## About The Pull Request
Depends on #94064 for the unit test but offers a better method for
finding an atom to mount on
- Finding a mount now takes into consideration the objects pixel x & y
offsets meaning diagonal mounting is now supported. Gives great
flexibility for mappers
- If you don't want to use pixel offsets but default back to using the
objects direction that behaviour is still preserved. Useful if your
object uses directional icon states(lights & cameras for now) AND don't
use offsets
- If no direction could be specified then as the last resort it defaults
back to the objects local turf for mounting
## Changelog
🆑
fix: all mounted objects on tables, fences, windows & walls should fall
of correctly when the atom it is placed on is destroyed
fix: security telescreen now falls off when their mounted wall is
destroyed
fix: defib wall mount falls off when their mounted wall is destroyed
fix: floor lights are mounted to the ground/catwalk/tram floor they are
sitting on meaning destroying it will destroy the light
fix: wall mounted plaques now fall off when their mounted wall is
destroyed
/🆑
## About The Pull Request
Empy lists. There are a lot of 'em.
<img width="981" height="512" alt="image"
src="https://github.com/user-attachments/assets/b94b041a-2904-466b-ab89-54bd1de11b4e"
/>
Going through ways to reduce memory I found a few easy ones here. Wires,
the edible component, the seethrough component. None of these are really
a concern when it comes to needing lists in memory for performance
reasons. Wires aren't going to be cut most of the time for each door. A
lot of food does not have any junkiness. Seethrough component lies
dormant most of the round. Etc.
Making lists lazy in these cases should be a no brainer.
Everything I tested still seems to work exactly the same.
## Why It's Good For The Game
Frees memory that is just taking up space a lot of the time.
## Changelog
Not player-facing, this is all under-the-hood stuff.
## About The Pull Request
Extends the part of the crafting unit test that ensures consistency
between the total mats of the components of a recipe (or rather, the
result of said recipe) and a generic instance of the same type as its
result, previously only implemented on food recipes.
## Why It's Good For The Game
This ensures a degree of consistency with the material composition of
various objects in the game. I couldn't do it in the original PR as that
one was too big already and it took months to get it merged, and have
the relative bugs fixed.
Currently a WIP as I slowly deal with the unit test reports.
## Changelog
🆑
refactor: Follow-up to the crafting/material refactor from months ago.
All objects crafted with stacks now inherit their mat composition (not
necessarily the effects and color) by default, while previously only a
few things like chair, sinks and toilets did. Report any object looking
or behaving weirdly as a result.
fix: The material composition of ammo boxes is no longer a 1/10 of what
it's supposed to be. It was a shitty hack to make it harder to recycle
empty ammo boxes. Instead, they lose materials as they're emptied now.
/🆑
## About The Pull Request
Closes#93635
`unique_reskin` is no longer a list on `/item`, now `/datum/atom_skin`
The actual reskinning behavior has been moved out to
`/datum/component/reskinable_item`
PKC reskinning is now handled via alt-click reskin, rather than via the
retooling kit. The retooling kit has been removed.
There's no limit on how many times you can reskin your PKC (though
perhaps we limit it to one reskin and keep the retooling kit as a way to
allow a miner to reskin it a second time?)
The Ashen Skull unique reskin is still a trophy, and instead unlocks its
unique reskin option in the alt-click radial.
## Why It's Good For The Game
I'm unsure why the retooling kit exists on its own, when it's relatively
cheap and just performs the behavior of alt-click reskinning.
So to keep it consistent with all other forms of reskinning I've just
made it baseline. To accomplish that I refactored reskinning.
The new form of reskinning allows for greater potential in adding
reskins, allowing far more than just an icon state change. Also we can
put it on turfs and mobs and structures now which is cool I guess
There's also the added benefit of being able to see an item's reskins
without needing to instantiate it, which the loadout menu uses to great
effect.
