 3-5 runs of each to get a good average, then compare the averages against each other.

Yeah with Andrew on this one. No idea that existed. Will be using that for some testing. Thanks Colen!