[MDB IGNORE] Deprecates Auxmos in favour of LINDA [PORT] #11165
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
JixS4v
changed the title
|
It's normal that this is failing, it's literally less than half done |
…(#56317) Basically when a group with a breakdown timer one tick away from settling merges with a group with hotspots in it, and the one with a high breakdown timer is larger, the group would settle without respecting the timer of the hotspot group, causing fwoosh levels of flame. There's two other ways to achieve this, if you had one group of tiles that are close to the same gasmix as another group, and they became inactive, we'd have the same issue. I've solved this by moving the hotspot subprocess to after active turfs and before excited groups, just for safety ya feel me? It's still in theory possible, but much much harder. The hard solution to this would be to integrate heat with how we reset excited group timers, and split excited groups into smaller portions, but I'm not sure I like that idea. I need to mull it over.
When I was in the process of developing my atmos fixes/optimizations, I focused heavily on maintaining the area of excited groups, since growing delayed self_breakdown(), and lead to shitty behavior which I thought was the result of my sleeping changes. This was not the case, those who came before me knew better, as always. Especially now, since I've unhooked breakdown from excited groups gaining new tiles, my concern was unfounded, and actually lead to shitty behavior and wasted time. I failed to recognize the value in random garbage_collect()s, they help prevent mass gas equalization over large spaces, they make gas appear to move more consistently when in these large spaces, and they lessen the amount of self_breakdown()s over large turf lists, which is very helpful for lowering the overall overtime of the subsystem. This fixes my mistake, and purges the excited cleanup subprocess from the air subsystem. It's free real estate babyyyyyyyy.
Why? delta_time is about maintaining behavior when changing the wait of subsystems SSair's wait is dynamic by design, we hardly ever hit below it. What is important then, is maintaining behavior across each process. The key point here is making sure excited groups and turfs share the same amoumt of gas each process, no matter how high or low wait is. This is why subprocesses are a thing in the first place, to maintain this consistency. delta_time fucks with this, and will end up changing behavior if wait is ever changed.
Makes both the act of starting a rebuild, and the full rebuild itself yield. This might? Expose some things that rely on parent existing, but that was a problem before, if a rare one. It'll need cleaned up at some point, but I'd like some feedback on how I'm acomplishing this. Oh and I changed a very slight detail about how volume is used, instead of storing it throughout the whole loop and applying it at the last moment, we just operate on the pipeline's volume step by step. This fixes like, pipes being wrenched up while a rebuild is in progress, and the behavior is older then git, but I figured I should mention it
* Raises the quantize threshold from 1E-7 to 1E-4. This makes gas dissipate as expected, and should help with the amount of useless gas floating around the station at highpop Adds a garbage_collect() call to the portion of pipeline code where all gasmixes are in one place, this should clean things up properly. Changes BREATH_VOLUME from 2 to 1.99. This is imperative Documents a FUCKING HELLBUG in quantize/breathcode that can lead to breaths just not working sometimes. I'm not sure how to fix this totally, so I'll document it and pray. See <https://www.desmos.com/calculator/5icdlnktus> Adds a unit test to check for this sort of failure. Addendum for people tweaking this value in the future. Because o2 tank release values/human o2 requirements are very strictly set to the same pressure, small errors can cause breakage This comes from QUANTIZE being used in /datum/gas_mixture.remove(), forming a slight sawtooth pattern of the added/removed gas, centered on the actual pressure Changing BREATH_VOLUME can set us on the lower half of this sawtooth, making humans unable to breath at standard pressure. There's no good way I can come up with to hardcode a fix for this. So if you're going to change this variable graph the functions that describe how it is used/how it interacts with breath code, and pick something on the upper half of the sawtooth NOTE: I've made this change with a focus on o2 requirements. Changing this will effect other settings, but most all of them can be ignored, as none will notice. * Thank you moth man Co-authored-by: Mothblocks <[email protected]> * Docs the purpose of the breath unit test, and better explains partial pressure Co-authored-by: Mothblocks <[email protected]>
Fixes mining not making turfs active, I forgot to add a type arg when I added one to AfterChange(). This isn't the only source of the issue, but it's the most prevalent one Cleans up total_moles_specific() slightly, there's no point iterating a list if you already have the gas id. Removes a null assignment and if check, they don't catch anything as things are now, and if someone passes null to the proc I want a runtime Oh and I reordered a bit of atmospherics.dm to make a comment make sense. Don't worry about it
|
After the PR run is finished this still needs to actually be made to work |
|
This pull request has conflicts, please resolve those before we can evaluate the pull request. |
|
wtf how is this shit not failing checks, this is the worst code ever |
|
This pull request has conflicts, please resolve those before we can evaluate the pull request. |
|
Seriously |
|
Wait, will you be the man to dust auxmos? the fuck? are you jesus? |
|
mapping is killing this guy, I suppose. |
|
You need to check the error details. |
I know about this yeah. I haven't had time to get to this. These are trivial to deal with anyways |
Barely, everytime a mapping PR gets merged I weep a little |
)
|
This pull request has conflicts, please resolve those before we can evaluate the pull request. |
|
This PR has been marked as stale due to being in an unmergable state for 7 days. Please resolve any conflicts and add testing evidence, then contact a project maintainer to have the stale label removed. |
|
Not on my watch you dont, bot! |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
About The Pull Request
What it says on the tin. Crossed requested this.
Progress:
share()canonical. Saves cpu time, makes the simulation proper tgstation/tgstation#63785Maints, feel free to add or remove PRs on this list to taste. Ping me on discord to request changes to the list otherwise.
Good news though, a lot of the work was in that initial commit.
Reverts #8264 (Sowwy bacon)
Why It's Good For The Game
Auxmos? It's terrible. Just terrible. You know who's behind it? A real bad guy. Kicked out of Beestation! Can you believe it? Total loser!
And lemme tell you, it's causing us nothing but trouble. Big trouble. The biggest! But LINDA? Now that's a system. So good. The best system, maybe ever in the history of systems.
TG - smart people, very smart - they're handling LINDA. They know their stuff, not like those Auxmos bozos. Trust me.
We're gonna throw Auxmos in the garbage, folks. Toss it out like a rotten orange. And we're bringing in LINDA. It'll be amazing. So amazing. We'll fix atmospherics like you wouldn't believe!
Testing Photographs and Procedure
Screenshots&Videos
Put screenshots and videos here with an empty line between the screenshots and the
<details>tags.Changelog
🆑
remove: auxmos
add: LINDA
/:cl: