Print

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

[Merendel]

Ahh so thats how you were doing it.  I went with a slightly different approach trying to minimize the footprint up higher but you expanded out and drained the everloving crap out of the place just using the up/downstairs as a floor which I didnt think of.  That would be faster with less pushing.  Nice work.

As a refinement you may just consider diging downstairs on the inner ring of stairs for A1 instead of up/down.  This would acomplish prety much the same thing but not need the removing of the upstairs later.

[Snaake]

Decided to do another test, using Merendel's method but with a larger cave-in. Embarked on a similar map, but only a pick, a proficient (5) miner, and 6 peasants this time.

The cart was caved-in down, and last ramp to surface (for the cave-in triggerer to get down, wasn't necesary this time, as the cave-in knocked the peasant selected down into the gopher hole) was removed the next day. Stupidly, the peasant that triggered the cave-in ran up to clean herself in a pond meanwhile, but even with waiting for her (one of only 2 females) to return, and some messing around with the "oh, I'm going to stand on the spot where I'm trying to construct a wall" bug, the dwarves and the yaks were walled in underground on the 8th of Granite. So 5th-6th is definitely doable.

Then I started digging a 11x11 area and a 13x13 square of upstairs on Z+0, so 2 steps larger than Merendel's original (reference is this post). I dug out Z+0 and Z+1 mostly identically to Merendel, but decided not to dig out the central up/down staircase on Z+1 (and if you don't dig that, you don't really need the central staircase on Z+0 to be up/down either, just down would be fine, but I kept it since I'd already dug it out)., and to already remove every second up staircase on Z+1 (for a bit less surface exposure later)Don't do this, dwarves can't channel up diagonally like I thought they could.. The miner and a peasant decided to take a drink of aquifer water on 22nd Granite. Z+0 and Z+1 were dug out on the 25th of Granite, the miner (very strong, agile and tough, this time) had reached 9 skill by now:

Spoiler (click to show/hide)

You should have a ring of up/down stairs, not alternating up/down and down, on Z+1.

I channel out 3 rings, one at a time, outside->inside, of the splash area already at this point, for useful training, just leaving a 5x5 in the center to do the chicken run + access to the 5x5. Also removed all the upstairs except the one just in front of the access tunnelDon't do this YET, either, or you'll have a more finicky time channeling out the stairs on Z+1 later. It's the 3rd of Slate, the miner skilled up to 10 from that: no training, just the initial 5x5 gopherhole, 10-tile tunnel from gopher hole to 11x11 meeting area, ~12-tile tunnels on 2 floors to the work areas above.

Spoiler (click to show/hide)

You should have a ring of up-stairs, not floor, on Z+0.

Did a succesful test chicken run elsewhere. Was worried about peasants or the miner wandering in for water just when I want to do the cave-in, so dug them a little water channel in the meeting area. Did the chicken run down in the centre of the work area. Saved for potential save-scumming, just in case I was wrong about the central staircase tile on Z+1. Which was useful, as I immediately failed at single-stepping for the chicken run. After successfully single-stepping, the chicken run was a success. Then channeled around the chicken run on Z-1 exactly like Merendel, and channeled out the remainder of the splash-down area on Z+0. Z+0 to Z-2 are now ready for the cave-in:

Spoiler (click to show/hide)

You should have a ring of up-stairs, not floor, on Z+0.

Next I just channeled away the ring of floor on the surface from Z+1, then channeled all but the trigger-stair away from Z+1. I hadn't checked earlier for trees and had to dfhack a saguaro out of existence (since I had no axe... and actually changed it to a dead sapling, because I figured out how to do that before I figured out how to magic it away). Did have a bit of trouble (the red text above), but got it done in the end. With all those issues, it was the 13th of Slate before I was ready to trigger the cave-in. As it turns out (did one run, savescummed, repeated), I'd recommend removing the central up stair on Z+0, or it will support the central cave-in tile. You can of course just mine it out later. No screenshots of this, should be pretty obvious. After the cave-in:

Spoiler (click to show/hide)

Note that there should, again, be a ring of up stairs on Z+0. On Z-1, that's as much as I can dig out, if I dug out a bit more, I'd open up to the aquifer again.

