Dwarf Fortress Bug Tracker - Dwarf Fortress
View Issue Details
0003802Dwarf FortressWorld Generation -- Warspublic2010-12-10 10:502018-12-18 04:14
0003802: World gen warriors make repeated uncontested attacks against empty sites
I have generated a world with an ongoing war that has lasted more than 900 years with dozens of different battles against a zero defense. The battles happen in most cases exactly a year after the previous battle, and it appears to be followed by an event where the site is conquered.

Gen a world until 1k and it could happen. I've only checked once, and only seen it once.

I'm attaching my world with this event. The war is called "The War of Targeting"

http://www.sendspace.com/file/sm2jom [^]
http://dffd.bay12games.com/file.php?id=11140 [^] (0.40.24)
Save Included
Issue History
2010-12-10 10:50Mason11987New Issue
2010-12-17 10:11Mason11987Note Added: 0014617
2010-12-18 11:00QuilNote Added: 0014623
2010-12-18 11:13FootkerchiefSummaryWars can last an indefinite period of time => World gen warriors make repeated uncontested attacks against empty sites
2012-01-12 15:52FootkerchiefNote Added: 0019363
2012-01-14 22:07DwarfuNote Added: 0019391
2012-01-14 22:47faiuwleNote Added: 0019392
2012-01-14 23:08faiuwleNote Edited: 0019392bug_revision_view_page.php?bugnote_id=0019392#r7268
2012-01-15 00:31DwarfuTag Attached: Save Included
2014-03-25 07:23DwarfuTag Attached: Fixed in 0.34.12?
2014-07-07 22:23FootkerchiefTag RenamedFixed in 0.34.12? => Fixed in 0.40.01?
2014-07-20 16:49SirPenguinNote Added: 0027078
2014-07-20 16:59DwarfuAssigned To => Dwarfu
2014-07-20 16:59DwarfuStatusnew => needs feedback
2014-07-26 11:57DwarfuTag Detached: Fixed in 0.40.01?
2014-07-26 11:58DwarfuNote Added: 0027635
2014-07-26 11:58DwarfuStatusneeds feedback => resolved
2014-07-26 11:58DwarfuFixed in Version => 0.40.01
2014-07-26 11:58DwarfuResolutionopen => fixed
2014-07-26 11:58DwarfuAssigned ToDwarfu => Toady One
2015-09-16 20:14lethosorNote Added: 0033052
2015-09-16 20:14lethosorAssigned ToToady One => lethosor
2015-09-16 20:14lethosorReproducibilityhave not tried => sometimes
2015-09-16 20:14lethosorStatusresolved => acknowledged
2015-09-16 20:14lethosorResolutionfixed => reopened
2015-09-16 20:14lethosorFixed in Version0.40.01 =>
2015-09-16 20:14lethosorAdditional Information Updatedbug_revision_view_page.php?rev_id=13306#r13306
2015-09-16 20:14lethosorNote Edited: 0033052bug_revision_view_page.php?bugnote_id=0033052#r13308
2018-12-18 01:29FantasticDorfNote Added: 0039025
2018-12-18 04:14FantasticDorfNote Added: 0039026

2010-12-17 10:11   
It appears this is caused by a civ without leadership that is holding a site without anyone located there. The site is repeatedly attacked, adn conquered, but for whatever reason it can't be captured.

So the civ without any living members continues.

I suspect the underlying bug is that civ's conquering sites, where the site is empty, and there isn't any leadership in the civ being attacked (everyone's dead) leads to the war never ending, and the site never changing hands.
2010-12-18 11:00   
I've seen something similar to this fairly often since updating to .18. A war will have several hundred instances of "attacker: umpteen squillion creatures. Attacker was uncontested" followed by "the 217th pillaging of Somewhere". Often there will be several attacks a year.

I'm thinking that the aggressor should be seizing or destroying these places, but isn't for some reason.
2012-01-12 15:52   
Reminder sent to: Mason11987

That save got removed. If you still have a copy, please upload it to http://dffd.wimbli.com/ [^] and post the link here.
2012-01-14 22:07   
rufio on the forums has reported about the same as Quil; I've asked him to upload a save.

http://www.bay12forums.com/smf/index.php?topic=99036.msg2904317#msg2904317 [^]

"For example, there's a war in my new world that has been going on for 150 years, and is basically just 80 repetitions of 1) goblin army attacks forest retreat, is completely uncontested and then 2) pillages forest retreat."
2012-01-14 22:47   
(edited on: 2012-01-14 23:08)
Yeah, that was me. Here is the save: http://dffd.wimbli.com/file.php?id=5367 [^]

If you go to the end of the age, there's a war with the last battle in 548 where a site was pillaged 79 times, and another one with the last battle in 546 where a different one was pillaged merely 28 times (though it took a few battles to initially wipe out the defenders and begin the endless pillaging). I don't know if either of the two victim elf civs are still extant outside of those sites.

2014-07-20 16:49   
Changes to both succession and site control have likely fixed this bug. I've genned a few worlds and have not noticed this bug - have others?
2014-07-26 11:58   
We'll go ahead and marked this fixed for the reasons specified by SirPenguin. If anyone finds this not to be the case, please PM a manager on the forums to reopen this report.
2015-09-16 20:14   
Witty managed to reproduce this in 0.40.24 - reopening.

http://dffd.bay12games.com/file.php?id=11140 [^]

2018-12-18 01:29   
I have a save from a modded world on 44.12 wherin two elven civs have driven a human civilization down to 19 human civilians total and raided between them, the same two sites between them 50 times, in a still yet unresolved conflict after breifly entering post-w.g and leaving it back to legends mode.

To make matters worse, the world was generated for 35 years, elves attacked the humans on year 1, while the second elf civ attacked them on year 2 and neither of them have made any progress in doing anything other than sending humans to extinction and pillaging.

http://dffd.bay12games.com/file.php?id=14155 [^]
2018-12-18 04:14   
Just to add, because of the nature of the modded save, you can change perspectives and play as the elven civilisation in fortress mode with the attached save raw's.

While the 'contact' (from hovering over a site in c view mode) currently reads the human civilization as 'war' (a recalculation post w.g maybe) on other generated worlds with ongoing wars at the beginning of a embark will have, 'Skirmishing' contact.

Toady might know more about the behaviour to comment, but post w.g war & conquering and during/pre worldgen skirmishing are vastly different. Since skirmishing by definition is meant to kill a few then usually peter out, like some examples of w.g wars do when the defenders win the first battle overwhelmingly, perhaps this is what happens when it oversteps and kills the diplomats (if any) too with no leader preservation but that's just reasonable speculation.

Siezing site leaders is on development goals, to keep them safe enough to agree to terms of defeat ought to address this.