Closed rogerburks closed 4 years ago
I determined that this decision is actually supposed to be spammed to "step up" from Gabon through Chad to occupy the Congo Basin. It was being spammed by the ai (and may have blocked other ai decisions) up until the invention of the Dark Continent. Therefore, I changed it to require that invention before it would be enabled.
Updating and modernizing this thread--also I recently discovered that it is not yet completely fixed. This decision becomes spammy by the ai when the allow
being activated during times when the effect
cannot do anything. This is a somewhat complex, iterative decision, and therefore has an increased tendency towards spamminess when unexpected situations occur.
Originally, the decision became spammy because the allow
was relatively permissive, most specifically not requiring invention = the_dark_continent
when the effect
requires it:
while the effect
was relatively strict:
I changed specific portions of the allow
to be more strict, because I wished for the decision to be visible but not be spammed pointlessly by the ai, with hopefully some indication of what would be needed to allow it. These are therefore relatively simple changes once the correct sections of the allow
are found:
Please see below for the parts that I missed (which are probably fairly obvious now that I have explained things like this).
The newly discovered problem looks like this, with correct initial colonies formed from the first click of the decision and prior events: colonized provinces 1972
= Libreville in Gabon, 1977
= Brazzaville in Congo, 1980
= Djambala in Congo, 1981
= Impfondo in Congo, 1962 = Bertoua
in Cameroon), but the button is still allowed without a valid effect
:
The problem concerns colonization of 1981
= Nola in Ubangi-Shari. The effect
requires invention = the_dark_continent
:
but the allow
does not:
Therefore I wish to change this part of the allow
to
AND = {
invention = the_dark_continent
owns = 1981
1966 = { empty = yes } #Ubangi-Shari
}
Additionally, at this stage I wish to revert a previous change, and make the decision visible to those who have not yet discovered the_dark_continent
, and therefore I have made the potential
more visible. This is for those who may be frustrated when discovery of the_dark_continent
is taking longer than usual, when the player may want to see what the decision does and why it is not enabled yet. The change in the potential
is from this:
back to this:
Additionally, I noticed that the effect
requires invention = colonial_negotiations
for the first click, which I now want to enforce in the allow
:
AND = {
invention = colonial_negotiations
owns = 1972
1976 = { empty = yes } #Gabon
}
AND = {
OR = {
has_global_flag = called_congo_conference
1962 = { empty = no }
}
invention = colonial_negotiations
owns = 1977
1976 = { empty = no } #Gabon
1980 = { empty = yes } #Congo
}
Additionally, there is a section in the my changed version's allow
that I am reverting, because it seems to do nothing and because it is not appropriate for the decision, since some of this is supposed to happen before 1890, it should not have invention = the_dark_continent
in the beginning of the allow
:
I am reverting to:
Test with the new settings worked (1890 version):
Test post-1880 pre-1890 with these settings also seems to be fine (unless I missed something):
Here is a post 1890 screenshot from the same run as the previous post, just to indicate that the Congo to Chad area was properly colonized once the_dark_continent
was discovered.
Here is a save with the 1880-1890 issue, especially vexing since I had France set as interesting.
https://drive.google.com/file/d/139zDXb9sYBHEWyritcNKFaAdMCuFL3iI/view?usp=sharing
Under certain circumstances, the Occupation of the Congo decision
occupation_of_the_congo
in FlavourMod_Africa.txt does nothing and does not clear from the decisions list. This may have consequences for other Congo decisions, and prevents the country from committing other decisions below it on the list. The problem seems to be due to thepotential
not being strict enough. Testing seems to hold up after equalizing thepotential
andallow
criteria.