## About The Pull Request
So I was looking for `attackby(...)` instances to kill and, hey look,
table frames- oh what the fuck is this.
Oh why does this have the same checks like five times, across the parent
type and subtypes.
So this pr's primary point is to refactor the table frame
`attackby(...)` into `item_interaction(...)`, and by extension lower the
amount of weird `attackby(...)` override jank going on with them.
Instead of having all of the subtypes define their own almost exactly
the same table construction interactions, we use a single generic
interaction chain and add a new `get_table_type(...)` proc to let
subtypes override what tables to construct for what stacks.
We also move the assigning of frame-and-stack-related things to the
actual table types themselves, instead of specifying all these
interactions on the frame.
We also add screentips for deconstruction and table construction, again
using the `get_table_type(...)` proc so screentips can differentiate
between which stacks can and can't make a table for our table without
needing to change the screentips manually for the subtypes.
Beyond that is mostly generic clean-up.
I'm a bit icky on my implementation of
`/obj/item/stack/proc/get_table_type()`, because of hardcoding for
`/obj/structure/table/greyscale`, but I think it's better than the
alternatives.
This lets us use the old method of letting all `/obj/item/stack/sheet`
subtypes use the generic material table, but limit it to working as long
as they define a `material_type`. I feel it's better than letting broken
tables exist, needing to hardcode it on the table frame, or needing to
expect people to add the generic material table `table_type` whenever
they add a `material_type` (no one would do this).
## Why It's Good For The Game
Screentips good 👍
Less jank encountered good.
Specifically the inability to deconstruct table frames on right click
when tables can be deconstructed on right click has thrown me off so
many times it's wild.
## Changelog
🆑
refactor: Refactored table frame interactions, please report any issues.
fix: You can no longer make material-less material tables out of certain
items.
fix: Fancy tables remember which carpet was used to make them, and no
longer magically transmute carpets into simpler types.
fix: You can no longer stack abductor table frame-using tables
specifically.
qol: Added screentips for deconstructing table frames and constructing
tables out of them if the material you're holding can do so.
qol: Table frames can now also be deconstructed on right click for
parity with tables, in addition to left click.
qol: Table frame interactions use balloon alerts.
/🆑
Hello everybuddy, your number three rated coder-failure here to clean up
some mess. This PR accomplishes some of the more major structural clean
up changes I wanted to do with /obj/ folder, but decided to wait on
until wallening gets merged, and so, time has come. Several things to
still be done, although I know these cleaning PR's are quite a load, so
will wait for this one to get done with first.
## Why It's Good For The Game
Saner spriters, better sprites, less annoyance. Also deleted a whole
load of redundancy this time around, a lot of sprites which existed
simultaniously in two places now got exit their quantum superposition.
## About The Pull Request
Using these search regexes:
Number ending in 0:
`do_after\((\w+), (\d+)0,`
Replace:
`do_after($1, $2 SECONDS,`
Single digit number:
`do_after\((\w+), [1-9],`
replace:
`do_after($1, 0.$2 SECONDS,`
Double:
`do_after\((\w+), (\d)([1-9]),`
Replace:
`do_after($1, $2.$3 SECONDS,`
## Why It's Good For The Game
Code readability
## About The Pull Request
When it comes to deconstructing an object we have `proc/deconstruct()` &
`NO_DECONSTRUCT`
Lets talk about the flag first.
**Problems with `NO_DECONSTRUCTION`**
I know what the comment says on what it should do
b5593bc693/code/__DEFINES/obj_flags.dm (L18)
But everywhere people have decided to give their own meaning/definition
to this flag. Here are some examples on how this flag is used
**1. Make the object just disappear(not drop anything) when
deconstructed**
This is by far the largest use case everywhere. If an object is
deconstructed(either via tools or smashed apart) then if it has this
flag it should not drop any of its contents but just disappear. You have
seen this code pattern used everywhere
b5593bc693/code/game/machinery/constructable_frame.dm (L26-L31)
This behaviour is then leveraged by 2 important components.
When an object is frozen, if it is deconstructed it should just
disappear without leaving any traces behind
b5593bc693/code/datums/elements/frozen.dm (L66-L67)
By hologram objects. Obviously if you destroy an hologram nothing real
should drop out
b5593bc693/code/modules/holodeck/computer.dm (L301-L304)
And there are other use cases as well but we won't go into them as they
aren't as significant as these.
