v50 Steam/Premium information for editors
  • v50 information can now be added to pages in the main namespace. v0.47 information can still be found in the DF2014 namespace. See here for more details on the new versioning policy.
  • Use this page to report any issues related to the migration.
This notice may be cached—the current version can be found here.

Talk:Variable positions

From Dwarf Fortress Wiki
Jump to navigation Jump to search

Hi! nice work so far!

A few points:

  • I really wonder where the differences between human and goblins are coming from. Do you have any idea? I almost know for sure that the difference in MARKET-positions is because of humans have the CITY- type in DEFAULT_SITE_TYPE. But the others, I dont know. perhaps because of their values or ethics?
  • There is a lot of vague language in this article ('Some religious entities(...)', which ones?), but also to much certainty, at places :-) . I would suggest to make it more tight. ChatGPT or another text AI can be of good use there, to remove any ambiguity
  • The number of -1 is very strange to me, because this is an invalid value when writing raws for entities. Do you have any idea what the circumstances are for these positions? I see for example with the priests. Can it be that PRIEST is REPLACED_BY HIGH_PRIEST and also by highest priest? In that case, number -1 means: "currently replaced". It should not have a current holder. For example, can you check a site with a mayor, does the expedition leader has -1? in that case, its caused by replacement.
  • Landholders are a strange bunch, I'm curious if you find out anything about those!

Joostheger (talk) 12:22, 12 February 2026 (UTC)

Thank you.

  • Not sure. The thing is even if the goblins do overtake another site, no market officials will be created. So it is more likely an ethics or values thing. The same for the reduced number of different officials.
  • Well no. And the thing about "Some religious entities" in that case is, that some religious entities do not have any positions (eg. due to no variable position being created during world gen for these religious entities), but all of these entities (with no positions) will probably not create an abbot (CUSTOM_LAW_MAKER), but priest and so on (if the world-gen could/would be restarted). But without checking all religious entities without positions in detail, it might be that in some cases an religious entity (which could create an abbot) does not have a position at the end of world gen. So while there is from one perspective some ambiguity, from another perspective it is very precise.
  • The number -1 means that it is unlimited, meaning for that position an unlimited amount of assignments can be created.
  • No there is no replacement between the priest. Look at the responsibilities for priest and the others. Also the highest priest is capped at one (ie. like the pope), while their is an unlimited number of high priests possible (like cardinals or bishops).
  • So far all entries in the variable position were empty. So it seems that neither humans nor goblins are landholders. And I still need a little bit more familiarity with Lua, for a few more complicated scripts (which also correctly query the different abstract building types).

In general some info is still missing. And also I am in a few cases not sure yet, what exactly is the case, so some ambiguity might also stem from that. Only at the last step I will try to remove (unneccessary) ambiguity.

ps. At some point the commented out text passages should be removed (deleted), but if I would try to do it, it will be detected as possibly harmful and not be allowed.91.49.240.46 13:15, 12 February 2026 (UTC)

  • Ok ill help you remove the commented lines, but you really should create an account :-)
  • The number unlimited is written in the raws as AS_NEEDED. is that correct? If you, i'll suggest to replace the -1 with AS_NEEDED, to make that consistent with other positions and understandability
  • I dont think you'll find any other landholders by the way. I think it isnt in variable positions to have that. Landholders arent also neccesary to let the game run and simulate an interesint world, while these variable positions are.

Joostheger (talk) 14:52, 12 February 2026 (UTC)

  • Well I might eventually create an account, iff no email address (etc.) needs to be provided.
  • The translation of data into understandable form is not a priority, but -1 will be changed eventually (or a note added).

ps. I will probably not edit anything in the next few hours on the article, so now or maybe tommorrow might be a good idea to remove the lines currently commented out.91.49.240.46 15:21, 12 February 2026 (UTC)

