Settler squad formed at gather point and don't do anything

[garrotte13]

Settler squad formed at gather point and don't do anything

[garrotte13]

[garrotte13]

one more squad. This time they shouldn't have pathfinding problems, because they are not on peninsula.

Frankly speaking, number of settler squads appearing on Migration AI state is something abnormal. 5 squads, then 1..2 more squads in 2..3 minutes. But hey, I set expansion settings for map once per 5..60 minutes and Rampant3.2 claims to respect it. And I have about 7 more squads generated by other clans at almost the same time. In total it's about 15 squads. It doesn't make any sense that settleCoolDown of 5..60 minutes will take place after that ... everything is populated already!

I suppose that this limit (maybe even say "unlimit") is killing entire idea of any reasonable limitation at some point when there are enough nests settled in generated chunks
base.maxExpansionGroups = mCeil((base.activeNests + base.activeRaidNests) / ALL_NESTS_PER_EXPANSION_GROUPS)
Maybe it's better to print out value base.activeNests + base.activeRaidNests in the line when printing points spending on settler squad.

[garrotte13]

This is console output for period of about 2 minutes. Too many settler squads really

I suppose something should be done with it. This is what Rampant.Fixed fork fans are talking about all the way.

[garrotte13]

Maybe it's better to print out value base.activeNests + base.activeRaidNests in the line when printing points spending on settler squad.

Found it in debug print out as "aN" and "aRN".
aRN typically goes from 250 to 1300. I don't know if it's big for other things, but for settler squads limitation it is dead end. Almost 10 settler squads allowed (e.g. 1300 divided by 150) for some clans. That's just a lot.

[garrotte13]

So when player sees vanilla map setting "Enemy expansion once per 10..60 minutes" he should expect 15+ real expansions once per 10..60 minutes.

[garrotte13]

GameSave for frozen settler squads:
https://drive.google.com/file/d/1y0WdGlwWqZvNZCVEOFLjEdhVau32LqiV/view?usp=sharing

[garrotte13]

Can it be the same problem, but after longer play and higher evolution?
https://mods.factorio.com/mod/Rampant/discussion/64289022eb306ae3d547c8a2

[veden]

At least one issue with the stuck biters has to do with the last unit not making it to the unit group.

[garrotte13]

Version e31c1e0

BOTH problems still exist:

1. sometimes settler squads form and don't go anywhere. Now I got situation when only one biter came to group gather point. Others lost their way or something like you wrote. Don't know how many biters were supposed to be in that group. Average settler squad size looks to be 7..8. This time I turned on both debug messages - settler squad formation spendings were printed for both, but settler settled message only for one of them.

2. A lot of settler squads per time period during MigrationAI state. For base.activeNests + base.activeRaidNests = 680 I got 4 settler squads formed by one clan in 5 minutes. Map enemy expansion settings are 5..60 minutes min..max interval, I don't suppose to be so lucky with exactly 5 minutes random outcome and divided by constant increased to 700 anyway it should be two, not four even if so "lucky" random outcome. Evolution = 40%.

[veden]

the second should be improved in the latest version and the first issue should be improved in the next version.

[garrotte13]

1. Great. Looks like fixed. Squad freezes for 3..7 real time minutes at gather point, then go to destination and settle. Sometimes it is about 10 minutes freeze. So a careful player has enough time to go and destroy this squad before settling attempt even starts.

[garrotte13]

2. Vengence settler group appearance is not limited by any biter expansion constants as I see.

[garrotte13]

2. Confirmed: both vengence and pollution-triggerred settler squads are generated ignoring expansion min time limitation.

Probably that is not a bug, but a feature, because it's a way settlers defend vs aggressive player, not an attack on peaceful player base. And both nesting actions are a product of low chance randomized calculations unlike high chance nesting during SiegeAI or MiratingAI.