* Minor underwear fixes and tweaks. * Allows ghost mentors to orbit around the target instead of changing their view to the mob's. * Ports some hud and update_icon code updates. * contamination nerf * Fix * Stops humanoids whose skin_tone variable is set to "albino" from showing up as pale when examined should their species not use skintones anyway. * Hello darkness my old friend. * Ports "Refactors can_see_reagents" * F * MKUltra made double opt-in * ROBUST OPTING MECHANICS * Adds a hacky framework for digitigrade polyclothes also adds a digitigrade state for polychromic shorts * also adds aphrodisiacs, slaps, all hypno as prefs * errors * Modernizes Polychromic Uniform Code No more modular_citadel, but the neck polychrome crap needs more work in another PR * oh god please don't mess up * git y u do dis ;-; * Revert "Makes HE pipes realistically radiate away heat" * more coat stuff yes yes * woops * Wew * Steos are bad * new wings and fluff * Forgot a sprite * another one * I swear I had this in here orginally. * Fixes mistake in less commit. * reattaches missing antenna * Effectively removes the long unrequired roboticist encryption key. * updates ratvar coat * magazine metal cost tweaks, speedloader in AIB * the very shitty idea: "muh speedloader" * Updating dynamic config, some balancing * Automatic changelog generation for PR #10282 [ci skip] * unfucks short hair 80s so the left facing sprite doesnt make you look like a crackhead * Automatic changelog generation for PR #10266 [ci skip] * ah, the wonders of atomization * [s] fixes a little door assembly glass dupe exploit * Update BoxStation.dmm * Update pai_defense.dm * fixes bad sprite states * Automatic changelog generation for PR #10242 [ci skip] * Automatic changelog generation for PR #10237 [ci skip] * Added a ghost role with its own spawner for the purposes of chillling in the ninja holding facility. * holy shit i hope this compiles * e * yee * Automatic changelog generation for PR #10128 [ci skip] * Nope, it didn't compile, Harold. * Automatic changelog generation for PR #10283 [ci skip] * Automatic changelog generation for PR #10281 [ci skip] * Automatic changelog generation for PR #10250 [ci skip] * Applied suggestions and minor tweaks. * f * yeah fair enough Co-Authored-By: Ghom <42542238+Ghommie@users.noreply.github.com> * Update engine_goggles.dm * changes * Automatic changelog generation for PR #10264 [ci skip] * Automatic changelog generation for PR #10298 [ci skip] * Good idea, tbh. * Small touch-up. * Nerfs armor and antistun * Automatic changelog generation for PR #10296 [ci skip] * let the adminbus commence * sprites * clothing add * fix * vending * loadout * FIXEdD * Automatic changelog generation for PR #10265 [ci skip] * Automatic changelog generation for PR #10289 [ci skip] * Automatic changelog generation for PR #10268 [ci skip] * Update CentCom.dmm * ehh fine but why would you use this proc for that * Update language_holder.dm * Automatic changelog generation for PR #10315 [ci skip] * cocooooonuuuuuuts * excludes certain ghost roles from being affected by health events * fixes the thing * probably fixes pandoras attacking their owners * Update stamina.dm * Update sprint.dm * Update food_reagents.dm * Automatic changelog generation for PR #10322 [ci skip] * does the fix? * What about this? * Wait no its actually the other way around. * weh * Automatic changelog generation for PR #10297 [ci skip] * Automatic changelog generation for PR #10299 [ci skip] * more weh. * even more weh. * all of the weh. * ah * Automatic changelog generation for PR #10217 [ci skip] * maybe fixes? * Don't use named arguments on new, please. * Made a line use the dang macro * Made sure ghost cafe visitors don't get midround antags * Automatic changelog generation for PR #10328 [ci skip] * Physiology so species changes dont fuck it up * Whoops * Aarggghh small fucking differences * Makes ass slap attempts fail more loudly on opted-out mobs * Automatic changelog generation for PR #10318 [ci skip] * Automatic changelog generation for PR #10091 [ci skip] * Update borghydro.dm * Update grass_carpet.dm * Ports reflector blobs Ports reflectors from /TG/ * mouth * Fixes UI for poly collars AltClick for collars was seemingly forgotten so I've added it back * auto ooc * roundend too * hnnnnnng code improvements * marshmallow * telescopic IVs * tiny fix * Automatic changelog generation for PR #10324 [ci skip] * Automatic changelog generation for PR #10323 [ci skip] * Update code/modules/hydroponics/grown/grass_carpet.dm Co-Authored-By: Ghom <42542238+Ghommie@users.noreply.github.com> * got it * Update engine_goggles.dm * [s] Fixes