Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  
Pages: 1 2 [3] 4

Author Topic: Dwarves at war  (Read 3007 times)

Urist McShire

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #30 on: September 05, 2014, 11:26:33 am »

Since it's his first time using minecarts, I think making a 100 z-level ramp track down is a little too ambitious a project until he's got the basics down.
Logged

Aslandus

  • Bay Watcher
  • Slowly descending into madness
    • View Profile
Re: Dwarves at war
« Reply #31 on: September 05, 2014, 12:07:28 pm »

So a ten block long carved track  down three or four z levels to haul rocks up slightly faster? That's how I learned about minecarts....

Urist McShire

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #32 on: September 05, 2014, 12:15:17 pm »

I learned about them to clean up a large workspace for my forges and smelters and create a number of quantum stockpiles so I wouldn't make massive stockpiles filled with rocks, blocks, or bars just taking up space.
Logged

khearn

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #33 on: September 05, 2014, 04:29:55 pm »

I've been trying to figure out how to make a drop chute safe for workers at the bottom. It's a non-trivial problem. I haven't been able to figure out a good way to make a fast inverter. Pressure plates take 99 ticks after the load has been removed before they send an 'off' signal.

The only problem is the 100 tick delay of bridges. I don't know, but how fast can an inverter be? Then you could hook up the pressure plate to an inverter and let that drive a hatch (no delay, I believe). An inverter could be super fast: A hatch with water on top that is dropped down onto a pressure plate triggering for 1-7 water (hook this up to the hatch to be driven) and then have a screw pump that pumps out the water from the room - or alternatively have a draining hatch that is triggered when the dorf walks out of the collection area (this _will_ lead to problems if more than one dorf is allowed into the collecting area!)
This will have your pressure plate send an 'on' signal, which opens a hatch, not closes it.

The best idea I've had is a minecart loop track set up to take just under 99 ticks to do a complete loop, with multiple pressure plates on it. Since the cart keeps crossing the plates, they never quite get to send their off signals. You'd also have multiple doors on the track, all links to the same input (a pressure plate on the way t the bottom of the shaft). When a dwarf triggers the plate, the doors immediately close, stopping the cart wherever it is on the track. The next presure plate on the track that the cart was just about to hit is only a few ticks from sending its off signal, so it does so in a timely manner.

The problem is getting the cart moving again after the doors all open again.

Here's a link to a post I just found by someone who managed to get the first part set up similarly to what I was thinking, only he used rollers instead of doors to stop the cart. But his system sends the cart flying off on a diagonal, so it's a one-shot system. We'd want something self-resetting.

I suspect something could be done with impulse ramps and Larix's Minecart Pathing Logic, but I haven't had the time to study it enough to work something out. It would end up pretty complicated and would probably take a bit of trial and error to get it working right.
 
Another option which I haven't given much thought is a loop of pumps moving one chunk of water around, like your typical mist generator, only with each spot being walled in so the water stays put when the pumps stop. This might work better since it would be easier to restart it. You'd just need a long enough loop to take just under 99 ticks. And a fair amount of power to run all those pumps, since it would probably take 90+ pumps. And I think the build order would be critical to get the timing right. I seems to recall that if you build it in the right order the water moves through one pump per tick, and with the wrong order it moves through all of them in one tick. Needs more thought and experiments, but it might be a better option.

Either way would be complicated to get working. just digging a minecart spiral to carry stuff is probably easier.

   Keith
Logged
Have them killed. Nothing solves a problem quite as effectively as simply having it killed.

Mimodo

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #34 on: September 07, 2014, 05:21:32 am »

Since it's his first time using minecarts, I think making a 100 z-level ramp track down is a little too ambitious a project until he's got the basics down.

More to the point, it's way too ambitious for the dwarf labour I'm allowing myself in this fortress. I think pirate bob's way works great for me... A little micromanagement is no huge deal for me.

