Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Radsoc]

Energy is what should be required to put the materials/mass together into its finished state, which also should be at a variable cost depending on tech level along the lines of this https://en.wikipedia.org/wiki/Kardashev_scale

I'm curious how you intend to scale energy costs "along the lines of" a scale of energy production.

I'm not being snarky here; there's probably a really good way to handle being able to trivialize energy costs as a matter of raw production at the planetary scale through the building of Dyson spheres (in the Stellaris case) and reduce it to a purely logistical problem, but it has to be done carefully to prevent it feeling like just having lots of energy on hand makes all your devices more efficient, which is what it would be mathematically.

To set the scale of energy production (or tech progress) to correlate with efficiency. With lower tech you would extract less work from the energy produced and so on. That would add progression to project sizes. In the beginning energy would more scarce, make it take long time to build big ships and so on. And if you make energy into a local resource, you would have star systems or civs only capable of producing frigates within reasonable time frames, unless at high tech.

[Damiac]

So they're going to make stellaris a different game again? This is the weirdest early access I've ever seen.  Would have been nice for them to label it as such.

[Trekkin]

Energy is what should be required to put the materials/mass together into its finished state, which also should be at a variable cost depending on tech level along the lines of this https://en.wikipedia.org/wiki/Kardashev_scale

I'm curious how you intend to scale energy costs "along the lines of" a scale of energy production.

I'm not being snarky here; there's probably a really good way to handle being able to trivialize energy costs as a matter of raw production at the planetary scale through the building of Dyson spheres (in the Stellaris case) and reduce it to a purely logistical problem, but it has to be done carefully to prevent it feeling like just having lots of energy on hand makes all your devices more efficient, which is what it would be mathematically.

To set the scale of energy production (or tech progress) to correlate with efficiency. With lower tech you would extract less work from the energy produced and so on. That would add progression to project sizes. In the beginning energy would more scarce, make it take long time to build big ships and so on. And if you make energy into a local resource, you would have star systems or civs only capable of producing frigates within reasonable time frames, unless at high tech.

Okay. That's not what the Kardashev scale is meant to quantify, thus my confusion.

Since energy logistics is an N² problem (for N sources/sinks) with people inevitably asking why they can't/have to control where the energy of every star goes, I had thought you were arguing for some kind of Civ-style ages system mapped to the Kardashev scale, which I could actually see working really well; if I'm harvesting the energy output of entire stars, I really shouldn't need to worry about how much energy I'm consuming to keep the lights running in a hydroponics bay somewhere. It's moot at those scales for a game this cinematic.

We could, for example, have technologies to build structures that transition the empire to planetary- and stellar-scale structural ages, perhaps with the planetary structure zeroing energy costs for everything in its system and the stellar structure doing so for everything in the empire, excepting the costs of building.maintaining planetary- and stellar-scale structures, respectively. Change the color of the energy icon to indicate what age we're in and it won't even need much of a UI change.

[forsaken1111]

I've actually been toying with such a system for a game I'm designing in my spare time (what little I get between work and class), where energy availability is sectioned off into ages. Once you crack fusion, for example, it has knock-on changes that would ripple through society making energy cheap and much more freely available. Successfully miniaturizing fusion reactors to the point where they can be used in small facilities the way we currently use diesel generators would revolutionize industry.

[pisskop]

Stellaris should be Masters of Magic -in space-.


in structure, not in actually having magic

[Telgin]

I can't tell if I'm glad that the AI is dumb.  I just fought two wars in heaven in separate save files, which ended up being relatively easy victories for me in the League of Non-aligned Empires despite my allies being pretty useless.

However, those would have very easily not been easy victories if the awakened empires were much smarter.  They frequently left their powerful fleets parked in remote systems instead of sending them to deal with my fleets, or using them to harass me by taking border systems.  By the end of the war, I was just jumping fleets into those systems and wiping their fleets out piecemeal to get enough war exhaustion to get a status quo peace.

