Thread: SpellEvent
View Single Post
Old 07-17-2009, 09:15 AM   #37
Themerion
Brahms-fan
 
Themerion's Avatar
 
Join Date: Jan 2006
Posts: 642

Submissions (2)

Themerion has a spectacular aura about (114)Themerion has a spectacular aura about (114)Themerion has a spectacular aura about (114)Themerion has a spectacular aura about (114)

Default

Quote:
Originally Posted by Anitarf
The benefit of having all spell event responses available on all spell events is significant, the speed gained by removing it is not.

I'm not quite sure if you mean variables or callback-functions. If you mean that I benefit from having the target unit on spell finish [for spell x], when [spell x] only use spell cast, then you are simply wrong. If you thought that I wanted you to remove a callback-function (such as spell finish), then you are wrong as well :)

Quote:
Originally Posted by Anitarf
I could write another paragraph on why the feature shouldn't be optional, but since the speed gain is not significant in the first place, I don't see a reason to spend more time explaining this.

Well. The speed gain is not significant, and it might never matter, BUT IT FEELS BAD WITH ALL THIS STUFF RUNNING WHEN I DON'T NEED IT TO!

Quote:
Originally Posted by Anitarf
If your particular map doesn't require this feature then you can always delete the event responses code from the library yourself, nobody is forcing you to only use the code as provided.

Some of the spells in map might need to have event response variables at spell finish, for them I want struct population. But certainly most of my spells won't (which is why I want to be able to bypass the struct in those cases).

Quote:
Originally Posted by Anitarf
Besides, the bj_wantDestroyGroup-esque syntax is ugly.

Yeah, but I wanted to make a suggestion that allows you to keep your current syntax (I hate when people tell me: Do that, and that! But not how it's actually supposed to happen).

Quote:
Originally Posted by Anitarf
The second optimization you suggested is meaningless, an if statement isn't faster than an exitwhen statement in any meaningful way.

I even wrote that it should be inlined/optimized away/whatever you will call it. I just assumed that if VexMapOptimizer finds if false then, it will remove that statement.
__________________
Submissions::
(v)JASS Tutorial
Covers how to make fast and safe coding for common tasks in vJASS/JASS.
Creep Respawn System
It has never been easier to respawn creep groups...
Themerion is offline   Reply With Quote