is_possible_vassal shortcomings

[moretrim]

Background

The wiki documents the is_possible_vassal trigger as follows:

Returns true if the specified country can be released as a puppet state (meaning do they own the target's capital?).

Worryingly this trigger does not appear to be used at all in the unmodded game. In HPM/HFM it is used sparingly in decisions and events, e.g. in the independence events of, well, events/Independence.txt.

The short of it

I have real trouble using the trigger dynamically. As an example, if you send a request for a diagnostic to every non-existing country of which you hold a core:

any_country = {
    limit = {
        exists = no
        any_core = { owned_by = THIS }
    }

    country_event = <diagnostic id>
}

(Note the use of any_core = { owned_by = THIS } over have_core_in = THIS, but that’s a story for another time.) Upon receiving the request, these countries perform:

random_country = {
    limit = {
        tag = FROM
        is_possible_vassal = THIS
    }

    # yes, THIS non-existing country is a possible vassal of FROM
    country_event = <positive answer id>
}

random_country = {
    limit = {
        tag = FROM
        NOT = { is_possible_vassal = THIS }
    }

    # no, THIS non-existing country cannot be a possible vassal of FROM
    country_event = <negative answer id>
}

(Do not forget allow_multiple_instances = yes for the answer events, as we are expecting many responses at once!)

As an aside, the fact that the code needs to perform a two-event jig before sending an answer illustrates a first shortcoming of the trigger: the lack of a swapped form. That is, at the start of this we cannot open a scope for all the non-existing countries for which <country that asks> = { is_possible_vassal = <candidate country> } holds.

When performed as 1836 Austria, the outcome is that every non-existing candidate answers no. But try the following in a decision trigger:

is_possible_vassal = TST
is_possible_vassal = CRO
is_possible_vassal = HUN

and watch the results:

(The lack of a localisation fragment is expected, though emblematic.)

Which is right: the event or the decision triggers? Hard to say, keeping in mind that TST is unreleasable.

Workarounds

In practice, what makes a country able to be released from the Politics screen? It all seems to hang on (with example checks):

• not existing: exists = no
• having cores present: e.g. any_core = { owned_by = <releaser> }
• being releasable: is_releasable = yes

Note in particular that there is only one requirement(ish) when it comes to capitals: the releaser never hands its capital province (not state) away. This leads to the following:

• the country is releasable even if the (historical) capital is held by someone else
• the country may be released even if it shares its capital with the releaser’s, as long as there are other cores
• a single core on the releaser’s capital is exploitable: the country can be “released” to free up 2 infamy, with no other effect (and choosing “Play as…” leads to an instant game-over)

Code that still wants to play nice with historical capitals can use capital_scope to select the province, e.g. putting it all together:

any_country = {
    limit = {
        exists = no
        is_releasable_vassal = yes
        any_core = {
            owned_by = THIS
            is_capital = no
        }
        capital_scope = { owned_by = THIS }
    }

    country_event = <id of event to release from THIS-that-becomes-FROM>
}

Except that is not quite it! Because the release event will need to be performing either release_vassal or release (secede_province is also possible), but that on itself does not quite work like releasing through the Politics screen. But that’s a topic for another discussion…

Code that wants or needs to hand out the releaser’s capital could consider moving it first.