## Changelog
🆑 Melbert
refactor: Refactored item reskinning (the alt-click way), report any
oddities with that
del: Deleted the crusher retool kit, now you can just reskin your
crusher with alt-click. The Skull skin is still locked behind having the
Ashen Skull trophy applied.
fix: Stunswords no longer have an incorrect lore blurb
fix: Fixed loadout item reskinning's UI
/🆑
## About The Pull Request
Depends on #94064 for the unit test but offers a better method for
finding an atom to mount on
- Finding a mount now takes into consideration the objects pixel x & y
offsets meaning diagonal mounting is now supported. Gives great
flexibility for mappers
- If you don't want to use pixel offsets but default back to using the
objects direction that behaviour is still preserved. Useful if your
object uses directional icon states(lights & cameras for now) AND don't
use offsets
- If no direction could be specified then as the last resort it defaults
back to the objects local turf for mounting
## Changelog
🆑
fix: all mounted objects on tables, fences, windows & walls should fall
of correctly when the atom it is placed on is destroyed
fix: security telescreen now falls off when their mounted wall is
destroyed
fix: defib wall mount falls off when their mounted wall is destroyed
fix: floor lights are mounted to the ground/catwalk/tram floor they are
sitting on meaning destroying it will destroy the light
fix: wall mounted plaques now fall off when their mounted wall is
destroyed
/🆑
## About The Pull Request
Passes a standardized define into some of the tgui to parody behavior
with vending machines.
Adds a few defines for money symbols to make it easier to rework types
of money.
If yall are for this I can also go through and find and replace a ton of
strings and other tgui with these, but dont want to sink 2 much time
into it if not.
<img width="966" height="646" alt="image"
src="https://github.com/user-attachments/assets/2d4753be-7e5d-4733-acf8-05f354127a6a"
/>
Proof of concept if Nanotrasen was not an evil conglomerate and the
dollar never went out of style
<img width="1009" height="661" alt="image"
src="https://github.com/user-attachments/assets/e68bfcd3-4053-4641-9b28-f384f2cb049b"
/>
some broken shit here but 99% thats cause minimal runtime station does
not support the supply console
<img width="795" height="767" alt="image"
src="https://github.com/user-attachments/assets/30dbd625-5338-4dc3-80db-922ff045f0f0"
/>
## About The Pull Request
I'm working on a pubby update for the map depot for admins to run, and
thought I should throw this here. Removes a varedit from Pubby and makes
the monastery telescreen replaceable by making it researchable and
printable through the service techfab.
Also re-adds the pubby map wall item, using the new map item outlines.
## Why It's Good For The Game
Repairability for a map that seems to be ran frequently by admins
(judging by it getting actual issue reports), this has no effect
otherwise.
## About The Pull Request
It's just a partial cleanup of
anti-[STYLE](https://github.com/tgstation/tgstation/blob/master/.github/guides/STYLE.md)
code from /tg/'s ancient history. I compiled & tested with my helpful
assistant and damage is still working.
<img width="1920" height="1040" alt="image"
src="https://github.com/user-attachments/assets/26dabc17-088f-4008-b299-3ff4c27142c3"
/>
I'll upload the .cs script I used to do it shortly.
## Why It's Good For The Game
Just minor code cleanup.
Script used is located at https://metek.tech/camelTo-Snake.7z
EDIT 11/23/25: Updated the script to use multithreading and sequential
scan so it works a hell of a lot faster
```
/*
//
Copyright 2025 Joshua 'Joan Metekillot' Kidder
This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program. If not, see <https://www.gnu.org/licenses/>.
