Print

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

[Glloyd]

So, while I guess this will be considered a roguelike (procedural generation, simple graphics, punishing) it will be more focused on the light simulation aspects. I've wanted to be able to run amok in a world that a) has interesting things going on whether or not I'm there to witness them, and b) will react to whatever shenanigans I come up with. Ideally it will be as much of a story generator as it is a game, much like DF.

This honestly sounds completely awesome to me. And, I completely agree with you about wanting to run amok in that sort of world.

I was just about ready to re-add some armies but I got thinking about governments a bit. Currently I have 2 "levels" of government - 1 for city level and 1 for state level. I'm thinking it will be much simpler just to make 1 level of government, the city, but let that government be subservient to others if they get conquered or whatever. I think it will both be simpler to set up for me, and also give the game more of a "city-states" kind of feel that I'm going for. Eventually I'd allow the government at the top of the food chain to enforce a certain government structure to those under them. Thoughts?

So, if i read this correctly, you intend to have cities as the seat of government, which can then extend their rule of government over other cities. So would you then have the king/emperor/despot/oligarchs/chief located in the city of origin, reigning over the other cities under their city's rule? And the other city rulers would be sort of governors, who individually rule over their city while under the overall command of the king? If so, that sounds absolutely amazing. A kind of Athens/Rome/Carthage etc. vibe going on, which i like indeed. That may just be my obsession with Roman history speaking, but i like this idea a lot.

Of course, raising levies means that you need to provide food for them,

Which could lead to some tricky situations, as much of the levy would be comprised of farmers i assume. Do you also intend for the levies requiring you to supply them with weapons and equipment as well?

EDIT: I found this link on another thread which could provide insight on how raising a levy could impact food production during different seasons. It's worth checking out, if just for information on medieval farming. http://www.penultimateharn.com/history/medievalfarmingyear.html

[Clownmite]

So, if i read this correctly, you intend to have cities as the seat of government, which can then extend their rule of government over other cities. So would you then have the king/emperor/despot/oligarchs/chief located in the city of origin, reigning over the other cities under their city's rule? And the other city rulers would be sort of governors, who individually rule over their city while under the overall command of the king? If so, that sounds absolutely amazing. A kind of Athens/Rome/Carthage etc. vibe going on, which i like indeed. That may just be my obsession with Roman history speaking, but i like this idea a lot.

Yeah, basically. It's actually set up that way already, but with an extra layer that kind of complicates things from a code point of view. I think it's the same difference in gameplay terms though. If you haven't already, you should check out Crusader Kings II. It's set in medieval times but it has the same sort of thing going on - it really feels like as much of a government and politics simulator as it is a strategy game.

Which could lead to some tricky situations, as much of the levy would be comprised of farmers i assume. Do you also intend for the levies requiring you to supply them with weapons and equipment as well?

Yeah, exactly. I just got it to the point where it stores food in granaries, so you'll need to provide that to any levies or professional soldiers you have. They'll eat while on the march too, so you'll be able to send them out with supply carts, but if you don't have draft animals it won't be as feasible to send out huge armies to conquer lots of land. I'm also thinking of letting them forage for themselves while marching, which will slow down the army but ease the cost of providing food. There'd need to be a limit on how much wild food you could forage from a particular area though.

And thanks for the link, it looks interesting.

[Lord Dullard]

This looks extremely cool. PTW.

If you run into any tricky spots with world generation algorithms, feel free to shoot me a PM. I may be able to offer some advice.

[Clownmite]

This looks extremely cool. PTW.

If you run into any tricky spots with world generation algorithms, feel free to shoot me a PM. I may be able to offer some advice.

Honored to have you comment here, you're one of the big inspirations that led me to start this. Of course if any of my code would be helpful to you, just say the word.

[Glloyd]

you should check out Crusader Kings II.

By far one of my favourite games of 2012. If this turns out anything like that, I'm more on board than i already am

And thanks for the link, it looks interesting.

No problem, I'm happy to help! If i run across anything else that could be useful, I'll be sure to pass it along.