ghost anti-ghostrole timers a little. * chestbursters deleting brains fix. * Wait a minute... * ...which var are you? * Update belt.dm * tiny fixes * makes it a trait * few fixes * wow divide by 0 very cool * Automatic changelog generation for PR #10095 [ci skip] * Automatic changelog generation for PR #10312 [ci skip] * Automatic changelog generation for PR #10243 [ci skip] * Automatic changelog generation for PR #10336 [ci skip] * Automatic changelog generation for PR #10335 [ci skip] * Automatic changelog generation for PR #10334 [ci skip] * Automatic changelog generation for PR #10239 [ci skip] * Automatic changelog generation for PR #10168 [ci skip] * Automatic changelog generation for PR #10331 [ci skip] * Automatic changelog generation for PR #10320 [ci skip] * Maybe he means this? * CE Locker Exodus: The Removaling Removes junk from the CE locker. Adds empty engineering lockers. I don't know how to do this as a child and not have populate contents fill the child. * ok * Automatic changelog generation for PR #10348 [ci skip] * Automatic changelog generation for PR #10333 [ci skip] * changelog.html * desc * hnnnng * weh * did this actually change anything * Automatic changelog generation for PR #10337 [ci skip] * yee. * yee2. * Automatic changelog generation for PR #10339 [ci skip] * Ghommies fixes * Automatic changelog generation for PR #10306 [ci skip] * amazing. * Automatic changelog generation for PR #10357 [ci skip] * Automatic changelog generation for PR #10301 [ci skip] * Update chem_master.dm * Automatic changelog generation for PR #10123 [ci skip] * Automatic changelog generation for PR #10165 [ci skip] * Automatic changelog generation for PR #10338 [ci skip] * Automatic changelog generation for PR #10344 [ci skip] * Oh god oh fuck * Update code/modules/hydroponics/seeds.dm Co-Authored-By: Ghom <42542238+Ghommie@users.noreply.github.com> * Update code/modules/hydroponics/seeds.dm Co-Authored-By: Ghom <42542238+Ghommie@users.noreply.github.com> * Update code/modules/hydroponics/gene_modder.dm Co-Authored-By: Ghom <42542238+Ghommie@users.noreply.github.com> * Update telescopic_iv.dm * Automatic changelog generation for PR #10364 [ci skip] * afwafwadjiwao * Automatic changelog generation for PR #10365 [ci skip] * Automatic changelog generation for PR #10319 [ci skip] * Automatic changelog generation for PR #10345 [ci skip] * Automatic changelog generation for PR #10362 [ci skip] * Automatic changelog generation for PR #10309 [ci skip] * Makes lockers into subtypes Adds new lockers cleanly at the bottom instead and as subtypes. * dab * dab 2x * Automatic changelog generation for PR #10349 [ci skip] Co-authored-by: Ghom <42542238+Ghommie@users.noreply.github.com> Co-authored-by: Putnam3145 <putnam3145@gmail.com> Co-authored-by: Michiyamenotehifunana <31995558+Michiyamenotehifunana@users.noreply.github.com> Co-authored-by: Seris02 <49109742+Seris02@users.noreply.github.com> Co-authored-by: Trilbyspaceclone <30435998+Trilbyspaceclone@users.noreply.github.com> Co-authored-by: shellspeed1 <46614774+shellspeed1@users.noreply.github.com> Co-authored-by: Hatterhat <31829017+Hatterhat@users.noreply.github.com> Co-authored-by: kevinz000 <2003111+kevinz000@users.noreply.github.com> Co-authored-by: CitadelStationBot <citadelstationcommunity@gmail.com> Co-authored-by: Detective-Google <48196179+Detective-Google@users.noreply.github.com> Co-authored-by: LetterJay <9606363+LetterJay@users.noreply.github.com> Co-authored-by: Lin <linzolle@gmail.com> Co-authored-by: Useroth <37159550+Useroth@users.noreply.github.com> Co-authored-by: kappa-sama <44128284+kappa-sama@users.noreply.github.com> Co-authored-by: Arturlang <arturpaavooskari.lang@gmail.com> Co-authored-by: loginsandylogout <nintendofan356@gmail.com> Co-authored-by: BlackMajor <henrydyer6@hotmail.com> Co-authored-by: keronshb <54602815+keronshb@users.noreply.github.com> Co-authored-by: KathrinBailey <53862927+KathrinBailey@users.noreply.github.com> Co-authored-by: nik707 <38332985+nik707@users.noreply.github.com>
Datum Component System (DCS)
Concept
Loosely adapted from /vg/. This is an entity component system for adding behaviours to datums when inheritance doesn't quite cut it. By using signals and events instead of direct inheritance, you can inject behaviours without hacky overloads. It requires a different method of thinking, but is not hard to use correctly. If a behaviour can have application across more than one thing. Make it generic, make it a component. Atom/mob/obj event? Give it a signal, and forward it's arguments with a SendSignal() call. Now every component that want's to can also know about this happening.
