Print

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

[Cobbler89]

So, we all know that sooner or later the new mechanics being added for Adventure Mode will be extended to Dwarf Mode in some way, shape or form, meaning that enemies will theoretically have the ability to scale cliffs and tear down our walls, not to mention jump across our magma-filled trenches. This means that we will need some new methods of separating above-ground areas from the rest of the world, for those who do not simply put their whole fort behind a military-grade booby-trap-lined Entrance Hallway o' Doom and restrict their dwarves to the fortress innards. You know, those of us who like being able to gather tree wood from the surface, or who for some stupid reason decided to embark with animals that need to eat the grasses that only grow up there, or whatever. Plus, if you can make the fortress cool in the process of securing it against climbing, jumping, wall-smashing enemies, you get Plump Helmet Points (Dwarven equivalent of Brownie Points).

So, I've come up ahead of time with a basic design that should do. Implementation will first require a certain amount of !!science!! to measure (theoretical or practical -- depends only on whether you feel confident in the practical measurement holding true for your world) maximum enemy jumping distances both vertically and horizontally, but the basic scheme should be foolproof. (It should also be unnecessary if, say, enemies are only able to jump or climb up one story, or some such practical limit that's easy to work around. But my planned method is cool anyway.)

Here's the gist of it:
Measure the maximum horizontal enemy jumping distance in tiles. Call this "X".
Measure the maximum vertical enemy jumping distance in z-levels. Call this "Y".
Around your fort or across a desired block-off area, dig a channel X tiles wide.
Once one level down, mine one tile toward the protected side.
Now from this space X+1 tiles wide, channel down Y more z-levels.
Pave the bottom so giant trees don't grow!

The result should be a trench just barely too wide to jump across, just barely too tall to jump out of, and with a natural (and presumeably for now therefore indestructible) overhang preventing scaling of the protected side of the trench.

You may, if you wish, carve out fortifications in the topmost of the underground levels facing out into this trench, allowing your crossbowdwarves to pick off enemies who for any reason venture into the trench.

Known complications:

This design does nothing to solve the problem of securing an entryway, and said entryway must go through the Trench o' Defense somehow. A bridge over it would be cool, but afford more risk in terms of having to build constructed gate/guardhouses on top that could be destroyed. It is, on the other hand, possible to leave a land bridge (or build a retractable bridge) at the bottom of the trench's necessary level and dig the actual trench level a bit deeper below that bridge. You could also fill the bottom of the trench below the entry bridge with magma, and then if the bridge is retractable...

You may wish to leave a way to climb back up the far side of the trench. If so, you will need to account for this space if, say, putting the trench as close to the edge of the map as possible. Note however that you only need one extra tile for a ramp that runs in the same direction as the trench rather than perpendicular (leading down toward it). On the other hand, if horizontal and vertical limits on jumping don't compound well for the jumper, it might be reasonable to have the entire trench instead be a slope down to the cliff with the overhang and all. !!Science!! will determine this once we have jumping, climbing, wall-shattering baddies.

Enemies can still shoot arrows over the top. Construct a wall atop the overhang just to block arrows. If X turns out to be less than the distance between the edge and where you can build walls, decide whether to build the walls back from the trench edge (hey, it may be uglier looking, but it gives you the most entrenched underground area), build the trench back from the map edge, or build the trench wider than necessary. (I'm for wider than necessary trenches, personally, but any of the three would work.)

Ground level differences. I don't have an exact solution for this except where the difference is that a slope lowers away from your fortress and so the area where the ground falls away anyway can be evened and built into your trench (basically, the outer side of the trench is naturally missing). Anyone with good ideas for handling notably higher ground or lower ground that would cut into your trench area, please post them.

Rivers. Oh man, rivers. Quite frankly, I also don't know how to handle rivers in the channel-next-to-map-edge-and-remove-the-ramps-to-block-off-the-edge-of-the-map exploit, except possibly to make your legit entrance be a bridge over the river anyway. So if anyone knows what to do about rivers for that older exploit, PM me or point me to a post on the matter, and if anyone has a suggestion for handling rivers in this new anti-climbing-jumping-wall-tearing-enemies trench design, please post it.

