Print

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

[atomfullerene]

The classical method of avoiding turning to stone is to use a mirror-or highly polished shield.  In greek myth, it was looking directly at Medusa that turned you to stone, not being looked at by Medusa.  Apparently reflections don't count.  Not sure how you could sim this in DF, though it would be pretty cool if you could.

[Vester]

Memetic mutation turned that into gaze attacks, though. I think a big part of it was that Greeks believed that the eyes weren't so much receptors as they were projectors, hence expressions like "being affixed by the beam of his eye". Eyes projected a sort of invisible beam that allowed you to see what they made contact with. [citation needed  ]

In Medusa's case, she was really, really, really ugly. Apparently that sort of thing turns you into stone.

Didn't the basilisk have an effect that was closer to the concept of the gaze attack, though?

[Martin]

On the ark issue, as a programmer I'd probably approach it a bit differently.

It's impossible to figure out what mods people will install. You can't work out all interactions if you can't know what the interactions might be, and in a game so focused on modding, it'd be silly to build a whole solution around the ark that doesn't work for a mythical ark, or an extinct ark, or the combination of two of them. Personally, I'd add tags to the species that describe their overall characteristics that have an ecological impact. We have size and damage, litter size, but no other reproductive rate (gestation, that kind of thing). I'd add herb-, carn-, omnivore, can_scavenge, and some things like that. Ignore for now the broader ecological predator/prey population modeling for something simpler.

Prior to world gen, I'd dump all these species the user has selected out onto the world and do a quick ecology model mostly independent of the civs placed down. It doesn't need to be global, maybe just one per biome combination, and only for the purpose of getting entity balance down. Something like a virtual arena.

In each setting, iterate over each carnivore and pit it against each other creature in the biome (including a dwarf peasant). If it's a pack animal, pit them as a pack. Work out a simple success ratio over a number of passes, normalize it, and use that to determine hostility toward those other creatures during world gen and the game. Let's call that the hunting table. Do the same for omnivores but add in the variable for them to go vegetation only during the game. On a mountain, they'll be carnivores, on a savanna they'll probably be herbivores. During the game have each wild animal seek food each season using the probability model and what they can actually find, during world gen, use the probability model to work out appropriate populations for each species rather than trying to manage it at the individual level.

The advantage to this is that once the tags are in there, the animal interactions can be established largely outside of gameplay just by having a reasonable model, which means that the community could help build and test the model for Toady. He'd have to do the in-game animal behavior, but the rest could be done as the OpenGL work was done. Plus, if you have an external simulator such as this, modders have a means to test the interactions. If big cats aren't willing to hunt a large hoofed animal, perhaps the hoofed animal is too strong, or the cat too weak. The simulator would help create better animal mods. I think this is what Vicomt is nudging toward.

[Lancensis]

It was never specifically explained why Cthulhu wasn't unleashed on the world that day, but my bet is that it had more to do with disrupting whatever the cultists had planned than the damage one little steam yacht could do.

The usual explanation is that Cthulhu will rise "when the stars are right."  When the stars come into an alignment that's sort of but not perfectly "right," Cthulhu becomes restless, R'lyeh rises a little bit, and humans the world over--especially artsy and imaginative types--have nightmares, go insane, and worship Cthulhu as a god.

Human cultists in the Lovecraftian cosmology have little or no effect on Cthulhu's awakening . . . they're byproducts rather than causal agents.

Wasn't it an earthquake that precipitated the events? An upheaval raised R'lyeh temporarily and roused Cthulhu slightly. I guess the cultists thought that this would be enough to raise their god, and travelled there to try and wake it fully. My guess is that Cthulhu wasn't properly awake, though. Like when you wake up in the middle of the night and have to go to the bathroom.

[Chthonic]

Prior to world gen, I'd dump all these species the user has selected out onto the world and do a quick ecology model . . .

I don't think it's necessary to do any explicit modeling.  Knowing pack size, group size, diet (carnivore or herbivore), and biome should be more than enough to produce a rule-of-thumb trophic level for each creature that usually works out well enough to populate a region according to only a few very simple rules.