//
*/
using System.Text.RegularExpressions;
class Program
{
static async Task Main(string[] args)
{
var readFile = new FileStreamOptions
{
Access = FileAccess.Read,
Share = FileShare.ReadWrite,
Options = FileOptions.Asynchronous | FileOptions.SequentialScan
};
FileStreamOptions writeFile = new FileStreamOptions
{
Share = FileShare.ReadWrite,
Access = FileAccess.ReadWrite,
Mode = FileMode.Truncate,
Options = FileOptions.Asynchronous
};
RegexOptions regexOptions = RegexOptions.Multiline | RegexOptions.Compiled;
Dictionary<string, int> changedProcs = new();
string regexPattern = @"(?<=\P{L})([a-z]+)([A-Z]{1,2}[a-z]+)*(Brute|Burn|Fire|Tox|Oxy|Organ|Stamina)(Loss)([A-Z]{1,2}[a-z]+)*";
Regex camelCaseProcRegex = new(regexPattern, regexOptions);
string snakeify(Match matchingRegex)
{
var vals =
matchingRegex.Groups.Cast<Group>().SelectMany(_ => _.Captures).Select(_ => _.Value).ToArray();
var newVal = string.Join("_", vals.Skip(1).ToArray()).ToLower();
string logString = $"{vals[0]} => {newVal}";
if (changedProcs.TryGetValue(logString, out int value))
{
changedProcs[logString] = value + 1;
}
else
{
changedProcs.Add(logString, 1);
}
return newVal;
}
var dmFiles = Directory.EnumerateFiles(".", "*.dm", SearchOption.AllDirectories).ToAsyncEnumerable<string>();
// uses default ParallelOptions
// https://learn.microsoft.com/en-us/dotnet/api/system.threading.tasks.paralleloptions?view=net-10.0#main
await Parallel.ForEachAsync(dmFiles, async (filePath, UnusedCancellationToken) =>
{
var reader = new StreamReader(filePath, readFile);
string oldContent = await reader.ReadToEndAsync();
string newContent = camelCaseProcRegex.Replace(oldContent, new MatchEvaluator((Func<Match, string>)snakeify));
if (oldContent != newContent)
{
var writer = new StreamWriter(filePath, writeFile);
await writer.WriteAsync(newContent);
await writer.DisposeAsync();
}
reader.Dispose();
});
var logToList = changedProcs.Cast<KeyValuePair<string, int>>().ToList();
foreach (var pair in logToList)
{
Console.WriteLine($"{pair.Key}: {pair.Value} locations");
}
}
}
```
## Changelog
🆑 Bisar
code: All (Brute|Burn|Fire|Tox|Oxy|Organ|Stamina)(Loss) procs now use
snake_case, in-line with the STYLE guide. Underscores rule!
/🆑
## About The Pull Request
Makes slime potions, injector, and cookies use item_interaction.
generizices behavoir that potions typed under
`obj/item/slimepotion/slime/` only affect slimes, moves the handful of
potions that did not do that out of that type.
removed a random type casting for slime cookies to verify there human
before it checks for `TRAIT_AGEUSIA` as there is no reason other mobs
wouldn't be able to taste it.
do to the refactor you wont "bash" a lot of items if the interaction
fails.
## Why It's Good For The Game
<img width="439" height="68" alt="image"
src="https://github.com/user-attachments/assets/f398fb6c-8cbd-4092-9a64-8cccce87f8b6"
/>
## About The Pull Request
This 1st has to be PR'd so the integration tests can point out all
wallmounts that could not find a support structure to mount on. I then
will do many map edits to align them onto the closest atom
Yes we no longer use wall mount but atom mounted component. All objects
that are mounted on windows/tables & fences now also fall off when
destroyed
It'll probably be a WHILE before I can fix all wall mounts. Long day.
Expect me to misalign many stuff to fix failing CI so make sure to
provide suggestions when possible
Improved wallmount code overall
- Fixes#93793
## Changelog
🆑
fix: fixes all incorrectly maploded wall mounts that aren't actually
hanging on any support structure
fix: objects mounted on tables, windows & fences also fall off now when
destoryed
qol: lights can be mounted on windows
qol: cameras can be mounted on windows
qol: buttons can be mounted on tables
refactor: improved how wall mounts interact with objects as a whole
report bugs on github
/🆑
## About The Pull Request
Various minor fixes and qol changes primarly related to sorting on the
crew monitor console (See changelog). Also simplifies how the UI cycles
the sorting options, as it was a bit bizarre before.
## Why It's Good For The Game
Improves the crew monitor user experience.
## Changelog
🆑
fix: binary sensors no longer report exact consciousness status to crew
monitor.
fix: crew monitor sort by name now respects ascending vs. descending
sorting.
qol: crew monitor sort by vitals now properly treats those with binary
sensors as either fully healthy or dead.
/🆑
---------
Co-authored-by: afonamos <d.splinter@live.com>
## About The Pull Request
When a camera update is triggered, it is instead added to a queue on a
background subsystem
An AI entering a camera chunk which is queued to update will force the
update immediately (bypassing the queue)
While the root problem of this is, ultimately, not addressed...
