Captain to the Rescue

[richvh]

Well then you should have said so.

 

[Korbah]

We seem to be having issues with Empires that gain de jure land via event are having that reset when loading from a save game. Unsure if it affects de jure drift. Is this a known issue or WAD, as with Duchies not being able to save their de jure status if changed?

 

[Rylock]

We seem to be having issues with Empires that gain de jure land via event are having that reset when loading from a save game. Unsure if it affects de jure drift. Is this a known issue or WAD, as with Duchies not being able to save their de jure status if changed?

Do you have any de jure kingdoms with no de jure empire? CK2Plus was having a similar issue, along with other empire-related problems, until we realized that was the source.

 

[Captain Gars]

I'm sure you're probably aware, but there are definitely a lot of missing random_* effects for corresponding any_* effects. Three that happened to be annoying me just today are the lack of random_attacker (war scope), random_direct_de_jure_vassal_title, and random_de_jure_vassal_title.

Also, on a related note, there are a ton of triggers that do not work as effects (and vice versa) when, by all reasonable interpretations, they absolutely should (either because the effects are just well-known to be bugged-- such as any_demesne_province, or because they just don't exist in the EXE). As a super-simple example, it would be tremendously useful if `any_de_jure_vassal_title` (not the DIRECT version, but all `de_jure_vassal_or_below` titles) was available as a trigger and not just an effect.

If you're interested, me and @Rylock can put together a quick list of the most offending / highest impact violators of the above two classes of problems (we'd already chatted about this briefly in response to your quoted comment). Most, of course, can be worked-around, but some-- all those mentioned except `any_demesne_province` included-- actually make certain classes of problems extremely difficult to solve efficiently or pseudo-elegantly.

It would be a long-term project, but I would at some point like to make sure there are a complete match between trigger and effect scopes (maybe something to put a new usermod programmer on when I get my minions). So by all means, feel free to compile such a list - just don't expect it to be implemented right away.

 

[ngppgn]

@Captain Gars : I think I've found an issue with now the engine treats randomness. Consider the following code:

namespace = test
character_event = {
 id = test.1
 is_triggered_only = yes
 hide_window = yes
 immediate = {
 any_title = {
 limit = { tier = count }
 location = {
 province_event = {
 id = test.2
 }
 }
 }
 }
}
province_event = {
 id = test.2
 is_triggered_only = yes
 hide_window = yes
 immediate = {
 if = {
 limit = {
 culture_group = indo_aryan_group
 }
 random_list = {
 50 = { religion = catholic }
 50 = { religion = orthodox }
 }
 }
 }
}

Once you fire event test.1 this should send event test.2 to every indoarian province, and give to each a 50/50 chance to turn catholic or orthodox. But that doesn't happenñ What happens instead is that either the whole set of provinces turns catholic or the whole set of provinces turns orthodox, which is not what the syntax of the events should do, if I'm getting that right.

Meanwhile the issue is looked at, could someone suggest a workaround to this?

 

[richvh]

That has been discussed up-thread (random seed the same for all daughter events in an any_* scope) and I think the Captain has fixed it for next patch.

 

[Rylock]

Meanwhile the issue is looked at, could someone suggest a workaround to this?

Send a province event to each province, and use the random_list command in that event. Since each is a separate event, the randomizer will function as you expect it to.

 

[ngppgn]

Send a province event to each province, and use the random_list command in that event. Since each is a separate event, the randomizer will function as you expect it to.

What do you mean? Surely, not code a different event for each province, as in, if I target X provinces, code X different events, right? Because I actually want to target all provinces in the map so...

And if that isn't what you mean, I don't understand, since I alrady am using the random_list within the province_event...

 

[Rylock]

What do you mean? Surely, not code a different event for each province, as in, if I target X provinces, code X different events, right? Because I actually want to target all provinces in the map so...

And if that isn't what you mean, I don't understand, since I alrady am using the random_list within the province_event...