And speaking of edge of map issues, if anyone could figure out some way to make this work at the edge of the map (hacky or not) we could potentially use the upcoming fort retirement mechanics plus this method of walling to build a Great Wall across a continent a la China, defending an entire dwarven civilization!

I may be back to post more complications that occur to me.

If anyone wants to draw ASCII pictures showing the actual DF implementation of this design, feel free; I'd rather make it in DF and take screenshots, myself.

[AutomataKittay]

Wouldn't scattering supports everywhere with plates on top of it to fall down on what's around it do the trick, and obsidian-cast bunker on topside?

[Splint]

Pit with spikes right below the walls, weapontraps accross the top. They dodge the weapontrap, and they have two choises: Into another armed trap or off the wall into the spikes/mid-air and thus breaking shit. Marksdwarves would finally have a permanent place amongst my armies as well.

[Cobbler89]

Wouldn't scattering supports everywhere with plates on top of it to fall down on what's around it do the trick, and obsidian-cast bunker on topside?

Support-with-clobbering-ceiling has been used effectively against FBs in the past and I imagine this may eventually extend to anything with an eye toward destroying supports (among other constructed things), though that depends if they do so by siege engine from a distance; I'm personally also hoping for something that would never have to be set back up -- passive barrier sort of defense -- hence my trench idea.

Obsidian casting is a good solution to the problem of constructed walls, if you can get the magma and water in place; thanks for reminding me.

Pit with spikes right below the walls, weapontraps accross the top. They dodge the weapontrap, and they have two choises: Into another armed trap or off the wall into the spikes/mid-air and thus breaking shit. Marksdwarves would finally have a permanent place amongst my armies as well.

You, sir, are a genius.

Keep in mind that constructed walls may at some point in the future be vulnerable to... whatever goblins will be using to destroy walls. But still, the essence here -- weapon traps at the top (maybe even weapon traps a few tiles thick in case they can dodge over to the "inside" rather than sideways or back into midair) plus upright spikes all over the floor... easy to do and sure to clobber a wide variety of enemies. As long as they don't get some way to dismantle the weapon traps before they come into the range of said traps, you get goblinite in your stocking every Dwarfmas and don't even have to do anything to earn it beyond setting up the traps in the first place.

[Splint]

The marksdwarves come into play regarding anything that has to get close to do damage, and possibly including a sealed Sallyport (A battering ram? Coming at my wall? AHAHAHA I don't think so greenskins! *unleashes the dwarven horde*)

Anyway lets have a situation of a wall set up that is 2 z's high, the second being the fortifications. Right in front of said fortifications are the weapon traps. The marksdwarves fire on the advancing enemy until they attempt to scale the wall. here, retracting spike they wouldn't have seen is useful as well, as a lever pull can greivously injur the would-be scalers. Those who manage to begin assent then have to face a fresh volly of bolts AND the weapontraps, and if they are unable to jump a sufficent distance, due to injury or dodging just screwing them, they end up greivously harmed no matter what, and possibly killed.

Scattered stonefall, cage, and weapon traps as well as some linging the roadsides save the wagon access can also assit in thinning the enemy onslaught.

[Cobbler89]

I just thought to wonder whether there will be any traps that work on the sides of walls... maybe that's for the suggestions forum, or maybe if it isn't too suggestiony and really just a legit general future gameplay question it belongs in Future of the Fortress?

[Splint]

What, and deny the enemy the chance to fling themselves headfist into spike and/or solid ground?

[Guedez]

build a bridge over the magma trench

have a noble Pull Lever [R]

then let the pathfind do it's job

[thiosk]

I was thinking of this.  Since a jump requires a running start, one could have a 1 tile chasm with walls every other space, staggered on two sides.

Code: [Select]

o,o,o,o,o
----------
,o,o,o,o,
[code]
Where o = wall, , =empty , -=gap

And thus you have what should amount to an uncross able gap.  Heights and distances need to de determined.

[Mitchewawa]

Only open entrance accessible via a huge scaled wall. At the top there's a raise-bridge to fling them back off.

