Print

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

[Warmist]

Also we could have a function that constructs a viewscreen from the helpstring. Then we could show (maybe interactive?) help screen.

[Dirst]

Also we could have a function that constructs a viewscreen from the helpstring. Then we could show (maybe interactive?) help screen.

At that point it might be easier to just launch a URL in the default browser?

[lethosor]

Yes, but not necessarily everyone has a web browser installed, or internet access, or bundled documentation. Also, getting a web browser to open on all systems is difficult (there's a good Python solution, but I haven't seen a Lua/C++ one).

[Warmist]

Also it's my quest to have everything df related in df
(like lua powered browser for wiki )

[Dirst]

Yes, but not necessarily everyone has a web browser installed, or internet access, or bundled documentation. Also, getting a web browser to open on all systems is difficult (there's a good Python solution, but I haven't seen a Lua/C++ one).

I was thinking browsing a local HTML file, but it sounds like even that would be difficult.

[mifki]

(like lua powered browser for wiki )

Oh, that's my dream.

[Bumber]

(like lua powered browser for wiki )

Oh, that's my dream.

And a directory navigator plus Nano text editor. (You probably couldn't mess with most DF files, but you could at least look at stuff.)

[FantasticDorf]

According to the df-structures it is possible to target fathers and mothers genes individually via a argument over egg fertilisation. (taking this out of context to force familiarial/inter-racial ties with similar attributes would be interesting)

You could have a setup in which you target egg laying parent A parent B (regardless of actual relation to each other) from the fortress population for their genes via cursor then implement into a selected individual or batch of eggs by forcing the 'fertilisation' and incubation time. For animals it doesn't matter due to no family trees so you could just find any animal of the eggs classification and essentially cheat-spore it as long as that animal is on the map.

It might be information you already know but three months time is equatable to 100800 ticks which is 2 months game time (even though the structure files reads it up as three, correct that discrepancy or my mistake please) 1200 is a day (1000 minutes on the worst possible lag and probably only a flat 100 or more on the best or good FPS, as time is disputably distortable)

Therefore with you can schedule the dfhack eggs or all eggs in general to be done early in more or less than a singular day similar to graze co-efficient in just speeding it up. Just negate - 10800 from the initial value of 0 for instantaneous results or something more conservative to have it done quickly. Elk birds can have their own script so they breed fast and spend less time starving on nest boxes without having buff all grazer co-efficient.

Other creatures seemingly can be contributed to egg results by changing the item references. A way to capture a creatures ID or reference and put it into a egg for a 're-birth' intelligent or not as a tame creature is a possibility

Here is the structure script

Spoiler (click to show/hide)

Code: [Select]

<class-type type-name='item_eggst' inherits-from='item_actual'>
        <int16_t name='race' ref-target='creature_raw'/>
        <int16_t name='caste' ref-target='caste_raw' aux-value='$$.race'/>
        <int32_t name='unk_7c'/>
        <compound name='egg_materials' type-name='material_vec_ref'/>

        <bitfield name='egg_flags'>
            <flag-bit name='fertile' comment='allows the incubation_counter to be checked/incremented'/>
        </bitfield>

        <int32_t name='incubation_counter' init-value='0'
                 comment='increments when fertile in unforbidden nestbox, hatch at >= 100800 (3 months)'/>

        <int32_t name='hatchling_civ_id' ref-target='historical_entity'
                 comment='hatchlings will have this civ_id'/>
        <int32_t name='hatchling_population_id' ref-target='entity_population'
                 comment='hatchlings will have this population_id'/>
        <int32_t name='hatchling_unit_unk_c0' init-value='-1'
                 comment='hatchlings will have this unit.unk_c0 value'/>

        <int32_t comment='v0.40.01'/> position uncertain

        <pointer name='mothers_genes' type-name='unit_genes' comment='object owned by egg item'/>
        <int16_t name='mothers_caste' ref-target='caste_raw' aux-value='$$.race'/>
        <int32_t comment='v0.40.01'/>

        <pointer name='fathers_genes' type-name='unit_genes' comment='object owned by egg item'/>
        <int16_t name='fathers_caste' ref-target='caste_raw' aux-value='$$.race'
                 comment='-1 if no father genes'/>
        <int32_t comment='v0.40.01'/>

        <compound name='hatchling_flags1' type-name='unit_flags1'
                  comment='used primarily for bit_flag:tame'/>
        <compound name='hatchling_flags2' type-name='unit_flags2'
                  comment='used primarily for bit_flag:roaming_wilderness_population_source'/>
        <compound name='hatchling_flags3' type-name='unit_flags3'
                  comment='not normally used, most/all do not stick'/>

        <int32_t name='unk_v42_1' comment='v0.42.01'/>

        <enum name='hatchling_training_level' base-type='int32_t'
              type-name='animal_training_level' init-value='WildUntamed'/>

        <compound name='hatchling_animal_population' type-name='world_population_ref'/>
        <int32_t name='hatchling_mother_id' ref-target='unit'/>
        <int32_t name='size'/>
    </class-type>


My own plant grow contribution i detailed some posts back is on hold until i get more information about the systems im learning about, time (which is more straightforward now as GROWDUR raws directly correlate to ticks i understand) and distance/co-ordinates (still working on learning about that, especially in relevance to workshops and fixed positions rather than globally) is helpful rather than a generic ++ to instantly grow.

