Print

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

[k33n]

The reason why PatrickLundall tries to shift the conversation to "anyone else" coding it is simple:

"If you absolutely want something done, do it yourself."

Something they themselves follow by coding scripts for dfhack.

Reworking the job management system and abandoning the VPL is scheduled. But it is not scheduled for next release - what's upcoming is artifact hunting, rework and myth generation.
But even if it was scheduled for next release doesn't mean it'd be available now, or that having a choice between it and non-vanilla UI wouldn't be nice, no matter the state of the UI now.

And even if I didn't know that, I have no real expectation for Toady One to do it, or for any updates tbh - I don't have a date to circle in calendar, that I'm eagerly waiting for. Though as relatively newer player, I suppose my perspective differs from someone who has done it all.

It is an issue because nothing has annihilated the player-base to this extent since I have been playing. Playing ostrich and attacking valid feedback is the least productive of all options.

[Fleeting Frames]

"Playing ostrich"? English as third language here, not familiar with it...

Regarding for your concerns regarding player-base...I dunno how you track it, but at least donations for 2016 are the greatest they've ever been, by a substantial jump: The Bay 12 Games Report, January 1st, 2017 - that, to me, suggests playerbase has increased, not decreased. What do you base the claim of DF losing players on?


As for the feedback being valid...Proper constructive criticism explains what to improve and the reasons behind it.
Your suggestion of baking in DT is valid, but probably half a decade+ old - it has been indirectly addressed (and rejected) in DF talk, even.

The rest you've posted has been hitting on "nothing is more important than good UI". If you want to perhaps address UI to Toady One, convincing could be easier with organized post introducing statistics on the importance of good UI, with testcases and guides on how to compare and make UI. He may not be working on jobs right now, but he is working on implementing conversations around artifacts.

Another more concrete idea, where one could perhaps actually help with suggestions to Toady...I don't expect his planned replacement of VPL to be a shining example of UI design, either. But one could suggest how to implement it well and what tests to use to compare different possible implementations. Haven't really thought about that myself, so it could be interesting and novel.

And if you feel attacked, I'm sorry - I've been trying to give tips to and push you towards getting you to implement your dream UI. None on the DFHack team are professional UI designers to my knowledge, so it could be an unbelievably massive improvement - at least that's what I hope.

[Sanctume]

There's some old cartoon where ostrich sticks their head in the ground, which is a way to ignore the world.
I think in this context, it's ignoring other's opinion? maybe. 

I actually learned to us u-z-p-l to toggle labors since playing 43.05 vanilla.  I don't even use recent alpha versions of dfhack.

It's a bit tedious, but I end up adapting my expectations and do ok after some play times.  It does get too much sometimes, but that is usually the time to save the game and take a break before I manually inspect the new wave of migrants. 

[PatrikLundell]

@k33n: The reason I tell YOU to implement your dream UI is that Toady most likely isn't going to create it for you, and the improvements he's going to make would appear sooner than expected if they appear within the next decade. So, if you want to see something happen in the near time you'll have to do it (or organize it) yourself.
We all know the DF UI isn't good and Toady has admitted to not being good at UI design. Yet another person saying it won't change anything, especially since all you say is that it's bad and should be done better, which isn't providing any useful information on HOW to make it better.

When it comes to bad UI design, just look at the vast majority of software out there: their interfaces and usability generally ranges between barely acceptable and horrible with the designers frequently not knowing how the thing is going to be used, targeting the wrong user base (often beginners at the expense of those who have to use it daily for hours, or the managers who are going to glace at what the peons have had to enter in the other end), and aiming for expressing their creativity rather than functionality (like far too many architects, really). Now, it's possible most software is produced by incompetents who don't know what they're doing and haven't got the smarts to employ UI professional designers.

[mikekchar]

how dare you

If you think it is a selfish thing to care about the game and its future, you can keep thinking that. I could say a lot about working on a team with someone who has your attitude, but I would rather not.

Back in the old days the sound <plonk> was used to show when someone went into your email filter so you wouldn't have to listen to them any more.  Too bad there isn't that option on boards like this.  In future, if you choose to quote me, *actually* quote me.  Don't make up your own stuff and attribute it to me.

I hope you will one day discover that your behaviour is not productive and will work to find ways to express yourself so that other people will take you seriously.  Like I said, I don't think you are completely wrong, just misguided.  I hope you'll work it out in the end, but I think I've reached my limit on this conversation.

[Thorfinn]

Not really the thread for it, but since people are asking for specific UI suggestions, something presumably easy I would really appreciate is if the Preferences screen showed what the cursor is resting on. For example, I'd like to be able to chop just the non-fruit-bearing trees first but that is too bleedingly tedious menu-hopping. Bonus points for showing a z-level indicator (relative or absolute) rather than making me < and > a half-dozen times trying to remember what the ceiling or floor was. Or maybe I should just stop making all those vaulted ceiling dwarfhomes with the balconies and whatnot. Naaah.

