Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - lookmeat

Pages: [1]
1
I do understand this, but on the other hand having a single small military outpost starting a huge war (though Dr. Strangelove shows us how much fun that is) doesn't make that much sense. It would still make sense to have an outpost of entirely military dwarfs, sustained by nearby defended outposts and raids, you'd still need to grow in size and influence I think though. You could just rename the baron's job as "General" and the king as "Warlord" or something like that for the effect you want.

2
I know that armies are on the way, and something like this (managing the whole empire) is already planned, I wanted to talk more about HOW it could be implemented. The log could be read as I "abandoning my fortress" guiding my military, and returning to my fortress later on. I personally believe that it shouldn't be done, I think that player fortress should be something special, and that certain things, as adamantium, can only happen in player fortress (if the player leaves the fortress will always collapse under the fun stuff). So there has to be a way to send invasions, start diplomacy, and create and protect colonies. I propose using a liason noble for that. I didn't want to go too much into detail as I'd have to assume much of what toady is going to release on the next patches.
On the actual dev-log only armies are in, and I myself barely go into the details of how those would work out, my interest is in having new colonies/settlements to be built from your fortress. It might have been as a bloat on the previous dev-log, I'd guess under the empire arc? I saw nothing like this, though, on the forum or recomendation page.
If it's already planned all the better.

3
The army and caravan arcs, in their current incarnation, have brought us a new element: sprawl.

Most sprawl was created as people reached for nearby resources for a central city normally. Hence cities would create farming villages, or mining villages, or such to get resources that were somewhat far-away but not enought to create a new city. Implementing this in world-gen shouldn't be hard (if a city can't sustain itself in any resource with local economy, it first searches if it could create a nearby new settlement). I, instead, want to talk about players being able to do this in Fortress Mode.

The concept is pretty simple. Through the Liason noble you can have an expedition made by apointing an individual dwarf with the leadership position and a request (from a list made depending on nearby resources) for specific resources (be they wood, food, metals, glass, pottery etc. also you should be allowed to be more specific if required (ask for specific kinds of wood, steel, metal or certain crops, etc.). When you do this the dwarf will then make a "request" it will ask for certain tools, certain items and a team of 6 dwarfs that fullfill certain skill levels (depending, again, on the type of colony that you sish to build). You then choose items, much like when you build anything, that fullfill the request. You also choose dwarfs (much like when you create a squad). When all requirements are fullfilled, you allow the dwarfs to leave, at which point they will build a caravan (or just line up in a file), grab all apointed items and leave of the edge of the map. You can give the caravan more resources and dwarfs, which increases the probability of a bountiful year.

After the first year, every year the liason (a noble you have) will go and visit all your protected colonies. (Let's make it one liason for simplicity and gaminess). Depending on if you've given help to the colonies (asked and unasked for), how dangerous the settled area is, how well trading has been working, and just plain luck diferent kinds of years might happen:

Bountiful year: The colony has done excelently, it has grown (which allows it to trade greater amounts of material and offer them at a lower price thanks to supply-demand mechanics) but also created excess riches. It will generally send a tribute back to the home city.

Great year: The colony has done greatly, it has grown. Depending on how much it feels it "owes" the player (sending help or gifts builds relationship, ignoring requests for help or making demands destroys it) it might or might not send tribute.

Stable year: The colony has done well, it remains the same size. Depending on the relationship it might or might not send tribute.

Bad year: The colony has done badly, it remains the same size though. It will not send tribute and will trade less that year.

Terrible year: The colony is suffering, it has shrunk in size (offers less to trade and prices increase through demand-supply). It will not send tribute and will also trade less that year.

Disconection: Due to a siege or simply because the colony is too small (maybe less than 7 dwarfs?) the colony could not be reached, or there was little to reach eitherway. No trading or requests possible. You can, though, send help to the colony so that it can grow again.

Dissapeard: (This can only happen with a year previous warning, allowing the player to send help and save the colony if necesary). The colony is gone, destroyed by invaders, the natural forces or maybe from within (just like player fortresses!) and is no more, the player can send a team of dwarfs to reclaim the lost colony (it works much like sending them to a new one, only they ask for more soldiers) and try again.


Moreover every year events may occur on the colonies:
Invasion: The settlement is attacked by invading forces, be they goblins, elves, etc. As expected the colony may become unreachable until helped.