Most of the "easy" work has been done. Next steps would be (not necessarily in the stated order): 1. replacing/removing the technical flags. 2. further research on what the concret requirements for the creation of different official and market official positions on site level is. 3. Figureing out the reason why goblins so far only create variable positions on site level, when the leader was a custom bandit leader (ie. at least in case of the forced admin it should be possible). 4. Figureing out why the goblins only create a subset of all possible variable official positions (both on site and civ level). 5. Researching why the goblins do not create custom market officials. 6. Figureing out why in some cases the dwarves use the generic forced admin and not the forced admin from the dwarven raws. 7. Removing any ambiguity (if possible) from the page and improve understandability. For that I suggest that with pointing to the unclear part a discussion entry is made, so that I am made aware of it and can then clear it up. 8. Improve formatting and table presentation: ie. possibly move some info to other tables to reduce table width (of the existing table). Or use smaller fonts (scaling) for some table entries. etc. 9. Figureing out (and testing) which other values in the tags [variable_positions] (and variable_site_positions) are valid. I do not think that it is only [variable_positions:all]. (This won't be done by me). 10. Add section(s) for modifing the game data, so that in an existing world a few more variable positions are created (and filled with historical figures). So that the variable positions then function as if the variable positions had already been created (and filled) during world gen.

I am probably missing a few things. But that is all for now, from my part. Further edits (by me) to the variable positions page will not be done before next week.91.49.240.46 17:32, 13 February 2026 (UTC)

a few thoughs on your questions: 3: I quess that the most important factors are their DEFAULT_SITE_TYPE and their values. 4: same. See: [[1]]. 5: Because there are no markets in the dark fortress. 6: I think that the entity MOUNTAINS (dwarves) always uses the raw-forced administrator. I think that what you have seen, are probably not entity type 1, but maybe like outcasts or villains or whatever. 9: I can see what i can do Joostheger (talk) 21:10, 17 February 2026 (UTC)

Well the dwarves are (probably) supposed to only use the forced administrator as defined in the raws, but they are not.

In one case the site ruled by the forced administrator was originally a human site, which was then taken over by a (dwarven) necromancer. Afterwards a dwarven civ defeated the necromancer forces and formed a new government with a forced administrator (with precedence 100). And the government still rules at the end of world gen.

The other was originally a dwarven site (hillock). It was then taken over by a necromancer (again dwarven). The dwarves (original owners) then took the site back and put a forced administrator in place (again with precedence 100). But the site did change its owner again afterwards.

To 3.: The goblins can take over other sites, so that is unlikely the cause. And would also not explain why the goblins with a custom bandit leader do create some officials, but the forced admin does not. (It is a totally different question, why the normal goblin site leaders do not create any officials.)

To 4.: Well for one thing the ethics (and not the values) might play a role (ie. most things being a "personal matter"). For some other things (chef, grain official) it is probably either that the goblins do neither drink nor eat or it is (indirectly) because of that as they might not have some labors/jobs (see [[2]]). The later is probably also the reason, why they do not have doctors. The values on the other hand will probably play a role, whether the custom law maker is a diplomat (or has the military strategy responsibility).

To 5.: Again the goblins can take over other sites, so unless they always destroy the markets, when they occupy such a site, this cannot be the reason. But needs to be something else.

To. 9: The question is, whether the definitions allows multiple variable_position tags (resp. site variable position tags) or if only one is evaluated (and used). Furthermore the question is, whether only ALL and the name of a responsibility (like MANAGE_WORK_ORDERS) is allowed or if the name of the custom officials (as used as code) is also valid. And then a follow-up question would be, if ie. MILITARY_STRATEGY is written as entry, whether this will always create the CUSTOM_MILITARY_STRATEGY. And what about the cases, when an official has usually two responsibilities (like the "chef"), will that official be created with only one resp., if only one is given. Do both responsibilities need to be written (ie. with a comma as separator) in one tag. But that are all modding related questions (and not questions for modifying an existing world).

