IsDestructableTree

[Moderator]

12th Dec 2015
IcemanBo: Too long as NeedsFix. Rejected.

17:51, 2nd Sep 2012
Magtheridon96: Yes, I know I said this would be the fastest way to detect trees, but this method over here is still quite fast:

Something I wrote just now on the spot

library IsDestructableTree requires OrderId

	globals
		private unit dummy = null
	endglobals

	function IsDestructableTree takes destructable d returns boolean
		return IssueTargetOrderById(dummy, ORDER_harvest, d) and IssueImmediateOrderById(dummy, ORDER_stop)
	endfunction

	private module Init
		private static method onInit takes nothing returns nothing
			set dummy = CreateUnit(Player(15), 'hpea', 0, 0, 0)
			call UnitAddAbility(dummy, 'Aloc')
			call ShowUnit(dummy, false)
		endmethod
	endmodule
	private struct InitS extends array
		implement Init
	endstruct

endlibrary

It also requires no work at all from the user making it a perfect system.

 

[Kingz]

Eh there is a better way afaik?
And already exists, it involves something with a certain ability (i think it was a ghouls harvest ability).

 

[PurgeandFire]

Yes, there is already:
http://www.wc3c.net/showthread.php?t=103927

This system requires more work from the user.

 

[defskull]

There is a better way (others has been using it - link by Purge) :/
1. Create dummy unit on mapInit (set it to variable)
2. Each time you want to check if the destructible or not, move the dummy to the destructible's location and order the dummy to 'harvest' it
3. Check if current order of the dummy is 'harvest', then it's a tree, else, it's not a tree.

Simple.

 

[maddeem]

I know this exists, but this method is much faster than the other.

 

[baassee]

Since when were hashtables fast?

 

[HappyTauren]

Since always. Hashtables will always be faster than ordering units to do anything. I don't know why there is all this propaganda against hashtables, usually because people are uptight about 'efficiency'.

 

[baassee]

Since always. Hashtables will always be faster than ordering units to do anything. I don't know why there is all this propaganda against hashtables, usually because people are uptight about 'efficiency'.

Arrays are faster ^^ - always.

 

[HappyTauren]

Most of the time it does not even matter, but that's a completely different thing, though. Speed penalties to hashtable use are much lower than speed penalties to bad code, and I am disgusted to see what kind of Array Nazis we have as moderators.

 

[Kingz]

Hashtables are on average 8 times slower than arrays.

 

[baassee]

I'm not a moderator :O yet mag96 was my padawan

Worst part is with hashtables that everyone creates a spell, each using one hashtable and when a user hits the 256 limit then boooom omgomgomgomgwtfwtfwtfspellnotworkingblablabla

 

[HappyTauren]

Hashtables are on average 8 times slower than arrays.

Benchmark it.

 

[Kingz]

Benchmark it.

I don't have to, i saw Bribe say it (tho i can't remember if it was 8 or 4 times).

Anyhow, not that it matters, arrays have their use, hashtables have other use. Preventing O(n) search is the primary use of a hashtable imo.

 

[maddeem]

I don't see why this isn't accepted, who cares if it adds more work to the user? It increases efficiency!
This would make knockback systems that check for trees so much smoother.
And hashtables are the fastest method for checking in this situation

 

[Kingz]

It's a pain in the ass to setup.
But it could be approved as an alternative.

 

[Ender]

you could think about using this instead as it might be a tad fasterHaveSavedBoolean(h,i,0)

 

[PurgeandFire]

We are not rejecting this because we think it is slow:

Yes, I know I said this would be the fastest way to detect trees, but this method over here is still quite fast:

This system requires more work from the user.

It just has less modularity. In this case, yes a simple O(1) retrieval is advantageous speed-wise (most likely very minimally though), but if someone uses custom trees, it requires extra implementation opposed to the harvester method which requires none.

 

[Cokemonkey11]

* You need to conduct a speed test to qualitatively state how much faster this is

* You should use Table instead of instantiating hashtables yourself.