Oh, and if I could save the precious candlenut (or even willow, on some embarks) to use only on things I want light, like bins. Maybe there's an easier way than switching forbid status on the various piles, something like the build menu does, but I haven't found it.

[PatrikLundell]

- I've made a script that orders the chopping of all surface trees except one of each boozable fruit bearing species (saves one of each thread producing species as well, but there are none of those) as I got tired of manually looking for which trees to cut and which to save.

Spoiler (click to show/hide)

function closer (x1, y1, x2, y2)
  local dx1
  local dx2
  local dy1
  local dy2
  local max_x, max_y, max_z = dfhack.maps.getTileSize()
  local center_x = max_x / 2
  local center_y = max_y / 2
 
  if x1 < center_x then
    dx1 = center_x - x1
  else
    dx1 = x1 - center_x
  end
   
  if y1 < center_y then
    dy1 = center_y - y1
  else
    dy1 = y1 - center_y
  end
 
  if x2 < center_x then
    dx2 = center_x - x2
  else
    dx2 = x2 - center_x
  end
   
  if y2 < center_y then
    dy2 = center_y - y2
  else
    dy2 = y2 - center_y
  end
 
  return dx1 + dy1 < dx2 + dy2
end

function spareone ()
  local valuable

  for i, v in ipairs (df.global.world.raws.plants.all) do
    if v.flags.TREE and
      not v.flags.BIOME_SUBTERRANEAN_WATER and
      not v.flags.BIOME_SUBTERRANEAN_CHASM and
      not v.flags.BIOME_SUBTERRANEAN_LAVA then
     valuable = v.flags.DRINK or
                v.flags.THREAD
     local found = false
     local skipped = nil
     local skip
     local sapling_found = false
         
      for k, plant in ipairs (df.global.world.plants.all) do
      if plant.material == i then       
        local cur = dfhack.maps.getTileBlock (plant.pos)
        local x = plant.pos.x
        local y = plant.pos.y
        local z = plant.pos.z
         
        if cur.tiletype [x % 16] [y % 16] == df.tiletype.TreeTrunkPillar or
           cur.tiletype [x % 16] [y % 16] == df.tiletype.TreeTrunkNW or
           cur.tiletype [x % 16] [y % 16] == df.tiletype.TreeTrunkNE or
           cur.tiletype [x % 16] [y % 16] == df.tiletype.TreeTrunkSW or
           cur.tiletype [x % 16] [y % 16] == df.tiletype.TreeTrunkSE then
         if cur.designation [x % 16] [y % 16].dig == df.tile_dig_designation.No then
              current = df.global.world.job_list
           skip = false
 
              while current ~= nil do
                if current.item ~= nil and
                   current.item.job_type == df.job_type.FellTree and
               current.item.pos.x == x and
               current.item.pos.y == y and
               current.item.pos.z == z then
              skip = true
              break
               end
   
               current = current.next
              end
                  
           if not skip then
            for l, u in ipairs (df.global.world.job_postings) do
              if u.job ~= nil and
                u.job.job_type == df.job_type.FellTree and
                u.job.pos.x == x and
                u.job.pos.y == y and
                u.job.pos.z == z then
               skip = true
               break
              end
            end
           end
         
              if not skip then            
             if skipped == nil and
               valuable then
              skipped = plant.pos
              found = true
               dfhack.println ("Skipping (" ..  tostring (x) .. ", "  ..  tostring (y) .. ") " .. v.id)
             
            else
              if valuable then
                if closer (x, y, skipped.x, skipped.y) then
                 cur = dfhack.maps.getTileBlock (skipped)
                  x = skipped.x
                  y = skipped.y
                 z = skipped.z
                  skipped = plant.pos            
                   dfhack.println ("Swapping (" ..  tostring (x) .. ", "  ..  tostring (y) .. ") for ("
                                                 ..  tostring (skipped.x) .. ", " .. tostring (skipped.y) .. ")")
                end
              end
             
               cur.designation [x % 16] [y % 16].dig = df.tile_dig_designation.Default
              cur.flags.designated = true
              dfhack.println ("Designated " .. v.id .. " at (" .. tostring (x) .. ", " .. tostring (y) .. ")")
              found = true
            end
           end
         end
         
        elseif cur.tiletype [x % 16] [y % 16] == df.tiletype.Sapling then
          sapling_found = true
        end
      end
     end
    
     if valuable and not found then
       if sapling_found then
        dfhack.println ("Found " .. v.id .. " sapling")
      else
        dfhack.println ("Failed to find " .. v.id)
      end
     end
   end
  end
end
 
spareone()

- You can zoom to desired locations by hooking them up to F1-F8 (and something beyond that) via the "H" command.
- A clunky way of saving a particular kind of wood (or stone, or ...) is to specify what material to use in the work orders.

