Bay 12 Games Forum

You pulled me back into playing DF again, heh.  One pick is I guess a rather drastic solution for my embark profile OCD!

I'm incredibly impressed at how long you've run this fort.  Most of mine don't get beyond year 5, though I guess that's partly because I get OC about all my immigrants.

So far my first attempt ended up in a savage biome, turned out only the ocean beach was terrifying.  After failing 2-3 times on the chicken run method I realised that it was failing because the miner was trying to dig stone, not dirt.  Then I realised that gabbro wasn't an aquifer stone.  Woops.

So I tried out methods to do a 1 layer aquifer punch w/o opening to the sky.  Since there was no lower aquifer layer to drain into, I had to undercut the plug in the lowest dry layer, so I needed 3 layers of dry soil to pierce the 1 deep aquifer instead of just 2 to pierce a 2 deep aquifer.

The chicken run drain aquifer pierce is really a very sweet design.  We should get this into the wiki.  I think it's faster then the doubleslit method even for a normal embark, especially on soil aquifers (building walls at least feels much slower then smoothing).

And if you do the setup for the cave-in right, you should be able to avoid getting any cave-in dust on the miner, so it's not like that's a big risk over the doubleslit either. (channel from top layer, use diagonal step in corridor to trap dust, give miner job elsewhere to make them leave before the dust rises up to their z-level.  This at least worked with my 1 layer pierce, which was basically a 3x3 version of the 2 layer method with a dry upper aquifer layer.)

This looks really interesting! What's the peak speed, ticks per cart sent through?

Is there an easier way to count ticks then pressing '.' repeatedly and losing count?
The main problem speed wise is the magma refill.  I ended up with a 1 wide channel 13 squares long from the magma tube.

In the design I listed at the start, I had 4 carts loaded into it waiting for the 7/7 pressure plate to trigger, 1 in the trench and 3 stacked above it.  All 4 slammed through the trench right after it filled, but only 3 of them were full of magma.  I expect what that means is that if the trench is full to 7/7 magma, it can fill 2 carts.  Which seems to fit with my testing of the (failed) reversed design, opening a door makes the magma next to it drop in half right away

I tried a reversed system (dry first, then magma), and it doesn't work anywhere near as well.  I was using dumping zones for testing this time, since setting up minecart routes each time I changed the dump location was a royal pain, ...  But the main problem is losing magma depth before the carts have a chance to fill.  There also seems be a time or speed requirement, so one needs to be careful to not over accelerate the carts.

More testing reveals that impulse ramps will cause minecarts to gain speed in 6/7 magma and lower.  Impulse ramps seem to be able to keep minecarts moving on the level in magma, but can't push it up a level (which we already knew).  Possibly they are hitting some maximum speed?  However it seems that all your really need is a way to keep the last couple of spaces at 6/7 or lower.


New Improved Design! only 1 magma safe mechanism, and even that's optional!

Well, I'm not counting anything you're using to create a magma drain.  I ended up just piping it off the map for this one, since it made testing simpler.  The basic setup is just having a 4wide input, and a 1 tile drain to the side of the last ramp, this should keep the first several ramps permanently at 6/7 while the drain is open.

Top view, z0
wwww
+dRw
wwRw
wwRw
MMRw
MMRw
MMRw
MMRw  <-- I was dropping on this ramp when testing
wwRw <-- I'd suggest adding this line when building it to drop on a ramp here instead, as I ended up with one cart knocked into the magma input area once
wwww

Top view: z1 (as tested, the open space '.' is over the south most ramp)
wwww
w.pbw
wwww

Top view: z2
www
w.w
www

Top view: z3
www
S.w
www

The 4 input channels (~11 tiles long) coming off the side of my magma pipe were enough to keep the first 4 squares at 6/7 solidly with a 1 wide output.  I then had a 4wide path off draining off the edge of the map through fortifications.  Other drain systems would work, but since my pipe is right near the map edge that was super easy to setup.

On z1 I have a pressure plate (dry!) that opens the door on z0 when a cart is about to go through the lava (it will still be open when the cart leaves).  The door can actually be left out, or manually controlled  It's mostly just there so the magma stops flowing if you don't use it for a while.

z2 is just the top of the minecart stack and z3 is the track stop, you can put the stop on z2 (or even z1) safely if you're not using the bridge, since you won't have minecarts stacking up in reach of your dwarfs.

The bridge is optional, I used it to prime 4 carts at once, which makes one gets stuck on the pressure plate when you launch all 4 together, however it does not seem to jam the system and it loaded the other 3 with magma.  I'm sure someone has a good quantum cart stack dispenser design on the forums somewhere..... hopefully...  For a basic setup I'd just skip the bridge and rely on fact that track stops can only be used by one dwarf at a time.  If it does get stuck, you do kinda need to remove the cart, since it defeats the point of the door.  Then again, you don't really need the door either if you don't care about magma updates.

Once I ended up with a cart knocked sideways into the magma entrance area, but I've been unable to reproduce it.  I'd change the design as I noted in the diagram though just to be safe.

A bad method for this could be done relatively easily. Just set pigs to have a skin that ignites at room temperature. It could be entertaining.

Would storing them in nethercap cages work?

