Add the baystation atmos

[Harmony]

Paradise atmos

this system is embarrassing, the gas expands very badly, It does not even help to create traps with atmos

you have a lot to learn from baystation or vg

 

 

[SkeletalElite]

 

On 7/25/2015 at 8:39 AM, necaladun said:

* Anything to do with atmos is trickier than it sounds.

[Birdtalon]

Paradise station github is here, where you're more than welcome to contribute to improving it

https://github.com/ParadiseSS13/Paradise

[Calecute]

Hey since we're here, whats' exactly the difference between bay atmos and para atmos?

[BottomQuark]

30 minutes ago, Calecute said:

Hey since we're here, whats' exactly the difference between bay atmos and para atmos?

As far as I know:
Our atmo pushes things around really fast DAMN YOU FASTMOS  but it transports pressure kinda slow
Their atmo, well basicly 1 tile breach and whole area is empty in less than a second, but it does not move items ALMOST at all

Their atmo makes inflatables REALLY usefull ,cuz if you move into depressurized area without inflatables, oh boi, god bless you cuz you just got another area to fill with pressure

I'd love to see para atmo transports pressure between two tiles faster, but I'm not touching that code, that's suicidal I tell you!

Saying that, I don't mind our atmo. It makes you more vigilant when you work with breaches.

[Harmony]

41 minutes ago, Calecute said:

Hey since we're here, whats' exactly the difference between bay atmos and para atmos? 

bay is area atmos, pushes you into space and throws you to the floor, open a canister to fill the entire hallway instantly

[ZN23X]

12 minutes ago, Harmony said:

open a canister to fill the entire hallway instantly

Does this also work in reverse where if there is a breach the room is devoid of any atmosphere instantly? If so, how would that work in something like a large hallway? Doesn't make sense for the room to fill up or empty instantly if its large.

Should have seen the system we had prior to the current one, or the one before that. This one IS an improvement.

[BottomQuark]

2 minutes ago, ZN23X said:

Does this also work in reverse where if there is a breach the room is devoid of any atmosphere instantly?

I belive so. Yeah

[Calecute]

So instead of atmos being handled by tiles it's handled by rooms? I'd like some more violent depressurizing instead of big rooms being able to be open to space sometimes even indefinitely without much problem (except of course poor folks being fastmos if they get close enough to the breach), but I think handling atmos with tiles makes that pretty much resource intensive, and handling by rooms may have some other problems. 
 

Quote

Doesn't make sense for the room to fill up or empty instantly if its large 

This depends on the flow rate from the can. I'm pretty sure IRL you could fill a big room instantly with enough air flowrate. The time it takes for pressure to normalize is pretty much negligible. Imagine filling a pool with water. The time it takes for the water to flow from one place to the other, so the surface is even is irrelevant, the only constrain is how much water are you pumping in it. (althou actually water is a bad example cuz its uncompressible I think you get the idea). 

Assume you have a kind of balloon, filled with air, it fits in your hand, but it has enough air to fill a 10mx10mx2m room (200kl) with one atmos pressure. The balloon bursts, how much time you think it takes for air to distribute evenly in the room so we have one atmos pressure everywhere?

[Calecute]

Quote

 

P = Po exp[-(A/V)t*(200m/s)]

This gives you a quick rule of thumb, the one-one-ten-hundred rule: 
A one square-centimeter hole in a one cubic-meter volume will cause the pressure to drop by a factor of ten in roughly a hundred seconds. 
(for quick approximations; only roughly accurate). This time scales up proportionately to the volume, and scales down proportionately to the size of the hole. So, for example, a three-thousand cubic meter volume will decompress from 1 atmosphere to .01 atmosphere through a ten square centimeter hole on a time scale of a sixty thousand seconds, or seventeen hours. (it's actually 19 hours by a more accurate calculation).

The seminal paper on the subject is by Demetriades in 1954: "On the Decompression of a Punctured Pressurized Cabin in Vacuum Flight."

 

From a random site  I found on the internet, but it makes sense, has formulas and citations.

If we assume one tile to be 2x2m, and the room height to be 2m, and trow this formula in Wolfram Alpha, then a one tile breach would vent (to 0,1atmos) one tile in 0.02s, 100 tiles in 2.3s and 1000 tiles in 23s (by my rough estimates the whole station is in this order of magnitude). It would them go from 0,1 atmos to 0,01 atmos in the same time. Looking at our old map, escape had 98 tiles so one window breach in it would vent it in ~2.3-4.6s depending on how low the pressure has to be for you to consider it vented.

Note the the same site says:

Quote

For reference, when the pressure drops to about 50% of atmospheric, the subject will be entering the region of "critical hypoxia"; when the pressure drops to about 15% of atmospheric, the remaining time of useful consciousness will have decreased to the 9-12 seconds characteristic of vacuum.