[Chthonic]

Wasn't it an earthquake that precipitated the events? An upheaval raised R'lyeh temporarily and roused Cthulhu slightly. I guess the cultists thought that this would be enough to raise their god, and travelled there to try and wake it fully. My guess is that Cthulhu wasn't properly awake, though. Like when you wake up in the middle of the night and have to go to the bathroom.

I think you're right . . . I don't remember whether the earth movements were supposed to be the result of a celestial near-alignment or not, though.  R'lyeh seems to pop up like a tempermental jack-in-the-box in later works that feature the big C :/  It's more like, when isn't the city on the surface?

[Martin]

I don't think it's necessary to do any explicit modeling.  Knowing pack size, group size, diet (carnivore or herbivore), and biome should be more than enough to produce a rule-of-thumb trophic level for each creature that usually works out well enough to populate a region according to only a few very simple rules.

Maybe, though it'd be a shame for dwarves to get killed by hoary marmots more often than lions in combat. And that is a problem with the game mechanics now - the creature damage/defenses are quite a bit out of whack and you just know that someone is going to want to have killer hoary marmots taking over the world. But I think it'd also be interesting to see better behavior of creatures during gameplay. If you have lions and leopards on your map and a bunch of ungulates, it'd be cool to see the lions and leopards avoid each other in favor of the ungulates (or dwarves) until they couldn't find food and then turn on each other, or see the lions be willing to go after the cheetah if they are sufficiently overpowering, but the cheetah be very deferential to the lion. I just think that would kick ass.

[Neonivek]

Well, while you don't want a Dwarf to lose to such a small animal (unless something horrible happened like it hit an Artery or the wound became infected) you do want a whole herd of small animals to be capable of taking down even a prepared dwarf.

[Chthonic]

I thought Toady was specifically working to mitigate the marmot threat to armored dwarves for a while?

[Martin]

Well, while you don't want a Dwarf to lose to such a small animal (unless something horrible happened like it hit an Artery or the wound became infected) you do want a whole herd of small animals to be capable of taking down even a prepared dwarf.

Oh, sure, and I'd make pack size part of the tags. Though on Chthonic's point, it might be easier to just multiply the animal stats by the pack size and see how that compares to one dwarf rather than run an actual model. The 'prepared' part is up to us in-game to deal with.

[Neonivek]

I thought Toady was specifically working to mitigate the marmot threat to armored dwarves for a while?

He did, thanks to the new materials the game no longer says "ok it got lucky and did actual damage" because 0 is no damage and 1 is damage

The game can now say "Alright the creature does cosmetic damage"

[Neruz]

I thought Toady was specifically working to mitigate the marmot threat to armored dwarves for a while?

He did, thanks to the new materials the game no longer says "ok it got lucky and did actual damage" because 0 is no damage and 1 is damage

The game can now say "Alright the creature does cosmetic damage"

So long as the groundhogs aren't biting people with their entire face.

[SirPenguin]

Thankfully Toady's new system is the best of both worlds. Small, weak creatures can only muster small, weak attacks...but said attacks can build up and accumulate, which will eventually kill whatever creature it is. It really paves the way for things like swarms of blood sucking swamp flies or whatever. They're very weak individually, but if they swarm you then you're in trouble.

[Martin]

Not to get further bogged in details, my point was to steer away from having the creature interactions be static, and instead be dynamic. Further, I think those interactions can be modeled outside of the game, and would benefit modders if they could (quickly) simulate the effect of dropping dragons in as pack animals.

I have no problem if the model doesn't scale well. Toady need not add the whole ark, just the subset he feels is best, and if we want to add more, we just have to be willing to pay the price at worldgen (or even pre-worldgen if we don't touch the mods, so worlds can run over and over). But that can be our choice.

And if a mechanism can be carved out that takes some of the burden off of Toady (particularly if it's a burden he wants taken off) then better yet.

[Chthonic]

Not to get further bogged in details, my point was to steer away from having the creature interactions be static, and instead be dynamic.

Dynamic how?