In the code
Slippery things
At the time of this writing, every object that is slippery overrides atom/Crossed does some checks, then slips the mob. Instead of all those Crossed overrides they could add a slippery component to all these objects. And have the checks in one proc that is run by the Crossed event
Powercells
A lot of objects have powercells. The get_cell() proc was added to give generic access to the cell var if it had one. This is just a specific use case of GetComponent()
Radios
The radio object as it is should not exist, given that more things use the concept of radios rather than the object itself. The actual function of the radio can exist in a component which all the things that use it (Request consoles, actual radios, the SM shard) can add to themselves.
Standos
Stands have a lot of procs which mimic mob procs. Rather than inserting hooks for all these procs in overrides, the same can be accomplished with signals
API
Defines
COMPONENT_INCOMPATIBLEReturn this from/datum/component/Initializeordatum/component/OnTransferto have the component be deleted if it's applied to an incorrect type.parentmust not be modified if this is to be returned. This will be noted in the runtime logs
Vars
/datum/var/list/datum_components(private)- Lazy associated list of type -> component/list of components.
/datum/var/list/comp_lookup(private)- Lazy associated list of signal -> registree/list of registrees
/datum/var/list/signal_procs(private)- Associated lazy list of signals ->
/datum/callbacks that will be run when the parent datum receives that signal
- Associated lazy list of signals ->
/datum/var/signal_enabled(protected, boolean)- If the datum is signal enabled. If not, it will not react to signals
FALSEby default, set toTRUEwhen a signal is registered
/datum/component/var/dupe_mode(protected, enum)- How duplicate component types are handled when added to the datum.
COMPONENT_DUPE_HIGHLANDER(default): Old component will be deleted, new component will first have/datum/component/proc/InheritComponent(datum/component/old, FALSE)on itCOMPONENT_DUPE_ALLOWED: The components will be treated as separate,GetComponent()will return the first addedCOMPONENT_DUPE_UNIQUE: New component will be deleted, old component will first have/datum/component/proc/InheritComponent(datum/component/new, TRUE)on itCOMPONENT_DUPE_UNIQUE_PASSARGS: New component will never exist and instead its initialization arguments will be passed on to the old component.
- How duplicate component types are handled when added to the datum.
/datum/component/var/dupe_type(protected, type)- Definition of a duplicate component type
nullmeans exact match ontype(default)- Any other type means that and all subtypes
- Definition of a duplicate component type
/datum/component/var/datum/parent(protected, read-only)- The datum this component belongs to
- Never
nullin child procs
report_signal_origin(protected, boolean)- If
TRUE, will invoke the callback when signalled with the signal type as the first argument. FALSEby default.