It would also work in reverse. If you opened a vented room to a full room it would fill it in the same time it takes to empty it, as long as the other room is big enough for the change in total volume of air to be negligible. Could also be done this quick with a can with enough flow rate, actually even quicker I think, but not in a practical way, and not quicker than the speed of sound. Of course a big wave of air hitting you at the speed of sound might not be very healthy, my guess is that it would also be noisy as fuck, but a good way to find the room's resonances. 

Edited September 17, 2018 by Calecute
Added a small parenteses on the size of the station.

[Harmony]

7 hours ago, ZN23X said:

Does this also work in reverse where if there is a breach the room is devoid of any atmosphere instantly? If so, how would that work in something like a large hallway? Doesn't make sense for the room to fill up or empty instantly if its large.

Should have seen the system we had prior to the current one, or the one before that. This one IS an improvement. 

yes, it's an area-based atmosphere, f you break a window, the space throws you to the flood and drags you out.

you can not create a fire, because of its way of expanding so slowly

 

[Calecute]

Not sure how fast plasma would spread in a pressurized enviroment thou. That's one of the problems with using area, tiles does that better (thou much slower than it should be I guess).

[ZN23X]

Still just doesn't make room for an entire room to fill or drain instantly regardless of how big the room is and how small the hole is. I don't understand why that is superior, or more realistic.

[Calecute]

Well I just showed through examples and some math of how explosive depressurization works IRL that it's more realistic. As I said under some plausible assumptions on the size of a tile old escape would vent in 2-5 seconds with a one tile breach. 

But gameplay wise I think it would be better to have rooms depressurizing quickly for lots of reasons. It would make so we lose more air, making  air more valuable, witch can create some interesting choices on what to pressurize as atmos and command. It would make people working in breaches work much more carefully so as not to vent another room. It would make so breaches actually need to be fixed. 

Right now breaches are a joke to be honest, people just open firelocks and go around them. If you don’t stand one or two tiles away from the breach you are ok. Just as an example I made a breach in escape in a private server, turned atmos back on, went to eat and came back to see how the breach was:

24 minutes with the room breached, and you can breathe ok in 2/3 of the room, and can go through the other 1/3 with no problems if you don't stay there and avoid the tiles right next to the breach. That keeps engies lazy, breaches in maint stay there forever, and even in some other parts sometimes. That makes people breach without regard for consequence. People breach perma all the times, never thinking they will suffocate their prisionmates. I sometimes breach into/out of the station so I don't go through a slow external airlock in places where the station is not reinforced. People have no discipline regarding firelocks, because there's no consequence to opening firelocks to areas with atmos alarms. This room is roughly 208 tiles, IRL it would vent in 5-10s with a one tile breach assuming tiles being 2x2x2m.

[Allfd]

Author of fastmos, which is an addition to Linda.

Our atmos is a few years behind the other code bases, in both features and performance.  The throwing mechanic makes it appear faster, but the reality is, its super expensive to do tile by tile calculations, which means that to make it spread faster, we would have to increase the rate at which it calculates all active tiles.

So essentially, short of porting Bay or TGs system, this is about as good as we will be able to get.

Considering that byond is single threaded, this is extremely unlikely to change in the future.  Its a known limitation of our atmospherics system.  Anything short of a port of a more modern system will not change this dynamic.

As far as the 24 minute thing goes,  before fastmos, our atmospherics system was 4 times slower.

So um, that would be 100 minutes.

Edited September 18, 2018 by Allfd

[Calecute]

Don't get me wrong Allfd, I know a lot of work went into fastmos, and that it's a hard problem. I also know that's very expensive computationaly to make tile atmos faster. And I also know there are trade-offs between area x tile atmos. As I said I think tile atmos probably deal with gases under ~same pressure mixing better, but there a probably other issues I didn’t think about, mainly for lack of experience in area atmos systems. 

I'm not very familiar with SS13 coding and the differences between code bases, I was just trying to figure out the differences and their consequences and pointing out we really have a problem simulating explosive decompression and why it would be better to simulate it better for gameplay and realism reasons because people raised doubt it was a problem. 

About the 24 min thing, it will eventually reach an equilibrium where the air that goes in is the same as the air going out because I turned the vent back on. So a fast system just lead to lower equilibrium pressure and more different pressure areas in the room. 

But as someone who mainly play atmos, I love fastmos even with its flaws, thank you for your work!

[Allfd]

Oh no worries,

I prefer Bay atmos myself,  I just don't want to deal with the difficulties of porting it, as I think I would be unable to overcome them.