<img width="554" height="58"
alt="467828777-eff3f0e5-49d6-4997-b4d7-05eff6432155"
src="https://github.com/user-attachments/assets/c2d6a5f5-d958-463e-959f-116bd0dab475"
/>
...the change will ultimately prevent update spam from consuming all of
the server's resources - instead allocating updates to the backburner in
times of high server stress (or on multi-z maps)
## Changelog
🆑 Melbert
refactor: Refactored the way camera updates are handled to hopefully
reduce some lag. Report any oddities
/🆑
## About The Pull Request
/area/station/ai_monitored's behaviour was isolated into a component,
`/datum/component/monitored_area`, which itself uses
`/datum/motion_group`s to query cameras. Functionally, it (should) work
identically to the old implementation. I'm sure that behaviour could
have been further cleaned up, camera code is quite dreadful, but it's
better to focus on isolating the behaviour first. Baby steps.
Areas that want to opt into monitoring can set `var/motion_monitored` to
TRUE (this approach was taken to make subtyping easier).
The following non-AI areas were changed:
- /area/station/ai_monitored/security/armory ->
/area/station/security/armory
- /area/station/ai_monitored/command/nuke_storage ->
/area/station/command/vault
- /area/station/ai_monitored/command/storage/eva ->
/area/station/command/eva
All other `/area/station/ai_monitored` subtypes were repathed into
`/area/station/ai` and cleaned up in a way that I thought made logical
sense. It is **much** more readable now. For example:
- /area/station/ai_monitored/turret_protected/aisat ->
/area/station/ai/satellite
- /area/station/ai_monitored/command/storage/satellite ->
/area/station/ai/satellite/maintenance/storage
- /area/station/ai_monitored/turret_protected/ai ->
/area/station/ai/satellite/chamber
## About The Pull Request
- Fixes#93392
- Replaces all custom callbacks with call to `deconstruct()`. The
callbacks weren't necessary as it did the same thing as `deconstruct()`
but in an round about way
- Removed duplicate `Initialize()` procs and the params `building` &
`ndir` from all wall mounts. Makes everything cleaner
## Changelog
🆑
fix: wall mounts placed by player now falls off when the wall they are
mounted on is destroyed
code: cleaned up wall mount code
/🆑
## About The Pull Request
`/proc/flatten_list()` -> `/proc/assoc_to_values()`, also changes the
code doc to mirror the language of `assoc_to_keys()`'s code doc
## Why It's Good For The Game
Having a proc called `assoc_to_keys` that takes an associative list and
returns a list of the keys, and also having a proc that takes an
associative list and returns a list of the values, and the latter not
being called `assoc_to_values` is very funny
## Changelog
🆑
code: renamed flatten_list proc to assoc_to_values
/🆑
---------
Co-authored-by: Ghom <42542238+Ghommie@users.noreply.github.com>
## About The Pull Request
- Moves the anchored check to when construction is being finalized
- Removes unnecessary "missing components" balloon alert when failing
finalization
## Why It's Good For The Game
Fixes a bug where you could have unanchored functional PCs which can be
pretty powerful depending on the PC. It's fun so I hate it and it needs
to die instantly.
## Changelog
🆑
fix: moves the computer frame's anchor requirement from circuit step to
finalization step, fixing an issue where you could construct unanchored
and FUNCTIONAL computers
qol: as a side effect, you only need to wrench the PC down when
construction is being finished
/🆑
## About The Pull Request
ports https://github.com/DaedalusDock/daedalusdock/pull/1144
ports https://github.com/DaedalusDock/daedalusdock/pull/1147
full credit to @Kapu1178 for the juice
instead of `reacher.CanReach(target)` we now do
`target.CanBeReachedBy(reacher)`, this allows us to give special
behavior to atoms which we want to reach, which is exactly what I need
for a feature I'm working on.
## Why It's Good For The Game
allows us to be more flexible with reachability
## Changelog
🆑
refactor: refactored how reaching items works, report any oddities with
being unable to reach something you should be able to!
/🆑
## About The Pull Request
- Added full support for RETA system
- Categorized all configs for search purposes
- Added channel check for most usages of get_announcement_system() calls
## Why It's Good For The Game
## Changelog
🆑
code: Little AAS code cleanup. Added full RETA support. Grouped configs
for search purposes.