This was 2 steps larger (maintining a single-tile shaft) than Merendel's, he had a 5x5 center area for the water to spread out to, this is 9x9. This level would end up with only 1/7 water (9*7 units of water spread over 81 tiles), if I had waited for the water to evaporate fully on Z+0 after the splash, instead of just channeling the spread-out area immediately. I still had to bucket out the central staircase tile at Z-2 before I could get to the tetrahedrite beneath. And had to channel out the stairs on Z-1, so 1 wood gets temporarily tied down to that too.

So yea, the no-wood method is better, since this still required 1 bucket + 1 wood for stairs, and at least for me, took longer. With good design (meeting area close to bucketing area) and the minimal footprint Merendel had, it's probably competitive with the no-wood pierce in terms of time, but exposes you to the surface while you do the final prep for the cave-in, trigger it, and empty the water with buckets.

[Merendel]

Just out of curiosity why do you cave in your wagon?  in the time it takes you to dig a cubby, dig a cavein and get everybody underground you could have just goten everything inside without the risk and unhappy thought of a cavein.   What I do is channle down 1.  next layer channle down on the next tile over (so the ramp stays accessable).  Layer below that dig over 1 then dig an upstair.  Above the upstair dig a downstair and then pick a direction and dig a strait hallway.

While this is happening I get someone deconstructing the wagon, started at the same time as the dig.  Designate a 1 tile meeting zone on one of the underground tiles your miner has opened up just in case your carpenter works faster than expected.  Once your miner has cleared 3 tiles past the stairs designate another meeting zone/pasture on those 3 tiles and designate a wood stockpile, your wagon should deconstruct right around now.  3 dwarves will haul in the wood 2 more will grab the animals. remove the first temporary meeting zone.  If someone runs back outside anyway or does not move inside quick use a civ alert burrow.  As soon as your miner has cleared a bit more room move the meeting/pasture zone over a bit so people are not standing right by the stairs and have the miner channle out the downstair(make sure nobody's outside or on the level below first)  your now safe and its probably the second or third of granite.  Only thing that can get in at this point is a flyer that will path to dwarves without LOS.  with this method you only need a minimum of 8 dig operations to get everybody and everything underground and only a few more that get done while people are moveing to seal the entrance.

update
Just tried your version of the undercut breach and it worked much better than what I was attempting.  2x2 stair through the aquifer by 15th of felsite and I honestly overleveled my miner.  I basicly had him mass diging till he got hungry and took his post meal nap mid slate.  he was up to 11 or so skill by then.  After that and after he washed that meal down with some water I had him do the chicken run.   One thing I've noted after doing that chicken run this many times. If at all possible start your miner in a shaft.   If the only place he has to stand when diging the first stair into the aquifer is on the stair itself he only has to move 1 step down to start diging into layer 2.  If you clear the level above the aquifer first he will tend to stand on the tile next to the stairs when diging into the first layer and has to move 2 steps to start diging into layer 2.  That extra step makes it much more likely that he'll fail the chicken run requireing higher mining skill to compensate.  I suspect if you designate your dwarf with the best avrage of STR/Agi as your miner you could get away with 7 or 8 skill in mining for a chickenrun if you start him in a shaft.

The other thing I like to do is when I'm carving out the upper levels I do it in such a way that the cavein triggering dwarf is standing in a hallway when he sets off the cavein. If you have stairs on either side or under him when he triggers it theres a chance he'll get knocked out and fall down one of those staircases and into the water with your initial design.  considering he'll be unconsious he'll probably drown with your pick.  If he's standing on a floor tile with walls on either side if he gets blasted by the dust he'll most likely either go strait back or slam into the wall.  Even if the wall impact kills the dwarf to bad luck (had this happen on a cavein while testing my initial method) at least the pick is in a recoverable spot.

