Immortal v0.10c

Upload your latest creations here.
Please note that by posting comments to posts in this forum you forfeit your personal copyright on whatever you post. AI Scripters reserves the right to re-use objective comments in script reviews without requiring the posting author's permission. Credit, where appropriate will be given in such cases.
TheMaximalBeing
Forumer
Posts: 91
Joined: Fri Jun 02, 2017 11:00 pm
Location: Australia

Re: Immortal v0.10c

Post by TheMaximalBeing »

venomus wrote:
Tue Nov 07, 2023 1:32 am
Yes, Immortal keeps sending troops to the enemy base without defending against attacks along the way, causing it to lose everything.
Do you have a screenshot?

TheMaximalBeing
Forumer
Posts: 91
Joined: Fri Jun 02, 2017 11:00 pm
Location: Australia

Re: Immortal v0.10c

Post by TheMaximalBeing »

gmiwol wrote:
Tue Nov 07, 2023 2:50 am
It seems to attack the player with the highest score first. In multiplayer games, the early interfering troops wasted a lot of time marching because he kept switching the priority targets.
It should prioritize the closest player. Were there more than 4 enemies for the Immortal player? (this will break the player selection)

TheMaximalBeing
Forumer
Posts: 91
Joined: Fri Jun 02, 2017 11:00 pm
Location: Australia

Re: Immortal v0.10c

Post by TheMaximalBeing »

gmiwol wrote:
Tue Nov 07, 2023 3:22 am
It will attack the city gate, but if I only use wooden fences and other opportunities to clean it myself, or use houses to surround it, it will have a high chance of conducting invalid patrol actions.
It's not really possible to deal with these exploits properly (where an invalid path is created). There is no AI command that tells us whether a unit can be pathed to for attacking purposes.

gmiwol
Forumer
Posts: 35
Joined: Tue Mar 07, 2023 6:36 pm

Re: Immortal v0.10c

Post by gmiwol »

TheMaximalBeing wrote:
Fri Nov 10, 2023 4:17 am
gmiwol wrote:
Tue Nov 07, 2023 3:22 am
It will attack the city gate, but if I only use wooden fences and other opportunities to clean it myself, or use houses to surround it, it will have a high chance of conducting invalid patrol actions.
It's not really possible to deal with these exploits properly (where an invalid path is created). There is no AI command that tells us whether a unit can be pathed to for attacking purposes.
I entered all the replies here.

1. In the extremely difficult mode of the old version, AI started with four resources increased by 500 by default and needed to be reduced to the normal value.

2. I know that it is the AI of land. What I said last time was not complete enough. It was attracted by fishing boats on a land map with many small fishing ponds, so I feel that this also needs to be revised.

3. Frequent player switching attacks have occurred in 2AI vs 2AI and 5human vs 2Immortal+1other AI.

CSAxIsxRecruiting
Forumer
Posts: 5
Joined: Fri Nov 10, 2023 9:27 pm

Re: Immortal v0.10c

Post by CSAxIsxRecruiting »

Hey....

DE version playing in voobly it warning you it has an error in line 16300 and AI stop playing... Doesnt created vills or build things, only use them to explore...

The error says "Rule is too long..." This was tested playing Immortal vs others AIs in hardest mode.... In several maps


In other game, playing also on hardest, 4 humans against 4 AIs, 2 Bambis, 2 Immortals(version 0.9)... The bambis played well, the immortal... one played good, at the point it was hard to kill in some point, but the other one started attacking in dark age, then spammed "kill tower" messaged, he reached feudal and then he was stuck in feudal for so long.... he never reached castle age nor did something else, when game ended he barely reached 30 vills...
Last edited by CSAxIsxRecruiting on Sun Nov 12, 2023 5:39 pm, edited 2 times in total.

neilkaz
Forumer
Posts: 47
Joined: Fri Jul 15, 2011 10:36 am

Re: Immortal v0.10c

Post by neilkaz »

This AI is clearly the strongest as it wins more than it loses 3v4 Arabia vs the game's AI or Bright Spark. However, it doesn't seem quite ready to play the two new civs yet since it doesn't make any mule carts! It also failed to make Sevars for Persians even though it had over 40 cavs and tons of resources at max pop in post_Imp.

So it needs to be fully updated for the new X-pack changes.

thx ... neilkaz ..

gmiwol
Forumer
Posts: 35
Joined: Tue Mar 07, 2023 6:36 pm

Re: Immortal v0.10c

Post by gmiwol »

There is another problem. When there are more than two immortals, there will be a terrible delay that will make it impossible to continue the game.

jadeeye
Forumer
Posts: 31
Joined: Tue Dec 17, 2019 6:04 pm

Re: Immortal v0.10c

Post by jadeeye »

TheMaximalBeing wrote:
Fri Nov 10, 2023 4:11 am
1. I'm not sure what happened without seeing it. Regarding the treb issue - right now Immortal will not attack lone trebs as it needs to meet the minimum number of enemies. This will likely be fixed in the next version.

2. If the enemy has a castle then I think they should always train trebs. Are you sure there were none at all? Regarding the trebs attacking regular units -- yes it is a problem that will likely be fixed in the next version.

3. Did those monks have any faith left? Were you using an older version or the most recent one? I think the behaviour should be slightly better on the most recent version (but not perfect).
2. In a match, some civs didn't train Ram or Treb, some civs did that. For example, I have seen Vietnam and Spanish just use their archer-line attacked each other under their Castle.

3. Even if monks run out of faith, they should not move to enemies. My AoE 2 DE version is Dynasties of India.

clc8901
Forumer
Posts: 9
Joined: Sat Jul 30, 2022 12:12 pm

Re: Immortal v0.10c

Post by clc8901 »

Amazing update. I had some games where Khmer didn't research halberdier, onager, and chemistry.

screamingkoos
Forumer
Posts: 41
Joined: Wed Feb 05, 2020 6:27 pm

Re: Immortal v0.10c

Post by screamingkoos »

neilkaz wrote:
Sun Nov 26, 2023 8:03 am
This AI is clearly the strongest as it wins more than it loses 3v4 Arabia vs the game's AI or Bright Spark. However, it doesn't seem quite ready to play the two new civs yet since it doesn't make any mule carts! It also failed to make Sevars for Persians even though it had over 40 cavs and tons of resources at max pop in post_Imp.

So it needs to be fully updated for the new X-pack changes.

thx ... neilkaz ..
Adding the following code from Promi allowed my AI to research the Savar upgrade in DE if anyone is interested:

#load-if-defined PERSIAN-CIV
(defconst ri-paladin 526)
#end-if

Savar's work with knight-line in regards to any other rules you may have (for researching upgrades or training). Not sure if you would also have to defconst just "paladin" as well for training, example: (train paladin) to train savars. I only use the -line unit identifiers recognized in DE and did not see any mentions of savar from the promi code.

For mule carts, they are recognized as a building, so use (building-type-count) and (build mule-cart), and using the following strategic-numbers (listed below) at different times dependent on how you want them to be placed will be helpful (remember mule-carts can only have wood, gold, stone, or food from animals be dropped off into them). These strategic numbers act like how the town-center placement strategic numbers works:

(set-strategic-number sn-mule-cart-dropsite-placement 0) ;default placement
(set-strategic-number sn-mule-cart-dropsite-placement lumber-camp)
(set-strategic-number sn-mule-cart-dropsite-placement mining-camp)

Or if you use DUC for lumber-camp and mining-camp placement, then these may not be needed.

Hope this is helpful to someone! :)

Post Reply