Please tell me this actually works. I remember briefly seeing that someone had said impulse ramps didn't work so well for minecart lava retrieval due to the fact that there is a lower limit on speed required to go up a ramp, and an upper limit on the speed a minecart can be traveling at in order to still collect any magma at all, and that there was no overlap between the two ranges.

I have actually built it, the original setup was much simpler (fewer magma-safe mechanisms) but would jamb if too many carts got dumped in quickly.

There isn't actually a upper limit on speed to collect magma as such, the cart must be immersed in 6/7 magma to pickup magma.  The problem is that if the cart is going too fast it will bounce across the top of the magma instead of sinking into it.  This limits the ability to preaccelerate  carts, though you just gave me an interesting idea to try out.

The other problem is that 7/7 magma has a HUGE amount of friction (60000 + regular tile friction to be exact, for comparison a highest friction track stop is 50000)
But 1/7 magma has no (extra) friction, and 2/7 magma is only 10000, the same as a high friction track stop.

data from http://www.bay12forums.com/smf/index.php?topic=112831.msg3536975#msg3536975

There is a much easier, tested way to do this. All you need to get a minecart out of the magma trench is another minecart.

Sideview:

Code: [Select]


_2
   \      >___
    |>>>1/

_ = track
> = impulse ramp
1,2 = minecarts

First, you push minecart 1 into the trench. It will fill, move up to the exit ramp and stop in the trench due to friction. Next, you push minecart 2 in. It will fill, move up to the exit ramp, push the first minecart out with it's momentum and stop where minecart 1 used to be. Minecart 1 will climb up and head where ever you want it to go.

There's no spillage because the collision of the two carts is so slow in magma and the minecart comes out at a very slow and manageable speed. For automated systems, dropping carts in through a pit seems more reliable than using slopes, as the latter might accelerate the cart too much and cause it to skip accross the magma trench.

When I  tested a newton's cradle setup I ended up with multiple minecarts stuck in the lava if several of them got dumped in too quickly.  It's possible my problem was related to an unfilled cart hitting a filled cart.  Minecart momentum has some weird quirks.

Of course the absolute easiest method, if you don't mind doing most of it manually, is to just drop 2 minecarts into a trench of magma, then dump water on it from the floor above, then dig out the obsidian.  2 carts worth of magma is enough for a magma forge (4/7 magma).

I especially designed this for a minecart in wheelbarrow magma transport system, so you can get magma workshops really early without having to setup a bunch of power reactors.

The key to the system is keeping the magma level low on the exit impulse ramps so they can accelerate the minecart out of the trench without excessive friction, but still have the magma level high where you're trying to load the cart.

Side View
s.
w......._
wbRpRdRRw

Top View of trench layer
wwwMwwwww
wwwpwwwww
wbRpRdRRw
wwwwwwXXw
wwwwwwXXw
wwwwwwwww

Legend:
s - track stop
b - retracting bridge over impulse ramp
R - impulse ramp
p - pressure plates
d - door
w - wall
X - raising bridge, into wall
. - open space
M - magma

Hopefully the bad ascii art is clear enough. The working bits are: a pressure plate set to 7/7 magma (top one) to trigger the retracting bridge, to launch minecarts.  The pressure plate set to open the door when a minecart goes through, and a drawbridge connected to the same plate to get rid of extra magma. The one space gap is critical, when I had the pressure plate next to the door the minecarts just stopped.

The minecarts will make a quantum stack just next to the minecart stop if the magma is not full.  When testing with 4 minecarts in a situation where they all end up stacked, they all to exit properly, however only 3 of them got filled.  Make sure you have enough room store your minecarts after the exit so they dont back up into the lava.   

In practice, you'll want the filled minecarts to end up sitting on a stockpile, so you can use give/take orders to get them where you want (be sure to use wheelbarrows in the taking piles!!)  the easiest way to dump these minecarts is to setup 2 track stops right next to each other, both with dumping orders, and 2 stockpiles, one next to the first stop set to only take from links and use wheelbarrows, and one next to the 2nd stop.  Setup a route with a stop on the first stop, set to take from the first stop's link only pile.  When a filled minecart is delivered, it will be placed into the route's (wood ok!) minecart and dumped right away onto  the second stop, which will dump the magma out!  Then a dwarf will show up and place the minecart into the 2nd pile's stockpile, since it's the closest matching stockpile.  Then send the cart back down to the depths! 

It would also be fairly straight forward to setup an automated system, if you make the first stop dump onto an impulse ramp, and use a lowest friction stop for the magma dumping.  Then route the track to a dropshaft to load them back into the magma loader, would only need dwarfs for the upwards route.  You could even setup pressure plates to sense the magma level and turn off the magma dumping track stops to have a multi point distribution system.  But that's not really on the topic of no power magma loading is it?

I'm going to see if I can make one using the "Moses effect" to clear the lifting ramp that will work with (intermittent) dwarf power, for a simpler starter system. It didn't work very well.

[Sorry to resurrect an old thread, but I felt posting info about how to fix the OpenGL window error was worth it]

It seems to be actually a problem with DRI and wine not falling back to software rendering automatically (probably a feature... but I wish the error message was clearer).

This link has instructions on how to fix the wine opengl window issue in Ubuntu: http://ubuntuforums.org/showthread.php?t=361136

I was able to get it working with just removing the fglrx modules and then reinstalling libmesa-dri in synaptec, YMMV.