Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[PatrikLundell]

What about just retiring then un-retiring? Would that help at all?

I just tried it, and DF died on saving the retired fortress (both with and without DFHack)... so there's no save to unretire. It's only one world, though, but still an indication. Wouldn't exactly be surprise if DF is unable to package up the corrupted data.

Abandoning worked, though, and reclaiming with default embark parameters worked (with the surface being an absolute mess of items once inside the fortress, as well as an immediate "ambush" from the locals). Initially the specific equipment in all categories except weapons were empty (as nothing was brought). Reclaiming the junk on the surface populated the lists with legal items (no corruption seen) in all categories except shield (empty) and weapons (standard tools only still). Presumably the shield and weapons are inside the fortress down in unrevealed areas. Thus, embarking afresh might work, as might abandoning/reclaiming (assuming you're prepared with the mess that's involved with reclaiming, but I guess you can send off all citizens except one to other places (unless they have offsite kids) before abandoning so you don't have to slaughter the locals on reclaiming.

Regarding the structures Roses mentioned, they seem to be used for squads, and I assume DF crashes when DF tries to place an unsuitable object in those structures. We probably need to get at the lists from which DF selects those items to DFHack around the issue.

[Clément]

I'd like to know if my Qt experiment works on macOS (I am worried by this). But, since I cannot use macOS myself, I need help from a macOS user who knows how to compile DFHack plugins.

Download and compile the plugin from here. You will need to install Qt (homebrew can help with that) and pass its installation directory to cmake with -DCMAKE_PREFIX_PATH (-DCMAKE_PREFIX_PATH=/usr/local/opt/qt for homebrew installation).

Try enabling the plugin ("enable qapplication"), then try commands like "qt-info" or "qt-log". Please tell me if windows appear and if DF survives it. Also check the content of "qt_log.txt" in DF directory (or the content of the log window if it works).

[fortunawhisk]

@fortunawhisk: There's no lack of saves with the bug (including a couple of mine). Most of the crash reports for the current version on the bug tracker is caused by this bug. Someone with a tag I can't catch (ruisuif or something like that) went through the crashing saves and verified the cause in the majority of those (finding the corruption of the data and verifying the repeatable crashes went away of all militia was disbanded). Thus, there's not much point in more saves containing this bug.

Adding a link to the equipment corruption / crash bug report. There ARE a lot of linked saves.
http://www.bay12games.com/dwarves/mantisbt/view.php?id=11014

[Roses]

Regarding the structures Roses mentioned, they seem to be used for squads, and I assume DF crashes when DF tries to place an unsuitable object in those structures. We probably need to get at the lists from which DF selects those items to DFHack around the issue.

I was worried that might be the case

[Rumrusher]

so I guess if you want to raid in this buggy state you have to remove all items from your military uniform and send them out?

[PatrikLundell]

so I guess if you want to raid in this buggy state you have to remove all items from your military uniform and send them out?

No, the bug probably doesn't have to do with uniforms in the sense of a set of equipment, but rather with uniforms in the sense of the equipment allocated to militia in their capacity of militia. A uniform is just a template of selection criteria for items, and as far as I know these criteria get copied to the militia member when the uniform is allocated, so there's no difference in the militia squad info between a selection of a metal helmet and a user selection of a metal helmet for the head slot. Sure, civilian clothing and no shield nor weapons might not provoke the bug (we don't know for sure that civilian clothing doesn't corrupt things as well: after all, it also gets removed from the fortress and then gets returned to it when the buggers return), but their survivability isn't that great should it come to blows (but a squad that doesn't return probably doesn't generate corruption). If I was to gamble, I'd try to never load a save where any troops are outside of the embark and see if the corruption is caused by restoring offloaded data, as it might be that data that's retained in the item structure is fine, but I haven't heard of that being tested. The only safe action currently is never to raid.

[Atarlost]

Is there some way to truly erase trees?  I tried using tiletypes to turn them into empty air and they reappear the next time plants are processed and "plants extirpate" seems to no longer be implemented. 

They're kind of a problem when they're in edge connected water where I'd need to drop my dam to drain the lake. 

