Skynet AI

[bitterhowl]

but can´t control the reinforcements army

I doubt that the deal is in campaign script, I think there's a settings about that.

[mekkermann]

With the skynet AI I can control the reinforcements, with liquid steel I can´t. There where two armies one was set to be direct controllable one was set to be indirect controllable via general commands and stance (like defensive,agressive,ranged), the indirect controllable army did not react to any of my commands the direct controllable army did react, also I haven´t changed anything in the game settings so it has nothing to with them. Just to be clear you throw out a seemingly untested Battle AI here and when I find bugs you teel me it´s a problem with the implementation or with the settings on my side. Please try the stuff you post yourself and fix it, don´t rely on others to figure out your mistakes.

[bitterhowl]

Well, you're quite right. Meanwhile I thought that bug fixing process looks a bit different, but if you want to speak in such manner then ok. I think you have to stop worrying, be happy and return to SkynetAI.

[mekkermann]

Hello again,

I´m sorry, that wasn´t the best day of my life and surely an over reaction. I managed to get it going now. I used the Battle AI version released in the "a medieval 1191 mod" patch from 12 of August 2023. However I found some errors with the script checker in the Battle AI. https://www.twcenter.net/forums/showthread.php?445393 Nothing deeply concerning, basicly just missing = entries in the I_BattleSettlementFortificationLevel entries, so I_BattleSettlementFortificationLevel 1 instead of I_BattleSettlementFortificationLevel = 1 and some missing = for BattleEnemyUnitCategory entries, you might also check the hardcore BAI as I haven´t looked at it yet and don´t plan to use it for the moment.

So sorry again and have a nice day.

[Curt2013]

@mekkermann From my understanding the bai from bitterhowl is in beta and I actually jumped the gun by installing it in my 6.0 version of my mod like i told him in a pm on different site I thought his bai was further along then it was, so my bad on that. The current ai in 1191 mod is basically g5 scrips with bitterhowl xml's. I tested it out for a few hours and now in campaign and it works well. Any way hope that helps.

[mekkermann]

@Curt2013 Thanks for clarifying that. I saw by comparing the MEDM 6.0 and 6.1 campaign_scripts with an SS 6.4 campaign_script that these errors are indeed old G5 errors. I just saw that you mentioned a new BAI in the latest update, and was a bit overexcited I guess. I thought it would be a further evolution of the Germanicu5 AI by Bitterhowl while I looked at it during implementing and didn´t bother to compare the files with older AI files. From my limited understanding the "Invalid number of arguments for condition" errors will not necessary lead to a crash. I had plenty of them in my own scripts, which somehow worked, before the script checker was shown me last week. I guess it´s more that either the line gets ignored ( so the script wouldn´t check for certain conditions ) or missing =/>/< sign is replaced by a blank by the program, maybe = ( so the script would maybe check for the wrong argument ).
I think either way it´s better save than sorry, even if it´s an old bug. There was also a rare "deploy stakes ... error" bug when I used G5 AI last time will see if it´s fixed now.

Since you are here right now I also would recommend that you check the "a medieval 1191 mod" campaign_script with the script checker as well, I saw some similar error codes in it.

You also mentioned that some of your users complain about to long turn times the last time I saw you and that you shortened your campaign script because of that. Gigantus showed me a nice script last week that limits the numbers of agents per faction rather then per building this should help in late game to reduce the calculations for the agent spam. https://www.twcenter.net/forums/showthread.php?643718