[Snaake]

...
As a refinement you may just consider diging downstairs on the inner ring of stairs for A1 instead of up/down.  This would acomplish prety much the same thing but not need the removing of the upstairs later.

To this, and as a partial answer to why did I cave in the wagon, I'm going to say that you need to train the miner up anyway, and I'd prefer not to dig out large empty areas just for the sake of training (I concentrate a bit too much on not digging out tiles I'm not going to use/being neat about digging in general). So it's not really wasted time/effort as such.

The main reason about the wagon cavein was that the first time, I also had a bunch of booze and food, and I remembered the wagon deconstruction taking longer than it probably does (since I don't usually deconstruct until I've got basic workshops going on the surface, on non-evil embarks). On the second time, I still think I got 6/7 dwarves underground faster than I would've without a cavein, and actually if I would've triggered the cavein from below rather than from the surface, walking access to the surface could be cut off as soon as the dwarves notice the new meeting area/obey a civilian alert burrow to get underground, really. Without testing, I'd say you only need to dig enough tiles underground to actually get your dwarves to move there (plus remove the ramp/down stair that gave you access from the surface), then move them there, to be safe from non-fliers. And then you can cavein the wagon in peace. All in all though, I'd say it really doesn't matter all that much either way: getting through the aquifer doesn't take so long that you wouldn't have time to do the cavein

The real reason is that I kinda wanted to test caving the wagon down, since I hadn't really ever done it. Or done proper forts on evil biomes, really, so I am a bit of a noob when it comes to the actual challenge . The risk seems pretty minor, even with the 5 caveins I did over all those tests, the worst was the first wagon one: the miner got an arm and his upper body bruised. Which is why I picked someone who was quick to heal.

tl;dr: I'm a noob on evil biomes, and thought it was sort of standard procedure (which it probably is, on evil-biome embarks that are not single-pick challenges).

P.S. I would give the challenge a go, after doing those aquifer test runs, but I decided, before I started it, to commit to my current fort until it dies/the new version comes/I've got adamantium/steel military + fully self-sufficient on everything, the caverns are conquered etc. at which point I may unleash the HFS, or first build a minecart water autocannon to kill them. Because that's my biggest failing as a DF player; I switch forts too easily, before they're really mature or meet their end naturally.

[Merendel]

Admitedly on a standard evil embark caveing in the wagon is standard pratice simply to make retrieval of equipment somewhat safer.  With only 3 pieces of wood and 2 animals to move you can just as easily deconstruct it grab the bits and run.  The only time I've resorted to a cavein to retrieve the wagon on a single pick challange was the one time I embarked with 2 corpses already spawned and only about 20-25 tiles away from the wagon.  Had to set a meetingzone/pasture on the opisit side of the wagon and issue an alert as soon as the miner grabed the pick off the wagon just to be safe. dug under from there and went back for the wagon a few days later.

I ran a speed test of my method earlyer and I had everybody inside and channled out the downstair locking the entrance on the 2nd.

Also the area's I'm clearing out during the skillup proccess are not wasted space.  You know what those expansive rooms in the dirt layer are? safe treefarms once I crack open the caverns.  While it may take a year for any saplings to grow to be harvested you'll start geting herbs withen a week.  A couple of 21x21 rooms can keep an herbalist realitively busy and get initial booze production going.  They also make a good place to let your dwarves idle while they have nothing to do.  Considering theres nothing for anybody but the miner to do till you breach asside from butchering the animals they spend a great deal of time with their thumbs up their arses.