Artifact: A legendary artifact is created at a colony, this though happens on a much rarer scale that at the player fortress. The colony grows in fame and will surely grow next year!

Attack: Much like invasion, the settlement is under siege, but this time from a mega-beast or such. The colony may or may not become ureachable but will have a terrible year if not helped.

Bandits: The fortress has a bandit problem. Part of it's production is lost and you may loose part of your trading or tribute.


On resources colonies would seek to always sell the most processed item, this is:
Sand < Raw Glass < Glass Blocks / Glass Furniture
Hematite < Iron Bars < Iron stuff < Steel Bars < Steel stuff
etc.
Generally colonies try to build as good an economy as they can and will try to proccess as much of their material internally before trading. This is actually good, since players must have a larger military to defend all their colonies and couldn't sustain all industries as now.

This requires that you help the fortress. The liason's noble screen informs you of all requests of nearby colonies and you can at any point offer to fullfill them or ignore them (after certain time they are considered unfullfilled). You can also send items and dwarfs to fortress, by arranging them, much like an expedition, but instead sending them to an existing colony, this items will boost the settlement's chance of success and the relationship with the fortress.

You can also send squads or such to defend, when this is done they will leave the map for a season (or more depending on the distance) and return, with tales of their success or failure (worst case no one returns). They gain levels and skill as you'd expect them too, but their fight depends on who you choose as squad leader and which dwarves you send. You will need a bigger military, but industry should be much smaller with the caravan arc forcing fortresses to specialize, you'd depend on the colonies for resources.

Also you can request, either squads or resources from cities, they'd give them to you for free if the relationship and status of the settlement allows for it.

OTOH your protecting city (be it a barony, the capital or such, depends on where you settle) may also send help if you are under siege or have a forgotten beast attacking. They'd decide to help depending on the evilness/savagery of nearby lands, relationship with protecting city, and how much help you need. This is mostly to help new players settle nicely, but it'd also help deal with non-player bad luck. Pro fun-loving players might choose to simply ignore the liason from the protecting city and not recieve help, or if you are a noob you can send large amounts of tribute but ask for a lot of help. Also protecting cities might do requests for squads or resources, much like you can, and relationships must be mantained.

When your fortress upgrades (when a baron or count or king comes) your area of influence grows. When you choose a settelement, it chooses the area relatively close to your fortress. When you become a baron you can send them further still. As a barony you deal with your colonies and the towns in your barony, the towns may grow their own sprawl to maintain themselves, and hence may give even more resources. When your fortress becomes the capital, you deal with your colonies, the towns of your barony and also with the cities of your nation which, again, are even bigger than towns.


Now for the military aspects of this:
You should also send forts or military camps to be made. They are smaller, don't make any resources, but host squads and gives them a bonus on nearby attacks. This forts and camps allow for sustain siege of enemy cities (more on this later), trade blockades (or tax charging), scouting and defense, etc. They are a lot like settlements, except they work mostly through tributes and requests, and don't send a trade caravan. They mostly work to mantain military control of certain areas that are away.

Also you can send raids, or even invasions of enemy cities. You can raze a city to the ground, or conquer it, and mantain it as a nearby settlement. The cities you can choose to invade (which may start wars) again depend on how important your settlement is for your nation.

Wow, that's a lot. I wanted to add some screens of how an interface would look, to also explain some of my points better. I might do that later, but for now I'll post all this.

4
DF General Discussion / Re: "Age" change in fortress mode? (40d)
« on: August 27, 2010, 07:42:32 pm »
I've triggered an age change in Fortress mode.

Well, that's where the pop up window appeared, anyway. It was related to unleashing hell in adventure mode, bringing the world back into a Second Age of Myth. My adventurer died down there, though, so I saw the window when I embarked. :3

That just reminds me of Berserk, where the world was a pretty mundane, realistic medieval one, until various events begin with an outpour of demons into the world and later dragons and faries appearing...

5
Dammit that paper is awesome. The thing is, the dwarf fortress maps change a whole lot more than your average map. The algorithm might require too much changing, I'd have to play with it to see...

I though of a system where certain points where marked as "central points". Basically whenever a dwarf (or something) passes through a tile, it starts marking it, when it's been passed a certain amount of times it's considered a "key point". A couple other things that might influence if a tile (or area) is a key point is how many places it comunicates (say if it's an up-down stair) and how close it is to other keypoints (the farther the easier it is).
The craziest case would be building an up-down stair that makes a tile become a key-point and makes another keypoint stop existing.