[Fleeting Frames]

Neat! I've so far used prospect and getplants myself, then latter d-ting the whole map and d-xing the orchard.

@Thorfinn. There are three z-level indicators on the right. Image that you're probably familiar with.

For forbidding all your candlenut, stocks screen would work, provided you have a bookkeeper. If not, dfhack stocks plugin would also work.

[Werdna]

Oh, and if I could save the precious candlenut (or even willow, on some embarks) to use only on things I want light, like bins. Maybe there's an easier way than switching forbid status on the various piles, something like the build menu does, but I haven't found it.

Just throw a wood stockpile (set to Willow/Candlenut only) next to and link it to ('Give') a Carpenter workshop.  That's what I do!  No forbidding needed.  Nobody poaches the Willow when it's linked to the Carpenter.  I use a separate Carpenter shop for non-willow items.


My main problem with DT right now is that the "hover" menu on the dwarves has gotten so long after 10 years of play that many of the dwarves have their top information scroll right off the screen. 

[Thorfinn]

Thanks for the quick replies.

I was holding off on dfhack for the moment. As a (recovering) programmer I'd much rather try to understand the software as the coder wrote it first. (So why am I posting comments about DF in a DT thread?)  For my purposes, if alder and ash used the same white "O" that birch does, rather than the olive "O" that cherry does, that would be plenty. Maybe all I need is to tweak my tileset...

Re: candlenut, I guess bookkeeper/work orders would suffice. Just do it with forbid until I can afford to have a dwarf sitting in an office to give permission to make a bed for the new guy? Werda, I take it the general wood stockpile gives to the general carpenter shop, too? I guess I haven't been tinkering enough with the give option. I'll give it a try. Might be just the thing.

Re: Z-levels, not exactly what I was looking for. When I'm on a given level without a floor, (k) says "Open Space". It would be nice if it said "Open Space (3)" telling me at a glance there are 2 Open Spaces below it to the next floor.

[Werdna]

I take it the general wood stockpile gives to the general carpenter shop, too? I guess I haven't been tinkering enough with the give option. I'll give it a try. Might be just the thing.

No.  My experience is that once you link a stockpile to a workshop (via Give, or Take at the workshop), that workshop will only take from that stockpile until you unlink them.  When I link a Willow-only stockpile my Carpenter shop, if I specify a "Build Door" and specify Alder, the job will fail instead of going to the general stockpile (I forget if you get a cancellation, or the order just sits in the workshop eternally). 

If I wanted to be able to build Alder or Willow doors, I'd need to add a Give-link from a 2nd Alder stockpile to the Carpenter shop - but then anything I build (without additional details) will be randomly Alder or Willow.  Linking additional stockpiles only works if they are all Giving to the workshop; using a Take at the workshop will use only a single stockpile.

Edit: please note one drawback of this "specialized workshop" system is that the job manager has no understanding of this, and complications will arise if you issue an order for say "Make 30 Oak beds".  By default the Willow Carpenter shop will still receive the orders, but fail to complete them.  You need I think to manually take the workshop out of the job manager's list though from the workshops' menu, I forget the sequence.

[Thorfinn]

Edit: please note one drawback of this "specialized workshop" system is that the job manager has no understanding of this, and complications will arise if you issue an order for say "Make 30 Oak beds".  By default the Willow Carpenter shop will still receive the orders, but fail to complete them.  You need I think to manually take the workshop out of the job manager's list though from the workshops' menu, I forget the sequence.

Is there a good reason to use the manager system rather than just going to the various shops and queue them up? My experience with it is that adding that step just adds a sometimes significant delay before I get those beds.

[PatrikLundell]

Linking a stockpile to a workshop will only force that workshop to take from that stockpile (or any other stockpiles linked), but it does NOT prevent any other job from stealing the precious kinds of wood. There's take from anywhere/links only option for stockpiles, but not any give to links only.

At the current state I find no use for the manager (as discussed in a parallel thread), but others do. I would like the manager to work "properly", at which time I'd definitely use it.

I believe the raws specify colors, so you could probably hack the trees to be color coded the way you want without resorting to DFHackery. DF doesn't use that many colors, though, so you'd still end up with rough groups.

@Fleeting Frames:
If you like the spareone script you may like the getone as well. It designates one of each boozable and threashable plant for which there are no seeds and checks that the boozable/threshable parts are present (not specifying gathering of a strawberry plant with no berries).