Update:
Well my current fort is up and runing.  Did stoch's method of the aquifer pierce.  First migrant wave arrived... right under a patroll of undead falcon men.  I thought great 7(1 was a child) more corpses on the surface but opend the door anyway.  Surprisingly 6 of them made it in.   Most of the corpses took off chaseing a lamb pet of one of the migrants will one chased after an animal trapper.  After they were in I had the trapper go military and see if he'd win the fight agianst his single opponant.  A phiric victory at best, he killed it but lost a leg in the proccess and I wasnt about to risk sending someone out to retrieve him so locked him out (with a meeting zone in the outer entrance just in case he could get up) and went back to work.  He died when the corpse reanimated.  Anyway dug down and cracked open the caverns got some plants so the booze is starting to flow and I have enough wood for beds for at least the short term. I also layed out my fortress workshop, meeting/dining area, and living quarters and am leting him dig it out a wing at a time.

Second wave arrived while I was seting up my trade depot area.  As there was only the corpse of my valiant but short lived soldier up there along with a couple land critters on the far side of the map they all made it in bringing my workforce up to 20.  Ironicly a dwarven child was almost imidatly possesed after this and made a tower cap amulet... little food waster is now a little wood waster as well lovely.

Caravan arrived and I had my depot area up just in time so even the wagons got in.  Not alot of danger on the surface yet. Couldnt afford much, used the last of my prepared meals to buy out all their food and an anvil.  After draging what loose furnature and crafts I'd managed to make so far I also managed to buy 2 cheep picks and the few logs they brought as well.  Ordered a bunch of stuff for next year if they can make it in by then.  Oddly enough the dwarven civ has no access to hemitite or magnitite (only limonite) for iron ore.  Oh well.  Think the humans and elves were wiped out on this world or they are just too far away as they did not appear at embark so I dont expect another caravan till next fall.  All but a pair of guards made it off the map safely.

Mid granite I got a giant blob of salt as an uninvited guest.  Somewhat irritating.  I let it go play with my trade depot then sealed him in.  I'll deal with him later.  Time to dig out another depot room.   Oh good lord on the 25th of slate 29 migrants showed up.  I'm up to 46 dwarves and a handfull of kids if I can get them all in.  Not sure how I'm going to put them to use but I did manage to make a safe fishing area down in the caverns and there's a couple of fishers in there so feeding them shouldent be a problem for a while.

All but a siege engineer and some animals made it in before I had to close the hatch.  Left a meeting zone in a realitively safe area to hopefuly let the animals gather where I can collect them if they can make it in off the surface but I'm not holding out alot of hope.  the eggs from the 2 ducklings would have been a nice addition and possibly that alpaca for wool but no big loss either way.

[Stochasty]

I have an idea for a modification of my method to pierce a 1-layer aquifer by draining the aquifer into the rock layer below, but I need to do some science to confirm whether or not my assumptions about aquifer water production are correct.

Some preliminary dwarven work indicates that each aquifer tile produces 1 unit of water in its neighboring tiles about ever 60 ticks or so (during the initial chickenrun, you have on average about 60 ticks to get the second stairway carved before the water hits 4/7; since there are four aquifer tiles contribution, that gives 1 unit per tile per 60 ticks).  Assuming this is correct, after the initial chicekrun you now have ~105 more ticks until the initial stairway on A1 is again full.  Since we've already estabilshed that our miner is faster than that based on the initial chickenrun, so long as we can keep digging drainage we can stay ahead of the rising water level.  Furthermore, that first penetration is by far the most painful, since there are four aquifer tiles producing water and only a single tile for drainage (this is the highest possible aquifer tile per drainge tile ratio), so as we enlarge the drainage area by channeling, we also slow the rate at which the water rises.

Thus, the method: initial chickenrun, as before (note that this will be more difficult than in the two level case, since the layer below the aquifer is rock; you'll likely need a substantially higher level miner).  Then, immediately channel out five of the aquifer tiles in a semiclrcle around your staircase for added drainage.  At this point, your aquifer to drain ratio should be down to 5/3 (ten aquifer tiles, six drainage tiles).  This step is relatively fast, since your miner doesn't need to move.  At this point, each of the drain tiles should have about 2/7 of water in them (this is a guess, based on mining rates and my assumption on water production rates), giving ~327 ticks until our miner is flooded and must stop working.