On the other hand, I'm kind of glad they are dumb because there was some serious BS in one of those saves.  An awakened empire fleet was flying through my space and was in a system with one of my planets when the war in heaven triggered.  Said fleet immediately flew over and started bombing the planet, and would have easily taken it if an army fleet had been with it at the time.  Or, honestly, if the fleet had committed to what it was doing at all, since after a year or so of bombing my planet it just left.  I was under the impression that fleets are supposed to go MIA if they're in hostile territory when you declare war, but I guess not.

Guess I should start closing my borders by default from now on.  Or at least to awakened empires.

Unrelated to the above, but the game definitely seems buggier after 2.1.2.  I currently have a scientist that can be assigned to multiple science ships or research positions.  I also just watched an awakened empire's fleet slowly drift across a solar system, cycling between trying to use jump drives to get to a system and just flying there the old fashioned way.  After the fleet finally made it to the hyperlane point, it flew over to the next system and then finally used its jump drive to jump to where it wanted to go in the first place.

[Trekkin]

I legitimately wonder how much the AI is allowed to see at any one time. Sometimes it acts like it's responding to threats it should not, per my checking via observer mode, probably be able to see, but not with enough frequency for me to definitively say it's cheating vision -- and I'd rather hand the AI a noisy picture of its surroundings than burden the game with the computations necessary to impute them anyway.

[Telgin]

I haven't seen any particularly suspicious behavior in Stellaris, but I'd be pretty surprised if the AI had any limitations on what it can see on the map.  It would significantly complicate the AI for little perceived benefit, so I doubt they implemented it.  The AI presumably does at least ignore any empire it hasn't contacted yet, but probably not much more than that.

Then again, it doesn't seem to take much into account when sending fleets around.  I just watched the AI send some fleets through a system with a stellartie devourer which they were not prepared for, with predictable results.

[forsaken1111]

I can speak from some experience now that making an AI that can only see a portion of the game state is ridiculously difficult and expensive compared to just letting it know everything but have it behave as if it can only see parts.

[Dunamisdeos]

I feel like the latter allows for better control and easier alterations if needed.

[forsaken1111]

it also leads sometimes to unintentional cheating bad boy AI like in my current learning project. Little fucker was accessing the world state and not his own limited scope. Cheater!

[Dunamisdeos]

Yeah I guess the downside would be having to set rules for literally everything that can happen, basically. Otherwise it'll start taking things into account that it shouldn't.

[forsaken1111]

For my own project, the AI script can 'see' the entire gamestate but only considers information within n nodes for decision making. This is not the best way to do it, and sometimes leads to really weird behavior or loops. I can only imagine how difficult it is to make an AI that can take an entire 1000 node galaxy into account and try to perform intelligent actions. Mine can hardly traverse a known set of nodes and avoid obstacles right now. He does know to hide when reloading though

[Loud Whispers]

I haven't seen any particularly suspicious behavior in Stellaris, but I'd be pretty surprised if the AI had any limitations on what it can see on the map.  It would significantly complicate the AI for little perceived benefit, so I doubt they implemented it.  The AI presumably does at least ignore any empire it hasn't contacted yet, but probably not much more than that.

Then again, it doesn't seem to take much into account when sending fleets around.  I just watched the AI send some fleets through a system with a stellartie devourer which they were not prepared for, with predictable results.

One of the wonkier AI quirks I've found is that the crises do not work unless they can path to a player if there is a player in game

[Dunamisdeos]

I haven't seen any particularly suspicious behavior in Stellaris, but I'd be pretty surprised if the AI had any limitations on what it can see on the map.  It would significantly complicate the AI for little perceived benefit, so I doubt they implemented it.  The AI presumably does at least ignore any empire it hasn't contacted yet, but probably not much more than that.

Then again, it doesn't seem to take much into account when sending fleets around.  I just watched the AI send some fleets through a system with a stellartie devourer which they were not prepared for, with predictable results.

One of the wonkier AI quirks I've found is that the crises do not work unless they can path to a player if there is a player in game

My own fleets do that. Thankfully they finally implemented the "restrict system" button.