[lemmily]

I've got a basic form of an "economy" (I use the term lightly) functioning now, where they sell and buy things, and have a believed price etc. One thing I'm actually struggling on how to approach is the part where when they go bankrupt they may choose to try a different job. How did you approach this? Or are you handling it differently as in not letting them go bankrupt?

Also, do cities trade with each other? Or do your people only trade within the cities between themselves?

Considering Crusader Kings, I tried the demo. I found it a little uhmm, off? - it just felt like a copy of Europa Universalis (3 is the one I played.),  with some of the features stripped out. Once I understood what everything seemed to be doing, I just felt like there was no real reason to do anything else? I liked the character system in it though, you get to feel more connected to them than in EU3.

P.s is there anything visual you haven't shown off yet? I'd totally like to see some more pictures

[Clownmite]

I've got a basic form of an "economy" (I use the term lightly) functioning now, where they sell and buy things, and have a believed price etc. One thing I'm actually struggling on how to approach is the part where when they go bankrupt they may choose to try a different job. How did you approach this? Or are you handling it differently as in not letting them go bankrupt?

In the paper, they dropped the bankrupt agent and added in a new one of the most profitable type. That's what I'm doing currently but it doesn't make sense from a simulation perspective. In mine, the bankrupt ones get added to a list who will eventually show up in the game world as homeless people, rebels, bandits, etc (or they'll just emigrate to another city). I'm not sure about how new agents will show up. I've added some information about how much food is being collected per turn, so other agents could migrate to your city eventually or the city's population could grow on its own. But I won't have it happen instantly, because there needs to be a downside to having a bad economy.

Also, do cities trade with each other? Or do your people only trade within the cities between themselves?

I had to rewrite most of my economy a few weeks ago, but now they trade with others. There's a screenshot of it below. They still can't seem to sustain themselves for longer than 20 rounds of simulation, so I'll need to dig into that at some point and see what's up.

Considering Crusader Kings, I tried the demo. I found it a little uhmm, off? - it just felt like a copy of Europa Universalis (3 is the one I played.),  with some of the features stripped out. Once I understood what everything seemed to be doing, I just felt like there was no real reason to do anything else? I liked the character system in it though, you get to feel more connected to them than in EU3.

It takes a while to get into. The key thing to understand is that you're playing as a character, rather than an entire nation. There's a great system that allows groups of people to plot for different things (for instance, kill a certain person) and a simple but effective laws system which leads you to some interesting choices (lower crown authority or face widespread rebellion). I hope to translate a little of that to this game.

P.s is there anything visual you haven't shown off yet? I'd totally like to see some more pictures

Not too much, but here are a few. Keep in mind these first two are extremely rough. They show the combat view that I currently have. This one shows my wonky combat system in a small scuffle. The gray and cyan in the background is the early work on a city street generator:



This one shows a heatmap of where the enemy (in red) wants to go. White means it has more desire (except in the top left corner outside the "building"). In this particular instance, it has several desires - the main 2 being "get to that room in the top left" and "run away from player". It looks across all these desires to figure out the best place to move. I was originally blocking the narrow path to the room, but the enemy wouldn't come near me. I moved towards it, and once I got into the open it darted past me to get into the passage. It will obviously need much much more work, but I hope to be able to create a fairly robust AI this way.

Also shown is the attack menu I currently have. It lets you choose how fast and how wide to swing your weapon, and lets you target a specific body part. Chances are this whole thing will get scrapped for something that makes a little more sense. That will be a long way off though.




Finally, here's a slight update to the world map. Some cultures now will build villages which are not full-fledged cities. Also, each "M" is a merchant caravan which is bringing goods from one city to another. There's way more caravans than are on screen, but most of them are in the cities trading. They'll load up on goods, travel to a target city, and stay there til they've sold everything, and then they return home for more.

[lemmily]

n the paper, they dropped the bankrupt agent and added.... *snip*

Ah ok, a reasonable way of dealing with it! I'm also not looking to just "drop" the agent and replace it either. I want it to try and do something else/move/turn to dishonest ways. Hmmm also makes me think about the idea of loan sharks (along with the awful happenings when you fail to pay them) being a risky step some agents might take before turning them into beggars and thieves. Hmm hmmm!

