-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Hey guys, me and 'an acquaintance' are working on a mod project that adds new nations. We've currently created a new U99 in the North of Japan and it launches fine into the scenario but after unpausing the game DH instantly crashes to desktop. Since we don't get an error message we can't say what's causing the problem, probably a conflict with the new files somewhere. Is there a method to get a crash report out of DH?
Check savedebug.txt. You could also go to settings.cfg and enable "Extra logs" to get more error reports in savedebug.
In DH, anti-tank brigades are turned off.
Look in the db/units folder.... division_types... and also in the divisions folder for anti_tank.... You need to re-add the details of the unit here.
Check the infantry / armour techs files to see if the anti-tank is still there?
Set your settings.cfg file in the main game folder to 1, save it, and try to reload the game. Look in the savedebug file for all the errors you need to fix.
I ran through the savedebug but it shows up like everything's fined, no errors whatsoever. Still we get a CTD on the first hour, very odd. We're going to rework from an earlier build, we must have created some conflict with the nation modding.
Have you checked out:
INC file for U99?
revolt.txt file in the db folder for U99.. is there already another U99...?
The provinces allocated to U99... are they owned/controlled ... by another state already.. is there a conflict in the files?
If it opens and lets you play for an hour.. I'd suspect events
As I understand it when the production AI is set to relative build scheme = no, the AI will build according to the individual percentage ratio that you select by unit. However, when the AI is set to relative build scheme = yes, I suppose this is some sort of dynamic build AI based upon the current situation???
Unfortunately, after doing some investigations it seems when the scheme is active the builds are HORRIBLE. I have factored out all of the builds made through events and stripped it down to just the bare builds and I saw Japan, for example, putting like 85% of all IC into consumer goods with a zero dissent and very little in the build queue. In the same game I saw the Soviets researching CAG 1940 of all things, but not a carrier build in sight???
So here is the question...can anyone explain how the relative build scheme works? It seems to me it may be better to turn this off and set hard percentages. WIth the switch AI function couldn't you trigger different percentage builds based upon events and flags? Thanks to any who reply.
Unfortunately, after doing some investigations it seems when the scheme is active the builds are HORRIBLE. I have factored out all of the builds made through events and stripped it down to just the bare builds and I saw Japan, for example, putting like 85% of all IC into consumer goods with a zero dissent and very little in the build queue. In the same game I saw the Soviets researching CAG 1940 of all things, but not a carrier build in sight???
So here is the question...can anyone explain how the relative build scheme works? It seems to me it may be better to turn this off and set hard percentages. WIth the switch AI function couldn't you trigger different percentage builds based upon events and flags? Thanks to any who reply.
Yeah, if you look in the AI files, all the build schemes are set to "no". I've done similar experiments and agree with your findings.
You can of course use events to change the build scheme as the game progresses. You just need to put that part of the code into a new AI file, and call it by event. You put the new AI file in the switch folder, as files put there are additions to the basic AI file that is loaded on game start.
This is my understanding.. I'm sure if I'm wrong, someone will be along shortly to correct me...
That could be because Japan AI already has too big army/navy/air fleet for its IC base. There's a misc.txt modifier that defines needed/max produced supplies ratio after which the AI do not produce more units.
Unit types that an AI country builds are defined in the AI file. If it has carrier = 0, then no CV's will be built by that AI country.
Thank for the info, MartinBG. So does this mean whether the build scheme is yes or no does not matter if a max build amount criteria limited by the two things you reference have been met? In other words, does those limiting parameters take precedence over the build scheme?
How does the AI utilize the build scheme differently when "yes" versus "no". Seems to me "no" is the way to go and, while very cumbersome to write a gazillion events to change AI behavior as Novapaddy suggested, it will allow me to create a better build than selecting "yes". Thoughts?