The most inefficient moment is when a key point appears. Keypoints "precalculate" the optimal path distance of all tiles a certain distance from them, say until three other keypoints are reached. This is done with a flooding algorithm*, the nice thing is that this algorithm is a cell-automata based algorithm, so any change in the map doesn't trigger a full recalculation, but only recalculates the changed parts. It might take some time for the change to occur, but we aren't expecting optimal pathing and all knowing from the dwarves, let's give them some time to "learn" the new best path. A single tile could have the distances to many keypoints (though you should wonder why it isn't a keypoint itself if it's so centric).

Now we can calculate the optimal paths between keypoints. Because we extend the distances we calculate from a keypoint A until we have reached keypoints B, C and D (for example), we can do a map with diferent distances between the keypoints. We can use Dijkstra to find that the best path from A to E (which connects with B, C and F) is through C, for example. Hell, if the number of keypoints is small enought (and I mean something less than 50,000) floyd's algorithm could work, on the expectation that changing the optimal path between keypoints, or changing keypoints themselves is relatively rare (and it should). There could be cases where it actually finds a non-optimal solution by much, but if the number of keypoints that need to be found before stopping the flow algorithm.

When we want to find the optimal path between A (say a dwarf) and B (a barrel of dwarven rum), first A and B would find their closest "keypoint" (lets call them Ka for the dwarf and Kb for the rum). We find the optimal path between Ka and Kb. The dwarf then chooses a Keypoint Ks from all the Keypoints A has a measured distance to the Keypoint along the path that is closest to Kb. The dwarf then uses a greedy algorithm, reading the distance in each tile, to travel to Ks. From there he uses the same greedy algorithm to travel to Kb. Once the dwarf reaches Kb, the optimal path from B to Kb is found (again greedy method), and then the dwarf travels this path "backwards" to get to his precious alcohol. Now the dwarf if finally happy.

The biggest issue is how much the flooding would take in frames. If it happens separately, it might be reasonable. Another solution is to have a "step" in the flow to occur, but this might make a drastic map change take a couple hundred frames. None the less if the game runs at 100 FPS (not graphical, but logical step frames) this might not be that much of a problem.

Again I guess I'd have to do some math, and then experiment a bit (because algorithm efficiency only works when all steps take the same amount of time, which isn't so) and see what happens...

*The flooding algorithm works like this:
//Given that all tiles start with a distance "infinite" except the starting tile (which has 0)
flooding(tile):
    for adj in adjacent_tiles(tile):
        if adj.is_passable and tile.distance < (adj.distance + 1):
            adj.distance = tile.distance + 1
            flooding(adj)

6
DF Suggestions / Re: Build over floors?
« on: July 01, 2010, 11:11:35 pm »
See, I do think that most people that complain about this are people that haven't played that much, on the other hand people who think this is useless have played the game so much that they've becomed blinded.

See, even if an engineer needs to destroy a floor to build a support, you don't go telling him "destroy the floor and then build the support", it'd make sense that the engineer would realize and destroy the floor himself. So maybe when you want to build a wall over a floor, the dwarfs realizes it's a special case and destroy the floor first. It wouldn't be that hard of a problem.

Yet this would lead us to ask why for so many other things. To point just 1 thing out: why can't I carve fortifications on a constructed wall? I could see the usefulness, say that I'm being attack and want to improvise some fortifications from constructed walls without allowing the enemy through by destroying the wall first. Sure you could argue that constructed walls are not as simple to just make a hole through, but that doesn't make sense at all.

Instead we have to understand that this is a really small feature request that is not trivially implemented. It's not just some changes in the raws, it's not just removing an if or adding a new case for construction. This would change how constructions are built and destroyed fundamentally. At this point I'd rather see improvement on either relatively lower hanging fruit (such as being able to select material on workshop orders) that adds nice details or on features that make the game so much more awesome (such as moving fortresses) than to see this fixed.
This game is in alpha stage, and so a lot of features are left to be added before we can focus entirely on polish such as constructions becoming other type of constructions. On the other hand, if at any point the addition of other features make this feature aparently easy or almost trivial to implement, then by all means it should be done.

Pages: [1]