I kind of forgot that I'd have to invert any pressure plate signal, so I'm glad someone pointed that out to me. A hatch would be preferable to a retracting bridge?
Logged
Bay12 Forums... Probably one of the last few sanctuaries on the internet. We're all equally insane in our own way, and it is that insanity that brings us together

Larix

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #35 on: September 07, 2014, 06:21:07 am »

Yes, hatches react instantly to signals and drop stuff lying on them directly - bridges actually "throw" stuff a bit, which can lead to scatter or differences in drop time. When timing is critical, hatches are a very good choice.

For better security, you could put a hatch in the drop chute itself and another one - over a pit - into the path dwarfs need to go to retrieve dropped stuff; it'll serve as an access bridge. Link both hatches to the same signal source and they'll open and close in unison: when the access hatch is closed and thus passable, the drop chute is sealed and nothing can fall on the haulers' heads. When the drop chute is open, the "bridge" is retracted and haulers can't go and get their heads smashed.

Inverters in DF tend to be quite slow since they have to wait for the 99-step timeout on "off" signals. There's no easy way around that. The method lined out above - a refresh cycle where the off signal comes up quite quickly when the refresh event is blocked - works but takes tons of space, power and machinery.
Logged

Snaake

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #36 on: September 07, 2014, 08:41:49 am »

I vastly prefer bringing magma to the near-surface, and you don't need a track to do it. Each minecart can hold 2/7 depth of magma for a single tile, so a magma smelter requires 2 magma filled minecarts and a forge requires 4.  6 total.
Step by step:
1) Dig down to the magma ocean and setup a 6x1 magma flooding room. Use the old ramp and bridges trick (http://dwarffortresswiki.org/index.php/v0.34:Bridge#Ocean_drains) to safely pierce the magma and provide you an on/off switch for the flood. Add a large cistern for the magma to drain, also controlled by a lever bridge.
2) Construct 6 magma safe minecarts with ore and coal/wood topside. Nethercap doesn't work for this purpose (tried it!)
3) Designate a stockpile that only accepts magma safe minecarts in your flooding room. Wait till room is filled with your 6 minecarts.
4) Seal the flooding room and hit the switch to fill it. When all carts are in 7/7 magma, hit the switch to close the flood and the switch to open the cistern for draining.
5) While magma is drying up in your flood room with the now filled carts, dig your rooms for smelter/forge wherever you want them, and channel the holes for magma to be dumped into.
6) Construct 2 track stops next to each channeled hole with dumping direction into the hole.
7) Create a route for each track stop and designate the vehicle to be one of your magma filled carts. Remove all departure conditions from the stop.
8) Dwarves will haul the magma filled carts up your 150z stairway, place them on the track stop, which then dumps 2z of magma into the channeled spot.
9) Remove routes, dwarves will carry minecarts away.
10) Remove track stops, build smelter/forge.

Bonus: Dwarves will bring the minecarts back to the flooding room, allowing you to repeat steps 4-10 as many times as you feel you need. I usually only ever use one of each, since they are so efficient if placed near your fortress.

I'd recommend this for Mimodo to be honest, over any dropchute designs, since it has the lowest labour requirements once it's running. No need to haul stuff down, push around minecarts, or especially to bring the smithy products up (requires constant manual hauling), just a few haul jobs and a bit of micromanagement to do the initial setup. This can be a lot of micromanagement instead, if you're looking to power, say, 20 magma workshops, but with just a smelter, forge, maybe a glass furnace, and with minimal civilian professionals in any case, you wouldn't need all that much, so it's definitely doable.