[Heretic]

Multi Tileset load.
Just in cause of 64x bit version, where we can access to much more memory...
It will be very useful for streams of DF, different players love/better understand different tilesets, so if we can make in-game dfhack command to change tileset it would be really great!
I'm understand, that try to update running game will be really dangerous, slow and useless. But - the tilesets loaded in memory such as collection of images linked to in-game symbols, yes? That if we will make two sets of images loaded and relink them by the command?

[FantasticDorf]

[EDIT] Thought of one, thanks for the patience

Could you create a binary patch/plugin that properly divides int32_t in the DF structures codebase (or wherever is relevant) away from emotions and/or non sapient beings? (or at least apply some syntax)

• Forgive me im not the most personally code proficient but after reading the syntax, using that global vector may be interacting unintentionally with other pieces of code. (in lack of a better vector rather, moreso i guess i should be asking to give it a new set of integers instead)
• A suggestion for a new integer may end up being int16_t if it isn't too tricky or complicated to rework it around [CASTE] rather than [RACE] (unless someone like the zach brothers or very technically prowessed gave us a alternative integer to work with)

• Also defining the only relevant flags of the mind screen to be [INTELLIGENT] in regards to mental mind screens would be a possible solution so that emotions and higher thoughts only apply to them.
  
  • Additional help in pinpointing towards myself the relevant integers/code for morale (no quarter/brawl/lethal) would be greatly appreciated if i haven't covered it already in my research.
    
    Some speculative problem areas where this is happening involves  df.units animal training levels which would be consistent with animals (that possess non interface minds but active living unit status to use the mind screen, as the mind screen only stops when you are physically dead, but the mind screen remains open because death is not finality to the mind, as exploitatively entering dead skeleton's mind screen shows & DFTherapist) acquiring personalities within the wild state (as seen with depressed and needy animals with desires for interactions like worship, and sadness/lonelyness/anxiety over being restrained) and here in DF map where local features (etc, the population that lives there & the wildlife are all features) are also exposed to this vector.
    
    > I personally think i have strong evidence of local map features, when playing a civilization fortress with openly hostile relations to others, in instances where the wildlife reflects the ethical attitude of the settlement above it.
    > Dwarves who are attacked by wild creatures without weapons, and succeed in defeating their opponent never actually finish off the animal because its not sanctioned, it is treated as a brawl and the animal accepts the status and hobbles around being ignored by all other creatures (annoyingly in the case of item thieves), in modded goblin fortresses, the response is always lethal & in observing behaviour of non-intelligent creatures, the wildlife actively KILLS intelligent unarmed (gorlaks/plump-helmet men) intelligent species rather than crippling them in a draw.
    > Creature minds also extends over to using hand based non-weapon combat via scratching in the absence of more lethal forms of combat, usually this type of combat tells the unit more or less to begin a fistfight which runs parallel. Never have i observed a beak dog when uncommissioned from war actually bite a opponent rather than scratch it with its grasp/latching attacks.
    
    
  Code: ("particularly observable over any course of time with most animals) [Select]
          <enum-item name='NeedsUnfulfilled'>
            <item-attr name='caption' value='after [varying]'/>
        </enum-item>
        <enum-item name='Prayer'>
            <item-attr name='caption' value='after communing with [deity]'/>
</enum-item>
  Ideally out of the binary patch if my suggestions DO work and my research is correct.
  > Should castrate needs for prayer and emotional responses from animals beyond their usual code, which does contain baseline fear
  > Should underline the crossover points between  int32_t & int16_t to understand semi-intelligent behaviour (which is seemingly undefined clash of two coding standards, as they are both animal like but overwhelmingly more sentient since 42. onwards, to the point where collective local feature populations retain memories of attackers like intelligent peasants do)
  > Allow plugin users to poke around more for workarounds relating to issues like solving butchery of sapients/fortress 'members' (pets included) if the initial changes to clearing a sapient status from animals not defined as huntable (wild and non sapient, as the F forbid screen dictates) is cleared and it works as intented (best case scenario unless there is another issue elsewhere).
  > Most of these changes to be done within seperate plugins if possible in order to have control over effects
  
  I've spoke about this bug a little bit before in other places but it is my firm belief by observing game behaviour that they are all linked by this or something similar being unintentionally caught in a mix.

[lethosor]

I'm not really sure what you're saying. Are you saying that you think there's a bug that's caused by DF converting an int32_t to an int16_t somewhere? If so, linking to the report on Mantis would be helpful. We definitely cannot create a patch to change the layout of a structure in DF. (We could change our version of a layout in the XML files, but then our version would be wrong and probably cause crashes.)

