Decompression

[PrototypeNitro]

 

Here is the suggestion for more realism in the game. The Decompression. I think the decompression is too slow.

 

Situation: Science wing. Scientist makeking bomb. The scientsit fail. The bomb made a breach. The scientist run out.

 

And I think it's not possible. like planes. It's almost impossible to stay up in planes which has a breach. Because you have plane pessure and planet pressure in the sky of course. Imagine Station pressure and SPACE ! The breach should become super vaccum, Air go in space with the stuff and personnel if they are near it. Like Baystation 12.

 

I see 2 good point.

1- Give real job to Enginnering.

2- Add challenge when you are near a breach.

 

[Vivalas]

We don't even have to implement ZAS to achieve some emulation of decompression. Just increase the tickrate of the atmos controller and we get something even just barely dangerous.

[Ansari]

Does anyone remember the day of lag and 10 minutes of processing every bomb happened when ZAS was here? No one remember..

[Dinarzad]

 

Does anyone remember the day of lag and 10 minutes of processing every bomb happened when ZAS was here? No one remember..

I remember.

But in all fairness, ZAS is not being asked for, but an increased tick-rate on LINDA and ALSO since then we've gotten a hardware upgrade.

It is not so easy to compare nowadays because the context and siutations have changed.

It's like saying "Well last time I tried to breathe underwater I choked. Now I have a Snorkel but if I try to breathe underwater again I'm going to choke because that is what happened last time."

 

[Rurik]

 

I would absolutely adore a change like this. I was thinking of making a post like this myself, not because it would be more realistic, but because I believe it would also make things much more interesting and fun.

 

Engineering and atmospherics would become significantly more needed, as in my mind, it should be. Breaches would become a higher danger, and therefore become higher on the list to be repaired. Replenishing oxygen and pressure would also become more important, giving more importance to atmos techs. There would also be more danger in general with breaches. You would be sucked towards the breach, as long as there is air of course, maybe even faster than you can outrun it. Being able to stand on the other side of the room and not feel the effects for a couple mins would no longer be the case.

 

I don't know much about ZAS or LINDA at all, but I know if there is some way to speed up decompression without a significant downside, it would be worth it.

 

[polkjm]

 

There's a REASON the atmos is like it is. Paradise has a purposesly slow as fuck atmos tick rate, and that's the reason the game is so smooth. Honestly I don't have a problem with how atmos is right now. If I wanted realism and immersion I'd go elsewhere. I'd rather have smooth gameplay.

 

I've played on Bay a lot, and the realistic Atmos there is very annoying. If you're in a room with a breach you're 100% dead (which is an understatement if we're talking about realism). Baystation would be super fucked if it wasn't for the fact that breaches are really rare because antags don't take breaching the station lightly.

 

[BiberDark]

 

Thats nonsense, Atmos right now is bullshit. As i mentioned 100times in other posts. Atmos has to be annoying. Is it the idiotic behaviour of people who try to force open sealed areas just to get to bar or somewhere else. They never try to go arround they allways go straight through the sektor with the wallbreach. For this they have to die a long and painfull dead. Best would be their keyboard would give them a electroshock....

 

Better Atmos would be good for engineering and atmostech, also give the ai back a posibility to kill a traitor.

 

You only have to add a few firelocks so not all of the station is fucked like you said. Besides the problem with what i mentioned before. But they die and learn, at least i hope.

 

 

See link here :

viewtopic.php?f=2&t=8924&start=10&hilit=zas

viewtopic.php?f=12&t=8868

viewtopic.php?f=12&t=8175

viewtopic.php?f=12&t=7588

and so on...

 

[Dinarzad]

 

There's a REASON the atmos is like it is. Paradise has a purposesly slow as fuck atmos tick rate, and that's the reason the game is so smooth. Honestly I don't have a problem with how atmos is right now. If I wanted realism and immersion I'd go elsewhere. I'd rather have smooth gameplay.

 

I've played on Bay a lot, and the realistic Atmos there is very annoying. If you're in a room with a breach you're 100% dead (which is an understatement if we're talking about realism). Baystation would be super fucked if it wasn't for the fact that breaches are really rare because antags don't take breaching the station lightly.

 

Slower atmos is NOT the reason the gameplay is so smooth. I encourage you to go and observe a few rounds on /tg/station, they use LINDA as we do, but they have a very VERY sped up version of it, and it runs smooth as silk over there. It's primarily a question of hardware that the server runs on.

 

Now, I don't think our hardware compares to /tg/ so we likely can't mimic their tick rate, even post-upgrade, but anything that's "faster then growing grass" is a benefit at this point, not only for increased lethality in the game, which is kinda low, but also it would make the Atmos Technician job actually relevant again, with decompressed rooms needing an air pump or the atmos system optimized so those rooms refill faster.

 

[Fox McCloud]

 

Slower atmos is NOT the reason the gameplay is so smooth. I encourage you to go and observe a few rounds on /tg/station, they use LINDA as we do, but they have a very VERY sped up version of it, and it runs smooth as silk over there. It's primarily a question of hardware that the server runs on.