Also I am going to say thanks again for showing me that paper. It's helped so so much, and probably saved me weeks of head scratching.

I had to rewrite most of my economy a few weeks ago, but now they trade with others. There's a screenshot of it below. They still can't seem to sustain themselves for longer than 20 rounds of simulation, so I'll need to dig into that at some point and see what's up.

Do your cities trade directly with each other and send their own merchants or do the agents themselves perform the trades? I've been thinking about how I would deal with this and I thought I would have the city being like an umbrella that then will trade surplus things with the other city. So then the agents inside the city can buy from them. much like a merchant guild for example, who are in charge of trade. It felt like it would be much easier for me to implement. I love the screen with all the little M's 

"for longer than 20 rounds of simulation" I already feel the exasperation of this. Stupid agents following the rules I've told them that make them go bankrupt and foodless... oh wait. damn.

It takes a while to get into. The key thing to understand is that you're playing as a character, rather than an entire nation..... *snip*

Ahh I see well, I will definitely try giving it more chance, it'll probably be better then to not play as a king character straight away? More "stuff" to do then!

Considering your city generation image - will you be looking to keep a fairly grid like street system?

I love seeing the "in-dev" pictures of things, and the different ways people do things. So thanks! I take it you are using influence maps for the AI in the building pic?

I have to say I really really enjoy the colours you use on your maps, but I have to say though the water stands out quite a lot and doesn't blend so well with the land. I'm not sure there's much to be done about that though - It still looks great!

[Mephansteras]

Also shown is the attack menu I currently have. It lets you choose how fast and how wide to swing your weapon, and lets you target a specific body part. Chances are this whole thing will get scrapped for something that makes a little more sense. That will be a long way off though.

Speaking as someone who's done a fair amount of sword fighting, you don't actually *think* all that much when you're fighting. It's pretty much all muscle memory doing the main swings and parries and whatnot. You might try something interesting or different to throw your opponent off, but the bulk of swordswork (and most other melee weapons) is all done automatically. You certainly aren't thinking about how wide to swing with each shot.

An interesting system might be to have the character's skill determine how often they automatically do the right thing in a fight, and just give the player the ability to influence different things through the fight. For instance, starting with an offensive or defensive position or changing that position during a fight. Movement could also be controlled, trying to, say, maneuver an opponent away from where you want to go so you can make a break for it. Or other actions, like throwing a dagger or kicking a chair at your opponent.

Hmm...watch some old Errol Flynn movies. There is lots of great combat stuff in there you could use for ideas.

The DF style 'opportunity' strikes could also be good, giving the player a chance to do something interesting every once in a while during a fight.

[Rumrusher]

so is it possible to murder every one and lead to an age of nothing?

[Clownmite]

Do your cities trade directly with each other and send their own merchants or do the agents themselves perform the trades?

Each city has its own economy. I made a special class of agent called a merchant which is able to buy goods in one city and sell them to another city after transporting the goods there. So the merchant has 2 sets of beliefs as well - how much he thinks he could buy his good for in his home city, and how much he can sell it for in the other town. Of course, merchants need to eat like everyone else.

"for longer than 20 rounds of simulation" I already feel the exasperation of this. Stupid agents following the rules I've told them that make them go bankrupt and foodless... oh wait. damn.

Haha. It's always little things I didn't think of, like "ohhh, they're not eating anything because their inventories are too full to buy any food."

Ahh I see well, I will definitely try giving it more chance, it'll probably be better then to not play as a king character straight away? More "stuff" to do then!

Yeah probably - although playing as a duke in a kingdom with low crown authority is fun too - you can declare war and annex territory and your liege can't do anything about it. Plus you can plot with other dukes for control of the kingdom... But honestly, starting as an Irish count in 1066 is a good way to learn the game. You start weak, with a bunch of weak rivals, you have a concrete goal (form the Kingdom of Ireland!) and you get a great introduction to game mechanics like claims and succession. Also, you can get the game 75% off right now on steam because of a sale.