/🆑
## About The Pull Request
Station alarm consoles now register to all Station Z levels+Areas if
they were created on a station Z level (this doesn't include mining, but
station ones will see mining areas), otherwise will be a "local" console
to that Z level only.
Mostly the same applies for Cyborg/AI, but if they were built on a
non-station Z level then they have their Z level added on top of Station
Z levels, so example Nukie saboteur borgs will see alarms they might
need, while a Golem/Oldstation AI will also be able to see their own
alarms.
Because of how alarms and spawning in worked, Cyborgs/AIs had CentCom's
Z level as something it would detect, so we simply create the mob at the
spawn location rather than on the title screen during Initialization,
fixing that issue as well.
## Why It's Good For The Game
Adds easy multi-z support for station alarms and removes the previous
bandaid of having to deconstruct your board and multitool it or map in
every single station alarm console to allow station alerts (which is
quite silly) which was done in
https://github.com/tgstation/tgstation/pull/88343.
Closes https://github.com/tgstation/tgstation/issues/61592
Closes https://github.com/tgstation/tgstation/issues/83042
Closes https://github.com/tgstation/tgstation/issues/69314
I find this solution better since it catches pretty much all use cases
without the need for special subtypes and needing players to actually
know the difference between these things.
## Changelog
🆑
fix: Cyborgs and AIs will now see alerts of the whole station rather
than solely the Z level they were spawned on.
fix: Cyborgs and AIs will no longer get alerts about Thunderdome.
/🆑
## About The Pull Request
Listen dude I just really hate those fucking sandbraine-
In order to auto-disable tracking if a mob is not well, trackable (read,
wrong z, out of camera sight), we need to well, check if the mob is in
sight of a camera.
This requires cameras to update if they're marked dirty (which they will
be since any opacity change (read: airlocks closing) does this), sooooo
once every 10 seconds when the ui updates we have to update what like 60
camera chunks pessimistically?
That eats a variable amount of cost, somewhere between 300% and 600% of
the tick on catwalk in the round I'm observing, tho it kind of depends
on both the amount of z levels and pop count and so on.
Anyway so instead of doing that let's just not tell them if it's
disabled (like we do for every other button that does this). It's a UX
hit but I cannot come up with an acceptable solution, and it can ALREADY
be wrong because it's behind a timer, so like...
## Why It's Good For The Game
Hopefully significantly less random lagspikes at anything above lowpop
<img width="901" height="29" alt="image"
src="https://github.com/user-attachments/assets/936a6d91-aab9-4686-98e6-cd93b586227c"
/>
<img width="895" height="51" alt="image"
src="https://github.com/user-attachments/assets/bb6782c9-9e67-42dd-a241-cc5bf795744b"
/>
## Changelog
🆑
balance: AIs viewing the crew monitor will no longer be told ahead of
time if they cannot jump to a player.
/🆑
## About The Pull Request
**1. Qol**
- Air sensors(maploaded & hand crafted) now auto connect to any input
ports & output valves that are within a 4 tile radius from it meaning
there is no need for these ports to have unique IDs attached to them &
you don't need a multitool to link them together if you first placed
those ports & then placed the air sensor close to them
- Atmos control monitors now finds all air sensors that are closest to
it within the same z level meaning you can have/place multiple air
sensors of the same type on the same z level & the monitor will locate
them correctly without any ambiguity
**2. Refactor**
- Removed var `chamber_id` from both injectors & vents meaning you don't
need to have unique ids assigned to them. As long as mappers/players put
them within 4 tile radius from the air sensor it will auto connect to
them
- Removed var `GLOB.map_loaded_sensors`
## Changelog
🆑
qol: hand crafted air sensors now auto connect to input & output ports
if they are located within a 4 tile radius from it
qol: atmos control monitors now locate all air sensors nearest to in on
that same z level meaning you can have multiple air sensors of the same
type on the same map
refactor: refactored how air sensors & atmos monitors connect to each
other. Please report any atmos computers that have missing input
valves/output ports on their connected air sensors or those that don't
list any air sensors at all on github
/🆑
---------
Co-authored-by: LemonInTheDark <58055496+LemonInTheDark@users.noreply.github.com>
## About The Pull Request
Previously, the arcade game win experiment succeeded upon the machine
calling `prizevend()`, meaning the experiment completed upon tickets
being inserted into the machine, counter to the experiment's text. This
pr changes the experiment to succeed upon the machine calling the new
`victory_tickets()`, which has replaced each instance of a won game
dispensing tickets.