[Bumber]

Is there some way to truly erase trees?  I tried using tiletypes to turn them into empty air and they reappear the next time plants are processed and "plants extirpate" seems to no longer be implemented. 

They're kind of a problem when they're in edge connected water where I'd need to drop my dam to drain the lake.

Fire a ballista at them.

[PatrikLundell]

Bumber beat me to it...

Is there some way to truly erase trees?  I tried using tiletypes to turn them into empty air and they reappear the next time plants are processed and "plants extirpate" seems to no longer be implemented. 

They're kind of a problem when they're in edge connected water where I'd need to drop my dam to drain the lake. 

It sounds that your problem is cavern trees standing in the lake. My standard method to get rid of those is to fire at them with a ballista: A ballista arrow hitting a tree will disintegrate it completely. It's even possible to fire at tree stems under water if you've got a drain to protect the ballista operator. Note that any part of the tree can be hit for the tree to disappear, so most of the time you don't need to attack them from beneath.

It's possible to remove them from using DFHack, but I don't know if there's a command for it. Changing the tile doesn't help, as you've discovered, because the tree exists in the plants list as well (there's a bug involving repeated cave-ins due to maturing trees that were caved in originally, as DF apparently missed to remove those trees, so when growing they pop into existence and grow in the air, and then falls as they find there's no ground supporting them).

[zaporozhets]

Changing the tile doesn't help, as you've discovered, because the tree exists in the plants list as well (there's a bug involving repeated cave-ins due to maturing trees that were caved in originally, as DF apparently missed to remove those trees, so when growing they pop into existence and grow in the air, and then falls as they find there's no ground supporting them).

That's really interesting, and quite pertinent to something I'm working on, couldn't you just remove the tree from the plants list to prevent this from happening?

[Rose]

A script request, if anybody feels up for it.

Add a clothing item sized for each member of your fort to the manager screen, to help with multiracial forts.

[PatrikLundell]

@zaporozhets: Here's the removingflyingtrees script written some time ago to deal with trees caving in.

Spoiler (click to show/hide)

Code: [Select]

function removeflyingtrees ()
  local count = 0
 
  for i, plant in ipairs (df.global.world.plants.all) do
    if plant.tree_info ~= nil and
   dfhack.maps.getTileType (plant.pos) == df.tiletype.OpenSpace then
  plant.tree_info = nil
  count = count + 1
end
  end

  dfhack.println (tostring (count) .. " flying trees removed")
end

removeflyingtrees ()


In that case the only thing that needed to be removed was the plant entry. If you want to remove a specific tree you need to both get rid of it from the plants list and get rid of the tile info. However, I don't know if there's something else that would have to be dealt with as well.

[zaporozhets]

Thanks. Appreciated.

[rel1c]

Not sure where to ask this as I've scoured google for an answer..

I'm looking to cure only infection from my adventurer. I tried forking full-heal, but alas I can't seem to figure it out. I can get a list of wounds in unit.body.wounds, but they just look like memory addresses. Example:

<unit_wound: 0000025C0216BC10>
<unit_wound: 0000025C0216BD60>
<unit_wound: 0000025C0216C540>
<unit_wound: 0000025C0216BCF0>
<unit_wound: 0000025C0216BF90>
<unit_wound: 0000025C0216D420>

Any pointers? Thanks

[lethosor]

Not sure where to ask this as I've scoured google for an answer..

I'm looking to cure only infection from my adventurer. I tried forking full-heal, but alas I can't seem to figure it out. I can get a list of wounds in unit.body.wounds, but they just look like memory addresses. Example:

<unit_wound: 0000025C0216BC10>
<unit_wound: 0000025C0216BD60>
<unit_wound: 0000025C0216C540>
<unit_wound: 0000025C0216BCF0>
<unit_wound: 0000025C0216BF90>
<unit_wound: 0000025C0216D420>

Any pointers?

Looks like you found the pointers just fine to me!

Anyway, those are pointers to objects. You can access properties of each individual object, e.g. wounds[0] will be a unit_wound object. I've done this enough that I'm not really sure how to explain it better than that. If that doesn't help, can you post exactly what you're doing to get that output?