**2. To stop an object from being wrenched ??**
Yeah this one is weird. Like why? I understand in some instances (chair,
table, rack etc) a wrench can be used to deconstruct a object so using
the flag there to stop it from happening makes sense but why can't we
even anchor an object just because of this flag?
b5593bc693/code/game/objects/objs.dm (L368-L369)
This is one of those instances where somebody just decided this
behaviour for their own convenience just like the above example with no
explanation as to why
**3. To stop using tools to deconstruct the object**
This was the original intent of the flag but it is enforced in few
places far & between. One example is when deconstructing the a machine
via crowbar.
b5593bc693/code/game/machinery/_machinery.dm (L811)
But machines are a special dual use case for this flag. Because if you
look at its deconstruct proc the flag also prevents the machine from
spawning a frame.
b5593bc693/code/game/machinery/_machinery.dm (L820-L822)
How can 1 flag serve 2 purposes within the same type?
**4. Simply forget to check for this flag altogether**
Yup if you find this flag not doing its job for some objects don't be
surprised. People & sometimes even maintainers just forget that it even
exists
b5593bc693/code/game/objects/items/piggy_bank.dm (L66-L67)
**Solution**
These are the main examples i found. As you can see the same flag can
perform 2 different functions within the same type and do something else
in a different object & in some instances don't even work cause people
just forget, etc.
In order to bring consistency to this flag we need to move it to the
atom level where it means the same thing everywhere. Where in the atom
you may ask? .Well, I'll just post what MrMelbert said in
https://github.com/tgstation/tgstation/pull/81656#discussion_r1503086862
> ...Ideally the .deconstruct call would handle NO_DECONSTRUCTION
handling as it wants,
Yup that's the ideal case now. This flag is checked directly in
`deconstruct()`. Now like i said we want to give a universal definition
to this flag and as you have seen from my examples it is used in 3 cases
1) Make an object disappear(doesn't dropping anything) when
deconstructed
2) Stop it from being wrenched
3) Stop it from being deconstructed via tools
We can't enforce points 2 & 3 inside `deconstruct()` which leaves us
with only case 1) i.e. make the object disappear. And that's what i have
done. Therefore after more than a decade or since this flag got
introduced `NO_DECONSTRUCT` now has a new definition as of 2024
_"Make an object disappear(don't dropping anything) when deconstructed
either via tools or forcefully smashed apart"_
Now i very well understand this will open up bugs in places where cases
2 & 3 are required but its worth it. In fact they could even be qol
changes for all we know so who knows it might even benefit us but for
now we need to give a universal definition to this flag to bring some
consistency & that's what this PR does.
**Problem with deconstruct()**
This proc actually sends out a signal which is currently used by the
material container but could be used by other objects later on.
3e84c3e6da/code/game/objects/obj_defense.dm (L160)
So objects that override this proc should call its parent. Sadly that
isn't the case in many instances like such
3e84c3e6da/code/game/machinery/deployable.dm (L20-L23)
Instead of `return ..()` which would delete the object & send the signal
it deletes the object directly thus the signal never gets sent.
**Solution**
Make this proc non overridable. For objects to add their own custom
deconstruction behaviour a new proc has been introduced
`atom_deconstruct()` Subtypes should now override this proc to handle
object deconstruction.
If objects have certain important stuff inside them (like mobs in
machines for example) they want to drop by handling `NO_DECONSTRUCT`
flag in a more carefully customized way they can do this by overriding
`handle_deconstruct()` which by default delegates to
`atom_deconstruct()` if the `NO_DECONSTRUCT` flag is absent. This proc
will allow you to handle the flag in a more customized way if you ever
need to.
## Why It's Good For The Game
1) I'm goanna post the full comment from MrMelbert
https://github.com/tgstation/tgstation/pull/81656#discussion_r1503086862
> ...Ideally the .deconstruct call would handle NO_DECONSTRUCTION
handling as it wants, but there's a shocking lack of consistency around
NO_DECONSTRUCTION, where some objects treat it as "allow deconstruction,
but make it drop no parts" and others simply "disallow deconstruction at
all"
This PR now makes `NO_DECONSTRUCTION` handled by `deconstruct()` & gives
this flag the consistency it deserves. Not to mention as shown in case 4
there are objects that simply forgot to check for this flag. Now it
applies for those missing instances as well.