I'm curious though, why would a magma forge require 4 carts of magma? If you're referring to the fact that a magma forge has 2 impassable tiles and you're placing 4/7 magma below both of those... I was under the impression that magma workshops are powered as long as there's at least 4/7 magma under *one* of it's non-center tiles, regardless of which tile that is, but it's just a safety measure to place the magma underneath an impassable tile.


edit: you can speed up the "waiting for the minecart-filler room to dry/cool down" bit, by having the minecart filler be a 2z deep, with the top z that is submerged in magma having grates for a floor. This comes at the cost of micromanagement though (and you need the drainage cistern to be 1z further down, and bigger), because you can't place stockpiles in magma, so you'll need to use the dump commanda to get the empty, magma-safe minecarts onto the grates, or setup routes and have dwarves push/guide them there/use tracks to get them there somehow. And then somehow stop dwarves from wandering over to pick them up again. The dump command is probably the best for this, since it auto-forbids by default, but then you have to remember to unforbid the carts when you want them hauled up etc... so the grates would more probably be useful for a case where you're actually hauling up more than 1-2 loads of magmacarts.
« Last Edit: September 07, 2014, 08:49:10 am by Snaake »
Logged

Pirate Bob

  • Bay Watcher
  • [ETHIC: TORTURE_FOR_SCIENCE: ACCEPTABLE]
    • View Profile
Re: Dwarves at war
« Reply #37 on: September 07, 2014, 01:22:42 pm »

Is there a reason you can't just make the minecarts travel on a track through a pool of magma to fill them?  The wiki is somewhat ambiguous about this - it says that it is difficult because the minecarts slow down greatly when filled, but it seems like you could easily get around this by building a ramp of 5 or so z levels to gain momentum before filling.  Is there some other difficulty that I am overlooking? 

It seems like you should be able to just create a track stop at the top of a (tallish) ramp going down into the magma set to push always, and another on a 1z ramp out of the pool set to never return (or just return after a really long time, if no return condition makes the route not work).  To me this sounds easier than filling and draining magma from a room (which admitted also sounds pretty easy), but I feel like I am missing something.

Aslandus

  • Bay Watcher
  • Slowly descending into madness
    • View Profile
Re: Dwarves at war
« Reply #38 on: September 07, 2014, 02:51:02 pm »

There's a thread about it... somewhere... It is possible (or at least it used to be) but it requires you to build the track in a basin then let lava fill the basin then have the minecarts run through it and go back up out of the basin and up to where you want it to dump out... yeah.... it's just easier to dump the lava onto the carts then move the lava carts to where you need it...

Snaake

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #39 on: September 07, 2014, 04:30:29 pm »

Is there a reason you can't just make the minecarts travel on a track through a pool of magma to fill them?  The wiki is somewhat ambiguous about this - it says that it is difficult because the minecarts slow down greatly when filled, but it seems like you could easily get around this by building a ramp of 5 or so z levels to gain momentum before filling.  Is there some other difficulty that I am overlooking? 

It seems like you should be able to just create a track stop at the top of a (tallish) ramp going down into the magma set to push always, and another on a 1z ramp out of the pool set to never return (or just return after a really long time, if no return condition makes the route not work).  To me this sounds easier than filling and draining magma from a room (which admitted also sounds pretty easy), but I feel like I am missing something.

It's a bit tricky; it's easyish to get a minecart accelerated to high enough speeds that it'll pass through a magma trench and come back out, but with such a system it's almost guaranteed that the minecart is going too fast to actually fill up with magma. And vice versa, if you use lower speeds that allow the carts to fill, then they usually end up with too little speed to be able to take the ramp out of the trench full of magma.

Keep in mind we're talking unpowered systems here, if you can route power down to the bottom of the map, it's fairly trivial to have a roller in the trench to shoot the carts out. But routing power to the bottom of the map... yea, easier to make a dwarven water reactor, and again we end up with a more complex system than what's necessary. Returning to unpowered systems, one guy had success with using several minecarts in the loop (finicky, until you learn it, like much else in DF), so that the next cart would also bump the one stuck in the trench, out of the trench, and get stuck there itself...

Buuut returning more on-topic:
There's a thread about it... somewhere... It is possible (or at least it used to be) but it requires you to build the track in a basin then let lava fill the basin then have the minecarts run through it and go back up out of the basin and up to where you want it to dump out... yeah.... it's just easier to dump the lava onto the carts then move the lava carts to where you need it...

