Z-Level meme protected AI core

[BurgerBB]

This PR isn't too complex. Makes the AI more protected by adding a reinforced Z-level layer to prevent memes from occuring without a struggle. I mean, you could still do this but you're going to need to drill through several layers of reinforced walls and regular walls.

https://github.com/Aurorastation/Aurora.3/pull/4668

[Saudus]

Yes please, +1.

[eab549]

i will say that this is a huge -1, according to the image yes it covers the core from a "rod drop" from an engineering stand point this would only just slow me down at most. As it also stands rod dropping is just not done because people are scared player complaints over it or being "cheesey" with their malf kills. in order to even exploit this weakness of the AI core one must bring a shovel, pick axe and an RCD as well as knowing the exact spot to dig before breaching less the AI realize its being attacked and have a borg simply move its core in a few seconds. Aside from your AI core changes i would say great job moving the bomb range im tired of fixing that damn maintenance tunnel and its piping.

[K0NFL1QT]

Needs more turrets on the main level installation to actually deter an engineer from dismantling their way in, and then down. Put them encased in reinforced glass, able to shoot the outside area as people approach, and able to focus fire anyone who gets inside.

[eab549]

issue with turrets on the top is that miners may get shot by lethal turrets if they are exposed, this also just puts a giant target on top of the core for anyone to see unlike now where only someone with the blueprints or knowledge of them would beable to know and get the equipment to do it.

[Guest Marlon Phoenix]

Needs more turrets on the main level installation to actually deter an engineer from dismantling their way in, and then down. Put them encased in reinforced glass, able to shoot the outside area as people approach, and able to focus fire anyone who gets inside.

 

That's so unnecessary imo

[Eve]

, according to the image yes it covers the core from a "rod drop" from an engineering stand point this would only just slow me down at most.

 

Slow you down? From cheesing the AI again?

 

As it also stands rod dropping is just not done because people are scared player complaints over it or being "cheesey" with their malf kills. in order to even exploit this weakness of the AI core one must bring a shovel, pick axe and an RCD as well as knowing the exact spot to dig before breaching less the AI realize its being attacked and have a borg simply move its core in a few seconds.

 

I honestly disagree. People do not 'rod-drop' the AI because they are scared that they'd get bwoinked for it. People do not do it because it's pretty much well-known at this point that it fits the definition of powergaming. There is pretty much no way an AI can protect itself from getting rods dropped on it, and the chance gets even smaller when you realize that the AI doesn't have camera coverage one z-level above it, so it'd also have no idea what is coming for it.

AI being able to get cheesed by rods isn't an intended weakness, it's a game exploit. AI gimmicks are the most entertaining, in my opinion, they also do require the most build-up. Imagine speaking to Command, hinting towards them that you may not be the automated-door-opener that they've expected you to be, slowly build up abilities, get your Androids to be upgraded before you reveal yourself, before you finally start taking control of the station and its systems, locking people away from each other, and maybe call an emergency shuttle and try to force them off it. Or maybe you wish to activate the nuke to end the round with a bang. Now, imagine that your entire gimmick was done for 65% through because you didn't realize an engineer cycle out of the airlock, north of security, and drop some random shit onto you, and now you're in ghost-chat!

Doesn't really sound that nice, right?

 

issue with turrets on the top is that miners may get shot by lethal turrets if they are exposed, this also just puts a giant target on top of the core for anyone to see unlike now where only someone with the blueprints or knowledge of them would beable to know and get the equipment to do it.

 

I dissent, undoing walls, digging and whatnot as a non-engineer will warrant you a nigh-instant bwoink, and I don't really believe that the spot above the AI would be marked with a red X. Given that, I don't think turrets are that necessary on the z-level above the AI core. But, what I think is necessary is to fill the circle of reinforced walls, with more reinforced walls, so it becomes harder to breach them..

Now, in terms of changes, I certainly do like them. Toxins blowing up Cargo Maint has been an issue for the longest time, and I can see how this'd be able to fix it with a simple solution.

In terms of the AI core, however, it's going to be doing its job partially, as the walls can pretty much still be undone, but it seems fine to me, to discourage newer CEs from committing themselves to the instant death that is getting the SS13 equivalent of a Pharah ultimate on you.

[eab549]

snip

 

im going to agree with your assessment of the bomb range being changed, on to the topic of the AI core main level defense it seems my point was missed during your fictional story of malf gimmicks being good. The point im attempting to make is that as a defense it is not a deterrent, it dosnt cover the upload area at all meaning you can bypass the majority of the doors and directly access the upload console, meaning you also have unrestricted access to the core itself if they are fast enough. The big circle of walls dos nothing but alert people of the area already there to be exploited because its a half measure at best to prevent the easy route to do things despite the accepted method of malf killing being honestly faster and easier to pull off.