2) No more copying pasting the most overused code pattern in this code
base history `if(obj_flags & NO_DECONSTRUCTION)`. Just makes code
cleaner everywhere
3) All objects now send the `COMSIG_OBJ_DECONSTRUCT` signal on object
deconstruction which is now available for use should you need it
## Changelog
🆑
refactor: refactors how objects are deconstructed in relation to the
`NO_DECONSTRUCTION` flag. Certain objects & machinery may display
different tool interactions & behaviours when destroyed/deconstructed.
Report these changes if you feel like they are bugs
/🆑
---------
Co-authored-by: san7890 <the@san7890.com>
Ladies, Gentlemen, Gamers. You're probably wondering why I've called you
all here (through the automatic reviewer request system). So, mineral
balance! Mineral balance is less a balance and more of a nervous white
dude juggling spinning plates on a high-wire on his first day. The fact
it hasn't failed after going on this long is a miracle in and of itself.
This PR does not change mineral balance. What this does is moves over
every individual cost, both in crafting recipes attached to an object
over to a define based system. We have 3 defines:
`sheet_material_amount=2000` . Stock standard mineral sheet. This being
our central mineral unit, this is used for all costs 2000+.
`half_sheet_material_amount=1000` . Same as above, but using iron rods
as our inbetween for costs of 1000-1999.
`small_material_amount=100` . This hits 1-999. This covers... a
startlingly large amount of the codebase. It's feast or famine out here
in terms of mineral costs as a result, items are either sheets upon
sheets, or some fraction of small mats.
Shout out to riot darts for being the worst material cost in the game. I
will not elaborate.
Regardless, this has no functional change, but it sets the groundwork
for making future changes to material costs much, MUCH easier, and moves
over to a single, standardized set of units to help enforce coding
standards on new items, and will bring up lots of uncomfortable balance
questions down the line.
For now though, this serves as some rough boundaries on how items costs
are related, and will make adjusting these values easier going forward.
Except for foam darts.
I did round up foam darts.
Adjusting mineral balance on the macro scale will be as simple as
changing the aforementioned mineral defines, where the alternative is a
rats nest of magic number defines. ~~No seriously, 11.25 iron for a foam
dart are you kidding me what is the POINT WHY NOT JUST MAKE IT 11~~
Items individual numbers have not been adjusted yet, but we can
standardize how the conversation can be held and actually GET SOMEWHERE
on material balance as opposed to throwing our hands up or ignoring it
for another 10 years.
imageadd: Metal, Reinforced, Reinforced Glass, Plasmaglass, Reinforced Plasmaglass, Titanium Glass, Plastitanium Glass, Wood and Poker tables have been updated!
Converts most spans into span procs. Mostly used regex for this and sorted out any compile time errors afterwards so there could be some bugs.
Was initially going to do defines, but ninja said to make it into a proc, and if there's any overhead, they can easily be changed to defines.
Makes it easier to control the formatting and prevents typos when creating spans as it'll runtime if you misspell instead of silently failing.
Reduces the code you need to write when writing spans, as you don't need to close the span as that's automatically handled by the proc.
(Note from Lemon: This should be converted to defines once we update the minimum version to 514. Didn't do it now because byond pain and such)
Done using this command sed -Ei 's/(\s*\S+)\s*\t+/\1 /g' code/**/*.dm
We have countless examples in the codebase with this style gone wrong, and defines and such being on hideously different levels of indentation. Fixing this to keep the alignment involves tainting the blames of code your PR doesn't need to be touching at all. And ultimately, it's hideous.
There are some files that this sed makes uglier. I can fix these when they are pointed out, but I believe this is ultimately for the greater good of readability. I'm more concerned with if any strings relied on this.
Hi codeowners!
Co-authored-by: Jared-Fogle <35135081+Jared-Fogle@users.noreply.github.com>
## About The Pull Request
Fixes Issue https://github.com/tgstation/tgstation/issues/56152 making wood tables deconstruct at they should be.
Bug vivisection:
Okay, see here? This is the proc for creating a table, we can introduce three arguments. One of them is _buildstack. _buildstack overrides Buildstack on initialize, a variable used for storing the type of raw "ore" that the table is supposed to drop in deconstruction. Here is supposed to be null unless we want to override the buildstack with another ore.