Yes, there's a thread or few, and I think I have a relevant one bookmarked, was just looking at it earlier today, but unfortunately I seem to only have the bookmark on my main pc (on laptop now), and I'm slightly too lazy to search it up just now, right before bedtime... Because in any case, like Aslandus said, you'd then have to dig out a track spiral, carve track, have dwarves guide them up. So for just filling a minimal setup of 2-5 1-tile holes with 4/7 magma, just manually dumping/stockpiling the carts in a basin that you then fill with magma&drain (and have dwarves haul the carts up) is simpler, and requires less dwarf-hours. Referring back to what I said earlier, once you want to power, say 20 magma workshops near the top, or design some sort of magma-based defence system, then more automated magma minecart systems, magma pistons, and pumpstacks (in rough order of throughput) start to become worth the dwarf-hours you need to put in to get them working.
Logged

Mimodo

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #40 on: September 08, 2014, 12:17:10 am »

Well... right now, it doesn't make a huge difference how I do this, because just after getting the chute set up, my civilisation crumbled to it's end... Something about a fire breathing giant bat I think
Logged
Bay12 Forums... Probably one of the last few sanctuaries on the internet. We're all equally insane in our own way, and it is that insanity that brings us together

gunpowdertea

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #41 on: September 08, 2014, 02:14:10 am »

I've been trying to figure out how to make a drop chute safe for workers at the bottom. It's a non-trivial problem. I haven't been able to figure out a good way to make a fast inverter. Pressure plates take 99 ticks after the load has been removed before they send an 'off' signal.

The only problem is the 100 tick delay of bridges. I don't know, but how fast can an inverter be? Then you could hook up the pressure plate to an inverter and let that drive a hatch (no delay, I believe). An inverter could be super fast: A hatch with water on top that is dropped down onto a pressure plate triggering for 1-7 water (hook this up to the hatch to be driven) and then have a screw pump that pumps out the water from the room - or alternatively have a draining hatch that is triggered when the dorf walks out of the collection area (this _will_ lead to problems if more than one dorf is allowed into the collecting area!)
This will have your pressure plate send an 'on' signal, which opens a hatch, not closes it.
Yes, thus the use of the inverter. The Dorf steps on a plate, the inverter... uh, inverts the signal and the ...
Shoot, you're right. So the inverting plate goes into the upper room that is full of water and then the Dorf-triggered one opens a hatch to clear out the water. Plate is unloaded -> hatch closes. Thanks. Also for reminding me about the 99 tick "unload"-delay of pressure plates, I never timed that but was (quite often) caught by it - and forget about it again.

Quote
The best idea I've had is a minecart loop track set up to take just under 99 ticks to do a complete loop, with multiple pressure plates on it. Since the cart keeps crossing the plates, they never quite get to send their off signals. You'd also have multiple doors on the track, all links to the same input (a pressure plate on the way t the bottom of the shaft). When a dwarf triggers the plate, the doors immediately close, stopping the cart wherever it is on the track. The next presure plate on the track that the cart was just about to hit is only a few ticks from sending its off signal, so it does so in a timely manner.

The problem is getting the cart moving again after the doors all open again.
powered rollers before every door? This is a bit... overpowered (well, needs overly much power) though (I don't really like the impulse ramps). I am currently building spiral ramps to get stuff down from the surface and relocate my fort down to the magma sea, though I might try and implement some of the musings above for laughs. I generally move the fort downstairs once I get the magma furnaces running. A chute to bring the charcoal and metal ores down would be useful, though.

I do second the magma piston suggestion by Snaake, that is actually 1) easy 2) relatively fast and 3) gives you a lot of magma. With some effort you can even make it reusable. The only "difficulty" is to go through the caverns - difficult if they are the lair of some beast.
Logged
I don't care. I have discovered that if you spawn elves this way, cats will chase them down and eat them.

khearn

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #42 on: September 08, 2014, 04:28:21 pm »

I just had another idea for a "safe" drop chute. It requires that you have a steady supply of whatever you are dropping, such as logs. You set up a long timer connected to the hatch in your chute and a door in the access hallway to the bottom. So stuff builds up in the chute for a while, then gets dropped all at once while access to the bottom is blocked. then the hatch is shut and access to the bottom is reopened. Since your haulers should have it all cleared away before the next drop, there is no reason for anyone to be underneath when the drop happens.