[K0NFL1QT]

Needs more turrets on the main level installation to actually deter an engineer from dismantling their way in, and then down. Put them encased in reinforced glass, able to shoot the outside area as people approach, and able to focus fire anyone who gets inside.

 

That's so unnecessary imo

 

The AI core hasn't been designed with multi-z defenses, and it needs to be adapted to reflect the totally legit approach of going in from the top. Hence, just make it more difficult to achieve in game. It's a better solution than surrounding the AI with unbreakable walls and and unbreakable roof, and doesn't require bloating the ooc rules with an addition that says 'despite the AI core having vulnerabilities, you must always try and enter from the front', because that is frankly enforcing stupidity. Everyone seems to agree that just putting walls in the way is not going to be enough to do more than slow someone down.

Regarding the practicality; the main level turrets can easily be linked to a turret control that's separate from the AI core turrets, so they can be turned on/off independantly as required; default setting is off for Code Green so that random miners don't get lasered. But you could also just clear out a broad section of rock in that area to make it a generally terrible area to be in for mining. Also, keep in mind that the main level isn't the surface level, it's two layers beneath the surface and thus the 'secret' of the AI cores location is not really more exposed than it already is.

Another idea that's less killy; instead of turrets, put a shield generator in Burgers proposed main level room, that's linked down the z-level to the AIs SMES. But that's going to drastically increase the AIs general powerdrain, so it would need an overall more more independant power overhaul.

Edited April 28, 2018 by Guest

[LordFowl]

This implementation is both incredibly ugly and quite clumsy. If the offense is such that administration feels the need to step in routinely, then having or not having reinforced wall spam will solve nothing.

Furthermore, if you have a problem in mapping, your solution should never be to throw more reinforced walls at it.

[Korinra]

Why not build a second story to AI's chambers something with no floor. That way it's an open space two-story room. Now you have to breach in the walls on EITHER z level, but if you breach the top z level, you still have a drop, but one that leads just inside the upload (granting no benefit to breaching from the main level. Add a few cameras around the wall (but no motion sensor), It would add the aspect of many entry points without the problem of cheese.

[K0NFL1QT]

Why not build a second story to AI's chambers something with no floor. That way it's an open space two-story room. Now you have to breach in the walls on EITHER z level, but if you breach the top z level, you still have a drop, but one that leads just inside the upload (granting no benefit to breaching from the main level. Add a few cameras around the wall (but no motion sensor), It would add the aspect of many entry points without the problem of cheese.

 

Making the second layer have no flooring would be bad. If it has no turret defenses, all you need to do is breach the walls and then build a catwalk over to directly above the exposed core. At least when there's reinforced flooring you need either thermite, an RCD or an explosive to get through it at all.

[Guest Marlon Phoenix]

No matter how dense you make the AI security, it is going to be defeated by engineers. We should not be in an endless arms race. The AI core should have vulnerabilities that allow it to be countered. The exploits you mention are rather difficult to reproduce (even if it is repeatable) but the potential mobility of the AI core (which seasoned AI players take advantage of) means that in-game there are already methods of countering or facilitating z-level shenanigans.

[Scheveningen]

You know, you can just furnish the level above with reinforced floors and it's practically impenetrable, right? Short of a RIPLEY drill.

You're also metagaming if you, as a station engineer, know the exact location of the AI core to drop rods on it from a z-level above. You should be permabanned from engineering if you're caught doing this. This is so utterly blatant in intent and execution what you intend in doing here.

[K0NFL1QT]

You know, you can just furnish the level above with reinforced floors and it's practically impenetrable, right? Short of a RIPLEY drill.

 

The area above the core already has reinforced flooring. You can breach it with a Ripley Drill, an RCD, an explosion and I'm told thermite.

[Scheveningen]

Huh, okay, didn't know that.

That still requires a bit of deliberate effort to do, then. This doesn't sound like it's done on a regular basis by multiple engineer players.

[K0NFL1QT]

It's only been done once ever, by a CE, that I know of. The problem is that people know about the strategy and some people really dislike it. An AI core is easy enough to kill just going through the frontal assault, and utilising the multi-z structure of the station makes it easier. Yet it's a legitimate tactic, despite really only a CE/RD/Captain being able to pull it off due to limited knowledge of the AI core. However, 'limited knowledge' doesn't stop every engineer knowing the exact spot to setup an emitter to guarantee an uncounterable core kill.