```DM
/obj/structure/table_frame/proc/make_new_table(table_type, custom_materials, _buildstack)
var/obj/structure/table/T = new table_type(loc, _buildstack)
T.frame = type
T.framestack = framestack
T.framestackamount = framestackamount
if(custom_materials)
T.set_custom_materials(custom_materials)
qdel(src)
```
What happened? The proc for building a wood table from a wooden frame, shown below, passed the "type" variable, used for storing the type of table_frame, as a _buildstack argument to the make_new_table proc. This overrides the buildstack variable of the final wooden table, causing it to drop a wooden frame as it was an ore on deconstruction.
```DM
/obj/structure/table_frame/wood/attackby(obj/item/I, mob/user, params)
[...]
if (toConstruct)
if(material.get_amount() < 1)
to_chat(user, "<span class='warning'>You need one [material.name] sheet to do this!</span>")
return
to_chat(user, "<span class='notice'>You start adding [material] to [src]...</span>")
if(do_after(user, 20, target = src) && material.use(1))
make_new_table(toConstruct, null, type)
```
This is funnier (not very much, to be honest) when we consider that deconstructing with a screwdriver would drop a frame normally, causing it to drop two frames. We could repeat this ad nauseam, essentially cloning wood frames in place as we pleased.
So TL;DR: this is another of those simple but hard to hunt bugs that would be prevented with testing and a null on its right place.
Tables could be placed on top of each other, leading to stacking of glass tables, and infinite sin. This doesn't cover anything close to all of it, but it's a good patch for now.
🆑 floyd
fix: You can no longer dupe mats by making tables from rods
/🆑
apparently you could use rods and floor tiles to finish tables, now you can only use sheets.
fixes#48547
* Nanotrasen fires the Wave Motion Gun at the Clock Cult
* Fixes a random changelog appearing from the reebe void.
* Fixes a wrong type reference regarding the bronze sheets in the Lavaland Ruin for it.
* Fixes redundant cells after CC removal in ruin by populating them with varied ore chunks, or destruction clutter.
* why is it /obj/item/stack/sheet/metal but not /obj/item/stack/sheet/iron
* Fixes the lavaland ratvar ruin to have fluff variant structures of deleted shit it referenced instead what I thought was reasonable.
removes materials list from items, uses custom_materials instead. This might introduce some bugs so we should testmerge this for a while (and Ill test stuff locally as much as I can)
this also adds material crafting to sheets. Test case being chairs. In the future we can add stuff like tables, walls, doors etc.
also applies materials to everything, with fixes, which can close#46299
Ports carpets from TauCetiStation
Original PR: TauCetiStation/TauCetiClassic#2989
Adds tables to match the carpets
Carpets Good. Tables Good.
Changelog
cl
imageadd: 8 new carpets and fancy tables
add: exotic carpet crate
/cl
About The Pull Request
Converts every single usage of playsound's vary parameter to use the boolean define instead of 1 or 0. I'm tired of people copypasting the incorrect usage.
Also changes a couple of places where a list was picked from instead of using get_sfx internal calls
This was done via regex:
(playsound\(.+,.+,.+, ?)1( ?\)| ?,.+\)) to match 1
(playsound\(.+,.+,.+, ?)0( ?\)| ?,.+\)) to match 0
full sed commands:
/(playsound\(.+,.+,.+, ?)1( ?\)| ?,.+\))/\1TRUE\2/ 1 to TRUE
/(playsound\(.+,.+,.+, ?)0( ?\)| ?,.+\))/\1FALSE\2/ 0 to FALSE
I'm not very good with regex and these could probably be optimized, but they worked.
Why It's Good For The Game
Code usability
* small changes
* Adds a use_tool helper and changes some tools to use it
* Ports most tool operations to use_tool
* Converts more tool operations to use_tool and tool_act
* Changes some things to default_unfasten_wrench
* Improves tool_behavior support in mech construction
* Code review memes
* Fixes all instant use_tool calls failing
* Code improvements
* merge fixes
* Fixed a few bugs with clockwork proselytizers
* condensed
* formatting
* proselytizer.
* i need more sanity checks I think this crashed something
* proper!
* better desc
* sanity! SANITY!
Please refer to #20867 and #20870 for a easier view of the changes. Those two PRs show all meaningful changes (hopefully) and doesn't show the files changed with just 3 lines changed.
This PR does three things:
It makes all children of /obj/ use the same damage system.