## Why It's Good For The Game
fixes#80218
Standardized ticket dispensing
## Changelog
🆑
fix: Arcade machine experiment now succeeds on output rather than input
of tickets
/🆑
## About The Pull Request
Changes "Couldn't" to "Can't" on Z-level move failures
## Why It's Good For The Game
Why is it in past tense? Every time I see this message I ask myself that
question and each time I search my soul but I find no answer to satisfy
me. I stare at the word and feel hatred burning within me. One of us
must go.
## Changelog
🆑
spellcheck: When failing to move up or down a Z-Level, you are now tied
to the present
/🆑
## About The Pull Request
This PR adds the ability to install a B.O.R.I.S. in a circuit that
contains an MMI component. These circuits can then be remotely connected
to by an AI by clicking on them or anything they are inside of. To
indicate that a circuit allows remote AI connection, an indicator is
given to the circuit and anything containing it.
Additionally:
- Refactors the MMI component to use `item_interaction`, since it was
pertinent.
- You cannot insert an MMI/B.O.R.I.S. into a locked circuit.
- You can no longer hotswap MMIs/B.O.R.I.S.es - you must manually eject
the inserted one.
Let me know what changelog labels I should use for the hotswap removal
and the prevention of insertion into locked circuits.
## Why It's Good For The Game
If you can put an MMI or posibrain in a circuit, why not allow an AI to
use it using a B.O.R.I.S.?
## Changelog
🆑
add: B.O.R.I.S.es can be installed inside of integrated circuits with
MMI components, allowing an AI to remotely interface with them the same
way an MMI or posibrain could.
refactor: The MMI component now uses item interaction behavior for
inserting MMIs/B.O.R.I.S.es.
/🆑
---------
Co-authored-by: Ghom <42542238+Ghommie@users.noreply.github.com>
## About The Pull Request
Fixes#92188
1. The Accounting console now has a new UI
<img width="757" height="486" alt="image"
src="https://github.com/user-attachments/assets/0c6ce73f-ae1c-4a54-b6ba-bcc3b3232d13"
/>
2. The Accounting console can now dish out advances on paychecks, giving
you for full paycheck early. You can give up to 3 advances. This of
course means your next paycheck is not paid out when relevant.
3. The Accounting console can now change paycheck sizes of crewmembers -
up to 1.5x and down to 0.5x.
## Why It's Good For The Game
- Gives the HoP some more duties, now being able to dish out money on
request, reward good behavior, or punish bad behavior.
## Changelog
🆑 Melbert
add: Accounting Console: New UI!
add: Accounting Console: Now can give advances to crewmembers
add: Accounting Console: Can now give pay raises or pay cuts
add: Accounting Console: Now only printable in the security lathe
add: Accounting Console: A spare board is now now found in secure tech
storage.
fix: Fix vending machines adding payments to audit log twice.
fix: Non-crewmembers are no longer shown in the accounting console
/🆑
---------
Co-authored-by: ArcaneMusic <41715314+ArcaneMusic@users.noreply.github.com>
## About The Pull Request
- Typechecks for Protean brain before running a check against it, so it
doesn't crash on an unexpected brain type.
