Ai, Ai, Ai!
Posted: Mon Apr 08, 2019 11:58 am
I was planning to just batch these up and report then under RIP but the list was getting long and I wanted to start the reports, so I am starting this new topic for AI, auto_*, and AGH. Many of these will be chronic - perhaps most.
AI 0. Fearless.
I originally reported this elsewhere but left out some of the details:
If for an attacking unit the products attack * hitpoints < 3 AND firepower * hitpoints < 3 then AI's do not bother building defenders. Renders the current fc2.6.0 distro for civ1 ruleset totally useless for SP use.
I have not fully characterised this bug. It is not entirely obvious (to me) what is going wrong. Superficially it looked like wrong simple test (x > 1 instead of x >= 1) but the evidence indicates products are involved (and maybe not just the ones reported above.) In any event the problem should not be too hard to find, and be fixed with a general rework of the section. This is no longer critical because of the general avoidance technique.
As for the "hardcoded 'feature'" that I referred to: I am not sure whether this is a bug or was a genuine feature. In case it's not obvious the problem is that non native attacks (e.g. ship against shore target) have the firepower of both attacker and defender reduced to 1. In civ1, as modified with HP & FP set to 2 the battle outcomes include the posssibility of the winner being left with HP = 1 typically requiring two turns for repair. There are, of course, implications for all other rulesets but they will not be so obvious.
AI 0. Fearless.
I originally reported this elsewhere but left out some of the details:
If for an attacking unit the products attack * hitpoints < 3 AND firepower * hitpoints < 3 then AI's do not bother building defenders. Renders the current fc2.6.0 distro for civ1 ruleset totally useless for SP use.
I have not fully characterised this bug. It is not entirely obvious (to me) what is going wrong. Superficially it looked like wrong simple test (x > 1 instead of x >= 1) but the evidence indicates products are involved (and maybe not just the ones reported above.) In any event the problem should not be too hard to find, and be fixed with a general rework of the section. This is no longer critical because of the general avoidance technique.
As for the "hardcoded 'feature'" that I referred to: I am not sure whether this is a bug or was a genuine feature. In case it's not obvious the problem is that non native attacks (e.g. ship against shore target) have the firepower of both attacker and defender reduced to 1. In civ1, as modified with HP & FP set to 2 the battle outcomes include the posssibility of the winner being left with HP = 1 typically requiring two turns for repair. There are, of course, implications for all other rulesets but they will not be so obvious.