Edit: actually, no, I don't know what you're saying at all.

[mcbucko]

Would it be possible to have an interaction run a script that checks the calender date and applies different effects based on what time of the year it is? ie, spring summer fall winter, dry season wet season

[lethosor]

Yeah, there's a cur_season global that should be useful for that. If not, there's also cur_year_tick, which runs from 403200. There's also World::ReadCurrentMonth(), or dfhack.world.ReadCurrentMonth() in lua, which is probably best, although it requires cur_year_tick to be available, which it isn't currently on 64-bit Linux 0.43.05.

[FantasticDorf]

I'm not really sure what you're saying. Are you saying that you think there's a bug that's caused by DF converting an int32_t to an int16_t somewhere? If so, linking to the report on Mantis would be helpful. We definitely cannot create a patch to change the layout of a structure in DF. (We could change our version of a layout in the XML files, but then our version would be wrong and probably cause crashes.)

Edit: actually, no, I don't know what you're saying at all.

I apologise if im not the um, perhaps most concise person. I'll wind myself down a bit, i do get a little bit tangled when i have a lot of information to relay at once, i get caught up brainstorming and researching for a hour or two readability goes out the window. I have a bug report going on here http://www.bay12games.com/dwarves/mantisbt/view.php?id=10057 which is equally muddled from flexibly interpretive point of view going over overlapping problems of creatures not properly executing deaths before turning into buggy corpse items, and because of the problems within this thread discussion contribute to the combined issue (This thread issue < That bug report issue, but the two play off each other, to expand the bug report's effects and make it harder to patch and approach)

Spoiler (click to show/hide)

Some speculative problem areas where this is happening involves  df.units animal training levels which would be consistent with animals (that possess non interface minds but active living unit status to use the mind screen, as the mind screen only stops when you are physically dead, but the mind screen remains open because death is not finality to the mind, as exploitatively entering dead skeleton's mind screen shows & DFTherapist) acquiring personalities within the wild state (as seen with depressed and needy animals with desires for interactions like worship, and sadness/lonelyness/anxiety over being restrained) and here in DF map where local features (etc, the population that lives there & the wildlife are all features) are also exposed to this vector.

Ok stripping it all back here's my basic arguement

Its my belief that the emotion system defined within int32_t (from the syntaxed used on creature_raw and also being a undefined place to put unassigned integers) more specifically is affecting local fortress populations by interacting with enums like (<enum-type type-name='animal_training_level' base-type='int32_t'>) to first enable creatures to feel emotions and give them a 'mind' screen leading to desires like prayer, which is exhibited behaviour by allowing those types of creatures to feel emotions, leading to distraction and adopting behaviours of intelligent animals in a integer overlap (so you have your cause, but alone it cannot create harm it needs a catalyst). And creatures unless they are hardcoded in as huntable animals (boars, etc) cannot ethically eat a creature with a sentient mind (semi sentients fall particularly into this category as trolls and BCO's are not responding as they should with [large predator] but also expressing normally unfufillable desires, a troll might have god but you would never know when it goes to pray in a friendly goblin fortress.)

Secondly the enums

Code: [Select]

        <int32_t name='local_feature' init-value='-1' comment='index into world_data.region_map'/>
        <int32_t name='global_feature' init-value='-1' ref-target='world_underground_region'/>
<int32_t name='unk2' init-value='-1'/>
Serve as a catalyst to carry this emotional behavior to all the animals and units in a site locally, be they wild or civilized, this works for sentient life well but at the same time it affects all the wildlife. Emergent emotional/sentient behaviours like 'fistfight' unarmed come out when animals fight with claws and paws, and the actual ethics of the settling civ, affects the attitudes of all the wildlife (in my example i state that my ultra aggressive goblin civilization, because it has Kill neutral acceptable, animals on the cavern layer slaughter to death neutral sentients, while on a dwarven embark they only non-lethally brawl and part ways)

> There are more implications but that's the core root from my observation and personal research into it. I may be foolishly mistaking it out of a well meaning misunderstanding but im pretty sure the two are connected.  int16_t is my alternative suggestion for a integer to put emotions into instead (or somehow seperate the cause and the catalyst from one another intelligibly with a new integer or method with the zach brother's input)

A plugin or patch request to abstract that code and experiment with its effects would see if it could remedy a number of bugs (The butchery bug in my bug report request as to creatures not actually dying is relevant but if this bug was removed i think it would fix problems surrounding ethics in eating 'thinking creatures' and make pets/certain creatures butcherable again as well as patch behavioural issues) and eventually pathe a route to a absolute fix or temporary relief (if it works)

[Putnam]

int_32t is a type defined in the C standard and does not have any emotion system defined within it. Your entire issue seems to be based on a fundamental misunderstanding of what enums are.