Sorry. That's what I get for not reading carefully -- you did send a separate event to each province, which I assume would have worked. Clearly not.

 

[Zarathustra_the]

random does seem to not actually always evaluate.. IF you run the same event with a random list, it appears that sometimes the random gets stuck. Like how 25 events with a random_list with one option for 3% can actually fire 25 times if that first evaluation is the one it triggers first.

 

[fanoI]

In this case too random is not really working (the new character in the end of the event has always the same dynasty! But the dynasty should be random):

character_event = {
    id = republic_rev.3
    title = EVTNAMErepublic_rev.3
    desc = EVTDESCrepublic_rev.3
   
    is_triggered_only = yes

    option = {
        name = EVTOPTOKrepublic_rev.3

        random_realm_character = {
            limit = {
                is_republic = yes
                NOT = { trait = senator }
            }

            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }

        random_realm_character = {
            limit = {
                AND = {
                    is_feudal = yes
                    is_landed = yes
                    is_female = no
                }
                NOT = { trait = senator }
            }

            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }

        random_courtier = {
            limit = {
                OR = {
                    has_job_title = job_chancellor
                    has_job_title = job_marshal
                    has_job_title = job_treasurer
                    has_job_title = job_spymaster
                    has_minor_title = title_quaestor
                    has_minor_title = title_censor
                    has_minor_title = title_tribune
                    has_minor_title = title_aedile
                }
                NOT = { trait = senator }
                # We can not elevate not Nobles to be Senators!
                NOT = { dynasty = none }
            }

            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }
       
        # OK there is no elegible character let's create a new one
        create_random_steward = { # Always the same character of "Tusculum" dynasty is created!
            random_traits = yes
            dynasty = random
            female = no
        }

        new_character = {
            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }
    }
}

 

[knuckey]

I've noticed another problem that has arisen since 2.4. Previously the commands join_attacker_wars and join_defender_wars would allow you to add mercenary rulers to another ruler's war. This no longer works, the following code for example does not actually join the targeted mercenary ruler to a war:

random_independent_ruler = {
limit = { primary_title = { mercenary = yes } }
join_attacker_wars = ROOT
}

 

[Rylock]

I've noticed another problem that has arisen since 2.4. Previously the commands join_attacker_wars and join_defender_wars would allow you to add mercenary rulers to another ruler's war. This no longer works, the following code for example does not actually join the targeted mercenary ruler to a war:

random_independent_ruler = {
limit = { primary_title = { mercenary = yes } }
join_attacker_wars = ROOT
}

My experience with that is that mercenaries are not included in the any_independent_ruler/random_independent_ruler scope, period...so it's not that join_attacker_wars won't work for them, it's that your scope doesn't include them.

 

[knuckey]

Ah, didnt realise that scope had that problem, but that was only an example I picked to demonstrate the problem. Even if you scope to a specific merc company it does not work. For example I have a merc company called d_golden_company, and the following does not work:

d_golden_company= {
holder_scope = {
join_attacker_wars = ROOT
}
}

The tooltip shows up, but the effect does not go through.

 

[zijistark]

Sorry. That's what I get for not reading carefully -- you did send a separate event to each province, which I assume would have worked. Clearly not.

The event either needs to be deferred (e.g., delayed by a day), or instead of using any_* to fire off events directly, the event should be refactored to be tail-recursive so that it invokes its own next instance (thus providing a new RNG seed) until it runs out of target things to invoke itself upon. At least, that's how randomness currently works with anything rooted in an any_* effect.

EDIT: Folks, remember to write recursive events / chains with much care unless you like corner cases that cause the stack to explode and CK2 to crash.

 

[ngppgn]

The event either needs to be deferred (e.g., delayed by a day), or instead of using any_* to fire off events directly, the event should be refactored to be tail-recursive so that it invokes its own next instance (thus providing a new RNG seed) until it runs out of target things to invoke itself upon. At least, that's how randomness currently works with anything rooted in an any_* effect.

