Print

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

[vorpal+5]

I have a lingering question, about a minor DF lack that is quite annoying (obviously for me, but I dare believe it annoys others too?).

Why the (no job) and the (on break) icons in DF are the same, the small red/orange dot? Quite often I have to check if they are really idle, or just taking a well deserved rest... 

More generally speaking, does any plan exist to further improve DF? Like for example a filter on idle dwarves (only showing them?)

[schismatise]

More generally speaking, does any plan exist to further improve DF?

Current plan is:

• Slight hiatus
• Bugfix release
• Feature releases

Of course, I'll continue to try to provide layouts ASAP for new releases.

[vorpal+5]

ok thanks. 

[splinterz]

I have a lingering question, about a minor DF lack that is quite annoying (obviously for me, but I dare believe it annoys others too?).

Why the (no job) and the (on break) icons in DF are the same, the small red/orange dot? Quite often I have to check if they are really idle, or just taking a well deserved rest... 

More generally speaking, does any plan exist to further improve DF? Like for example a filter on idle dwarves (only showing them?)

don't worry, we're still reading the threads and keeping notes. it might not be done in the next 24 hours, but we're still aware of issues posted.

[thistleknot]

Thanks for splinterz for his patience & support!

Here is a script filter where you can use normalized probabilities to get results out of dwarf therapist!

By the way, this is exactly what Splinterz is working on for his next release of DT, except his means will be population based, and it will be WAY MORE user friendly than modifying a script.

Spoiler (click to show/hide)

Code: [Select]

var strAvg=1250; var strDev=380;
var agiAvg=900; var agiDev=300;
var tuffAvg=1250; var tuffDev=385;
var endAvg=1000; var endDev=372.5;
var recAvg=1000; var recDev=385;
var disRAvg=1000; var disRDev=380;
var anaAAvg=1250; var anaADev=363;
var focAvg=1000; var focDev=365;
var willAvg=1250; var willDev=390;
var creAvg=1000; var creDev=350;
var intAvg=1250; var intDev=392;
var patAvg=1000; var patDev=370;
var memAvg=1500; var memDev=390;
var linAAvg=1000; var linDev=355.5;
var spaSAvg=1000; var spaSDev=368;
var musAvg=1000; var musDev=386;
var kinSAvg=1000; var kinSDev=365;
var empAvg=1250; var empDev=393;
var socAAvg=1250; var socADev=393;

function findZ(x, mean, stdev){
  var result = ((x-mean)/stdev);
  return result
}

function normalcdf(X){   
var T=1/(1+.2316419*Math.abs(X));
var D=.3989423*Math.exp(-X*X/2);
var Prob=D*T*(.3193815+T*(-.3565638+T*(1.781478+T*(-1.821256+T*1.330274))));
if (X>0) {
Prob=1-Prob
}
return Prob
}

((
(normalcdf(findZ(d.strength(), strAvg, strDev))*.5)+
(normalcdf(findZ(d.agility(), agiAvg, agiDev))*.5)
)/1)
>.5

(
(normalcdf(findZ(d.strength(), 1250, 380))*.5)+
(normalcdf(findZ(d.agility(), 900, 300))*.5)
)/1
>.5

[splinterz]

well i've nearly got role columns finished:



it will rank dwarves based on the aspects for the role: attributes, skills and traits. each aspect can have it's own individual weight, and the groups themselves can also have a weight applied. this should give maximum flexibility in creating precise roles. additionally it's possible for traits to be rated inversely. for example if you want to find a very good liar, the less honesty a dwarf has, the better a liar they would be.

the roles are drawn the same as attributes, with dwarves better fitting roles as large black squares, and the worst dwarves with large red squares.

all roles are specified in the game_data.ini so it's very easy to add new ones once you have the syntax. for the role above the ini looks like this:

Spoiler (click to show/hide)

template:
name
attributes="<attribute group weight>::<attribute>:<individual weight>"
skills="<skill group weight>::<skill id>:<individual weight>"
traits="<trait group weight>::<trait id>:<individual weight>"

1/name="Chief Medical"
1/attributes="1::analytical ability:0.8, memory:0.3, intuition:0.3"
1/skills="1::60:0.8, 63:0.2"
1/traits="1::20:0.5, 23:0.5"

all weights are assumed to be 1.0 unless modified and are optional

finally against my better judgement i've also allowed a script="" entry for roles which will allow users to specify a custom script to generate values. however these are more difficult to draw and values outside of 0-100 will look funky.

i'd like to add a set of packaged roles as well, so if anyone wants to take the initiative and use the template provided (skill and trait id numbers are in the game_data.ini) i'm also still not sure on what the best weights would be. someone previously mentioned that skills > traits > attributes (i think?) would be the way to go. any other opinions or is this sound advice?

your input is greatly appreciated!

[walberg]

i'd like to add a set of packaged roles as well, so if anyone wants to take the initiative and use the template provided (skill and trait id numbers are in the game_data.ini) i'm also still not sure on what the best weights would be. someone previously mentioned that skills > traits > attributes (i think?) would be the way to go. any other opinions or is this sound advice?

your input is greatly appreciated!