Next step: mine out the three remaining tiles around the staircase.  These three tiles will be on edge of the 3x3 plug we need.  This step is a net loss, since it doesn't increase our drainage and, in the process, increases the A/D ratio to 3/2 (nine contributing aquifer tiles, still only six drainage); you get a little extra in the form of the 3/7 you're allowed within the cave until work stops, so if we assume ~30 ticks per tile, we've still gained considerable time: approximately 597 ticks before we're flooded.

Next step: we need to be careful about water flow knocking our miner into the drainage pit, so even though we have time we're probably not home free yet.  The solution is probably more drainage: so, pick a direction (clockwise or counterclockwise) and start working your way around, channelling away the ring outside the plug and mining away undeneath the plug itself.  Each step should improve the A/D ratio and add more drainage, giving plenty of time to finish mining underneath the plug.

There are three main concerns and one big caveat here.  The first is my assumption regarding flow rate coming from the aquifer.  I'm guessing that the process is that, after a somewhat random number of ticks (average about 60) each aquifer tile produces one unit of water in each exposed neighboring tile; this is the basis for my time calculations. 

The second is the problem of water flow across the floor underneath the plug; I don't really know what the danger level for flow displacing the miner would be.  1/7 water won't flow into neighboring tiles, but that just means that the tiles underneath the plug will keep accumulating until the hit 2/7, at which point one unit will flow towards a drain (is flow direction random?).  Is that enough to displace the miner?

The third concern is the race against time nature of this.  Make no mistake: what we are doing is constructing an elaborate drowning chamber for our miner, and hoping he can complete it before it kills him.  This means we'll need to single-step the entire process, and hope that our miner doesn't suddenly decide to take a go on break, eat, drink, or take a nap in the middle of the dig.  The first two of those probably just mean a failed attempt; the miner should escape the pit before it fills completely.  The last two are potentially deadly, if the miner decides to drink from his newly dug drainage trench or sleep where he stands.

The caveat is that this is probably a stupid way to go about things anyway.  If the miner is fast enough to pull off this method, he should be fast enough to pull of a straight forward chickenrun.  (Just have him keep channelling out the area around his stairway, and every so often make a short jog up a couple of z-levels and then over into a new shaft for additional drain capacity.)  Thus, establishing a cavern drain is probably easier and more effective.

[WanderingKid]

Stochasty, I think you're using an inaccurate presumption, but this is based on anecdotal evidence.

If you've ever done the double slit, you know that the last 'drain' aquifer tile in your walled area produces just as much water as before you walled up the area.  I believe it's computed not by number of aquifer tiles exposed, but by # of empty tiles exposed to aquifer tiles.  This would mean a square with four surrounding aquifer tiles would produce water at the exact same rate as a square with 3 of those cardinals walled off.  A lot more science in the rate of fill for aquifers would be necessary here for me to confirm that one way or the other.

The other concern I'd have with your method is that all the water above the miner goes down.  As you open up two layers to the aquifer you've got all that water pouring down into the drain.  I can't see how a miner could stay ahead of it, but I may simply be misunderstanding your intent.

[Stochasty]

Hmmm.  Never paid enough attention to the water flow rate when doing the double slit.  Based on my testing for the double-layer pierce, I had thought it was based on the number of aquifer tiles that were exposed, rather than the number of empty tiles exposed to aquifer, but I could easily be wrong.  I'll need to do some testing, then.

This possibility makes things harder, giving you much less time, but it should still be possible to stay ahead of the water; things should never be harder timing wise than the first chicken run, but I'll need to rerun my calculations to be sure.