EDIT: Folks, remember to write recursive events / chains with much care unless you like corner cases that cause the stack to explode and CK2 to crash.

Could you put an example of such recursove event, and how could it be basly coded?

 

[Captain Gars]

In this case too random is not really working (the new character in the end of the event has always the same dynasty! But the dynasty should be random):

character_event = {
    id = republic_rev.3
    title = EVTNAMErepublic_rev.3
    desc = EVTDESCrepublic_rev.3
  
    is_triggered_only = yes

    option = {
        name = EVTOPTOKrepublic_rev.3

        random_realm_character = {
            limit = {
                is_republic = yes
                NOT = { trait = senator }
            }

            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }

        random_realm_character = {
            limit = {
                AND = {
                    is_feudal = yes
                    is_landed = yes
                    is_female = no
                }
                NOT = { trait = senator }
            }

            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }

        random_courtier = {
            limit = {
                OR = {
                    has_job_title = job_chancellor
                    has_job_title = job_marshal
                    has_job_title = job_treasurer
                    has_job_title = job_spymaster
                    has_minor_title = title_quaestor
                    has_minor_title = title_censor
                    has_minor_title = title_tribune
                    has_minor_title = title_aedile
                }
                NOT = { trait = senator }
                # We can not elevate not Nobles to be Senators!
                NOT = { dynasty = none }
            }

            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }
      
        # OK there is no elegible character let's create a new one
        create_random_steward = { # Always the same character of "Tusculum" dynasty is created!
            random_traits = yes
            dynasty = random
            female = no
        }

        new_character = {
            # Let's elevate this character as Senator
            character_event = { id = republic_rev.4 }
            break = yes
        }
    }
}

How do you mean? Do you spawn X version of this event from another event and they all get the same dynasty, or do you spawn this event independently but every time you do the dynasty is always the same?

 

[fanoI]

Yes I'm calling it using debug console as "testevent republic_rev.3" the character dynasty is always the same the COA is random generated any time I test the event. This would change if I would change from another event or with a real trigger? This event should happen when a Senatorial dynasty dies out and so a seat on the Senate is empty so in future probably it would have a trigger.
However how random dynasties are generated? Using place names (so I get always Tusculum because in Rome there are not a lot of Baronies), using culture based names transformations (Claudius -> Claudianus) or using dynasties that are in history (and there are not a lot of dynasties in history with Roman culture)?

 

[nicowaves]

Hey, I think I found a bug with the end_war command. richvh suggested I should post it here instead of the bug forum, so I'll just paste it then.

[2.4.5] [HONO] Ending a war via "end_war = success" in decision makes the war end inconclusively

Description:
I wrote a decision that ends any war my character is an attacker in with a win. But, when I execute the decision, I get a notification that the war ended inconclusively and I get no war reparations, contested titles etc.

DLC:
Literally every one.

Mods:
I put together a version of my mod that only consists of the mentioned decision and nothing else, here is the code of the decision:

decisions = {
    win_war = {
        potential = {
            ai = no
            war = yes
        }
        allow = {
            is_ruler = yes
        }
        effect = {
            any_war = {
                limit = {
                    attacker = { character = ROOT }
                }
                end_war = success
            }
        }
    }
}

With this decision the checksum should be HONO.
No other mods were used.

Reproduction Steps:
1. Put the decision above in the game
2. Start the game
3. Load the game with any character & bookmark
4. Declare war on a character
5. Execute the decision

Observe: The "War has ended inconclusively" notification pops up and even though it said in the decision that your character would win, it is as if the war never happened.

 

[illathid]

@Captain Gars
It appears that the local_speed_modifier from patch 2.1.5 isn't working for provinces. When we put "local_speed_modifier = 10.0" in a province modifier that gets assigned on start up it shows this:

Spoiler

and it has zero effect on travel times to and from the province. Did we do something wrong or is it not working? Any help would be appreciated!