I think I would recommend putting the weights in the .ini as well, and provide a "reasonable" (for some completely random definition of "reasonable", maybe) set of defaults in the distribution. That way anyone who wants to lean more on skills can tweak them that way, but if someone's more into attributes, that can be set up as well. Of course there's a couple downsides - the code will have to be considerably more forgiving and probably have a lot more (in-)sanity checks to write. Also, putting too much in the .ini might just get ridiculous at some point - maybe split the file into a couple different ones for different purposes (settings vs model parameters vs layouts, etc.)... And documenting this largish set of weights and what each one does would likely be a tad tedious.

[sumofallwars]

Wow That will be very useful.  I hope it can make it's way into the main project.
I am always amazed at the skills in this community.

[CatalystParadox]

This is brilliant!

Of course, my pie-in-the-sky +1 wish would be... would it be possible for this/some other utility to scan your initial 7 pre-embark too?  You know, so you know who to assign the medical skills to, who to give the brokerage to, who to make the leaders, who to give the weapon & armor skills, etc.?

[Kraggh]

What "trought auto-update" means?
 
I think there´s any button to do that.

So, what can I do to play 0.34.5, with Therapist?

Thanks!

[SalmonGod]

Your work is very exciting splinterz!

[MrX]

Hey everyone.

I'm having problems running DT on my 64bit Fedora 15 box. When I start the program I see the window for a second and then it just disappears. I have tried with both DF 31.25 and 34.05. The log seems to be saying something about memory layouts, but I have files in etc/memory_layouts/linux which have the same version number that DF has.
Here's the run.log:

Spoiler (click to show/hide)

2012-Mar-14 20:14:58.988 INFO   core   Dwarf Therapist "0.6.12" starting normally. [src/dwarftherapist.cpp:100] (setup_logging)
2012-Mar-14 20:14:59.041 INFO   core   Loaded 4 views from disk [src/viewmanager.cpp:128] (reload_views)
2012-Mar-14 20:14:59.053 DEBUG   core   "redrew views in 12ms" [src/viewmanager.cpp:160] (draw_views)
2012-Mar-14 20:14:59.055 DEBUG   core   setting up connections for MainWindow [src/mainwindow.cpp:105] (MainWindow)
2012-Mar-14 20:14:59.056 DEBUG   core   group_by now set to 0 [src/models/dwarfmodel.cpp:456] (set_group_by)
2012-Mar-14 20:14:59.056 DEBUG   core   beginning to read settings [src/dwarftherapist.cpp:122] (read_settings)
2012-Mar-14 20:14:59.056 DEBUG   core   finished reading settings [src/dwarftherapist.cpp:168] (read_settings)
2012-Mar-14 20:14:59.128 DEBUG   core   attempting connection to running DF game [src/mainwindow.cpp:216] (connect_to_df)
2012-Mar-14 20:14:59.128 ERROR   core   No valid memory layouts found in the following directories... () [src/dfinstance.cpp:106] (DFInstance)
2012-Mar-14 20:14:59.160 DEBUG   core   base_addr: 134531376 HEX 804c930 [src/dfinstancelinux.cpp:324] (find_running_copy)
2012-Mar-14 20:14:59.202 DEBUG   core   DF's checksum is "0xfc15065c" [src/dfinstancelinux.cpp:328] (find_running_copy)
2012-Mar-14 20:14:59.202 DEBUG   core   DF's checksum is: "0xfc15065c" [src/dfinstance.cpp:876] (get_memory_layout)
2012-Mar-14 20:14:59.202 DEBUG   core   Could not find layout for checksum "0xfc15065c" [src/dfinstance.cpp:883] (get_memory_layout)
2012-Mar-14 20:14:59.202 INFO   core   Checking for layout for checksum:  "0xfc15065c" [src/mainwindow.cpp:396] (check_for_layout)
2012-Mar-14 20:14:59.203 INFO   core   Dwarf fortress path: "/home/box/Games/df_linux" [src/dfinstancelinux.cpp:335] (find_running_copy)

Any clues?

[vityav]

Any clues?

I'm also running FC15 x64. Did you compile DT? And are you running it from in the DT directory? I had to copy the DwarfTherapist executable from <DT>/bin/release to <DT>/ and then run it from in the directory, though I don't know if that was because of your error or something else. I've just always done it since whatever my problem was popped up a year or so ago. Might be worth a shot?

[fredrol]

This is amazing, thank you all for your hard work! Can't wait for the next release!

Of course, my pie-in-the-sky +1 wish would be... would it be possible for this/some other utility to scan your initial 7 pre-embark too?  You know, so you know who to assign the medical skills to, who to give the brokerage to, who to make the leaders, who to give the weapon & armor skills, etc.?

I agree, would be an awesome feature.

[Kento]

Are there two places where nicknames are stored now? I keep on seeing old nicknames in my announcements, as if when I delete nicknames through Dwarf Therapist, they are stored somewhere.

Old nicknames also appear in dwarf descriptions and relations, but do not appear on the unit list. Changing the nicknames through the game itself seems to not have this problem.