counting zombies

[battlestar]

This counts number of zombifications, it works most of the time (adds 1 to $zombie_number per kill) but occassionally it flips out and loops couple of times thus adding more than 1 per kill. Occassionally it even fires when the attack special isn't plague... Can't quite put my finger on why, and how to fix.

Code: Select all

# if turned zombie, +1.
	[event]
		name=attacker hits
		first_time_only=no
		[filter]
			side=1
		[/filter]
		[filter_attack]
			special=plague
		[/filter_attack]
		[filter_second]
			[or]
				side=2
				side=3
				side=4
			[/or]
		[/filter_second]
#		[message]
#					id=Valia
#					message= _ "$zombie_number"
#		[/message]
			[event]
				name=die
				first_time_only=yes
				[filter]
					[not]
						side=1
					[/not]
				[/filter]
				[set_variable]
					name=zombie_number
					add=1
				[/set_variable]
				[message]
					id=Valia
					message= _ "We have zombified $zombie_number humans."
				[/message]
				
			[/event]
	[/event]

[Crendgrim]

Well, the first event (attacker hits) is fired each time when a unit of side 1 is hitting another unit (I don't know if the filters are correct, they're looking weird, but I'm not that good in WML. Leaving this for someone else). In this event, you register another event - a die event which only triggers once.
However, imagine the following situation: You need two hits to kill your opponent - then two of the latter events are registered and triggered on that unit's death. This even works, if you hit one unit, don't kill it and then kill another one.
I don't think it's necessary to use nested events here, you should also be fine with only the die event (modified, of course).


Crend

[Sapient]

this is definitely wrong:

Code: Select all

         [or]
            side=2
            side=3
            side=4
         [/or]

Assuming for a moment that side= in Standard Unit Filter doesn't support lists, the first condition should be outside the [or] tag and the others should have their own seperate [or] tags. However, since side= in SUF *does* support lists, you don't need the [or] tag at all.

[battlestar]

Thanks for your inputs. Portion of the code fixed according to Sapient and the problem persist as Crendgrim pointed out. It added multiple times if several attacks using plague was done.

Using only name=die didn't work, it didn't seem to accept the weapon filter (the wiki also didn't list weapon filter can be used in die event). It needs an if-else I think, either inside the nested event: if another unit was selected -- nothing happens, else -- variable plus 1. Or no nested event: if secondary unit dies -- variable plus 1, else -- nothing.

Anyways, close to figuring it out.

[Brilliand]

Check the target's HP - it should be zero or less if the unit is dying.

Alternatively, check whether the unit's tile is occupied in the Die event. The unit should be gone by then, so if the tile is occupied, then it's probably due to a plague attack (you can check the unit type for more insurance).

[Crendgrim]

Alternatively, check whether the unit's tile is occupied in the Die event. The unit should be gone by then, so if the tile is occupied, then it's probably due to a plague attack (you can check the unit type for more insurance).

No, this shouldn't work; at least not if the Wiki is right. It says that the unit isn't removed until after the die event finishes and that it can still be adressed. So this won't work.


Crend

[monochromatic]

If I'm understanding what you want:

Code: Select all

[event]
    name=prestart
    {VARIABLE zombie_number 0}
[/event]

[event]
    name=attack end
    first_time_only=no
    [filter]
        side=1
    [/filter]
    [filter_attack]
        special=plague
    [/filter_attack]
    [filter_second]
        [not]
            race=undead
            [and]
                race=mechanical
            [/and]
        [/not]
    [/filter]

    [if]
        [variable]
            name=second_unit.hitpoints
            less_than_equal_to=0
        [/variable]
        [then]
            {VARIABLE_OP zombie_number add 1}
            # Messages.. and the like.
        [/then]
    [/if]
[/event]
# And repeat backwards in the case of a retaliatory plague kill.

[battlestar]

thanks, that does exactly it.

[monochromatic]

Did you repeat it backwards? At this stage, it won't count a kill if the WC kills a unit in retaliation.

[Ceres]

Code: Select all

[filter_second]
        [not]
            race=undead
            [and]
                race=mechanical
            [/and]
        [/not]
    [/filter]

Wouldn't [or] be more senseful?
Also, the closing tag should be [/filter_second] (though [/filter] would've created an error, which it apparently didn't, so I guess it's fixed by now).

[monochromatic]

Code: Select all

[filter_second]
        [not]
            race=undead
            [and]
                race=mechanical
            [/and]
        [/not]
    [/filter]

Wouldn't [or] be more senseful?
Also, the closing tag should be [/filter_second] (though [/filter] would've created an error, which it apparently didn't, so I guess it's fixed by now).

He must have fixed it. Sorry, my bad. And on the [or] things, I'm not sure.

[Crendgrim]

IMO Ceres is right; and if he isn't, it should be changed. You have to read "from inward to outward", so you first see "race=undead [logical and] race=mechanical" which cannot be true anytime, and then you negate this which will cause the event to trigger always.
At least that's how it makes sense to me.


Crend

[battlestar]

For the purpose of the current SP scn, it doesn't need to consider retaliation, I'll keep it in mind if another map has something similar. The only kink is when the enemy is on a village it doesn't get zombified... but I thought the player should get the bonus for the motion anyway, so letting it slide I guess.

Thanks for the help.

monochromatic's code syntax adjusted:

Code: Select all

[event]
    name=prestart
    {VARIABLE zombie_number 0}
[/event]

[event]
    name=attack end
    first_time_only=no
    [filter]
        side=1
    [/filter]
    [filter_attack]
        special=plague
    [/filter_attack]
    [filter_second]
        [not]
            race=undead
            [or]
                race=mechanical
            [/or]
        [/not]
    [/filter_second]

    [if]
        [variable]
            name=second_unit.hitpoints
            less_than_equal_to=0
        [/variable]
        [then]
            {VARIABLE_OP zombie_number add 1}
            # Messages.. and the like.
        [/then]
    [/if]
[/event]
# And repeat backwards in the case of a retaliatory plague kill.

[monochromatic]

For villages, this should work (untested):

Code: Select all

# Replace the [filter_second] with
[filter_second]
    [not]
        [filter_location]
            terrain=*^V*
        [/filter_location]
        [and]
            race=undead
            [or]
                race=mechanical
            [/or]
        [/and]
    [/not]
[/filter_second]

[Dixie]

It just occured to me that isntead of:

Code: Select all

    [not]
            race=undead
            [or]
                race=mechanical
            [/or]
    [/not]

You could probably use a more general/exhaustive code (I haven,t done this in a while, the exact syntax may be a bit flawed):

Code: Select all

[not]
    [filter_wml]
        [status]
            not_living="yes"
        [/status]
    [/filter_wml]
[/not]

Basically, undead/mechanical/whatever races are made unaffected by plague/poison/drain/etc. via giving them the not_living status (aka. making not_living=true, or "yes")

But you could also just filter every non-living as you are doing, it's just a different way of doing it

And by the way I don't know why you wouldn't want to check for zombie creation in retaliation, it souldn't be too difficult to add...