 

You're presuming a lot, here. Have you sat down and profiled our server, once--ever? Have you actually taken the time to see what is and is not expensive on Paradise, proc wise? Could you please name the most expensive loops for the game? What are the current major bottlenecks? What are some current inefficiencies?

 

 

My point is, please don't go around making claims like this unless you've actively coded on Paradise and have some inclination of how much overheaded we do/do not have and just how expensive things are.

 

It's not just a simple matter of "turning up the tick rate" as you're calling it. We have to unify the atmospherics system into one gigantic controller (technically the full atmos system is controlled by three separate controllers) before we could move up the speed at which it processes. In addition to this, it's not just LINDA itself we have to take into consideration---it's the entire game. Like it or not, Paradise's coding team is relatively small--we don't have the ability to make as many major leaps and bounds in performance improvements that TG does--as such, we're nowhere near as efficient as TG in terms of overall operating capability.

 

When you're more efficient, you have more overhead...and when you have more overhead you can afford to run things faster or play a little loose with some feature's performance capabilities---we have to put more thought/consideration/performance testing into things like this before we can remotely consider them.

 

It's not a simple matter of "maintainers/coders don't want to turn up LINDA's speed because they're a bunch of selfish sourpusses that don't give a flying fart in space about the community". No, we've profiled it, we've looked into it, and at the current level of optimization of Paradise, we really can't afford it---I'd even say that just turning it up because a few people want it fast would be acting against the community, because it would degrade overall gameplay for far more people and generate a far more negative game experience than LINDA being a bit on the slow side, as it is.

 

tl;dr: Don't go around making claims unless you're actually familiar with/have worked on SS13's code. Also stop claiming how simple things are to change without understanding consequences of those changes.

 

[NethIafin]

 

@Fox McCloud

Add to that complete inability to test any such major changes with high loads BEFORE publishing it on the server, and you get on the really dangerous path of playerbase disintegration (if something bad happens for even a week)

 

[Dinarzad]

 

-snip-

 

Yeah, all due respect, Fox, but half the things you're saying here, are not words that I said. I never said "Just turn it up lol" nor did I proclaim that "Maintainers don't wanna do it cuz fun nazis"

That's putting words into my mouth. What I said was, Slow atmos was not the reason the game was suddenly smooth, it was another very big factor in server hardware upgrade, and that a fast atmos can still run smooth, a la /tg/. In fact I even -expressly said- I don't think Paradise can run LINDA the same way /tg/ does. Not right now anyway.

Now, you wanna just make a general statement on "It's not that easy to fix", go for it, you're the maintainer. But don't make it sound like I just lolled on the floor describing how coders are anti-fun.

 

tl;dr: I will stop going around making those claims, as soon as I start making them to begin with.

 

[NethIafin]

Last time we were talking about, someone posted idea about knocking people down if they are pulled by pressure. And depending on the pressure difference, the time of knockdown should be different.

[shazbot194]

 

I had a thought. It would be most useful if checks for pressure and contents is what is bottle necking the code. What if we were to combine elements of both linda and zas. Where it checks room by room for a pressure difference, contents, if the room is connected to space, and floors missing, somewhat like saz. There could be a small threshold (+-2.5 pKa/+-1%) to keep CO2 from humans making it go crazy. If it detects any change it calculates air flow like linda. The benefit would be it could let areas that are mostly unchanged be calculated with less checks.

 

The idea is that if you had a 5x5 interior it would only have to check the pressure and contents 25 times, as opposed to 4 checks each tiles which is 100 checks for our example. While I could be wrong on how linda works, this is what I have gathered/told how it works.

 

It would probably take a while to make and may not be much better if at all, but if it is the checks that are holding the code back, this could be one way to speed it up some.

 

[GelatinousGlob]

 

There certainly are some improvements to be had but a big part of the problem is atmospherics code is fucking weird and most people aren't that comfortable messing with it and doing any real changes.

 

I do think atmospherics should be a much higher priority though. It doesn't feel much different now than when LINDA was first implemented and most of the early complaints about switch to LINDA where countered with the promise of future tweaks and upgrades to get a better balance between the efficacy of the system against the performance costs.

 

[Shadeykins]

 

Fast LINDA exists on /tg/ and works on a system with specs that are on par with ours.

 

From my understanding though our code is also incredibly inefficient when compared against /tg/'s due to the sheer number of maintainers they have over us.

 

As to Ansari, it wasn't ten minute lag. I've been around since ZAS and I've played on servers that still utilize it. It's not as bad as everyone likes to make out. Is it more costly? Sure, but that comes hand in hand with being more reactive and having far greater scope.

 

[Spacemanspark]

Fast LINDA is more tedious than it is dangerous, anyways.

[BiberDark]

Then why not port the faster linda ?

[Urlance Woolsbane]

 

Then why not port the faster linda ?

Because, I imagine, it's not 100% compatible with Paracode, and as Fox said, Paradise doesn't have enough coders to do so willy-nilly.