Double destroying

[Light Alkmst]

I've heard that destroying a variable that's already been destroyed, but not re-assigned is bad. How bad is it and what would it do?

 

[Samael88]

It would probably crash the game

If you where told to hit something and it just turned out to be nothing there, you would probably get confused aswell

 

[Light Alkmst]

I mean that let's say...

local effect fx = // i really don't want to code this on top

call DestroyEffect(fx)
call DestroyEffect(fx)

but you don't use it. It's just the double destroying itself I was wondering about.

 

[Cookiemaster]

Just ran "Custom script: call RemoveLocation(udg_TEMP_POINT[0])" twice in a row.

Nothing happened.


You're just wasting some memory for executing the command, nothing more.

 

[Infinity.4Sk]

I'd say it will do nothing but maybe create more leaks. (still think 1 is better)

 

[Exide]

Better still (for some special effects, at least), use:

call DestroyEffect(AddSpecialEffect("model", x, y ))

-Then you add and destroy the effect all in one call. :thup:

 

[Azlier]

*claps*

 

[Kazuga]

I'd say it will do nothing but maybe create more leaks. (still think 1 is better)

Removing a leak won't create a new leak...

No I don't think anything happens. When removing a variable you more like wash and clean it, if it already is washed and cleaned then it won't really make any difference. (Note, this is not a fact, I might be wrong.)

 

[Flare]

I'd assume it does in fact do something (because Grimoire notifies you of double frees), but that could just be to highlight the fact that "you're doing it wrong" If it does do anything, I have no clue what it is

 

[Exide]

I'd assume it does in fact do something (because Grimoire notifies you of double frees), but that could just be to highlight the fact that "you're doing it wrong" If it does do anything, I have no clue what it is

How about we make a simple test?
Let's loop: call DestroyEffect(udg_Variable) some 10 000 times every 10 seconds or so, wait five minutes and see what happens?

 

[Flare]

10000 per 10sec only? Pff, you could easily get 100,000 executions, if not 200,000, per sec on a .01 timer (although things might start sizzling then =D)

EDIT: Hmmm, it definitely doesn't cause lag (well, at least for me) - running at 400,000 executions per seconds, my FPS dropped during the test period (10s, so about 4,000,000 executions of call RemoveLocation (l)), but it went straight back up afterwards.

 

[Romek]

You could try it for different variable types.
I'm quite sure Lightning could cause problems.

Something similar was causing me problems, though I don't remember the cause. It might've been destroying a null lightning.

 

[Exide]

10000 per 10sec only? Pff, you could easily get 100,000 executions, if not 200,000, per sec on a .01 timer (although things might start sizzling then =D)

I figured 10 000 per 10 sec for a longer amount of time would give a more accurate result, but I guess it doesn't matter.

Conclusion: Do it right the first time, and you don't have to bother with these things.