Previously to make your new machine/structure be destroyable you needed to give it a var/health, and its own version of many damage related proc such as bullet_act(), take_damage(), attacked_by(), attack_animal(), attack_hulk(), ex_act(), etc... But now, all /obj/ use the same version of those procs at the /obj/ level in code/game/obj_defense.dm. All these obj share the same necessary vars: obj_integrity (health), max_integrity, integrity_failure (optional, below that health level failure happens), and the armor list var which was previously only for items, as well as the resistance_flags bitfield. When you want your new object to be destroyable, you only have to give it a value for those vars and maybe override one proc if you want a special behavior but that's it. This reorganization removes a lot of copypasta (most bullet_act() version for each obj were nearly identical). Two new elements are added to the armor list var: fire and acid armor values.
How much damage an obj take depends on the armor value for each damage category. But some objects are INDESTRUCTIBLE and simply never take any damage no matter the type.
The armor categories are:
-melee(punches, item attacks, xeno/animal/hulk attacks, blob attacks, thrown weapons)
-bullet
-laser
-energy (used by projectiles like ionrifle, taser, and also by EMPs)
-bio (unused for this, only here because clothes use them when worn)
-rad (same)
-bomb (self-explanatory)
-fire (for fire damage, not for heat damage though)
-acid
For machines and structures, when their health reaches zero the object is not just deleted but gets somewhat forcedeconstructed (the proc used is shared with the actual deconstruction system) which can drops things. To not frustrates players most of these objects drop most of the elements necessary to rebuild them (think window dropping shards). Machines drop a machine frame and all components for example (but the frame can then be itself smashed to pieces).
For clothes, when they are damaged, they get a "damaged" overlay, which can also be seen when worn, similar to the "bloody" overlay.
It refactors acid. See #20537.
Some objects are ACID_PROOF and take no damage from acid, while others take varying amounts
of damage depending on their acid armor value. Some objects are even UNACIDABLE, no acid effect can even land on them. Acid on objects can be washed off using water.
It changes some aspect of damage from fires.
All /obj/ can now take fire damage and be flammable, instead of just items. And instead of having just FLAMMABLE objs that become ON_FIRE as soon as some fire touch them (paper), we now have objects that are non flammable but do take damage from fire and become ashes if their health reaches zero (only for items). The damage taken varies depending on the obj's fire armor value and total health. There's also still obj and items that are FIRE_PROOF (although some might still be melted by lava if they're not LAVA_PROOF).
When a mob is on fire, its clothes now take fire damage and can turn to ashes. Similarly, when a mob takes melee damages, its clothes gets damaged a bit and can turn to shreds. You can repair clothes with cloth that is produceable by botany's biogenerator.
It also does many minor things:
Clicking a structure/machine with an item on help intent never results in an attack (so you don't destroy a structure while trying to figure out which tool to use).
I moved a lot of objects away from /obj/effect, it should only be used for visual effects, decals and stuff, not for things you can hit and destroy.
I tweaked a bit how clothes shredding from bombs work.
I made a machine or structure un/anchorable with the wrench, I don't remember which object...
Since I changed the meaning of the FIRE_PROOF bitflag to actually mean fire immune, I'm buffing the slime extract that you apply on items to make them fire proof. well now they're really 100% fire proof!
animals with environment_smash = 1 no longer one-hit destroy tables and stuff, we give them a decent obj_damage value so they can destroy most obj relatively fast depending on the animal.
Probably a million things I forgot.
If you want to know how the damage system works all you need is the three obj vars "obj_integrity", "max_integrity", "integrity_failure", as well as the armor list var and the resistance_flags bitfield, and read the file obj_defense.dm
This just adds some vars that the refactor will use. Putting this in a separate PR prevent the damage refactor PR from becoming too big (number of files changed) and hard to read and review.
Introduces the resistance_flags bitflag that replaces unacidable and burn_state.
Moves the armor var from item to /obj level and gives specific armor values to many objects, the armor list also gets two new armor types: fire and acid, which will be used in the refactor. the new fire and acid armor values are given to plenty of items.
* Adds brass, used to construct various brass objects
* brasshole
* give em the brass
* give em more brass
* DYNAMIC
* VAR MURDER SPREE
* brassy
* acceptaphil
* this is super against my general preferences but eeeeeh sure okay i guess