- If
Procs
/datum/proc/GetComponent(component_type(type)) -> datum/component?(public, final)- Returns a reference to a component of component_type if it exists in the datum, null otherwise
/datum/proc/GetComponents(component_type(type)) -> list(public, final)- Returns a list of references to all components of component_type that exist in the datum
/datum/proc/GetExactComponent(component_type(type)) -> datum/component?(public, final)- Returns a reference to a component whose type MATCHES component_type if that component exists in the datum, null otherwise
GET_COMPONENT(varname, component_type)ORGET_COMPONENT_FROM(varname, component_type, src)- Shorthand for
var/component_type/varname = src.GetComponent(component_type)
- Shorthand for
SEND_SIGNAL(target, sigtype, ...)(public, final)- Use to send signals to target datum
- Extra arguments are to be specified in the signal definition
- Returns a bitflag with signal specific information assembled from all activated components
- Arguments are packaged in a list and handed off to _SendSignal()
/datum/proc/AddComponent(component_type(type), ...) -> datum/component(public, final)- Creates an instance of
component_typein the datum and passes...to itsInitialize()call - Sends the
COMSIG_COMPONENT_ADDEDsignal to the datum - All components a datum owns are deleted with the datum
- Returns the component that was created. Or the old component in a dupe situation where
COMPONENT_DUPE_UNIQUEwas set - If this tries to add an component to an incompatible type, the component will be deleted and the result will be
null. This is very unperformant, try not to do it - Properly handles duplicate situations based on the
dupe_modevar
- Creates an instance of
/datum/proc/LoadComponent(component_type(type), ...) -> datum/component(public, final)- Equivalent to calling
GetComponent(component_type)where, if the result would benull, returnsAddComponent(component_type, ...)instead
- Equivalent to calling
/datum/proc/ComponentActivated(datum/component/C)(abstract, async)- Called on a component's
parentafter a signal received causes it to activate.srcis the parameter - Will only be called if a component's callback returns
TRUE
- Called on a component's
/datum/proc/TakeComponent(datum/component/C)(public, final)- Properly transfers ownership of a component from one datum to another
- Signals
COMSIG_COMPONENT_REMOVINGon the parent - Called on the datum you want to own the component with another datum's component
/datum/proc/_SendSignal(signal, list/arguments)(private, final)- Handles most of the actual signaling procedure
- Will runtime if used on datums with an empty component list
/datum/proc/RegisterSignal(datum/target, signal(string/list of strings), proc_ref(type), override(boolean))(protected, final)- If signal is a list it will be as if RegisterSignal was called for each of the entries with the same following arguments
- Makes the datum listen for the specified
signalon it'sparentdatum. - When that signal is received
proc_refwill be called on the component, along with associated arguments - Example proc ref:
.proc/OnEvent - If a previous registration is overwritten by the call, a runtime occurs. Setting
overrideto TRUE prevents this - These callbacks run asyncronously
- Returning
TRUEfrom these callbacks will trigger aTRUEreturn from theSendSignal()that initiated it
/datum/component/New(datum/parent, ...)(private, final)- Runs internal setup for the component
- Extra arguments are passed to
Initialize()
/datum/component/Initialize(...)(abstract, no-sleep)- Called by
New()with the same argments excludingparent - Component does not exist in
parent'sdatum_componentslist yet, althoughparentis set and may be used - Signals will not be received while this function is running
- Component may be deleted after this function completes without being attached
- Do not call
qdel(src)from this function
- Called by
/datum/component/Destroy(force(bool), silent(bool))(virtual, no-sleep)- Sends the
COMSIG_COMPONENT_REMOVINGsignal to the parent datum if theparentisn't being qdeleted - Properly removes the component from
parentand cleans up references - Setting
forcemakes it not check for and remove the component from the parent - Setting
silentdeletes the component without sending aCOMSIG_COMPONENT_REMOVINGsignal
- Sends the
/datum/component/proc/InheritComponent(datum/component/C, i_am_original(boolean))(abstract, no-sleep)- Called on a component when a component of the same type was added to the same parent
- See
/datum/component/var/dupe_mode C's type will always be the same of the called component
/datum/component/proc/AfterComponentActivated()(abstract, async)- Called on a component that was activated after it's
parent'sComponentActivated()is called
- Called on a component that was activated after it's
/datum/component/proc/OnTransfer(datum/new_parent)(abstract, no-sleep)- Called before
new_parentis assigned toparentinTakeComponent() - Allows the component to react to ownership transfers
- Called before
/datum/component/proc/_RemoveFromParent()(private, final)- Clears
parentand removes the component from it's component list
- Clears
/datum/component/proc/_JoinParent(private, final)- Tries to add the component to it's
parentsdatum_componentslist
- Tries to add the component to it's
/datum/component/proc/RegisterWithParent(abstract, no-sleep)- Used to register the signals that should be on the
parentobject - Use this if you plan on the component transfering between parents
- Used to register the signals that should be on the
/datum/component/proc/UnregisterFromParent(abstract, no-sleep)- Counterpart to
RegisterWithParent() - Used to unregister the signals that should only be on the
parentobject
- Counterpart to