Hallelujah, it's-a-rainin' gobs!

[Cobbler89]

I was thinking of this.  Since a jump requires a running start, one could have a 1 tile chasm with walls every other space, staggered on two sides.

Code: [Select]

o,o,o,o,o
----------
,o,o,o,o,
[code]
Where o = wall, , =empty , -=gap

And thus you have what should amount to an uncross able gap.  Heights and distances need to de determined.

Hmm, I forgot about running starts! Find a way to make it with natural or obsidian-cast pillars and it should do nicely; then mine into the protected side so they can't just climb up because "up" is the ceiling above their heads, and it should be a much more compact invincible wall! Nice! Will try it first thing whenever we get enemies with jumping and climbing abilities, if I can figure out how to make it natural and all.

[Xob Ludosmbax]

Obsidian casting is a good solution to the problem of constructed walls, if you can get the magma and water in place; thanks for reminding me.

With minecarts, you don't need to "get the magma and water in place".  You can fill a minecart with magma, have your dwarves carry it into position and dump it.  Same goes for water.  It's doesn't scale up very well, but huge pump stacks or magma pistons don't scale down very well.  I wrote a post describing it in detail at some point.

[darkrider2]

Honestly no defense is perfect unless it is a pump stack producing a curtain of magma around your entire fort. Then have a retractable bridge that puts itself above the front door, so the curtain opens to let people in.

Plus magma mist and a charred barren landscape.

[Xob Ludosmbax]

Actually, magma curtain may not be a perfect defense against magma-proof fliers, such as fire imps.  What you're looking for is a falling obsidian curtain.  (Magma curtain and water curtains being mixed together and falling off of a bridge so it doesn't "stick" when it mixes.  Then you need to drop it into SMR at the bottom.)

[Cobbler89]

Obsidian casting is a good solution to the problem of constructed walls, if you can get the magma and water in place; thanks for reminding me.

With minecarts, you don't need to "get the magma and water in place".  You can fill a minecart with magma, have your dwarves carry it into position and dump it.  Same goes for water.  It's doesn't scale up very well, but huge pump stacks or magma pistons don't scale down very well.  I wrote a post describing it in detail at some point.

D'oh. I clearly missed the full scale of the obsidian-casting implications of the minecart system! Should'a thought of that... massively expands the feasibility of obsidian casting.

Honestly no defense is perfect...

Well, no, naturally. Even the method of turning your fortress into a giant bottle submersed in water or magma, designed by some genius in preparation for the day when enemies tunnel, isn't going to work real well in terms of reuse (I mean, anyone want to go out and patch the wall so the space can be refilled?) unless somebody can find some way to ensure that the water keeps flowing into the top and out the bottom automatically. Or maybe that was already in that method, it's been a while since I read it. And in any case there's always the door you leave for the handful of migrants and traders you actually want, which you can make pretty darn secure -- but never perfectly secure. That's why losing is fun.

But with all that said, with new means in the pipeline for enemies to overcome some of the more basic defenses, it's worth some thought to find good, relatively convenient, relatively passive ways to thwart enemies who can climb or destroy walls and jump gaps. Sort of the jumping-climbing-destructing-enemy-proof version of your basic wall-off-the-area trick (or, for those of you who don't expend stone, equivalent of the basic channel-with-ramps-removed trick that achieve the same thing of blocking out non-flying enemies).


...
So, I was thinking about ways you could try to get around the pillar method (climbing the pillars and jumping from atop them? climbing around the other side of the pillars and jumping away from them if that works and can get across a tile?) and realized that when the time comes the best way to test, beyond arena mode !!science!! for raw capability factors, will probably be to create an example barrier, retire a fort and come back as an adventurer and try to get past the example barrier.

It'll be... interesting, in any case; for now, there are still ideas to be discussed that should help work out some of the kinks ahead of time and, of course, turn up alternate solutions to the basic problem. I'll be around, in any case, though probably in evening-long bursts of frequent reply with several days of not having time for this in between.

(Kinda-Off-Topic: Think we'll have the next version by Christmas at this rate? If so will we call that version Goblinite or Goblin Clause or some such reference?)