Since no one has any possible reason for being there, you will of course lose an occasional dwarf who decides to just hang out where he has no reason to be. But that sort of dwarf would otherwise decide to hang out somewhere on the surface when an ambush is arriving, so there is only so much you can do. It is inevitable. But the casualty rate should be far, far lower than just dumping stuff down with no protection. And the ones you lose are morons with death wishes, so it's probably for the best.

One possible timer is a captured were creature sitting on a pressure plate. Since were creatures are trapavoid, the plate will send an 'off' 99 ticks after it converts to were mode, and an 'on' when it converts back to normal, for a very reliable once a month timer. An alternate would be a cistern that slowly fills, with a plate that triggers when it is full that starts it draining, and another plate that closes the drain and opens the fill. It might need to be a pretty big cistern. You could also use a smaller cistern hooked to some counting logic.

Whatever you use, it needs to send an 'on' signal to the hatch, followed shortly by an 'off' signal, then a long delay before starting over. So you'd need to invert a were-clock, but the 99 tick delay in an inverter wouldn't matter, since we're dealing with much longer timing cycles. You'll need an inverter anyways, since the door needs to be connected to a signal that's the opposite of the hatch. You want the door to close when the hatch is open, and vice-versa. Ideally you'd want to close the door no later than the hatch opens, so there is no chance for Urist McAgile to notice the stuff at the bottom of the chute and get past the door before it closes. But it's probably not that big of a risk as long as you don't have a meeting room right next to the door for idle haulers to hang out. Or you could just have a hole in the floor in the access hallway, with a hatch cover over it. An 'on' signal opens it (and thus prevents access) at the exact same time that the 'on' opens the hatch in the chute.

So stuff gets dropped down the chute whenever, and stacks up on the hatch, which is closed most of the time. Then the timer goes off and the access door at the bottom closes and the hatch opens, dumping the accumulated stuff to the bottom. A little while later the hatch closes and the door opens and your haulers can come in and safely grab the stuff and haul it to a nearby stockpile. They should be done long before the timer triggers the next dump, so nobody has any reason to be at the bottom of the chute when it triggers again.

The downside is that you have to make sure you plan far enough in advance so that you have a month's worth of stuff in the chute every time the hatch opens, since it's not easy to get more down quickly if you run out. But it's easy to cut or mine faster than you use logs or ore, so this shouldn't be a problem as long as you pay attention.

You can also have a lever hooked up so you can open the hatch (and close the door) whenever you want to. But then you should also rig up a way to disable the timer so you don't have the next scheduled dump happen while your haulers are still in there clearing out your emergency dump.

   Keith
Logged
Have them killed. Nothing solves a problem quite as effectively as simply having it killed.

Urist McShire

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #43 on: September 08, 2014, 05:29:22 pm »

Well... right now, it doesn't make a huge difference how I do this, because just after getting the chute set up, my civilisation crumbled to it's end... Something about a fire breathing giant bat I think

So your fortress fell? How did it happen?
Logged

Mimodo

  • Bay Watcher
    • View Profile
Re: Dwarves at war
« Reply #44 on: September 08, 2014, 05:35:16 pm »

Well... right now, it doesn't make a huge difference how I do this, because just after getting the chute set up, my civilisation crumbled to it's end... Something about a fire breathing giant bat I think

So your fortress fell? How did it happen?

It started with a deadly dust werebeast, which hospitalised my entire military, before killing them one by one. In my frantic struggle for survival, I forgot to wall off the caverns, and then a giant bar who breathes fire came in and slaughtered my completely untrained military. It's the first time I've been completely destroyed by losing every dwarf
Logged
Bay12 Forums... Probably one of the last few sanctuaries on the internet. We're all equally insane in our own way, and it is that insanity that brings us together
Pages: 1 2 [3] 4