BTW: In many cases, when I looked into the history, a forced administrator was created usually after a site had been taken over by a necromancer (some time before). But I have not figured out yet, how I can easily check (from the entity definition), which group belongs to a necromancer and which not.91.49.240.46 16:44, 18 February 2026 (UTC)

Variable Positions and Quests[edit]

Quiet a few variable positions do have the same flags set as ie. the CUSTOM_BANDIT_LEADER (who is according to the Quest page) able to give out quests to adventurers, allthough some of the variable positions do not have any squads (which again according to the Quest page is necessary to be able to give out quests), but then again the question is, whether that information on the Quest page (which seems to have been more or less only been migrated from an older version to the current v5x version) is current (or has even been correct for the older version). It is probably a good idea if someone tests out in the adventure mode, if ie. the guild leader (dean/doyen), merchant corporation leader (and representative) and other variable positions (who are mentioned on this page to have the quest_giver and/or kill_quest flag set) are able to give out quests to adventurers and then ie. update the Quest page with the results.91.49.240.46 14:05, 7 March 2026 (UTC)

Holding multiple variable positions in world-gen.[edit]

Well while checking something else, I did come across a human necromancer, who currently holds 4 (variable) position assignments (and which are not former position assignments). The first three positions belong to an OUTCAST group, where the human holds the CUSTOM_OUTCAST_LT position (first assignment still active) and then two different CUSTOM_OUTCAST_FACTOR (assignments). The last position is the CUSTOM_LAW_MAKER (law-giver) of a human civilization (gained afterwards). The information is via dfhack and legends of a world directly after finishing world-gen (for the dfhack info a new game was started in the world and either before selecting the type of the game dfhack was used for analyzing the current game data or after selecting legends mode and analyzing the current game data with dfhack in addition to using the legends). In how far the necromancer is fullfilling all assignments is - of course - a totally different matter. An indepth analysis, how seldom/rare it is and which entities and position assignments are possible, will be done at a later date (possibly end of this month).91.49.243.26 11:52, 14 March 2026 (UT

Hello nameless one, love your work so far! Very impressive. About CUSTOM_OUTCAST_LT being only a single one, there is another possible reason discussed in the forums. Search for 'army' and 'lieutentant'. It is, that there is (almost) never a reason to create more than one army commander, despite the civ being able to do so. Joostheger (talk) 13:23, 15 March 2026 (UTC)

The squad size of the CUSTOM_OUTCAST_LT is zero (unless you do doubt the table entry), ie. in contrast to the usual army commanders. So the reason for no more than one CUSTOM_OUTCAST_LT has nothing to do, with being an army commander. To see for yourself, just ie. set the number for the OUTPOST_LIASION to AS_NEEDED and look whether you will get an OUTPOST_LIASION created (ie. the problem has not really anything to do with being an army commander or not, but with the AS_NEEDED or unlimited number, without the game knowing what condition to check, whether to create an (or another) assignment for the position, at least if the position is not tied to a number of locations/sites or the position being a landholder, which is again tied to the number of sites of a certain kind). And the reason for one CUSTOM_OUTCAST_LT getting created, instead of zero is already in the note.91.49.243.26 12:28, 16 March 2026 (UTC)

Addition: The only discussion I could find in the bay12 forums are from 2021 and thus concern v47 and not the current version. Besides considering that in my current game one human civ attacks itself nearly each year, shortly after their caravan returns home, in some cases with the attacker only a yak bull or yak cow (otherwise usually a single human), I do not think that any previous info about creation of army commanders is relevant, as yak bulls (and yak cows) are doing just fine as army commanders (of their one-yak-army). Or to put it in other words army commanders are currently created on the fly (otherwise in a world I did analyze there would be no way that a necromancer, with a tower, even if the necromancer would draft all position holders and itself as an army commander, could attack 20-30 different sites all at the start of a season of the same year; this btw. happened at least twice in the span of the 250 years of history in world-gen allthough with different necromancers). ps. If you are wondering the yak bulls/cows and the humans are previous traders, who visited my fortress.91.49.243.26 17:02, 16 March 2026 (UTC)

Here are the posts I was referring to. Do with it what you want.

It surprises me that CUSTOM_OUTCAST_LT has a squad of 0, despite its obvious responsibilities. Since this game is far from finished, my guess is that this is just an oversight. I would be cautious in drawing conclusions, because a lot of what you're working with is outdated, and hasn't been seen by the developers for years. another tip: in de old dev-diaries are hidden gems of information. See for example https://www.bay12games.com/dwarves/dev_2011.html and look for 'outcast'. Joostheger (talk) 09:56, 17 March 2026 (UTC)

Retrieved from "https://dwarffortresswiki.org/index.php?title=Talk:Variable_positions&oldid=315531"