[necaladun]

3 hours ago, Calecute said:

It would make so we lose more air, making  air more valuable, witch can create some interesting choices on what to pressurize as atmos and command

Currently the amount of air on the station is basically unlimited. I've tried to empty atmos tanks (without just breaching them) before and barely managed to get half of it out over the course of a round. Combined with the amount of air in various tanks and cannisters, there's enough spare air on the station to refill it a few times over. 

3 hours ago, Calecute said:

24 minutes with the room breached, and you can breathe ok in 2/3 of the room, and can go through the other 1/3 with no problems if you don't stay there and avoid the tiles right next to the breach. That keeps engies lazy, breaches in maint stay there forever, and even in some other parts sometimes. That makes people breach without regard for consequence.

Make the breach bigger, and it'll be faster. Escape is a fairly large room - add a floor breach or take out a good number of the windows, and it'll be much faster. 

I like my atmos dangerous, but a single window breach causing the deaths of basically everyone in escape would...not be fun - especially with generally how unavoidable it is if you're just sitting there waiting then are suddenly sucked out a window.

2 hours ago, Allfd said:

Considering that byond is single threaded, this is extremely unlikely to change in the future.  Its a known limitation of our atmospherics system.  Anything short of a port of a more modern system will not change this dynamic.

[Harmony]

tg has a system similar to a paradise, but there are no problems to create fires, the gas expands considerably better

[Calecute]

I dont feel we have a problem to create fires tbh. 

[Calecute]

Quote

 Currently the amount of air on the station is basically unlimited. I've tried to empty atmos tanks (without just breaching them) before and barely managed to get half of it out over the course of a round. Combined with the amount of air in various tanks and cannisters, there's enough spare air on the station to refill it a few times over. 

I suspected that we had too much air but never tested. 

 

Quote

 

Make the breach bigger, and it'll be faster. Escape is a fairly large room - add a floor breach or take out a good number of the windows, and it'll be much faster. 

 I like my atmos dangerous, but a single window breach causing the deaths of basically everyone in escape would...not be fun - especially with generally how unavoidable it is if you're just sitting there waiting then are suddenly sucked out a window.

 

I do know that a bigger breach will take air out faster, but even small breaches shouln't be just a minor inconvinience. They should be lethal. While it would indeed no be fun being insta killed by atmos if you just sitting in escape, is not fun to be insta killed by the engine, and its not fun to be killed by a plasma fire in escape, but we allow it in some circunstances.

 

[ZN23X]

Curious how much quicker a room as large as escape would empty if it didnt have the vents working thier asses off to replace the lost air, cuz thats a factor as well when you are talking about a small breach.

Edited September 18, 2018 by ZN23X

[BottomQuark]

As far as I know, air alarms never turn to "refill" mode own it's own.

It's always filtering or syphoning if there is plasma in the air.

And filtering does not provide much air if it even does that, I never noticed any pressure change due to air alarm beeing on filtering, but again, never really tested it on private server, so I might be mistaken.
If it does refill the area, it does that REALLY slowly, so can be easly ignored.

[Calecute]

Filtering does refill the room. On filtering the vents will pump air into the room until pressure is 1atmos. Filtering is pretty adequate for refilling rooms most of the time to be honest. Refill will frequently cause space wind, and overpressurize the room, I just use it when I have to repressurize a pretty big area, and usually turn it back to filtering when pressure is 0.8atmos. In my example in escape the vent was set to filtering. 

Also air alarms does not set to syphoning if they detect plasma on their own, only if they detect a pretty big temp and pressure (usually plasma fire). They set to off if they detect rapid loss of pressure. They're not even set to contaminated on detecting plasma. 

[Fursamie]

On 9/18/2018 at 8:25 AM, necaladun said:

 I like my atmos dangerous, but a single window breach causing the deaths of basically everyone in escape would...not be fun - especially with generally how unavoidable it is if you're just sitting there waiting then are suddenly sucked out a window.

Adding to what calecute said, unless there is an obstacle infront of you (window, wall, etc), you SHOULD be flying out. To put this into perspective: if you have ever been skydiving, have you noticed the slight pull when the jet/helicopter/plane door opens? That is the difference of pressure interacting and you being directly effected by it. The higher you are in the atmosphere, the lower the pressure is. Now, imagine this very same thing happening but 

1: The helicopter is hooked with oxygen

2: The windows are square and near impenetrable

3: The material which the helicopter is made of is near impenetrable and unbreakable

4:You are in space, in this very helicopter

Now, if the pressure is so low in the ocean which your body would be hopelessly crushed and broken, you would be absolutely decimated in space where there barely is pressure. That large of a pressure difference would most likely gurantee that if a breach had taken place in the bar, realistically you would most likely be dragged to there from the medlobby(assuming no firelocks). Granted it's no fun but being killed by a metal rod because of fastmos isn't fun either.