Considering your city generation image - will you be looking to keep a fairly grid like street system?

Not sure, city maps are a little way off, so whatever ends up being easiest for me to lay out. I do have some parameters which make the streets a little more windy or wider, but I forgot how to work them and just left them in default for that screenshot.

I take it you are using influence maps for the AI in the building pic?

Yeah, exactly. The concept comes from this article: http://roguebasin.roguelikedevelopment.org/index.php?title=The_Incredible_Power_of_Dijkstra_Maps

I have to say I really really enjoy the colours you use on your maps, but I have to say though the water stands out quite a lot and doesn't blend so well with the land. I'm not sure there's much to be done about that though - It still looks great!

Thanks for the feedback. I feel the same way too. I'm also making a distinction between shallow and deep water, which is why there's a dropoff between the lighter and darker shades. If you have any suggestions, you can try coloring my map with a different shade and I can grab the rgb values from paint and change them in the game.

Speaking as someone who's done a fair amount of sword fighting, you don't actually *think* all that much when you're fighting. It's pretty much all muscle memory doing the main swings and parries and whatnot. You might try something interesting or different to throw your opponent off, but the bulk of swordswork (and most other melee weapons) is all done automatically. You certainly aren't thinking about how wide to swing with each shot.

An interesting system might be to have the character's skill determine how often they automatically do the right thing in a fight, and just give the player the ability to influence different things through the fight. For instance, starting with an offensive or defensive position or changing that position during a fight. Movement could also be controlled, trying to, say, maneuver an opponent away from where you want to go so you can make a break for it. Or other actions, like throwing a dagger or kicking a chair at your opponent.

Hmm...watch some old Errol Flynn movies. There is lots of great combat stuff in there you could use for ideas.

The DF style 'opportunity' strikes could also be good, giving the player a chance to do something interesting every once in a while during a fight.

Yeah, you're totally right. Mostly I just put that system in as a placeholder to test my wonky physics system. I want any object to be usable as a weapon, so I put in a silly physics system that tries to calculate damage based on mass and velocity. But for the real game, I was thinking along the lines of what you mentioned. Basically passively going through with the standard parts of combat and then allowing some choice on the user's part when a big "opportunity" comes up.

so is it possible to murder every one and lead to an age of nothing?

Probably not. There will be lots of tribes of hunter-gatherers and chiefdoms in addition to the civilizations you encounter, so there will probably always be some living sapients. But you should be able to murder all of the important figures and cause civilizations to collapse if you are that capable.

[Clownmite]

Hi Clownmite,

the concept of your game sounds very impressive. I found it as I googled for roguelike economy. I was looking for a roguelike game with a strong economy part. Would love to see you post a demo sooner or later. Even if it's not playable yet. It's definitely a very interesting project you are working on!

Andreas

Thanks, I'd like to do a little more work before I post anything. But believe me when I say it needs more work!


Right now, I'm stuck on governments actually doing stuff. Does anybody have any links for how to structure a turn-based strategy AI? Even things as simple as raising a garrison or an army. They need to check resources and whether or not they already have raised the garrison, but they're going to have to check it every turn, which seems pretty wasteful to me. I can code something pretty simple for now, but as more government functions get added I can see it getting pretty messy.

[GalenEvil]

Gamasutra - Designing AI Algorithms For Turn-Based Strategy Games -- Should help a little.

GameDev - Turn Based Strategy AI -- The two links in the OP are from a dev log with some example code to read through. A little brute-force-y in the approach though.

Just a cursory search for stuff, but hopefully it gets you started well

You can do timed functions for some things; movement and immediate tactics would need to be checked every turn, but higher level planning like checking garrisons and telling the place to raise one if necessary would be lower priority and have longer between checks, or escalated to a higher priority if something happens to demand its immediate attention (eg. Enemy unit spotted nearby)

[Skyrunner]

Well, you can never have enough brute force.

[Glloyd]

I'm really liking the new screens. No real indication of what the UI will look like though. Is there any of that implemented thus far?