As for all the water going down: yes, that's the entire point.  After that first chicken run, you'll have 3/7 water in the tile below the aquifer.  This gives you another 7/7 to work with before you have to stop mining.  As you channel out the semicircle, this gives your drain more capacity.  As long you can channel new drainage faster than the water fills, you'll be fine.  Granted, the A/D ratio is much, much worse during this process if you are correct about the water production rate, so you have a lot less time than I was thinking you would have.  Disappointing, but nice to know.

[Merendel]

Well for a 1 layer aquifer I'd say do the chickenrun sideways.  You only need to drop a 3x3 to get through the layer.  On one side dig down then dig sidways.   We already know our miner can go down through 2 layers.  Instead of going down 1 then down agian go down 1 then sideways.   If even once the miner can manage to dig out 3 squares before the water gets too high you could make it work.  Basicly you need to manage to take out the center block of the 3x3 and you should have up to 4 chances (if you go corners first, sides you probably only get 2-3 chances).  Most likely you'll have to go down from all 4 sides and all 4 corners to knock out the 3x3.  Since dirt is much faster to mine its more likely that he can keep ahead of the flooding that way than through stone.


One thing you'll probably want to start the work from at least 1 layer above the aquifer so the dwarf is standing in an up/downstair each time he cracks the aquifer so he needs to take 1 fewer steps.  If at all possible try to hit the center tile from the side not the corner.  On further reflection while the corners give you more opertunitys to hit the center tile comeing in from the corners first may make further digs harder as you'll get diagonal leakage into your stairway as well as the aquifer fill rate that may prevent you from knocking down a side wall in time

[WanderingKid]

Just a note: Furnaceclans is not dead, I'm just braindead.  Work has thoroughly been kicking my arse the last few weeks and hasn't left me time to sit down and record/edit/play/rinse/repeat/etc.  Future updates are forthcoming, just... eventually.  You know how life goes.

That and it's giving me a chance to let go of a bit of burnout, so it's probably a good thing.

[WanderingKid]

FurnaceClans 12 is posted.  Sorry it took so long.  Life, Health, Work, and PC issues compounded themselves in series to make DF be a low-concern item in my world for a while.

In this installment I deal with parts of exploded corpses that have gotten into the magma loader system, a few random odds and ends, and a migrant wave gets blended.

Link: https://www.youtube.com/watch?v=Flq5xkwgpms

The path to becoming the Mountainhome continues!

[WanderingKid]

FurnaceClans 13 is up!

Link: https://www.youtube.com/watch?v=03uCQBdIUx0

Sarvesh has died.  Migrants die in droves.  Airborne critters avoid the OMG and get into the fortress with minimal casualties.  Our first child is now an adult!  Fishybang joins the slave labor useful members of society!

[WoobMonkey]

I admit to a certain level of giddiness when

DFWanderingKid 1

showed up in my subs list.

Spoiler: Watch the video first! (click to show/hide)

 What inspired the multiple bridge/drop pit design?  It's an interesting, convoluted, and dwarfy solution to the fire resistance of the undead, to be sure!

[WanderingKid]

Spoiler: Watch the video first! (click to show/hide)

 What inspired the multiple bridge/drop pit design?  It's an interesting, convoluted, and dwarfy solution to the fire resistance of the undead, to be sure!

Answer:

Spoiler (click to show/hide)

It started because I needed to clear out the guys in the near corner so I could get the grates installed to keep 'exploding bodies' from entering the magma refilling system, and their presence was freaking my dwarves out installing the last grate.

From there, I was curious as to how long it would take to expand it and how effective it would actually be.  It's highly effective, but takes an age to construct with only a few dwarves.  But at this point I can selectively remove anything in the lava runs.

[Dolwin]

I've enjoyed the series so far and have learned a bunch from it.

Spoiler: Undead Flyer (click to show/hide)

Out of curiosity, what is preventing the undead flyers from rezzing and flying back up the original dump pit?

One suggestion is that when you turn off the grinder to clean house, I'd recommend flooring over the grassy parts to prevent tree growth blocking the path to the grinder. 

Here's to hoping that some migrants survive to enter the fort sometime soon!