[Itanimulli]

It's only been done once ever, by a CE, that I know of. The problem is that people know about the strategy and some people really dislike it. An AI core is easy enough to kill just going through the frontal assault, and utilising the multi-z structure of the station makes it easier. Yet it's a legitimate tactic, despite really only a CE/RD/Captain being able to pull it off due to limited knowledge of the AI core. However, 'limited knowledge' doesn't stop every engineer knowing the exact spot to setup an emitter to guarantee an uncounterable core kill.

 

Now that, kiddo, is a lie. I've had it done to me the first time I even tried malf. I've seen it done to other Ai players who do get subverted. Digging in from the top happens all the time. Not saying it should be prevented, it's just, it happens more than you'd like to think.

[K0NFL1QT]

It's only been done once ever, by a CE, that I know of.

 

Bold for emphasis.

[Eve]

it seems my point was missed during your fictional story of malf gimmicks being good.

 

Seems like a personal issue that you do not enjoy Malf AI gimmicks, really. I have quite enjoyed most of them.

The point is, I do not want to go deep into it, but as a CE main myself, I strongly believe that rod-dropping is an exploit, and should be banned all together.

Now that I think of it, why are we even blocking the AI's topside to prevent rod-cheese, instead of just treating rod-dropping like the act of powergaming it is?

Most of the time, AIs won't be able to see the rod-drop coming, as it's outside of their camera coverage, and even if they do, they'd need to alert an Android, so that they could be moved. But here's the thing:

Unless there is an Android who exists to move the AI for the duration of the round, the AI will get killed by dropping rods.

Cheesing the AI with rods isn't be a strategy, neither should it be one unless the AI gains a way to counter it efficiently.

[Scheveningen]

Now that I think of it, why are we even blocking the AI's topside to prevent rod-cheese, instead of just treating rod-dropping like the act of powergaming it is?

 

I agree. It is a form of powergaming and is not a practical way to do things in the roleplay sense. It is deliberately done to exploit the throwforce of rods and the vulnerability of the AI core from vertical attacks. Which an average engineer should not know the exact GPS location of the AI core.

[K0NFL1QT]

It is deliberately done to exploit the throwforce of rods

 

The 'Rod Dropping' strategy in actuality is only tangentially related rods at all, and basically nothing to do with exploiting throwforce of rods. If rods had a disproportionate throwforce, you'd see people throwing them at each other on the station; like how you see people throw glass shards. Rods are just easy to make, and plentiful to anyone with a few sheets of metal. The issue is the fallforce of all objects; you could, if you were so inclined, kill the AI with sand, I'm fairly sure. If coders do want to adjust the fall force of objects to vary with their actual/approximate weight, they're welcome.

 

the vulnerability of the AI core from vertical attacks.

 

The vulnerability of the AI core to vertical attacks should be valid, because the core is clearly designed to not have any multi-z defences, beyond a metal roof. The AI core has one layer of 'reinforced flooring' for its above and below structure; about as 'secure' as one layer of reinforced wall, but without any turrets to defend it vertically.

 

Which an average engineer should not know the exact GPS location of the AI core.

 

True, yet literally every engineer already knows the exact location of the AI Core as shown by literally every Malf round ever in which engineers set an emitter to kill the AI, yet this is still permitted.

[Eve]

True, yet literally every engineer already knows the exact location of the AI Core as shown by literally every Malf round ever in which engineers set an emitter to kill the AI, yet this is still permitted.

 

I'm pretty sure you can snipe where the AI core is by using mesons, after digging towards its general location, but again, I've never seen an engineer (not CE), do the exact thing.

The AI core is designed to not have multi-z level defenses because it cannot have multi z-level defenses. The defense mechanisms of the server, especially for AI, are extremely poor, and the issue with digging its top and dropping random shit in it is more of an OOC exploit than anything, at least from how I've seen it, and how people have talked about it.

[LordFowl]

The 'Rod Dropping' strategy in actuality is only tangentially related rods at all, and basically nothing to do with exploiting throwforce of rods. If rods had a disproportionate throwforce, you'd see people throwing them at each other on the station; like how you see people throw glass shards. Rods are just easy to make, and plentiful to anyone with a few sheets of metal. The issue is the fallforce of all objects; you could, if you were so inclined, kill the AI with sand, I'm fairly sure. If coders do want to adjust the fall force of objects to vary with their actual/approximate weight, they're welcome.

 

 

This is, more or less, completely incorrect. The formula for fall impact of objects takes into account pretty much every relevant variable of an object, including their approximate weight. Dropping a rod from 2 stories up would deal 27 damage, largely due to its high throw speed and throw force and not insignificant weight class, whereas dropping sand from the same height would only deal 6 damage. And this is only because sand already deals 10 damage when you throw it normally.

[K0NFL1QT]

I stand corrected then. Is the rod the most optimal dropping weapon in terms of damage then, or are there other objects that do more damage?