- Deletes unused Skyrat crew monitor TGUI
- Removes duplicate datum override
- Updates crew console TGUI for Bubber health
- Broken sensor mobs don't bounce around the list every update
Closes#4372
## Changelog
🆑 LT3
fix: Fixed edge case of Proteans causing crew health monitor to
bluescreen
qol: Mobs with broken suit sensors now stay in order each screen update
/🆑
## About The Pull Request
fixes#90586
## Why It's Good For The Game
the poster isnt _that_ wanted
## Changelog
🆑
fix: Wanted posters now correctly return to your inventory if you move
while placing them
/🆑
---------
Co-authored-by: Xerseon <138307753+Xerseon@users.noreply.github.com>
## About The Pull Request
fixes#90586
## Why It's Good For The Game
the poster isnt _that_ wanted
## Changelog
🆑
fix: Wanted posters now correctly return to your inventory if you move
while placing them
/🆑
---------
Co-authored-by: Xerseon <138307753+Xerseon@users.noreply.github.com>
## About The Pull Request
**1. Makes the machine creation process from both efficient.**
The machine frame no longer creates & deletes the machines old component
parts before moving in the new parts. It instead now passes its parts
directly to the circuit boards `apply_default_parts()` meaning the old
parts are never created and the new ones get applied directly. This also
mean it does not have to again call `RefreshParts()` as the machines
`Initialize()` proc will do that for us by default so the performance is
just faster now.
Flatpack box also benefits from this and it will need it as stated below
- Fixes#86514. As the circuit board is passed to the machine's
`Initialize()` proc during reconstruction so it sets it's
`all_products_free` from the board both during mapload & when player
constructed from either machine frame & flatpack
**2. Flatpack enhancements**
- The flatpack box now passes its stored circuit board into the machine
meaning any changes made like with a screw driver or such are preserved
in the newly created machine rather than the board getting deleted & the
machine creating the default board
- Flatpacker now accept custom component parts that replaces its stored
circuit board default parts. This implements
https://github.com/tgstation/tgstation/pull/91670#issuecomment-2977259973
meaning instead of the vendor created from the flatpacker starting out
empty it now accepts the refill canister and will package both of them
together. Then during deployment that stored canister is stored back in
the vendor meaning instead of starting out empty you get to carry over
all your previous stock
Depends on https://github.com/tgstation/tgstation/pull/91512 because of
named arg conflict. Keep this drafted till that gets merged
## Changelog
🆑
qol: flatpacker now accepts custom component parts(like vendor refill
canister) along with the circuit board so you can carry over your
previous stock during deployment rather than starting out empty
fix: flatpack boxes now passes its stored circuit board into the machine
thus preserving its properties instead of creating a default board in
the machine
fix: vending machines reconstructed on station from circuits loaded off
station(having free products) will continue to sell free products
code: machine construction via flatpacker/machine frame is faster
performance wise
/🆑
## About The Pull Request
**1. Makes the machine creation process from both efficient.**
The machine frame no longer creates & deletes the machines old component
parts before moving in the new parts. It instead now passes its parts
directly to the circuit boards `apply_default_parts()` meaning the old
parts are never created and the new ones get applied directly. This also
mean it does not have to again call `RefreshParts()` as the machines
`Initialize()` proc will do that for us by default so the performance is
just faster now.
Flatpack box also benefits from this and it will need it as stated below
- Fixes#86514. As the circuit board is passed to the machine's
`Initialize()` proc during reconstruction so it sets it's
`all_products_free` from the board both during mapload & when player
constructed from either machine frame & flatpack
**2. Flatpack enhancements**
- The flatpack box now passes its stored circuit board into the machine
meaning any changes made like with a screw driver or such are preserved
in the newly created machine rather than the board getting deleted & the
machine creating the default board
- Flatpacker now accept custom component parts that replaces its stored
circuit board default parts. This implements
https://github.com/tgstation/tgstation/pull/91670#issuecomment-2977259973
meaning instead of the vendor created from the flatpacker starting out
empty it now accepts the refill canister and will package both of them
together. Then during deployment that stored canister is stored back in
the vendor meaning instead of starting out empty you get to carry over
all your previous stock
Depends on https://github.com/tgstation/tgstation/pull/91512 because of
named arg conflict. Keep this drafted till that gets merged
## Changelog
🆑
qol: flatpacker now accepts custom component parts(like vendor refill
canister) along with the circuit board so you can carry over your
previous stock during deployment rather than starting out empty
fix: flatpack boxes now passes its stored circuit board into the machine
thus preserving its properties instead of creating a default board in
the machine
fix: vending machines reconstructed on station from circuits loaded off
station(having free products) will continue to sell free products
code: machine construction via flatpacker/machine frame is faster
performance wise
/🆑
