Print

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

[Kat]

Hi people great seeing DFHack still being updated.

Anyway I have something I am struggling with in my game, and I am wondering if any of you have ever come across this, and what solution if any, you found.

So, I have temples in my fort. I want to assign at least one performer to each of the temples.

I can find citizens with dancing skills etc via Dwarf Therapist. But... I can't tell which gods those citizens worship, except by viewing their unit info. So I can't easily tell which citizens to assign to which temples.

is it possible, or would it be possible in future, to display or sort the available performers, so that you can find ones that worship the god of the temple ?

Here's a picture of what I mean:
https://i.imgur.com/F6K2uNt.png

[PatrikLundell]

@Kat: It's possible to write a DFHack script that does what you want, but you'd likely have to (learn to, if needed) write it yourself, given that you'd need a fair bit of luck to find someone else who shares your particular needs and who's interested in writing it.

[lethosor]

Couple questions;

1. Does the defclass() support overloading operators? I know with "normal" metatables you can put functions for things like __add, __eq, etc... Just making sure that I can still do that.

Looks like it:

Code: [Select]

[lua]# x = defclass()
[lua]# function x:__add() print('adding') end
[lua]# print(x() + 2)
adding
nil


2. I'm trying to decide how to organize my scripts and functions collection and am wondering what some of my options might be. Currently I am adding expanding on some of the current df-structures objects by using classes. As an example, I have a UNIT class which adds functions like UNIT:makeProjectile(vx,vy,vz) which turns the unit into a projectile. Now for the question, my current thoughts for organization are 1) Putting the defclass and all the functions associated with that class in their own file, or 2) put the defclass for each of the different classes in a single file with descriptions of each class, and then put the functions for the class in a seperate file, 3) group classes and functions into categories and create files for each category. Is there a "best" way to handle this in terms of performance (I'm guessing #3)? If they are all roughly equal I was thinking of using #2 because it would be the easiest to read and create help files for. Is there another way that I'm not thinking of?

I wouldn't split up the defclass and function definitions - defclass is really just one line. Personally, I would go with #1, unless you have a lot of classes, and then #3 might be a bit faster. One disadvantage of #2 is that if you forget to include the file that defines all the functions, you'll just end up with a class that does nothing.

3. I noticed in some of the scripts in the repository there is the line

Code: [Select]

if not dfhack_flags.module then
    main(...)
endJust wanted to make sure I understand this, if a script gets called from the command line or with run_script/run_command it will call main, if it gets access with reqscript() it will not call main, correct?

Yup! It's the same as the "moduleMode" check you might see in some scripts, but that's deprecated.

[Rumrusher]

just checking dfhack r3 and oh finding out units just have caste flags in their enemy section is a deep life savor. this opens up for like getting creatures and reanimated beings that don't talk to uhh talk.

[darkhog]

Any timetable on how soon dfhack can be updated for the new version?

[lethosor]

We can't make any guarantees - we don't know exactly how much has changed that concerns us yet. It will probably be at least a week before we have something partially usable, but even then it will likely be unstable for a while longer.

If anyone's interested, we do have an IRC channel where you can see more up-to-date activity (see the first post for details - just please don't bug us there with questions about when DFHack will be updated).

[jecowa]

Thanks for keeping us apprised.

[PatrikLundell]

I wouldn't mind if a post was made in this thread when the first cut of the (re)mapping has been made so "amateur" mapping activities can be started.

[HungThir]

is it possible to build dfhack usefully in an msys2 shell?  i think this is an abi compatibility question, i don't really understand how the mingw/msys2 posix/win64 translation layers work or to what extent stuff built with the mingw toolchain is able to interact with native win64 apps.  i'm a linux/posix programmer, and would like to be able to contribute, but i don't run linux on my gaming pc, i don't run dwarf fortress on the work laptop, and i don't know thing one about the windows toolchains, really

i could potentially figure out (and document) how to build it under msys2, if i knew before sinking time into it that it would actually work. but my current understanding is that it wouldn't work. am i right?

[Clément]

The issue is that MSVC and GCC C++ ABIs are incompatible (there is no standard C++ ABI). It is not about posix/win32 translation. You are stuck with whatever compiler Toady uses.

[HungThir]

The issue is that MSVC and GCC C++ ABIs are incompatible (there is no standard C++ ABI). It is not about posix/win32 translation. You are stuck with whatever compiler Toady uses.

gotcha, cheers

[Quietust]

The issue is that MSVC and GCC C++ ABIs are incompatible (there is no standard C++ ABI). It is not about posix/win32 translation. You are stuck with whatever compiler Toady uses.

My understanding was that it was mainly a runtime library issue, specifically the memory layouts of std::string/std::vector and the fact that DF's heap is managed by the VC++ 2015 runtime library (which MingW does not use).

[Clément]

By C++ ABI, I meant both compiler and standard library ABIs. On the compiler side you'll get issues with name mangling, class layout, call conventions, ...

[Roses]

Is there a way to check if a dfhack structure has a specific member? What I mean is that if I have a normal table I can do "if rawget(table,key) then..." or "if table[key] then..." but trying the first one results in the error that a table is expected but got userdata, and the later gives the error that the field is not found.

[PatrikLundell]

The method I've used to detect if a field is present (to determine if the DF version is the current one or an older one) is to iterate over the fields with "pairs". If it's present I'll eventually get a match.

Code: [Select]

  --  Backwards compatibility detection
  -- 
  local river_type_updated = false
  if true then  --  To get the temporary variable's context expire
    local river = df.world_river:new()
    for i, k in pairs (river) do
      if i == "flow" then
        river_type_updated = true
        break
      end
    end
    river:delete ()
  end