Spoiler (click to show/hide)

function getone ()
  for i, v in ipairs (df.global.world.raws.plants.all) do
    if (v.flags.DRINK or
       v.flags.THREAD) and
      not v.flags.SAPLING and
      not v.flags.TREE and
      not v.flags.GRASS and
      not v.flags.BIOME_SUBTERRANEAN_WATER and
      not v.flags.BIOME_SUBTERRANEAN_CHASM and
      not v.flags.BIOME_SUBTERRANEAN_LAVA then
      local found = false
     local designated = false
     local skip = false
    
     for k, seed in ipairs (df.global.world.items.other.SEEDS) do
       if seed.mat_index == i then
        skip = true
        dfhack.println ("Skipping " .. v.id .. " as seeds are already available")
        break
      end
     end
    
     if not skip then
        for k, plant in ipairs (df.global.world.plants.all) do
        if plant.material == i then
          local cur = dfhack.maps.getTileBlock (plant.pos)
          local x = plant.pos.x % 16
          local y = plant.pos.y % 16
          local skip
         
          if cur.tiletype

• [y] == df.tiletype.Shrub then

           if cur.designation

• [y].dig == df.tile_dig_designation.No then

                current = df.global.world.job_list
             skip = false
 
                while current ~= nil do
                  if current.item ~= nil and
                     current.item.job_type == df.job_type.GatherPlants and
                 current.item.pos.x == plant.pos.x and
                 current.item.pos.y == plant.pos.y and
                 current.item.pos.z == plant.pos.z then
                skip = true
                break
                 end
   
                 current = current.next
                end
                  
             if not skip then
              for l, u in ipairs (df.global.world.job_postings) do
                if u.job ~= nil and
                  u.job.job_type == df.job_type.GatherPlants and
                  u.job.pos.x == plant.pos.x and
                  u.job.pos.y == plant.pos.y and
                  u.job.pos.z == plant.pos.z then
                 skip = true
                 break
                end
              end
             end
         
             if not skip then
               found = true
               designated = true
               for l, growth in ipairs (v.growths) do
                if growth.id == "FRUIT" and
                   (df.global.cur_year_tick < growth.timing_1 or
                   df.global.cur_year_tick > growth.timing_2) then
                  designated = false
                end
              end
            
              if designated then
                 cur.designation

• [y].dig = df.tile_dig_designation.Default

                 cur.flags.designated = true
                 dfhack.println ("Designated " .. v.id)
                 break
              end
             end
           end
          end
        end      
       end
     end
    
     if not skip then
       if not found then
         dfhack.println ("Failed to find " .. v.id)
       elseif not designated then
         dfhack.println ("Found " .. v.id .. " but it's not ripe")
       end
     end
   end
  end
end

getone()

It can also be mentioned that getplants has become slighly bugged by changes since it was written: if can add additional orders to cut trees even if orders are already present, leading to the woodcutter cutting the tree, walk away, walk back "cut" it again, walk away... for as many times as you've got extra orders.

[Thorfinn]

I believe the raws specify colors, so you could probably hack the trees to be color coded the way you want without resorting to DFHackery. DF doesn't use that many colors, though, so you'd still end up with rough groups.

Awesome goodness! Exactly what I was looking for. At least for this task. Thanks!

For anyone who would like to do this and doesn't know how: (it's really easy)

Spoiler (click to show/hide)

Go into your save folder and use your favorite text editor to open:

raw\objects\plant_standard.txt
   and
raw\objects\plant_new_trees.txt

Find whatever trees you want a different color (non-edible, non-fruit, whatever), move down a few lines to find:
   [BASIC_MAT:LOCAL_PLANT_MAT:STRUCTURAL]

and immediately after that, insert the line
      [DISPLAY_COLOR:7:0:1]


That's it.

[Werdna]

Is there a good reason to use the manager system rather than just going to the various shops and queue them up? My experience with it is that adding that step just adds a sometimes significant delay before I get those beds.

I tend to be in Patrick's camp, I don't frequently use it.  I think it can be useful, I've read some threads where it has been used to automate some really cool stuff.  There's a learning curve to it though if you want to avoid constant work cancellation spam.  I feel I've already got a system that works for what I need, so I rarely use it except for queueing up > 10 large orders that don't lend themselves to being placed on Repeat.  For instance, kitting up a new military squad with weapons and armor is at least 70 jobs or so, I throw that in the manager because if I do it through the workshop I eventually forget if I'm issuing orders for suit #6, or is it #7 now?

Pat, in my experience workshop A will never take from a stockpile that is linked to Give to workshop B.  For instance I have a Mason shop with a marble stockpile linked, in the same vicinity of my Smelters, and I could swear the Smelters regularly run out of marble for flux without ever stealing from that Mason's pile.  I'll double-check that when I get home.  But I forgot, outside of workshop jobs, it is still possible for those items to be used for other jobs, like being hauled to a Depot for trade, or being tasked to use in building a wooden stockade.