As I thought I understood this would only be the case if the build scheme = no, correct. As I understood it the ratio of builds listed in the AI file would only be followed if build scheme = no based upon the comments in the modding documentation for AI files. With build scheme = yes those follwoing lines of specific AI build ratios are ignored I thought. Can anyone confirm or debunk my thoughts?
AI build ratios are always considered. Same applies to the misc.txt supplies ratio.
The difference between relative_build_scheme set to yes or no lays in the way a new build is ordered. IIRC when set to "no" a random unit (based on the build AI ratios; ratios set there define the chance for the unit type to be produced compared to all other unit types) will be produced, while on "yes" the AI will try to maintain the set build AI ratio (if you have infantry and militia set to 50 and currently the country has 10 inf and 9 mil, then a mil. will be produced; IIRC this calculation is messed up by taken/given expeditionary forces and that's why it is rarely used).
The difference between relative_build_scheme set to yes or no lays in the way a new build is ordered. IIRC when set to "no" a random unit (based on the build AI ratios; ratios set there define the chance for the unit type to be produced compared to all other unit types) will be produced, while on "yes" the AI will try to maintain the set build AI ratio (if you have infantry and militia set to 50 and currently the country has 10 inf and 9 mil, then a mil. will be produced; IIRC this calculation is messed up by taken/given expeditionary forces and that's why it is rarely used).
Based on your suggestions we got the first and second country working, thanks for that. Now, when trying to add a third country in the South of Japan we crash the secnario with an error 'provinces X Y Z are owned by Republic of Japan while Republic of Japan wants it'. We went through all the relevant misc-files and found no province conflicts, so the country seems to be conflicting with itself?
It seems to be that are .INC files of two different countries where these provinces are listed as "owned" / "controled"
It seems to be that are .INC files of two different countries where these provinces are listed as "owned" / "controled"
Yes, but it lists the conflicting nation as being the same as the owner, 'Republic of Japan'. The savedebug also shows that the third revolter 'Southern Free States''s provinces are all not bound to an owner, although the rest of the files should be fine. Oh god my head :blink:
EDIT: Nevermind, we found the issue. Our coder had put the wrong tags in the INC file causing the southern revolter to load the Republic of Japan tag aswell, causing the conflict. Thanks for the help! And as for our main coder, we shall get the flog and deal with this...
hmy:
Last edited:
That opened my eyes, thanks. I thought the build ratio was based on IC invested, meaning if you had the USSR Carrier = 2 that would indicate 2% of the IC would go towards building a carrier. Of course, your post just made me realize why my assumption was impossible because the only "partial" build of a unit is the last unit in the build queue once IC runs out. Thus my thought that there would always be slow progress towards building a carrier proved false. With that said, after my bitching about the AI researching the 1940 CAG, lo and behold a Soviet Carrier was added to the queue. Guess that 2% chance hit.
I cannot thank you enough for explaining this. I am re-energized to move forward with my plans.
Can anyone explain how the speed of an air unit is used? Is it a factor that modifies combat in some way?
Is there any way to send expeditionary forces via event? I could use switch_allegiance, but that's not quite... expeditionary.
As of now air speed is used only for travel between provinces, but there are some specifics in air movement (unit can travel with speed of 1 province per hour at most).
Let's say that we want to move air unit from prov. A to prov. B, distance 500km, both provinces are next to each other.
If the air unit has a speed of 100km/h then it will take 5 hours to complete the travel, with speed of 250km/h - 2 hours, with speed of 400km/h - 2 hours, speed 500km/h or more - 1 hour.
Now lets assume that there's province C between A and B (A -> C = 250km, C -> B = 250km, A -> B = 500km).
If the unit is with speed of 250km/h it will complete the movement in 2 hours (first from A to C and then from C to B), but the same will happen if the unit is with any other speed above 250km/h because it has to move first to province C (1 hour) and then to B (on the second hour).
Fnatastic!!! I did not think the value was used as I had always thought one province per hour was the norm. Good to know. Thank you for your multiple answers to my questions lately.