Print

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

[Nargorth]

Dunno if its useful information, made little test to check how efficient is wheelbarrow/minecart stone hauling. Chosen distance is probably longer than usual rock haul, though.

Setup:
*63x31 mined stone area => 4-tile wide 128 tile long area => stockpile area
*fresh embark, seven starting dwarves with wheelbarrows
*raw stone from digging areas left untouched (marble)
*all stone except marble in 63x31 area forbidden
*there are no other labors on map
*migrants were exterminated with DFhack and their bodies/items forbidden
*miners got 7 beds, food and alcohol in nearby room for their use
And either:
*use wheelbarrows to transport stone (128+ tiles) to final stockpile; two 11x11 stockpiles with 5 wheelbarrows each enabled, cleared from stone with DFhack autodump when ~1/2 full
*use wheelbarrows to transport stone to feeder stockpile (center of 63x31 area), load onto minecart and guide it to final  (quantum/1tile) stockpile


Start: 30-04-27
*1 minecart setup: dwarves filled feeder stockpile in 30-05-08, abort
*2 minecarts setup: dwarves filled feeder stockpile in 30-05-28, abort
*4 minecarts allowed dwarves to work full-time (except sleep, drink, eat). Task was finished in 30-08-18, with 4 stones left in one of minecarts, three other empty and returned to start position.
466 marble stones were moved in (28*4-10)*7=714 dwarf-days

Start: 30-04-29 [one miner started in 30-05-10 to 11 due to small error]
*wheelbarrow setup finished in 30-12-23 with 4 stones left and 1 job started.
466 marble stones were moved in (28*8-6)*7-11=1515 dwarf-days

It took about 2.1 times longer without minecarts; time for track carving was not counted, so it may be less than 2 times faster. Distance is pretty long – up to ~160 tiles from farthest stones, so I think that test should be done again with ~80-100 tiles.

[NESgamer190]

Ah, I will admit that this proved rather interesting to see the efforts unfold as you have done, though I assume you had the dwarves in question guide the carts with the minecart team?

Perhaps, a couple ideas could be had for alternative means of transferring stones via catapults or doing a variant of the minecart system you used, but utilizing a means to mechanically move the carts as opposed to using plain dwarfpower on its own?

Definitely looking forward to just how much of a reduction in return one'd get in getting things done with the reduction of distance.

[anewaname]

In the test you are performing, your minecart system is under ideal circumstances, so it is working at maximum efficiency.

There are two other considerations for the experiment.

"The comparison of the rates of job creation verse job completion"
Your example has the 466 rocks available all at once. Consider what happens if the rock is being dug slower than the loading is completed.

"The availability of other jobs to be done"
Your example has few, if any, other jobs to be done. Consider what happens when a hauling job is available at the farm and the jobless dwarf is at the dig site.

When combined, these two considerations will lead to this event... Urist walks to the dig site and puts a rock into the minecart. There is no other rock to load, so he is tasked another hauling job and walks back. A moment later, the miners produce another rock, so Mathol is assigned the job and walks to the dig site to load the rock into the minecart. Urist and Mathol's walk time is the cost of having a lower job creation rate verse job completion rate.

[PatrikLundell]

In the test you are performing, your minecart system is under ideal circumstances, so it is working at maximum efficiency.

There are two other considerations for the experiment.

"The comparison of the rates of job creation verse job completion"
Your example has the 466 rocks available all at once. Consider what happens if the rock is being dug slower than the loading is completed.

"The availability of other jobs to be done"
Your example has few, if any, other jobs to be done. Consider what happens when a hauling job is available at the farm and the jobless dwarf is at the dig site.

When combined, these two considerations will lead to this event... Urist walks to the dig site and puts a rock into the minecart. There is no other rock to load, so he is tasked another hauling job and walks back. A moment later, the miners produce another rock, so Mathol is assigned the job and walks to the dig site to load the rock into the minecart. Urist and Mathol's walk time is the cost of having a lower job creation rate verse job completion rate.

And this ignores the fact that harvesting seems to be at the very top of the job priority list, so even if there are multiple stones to load, the loader will take the harvesting job as soon as it is available (the blasted job even causes miners en route to a digging site to turn around, which leads to very slow digging if the digging site is far away and you're at full employment, as that's about the only job that's interrupted by the posting of a harvesting job).

[Kametec_Housen]

Which is why you change orders and only allow growers to harvest.

[Insert_Gnome_Here]

Would burrows help? 
It'd be micromanagey, but it might work to keep dwarves from running off to do other jobs. 
 
How many rollers would it take to get a cart 160 tiles away? Would a modest height drop be enough?

[anewaname]

I think the thing that helps to improve the efficiency of long-distance minecart transport, above all else, is to trigger many hauling jobs at once, so the dwarfs loading the minecart will stay in the area to load again. This is a method of increasing "The comparison of the rates of job creation verse job completion".

Some method to prevent dwarfs from taking distant harvesting jobs should be used in a fort. I habitually placed meeting areas near the farms, so the kids are always available for harvesting, but noting how they are fixated on playing, I may need to change methods.

Burrows... If you wanted to specifically prevent miners from harvesting, would "an embark-sized burrow with the farm plots erased from the burrow" work?

[Greiger]

I am very interested in this experiment and wish to thank you for continuing the dwarf fortress tradition of SCIENCE.  So I am posting to watch.

As for your concerns about the distance being too far, I for one often want to make surface walls out of materials that the average joe considers to be strong(even if most traditional walls were just made of what stone was on hand) and I won't mix stone types in the construction, so I'll often dig quite deep and out of the way to find large amounts of granite or some other well known stone. In those cases 180 tiles and even upwards of 200 tiles does not seem out of the question.

That said more data is always good.

[Kametec_Housen]

Burrows... If you wanted to specifically prevent miners from harvesting, would "an embark-sized burrow with the farm plots erased from the burrow" work?

I can imagine it would, but I'd suggest altering [ o]rders so that only farmers [h]arvest instead.

[Nargorth]

Hey, RealLife [TM] managed to take all my time :|

@anewaname
@Greiger

Yes, the goal was to work at almost 100% efficiency. In my fortresses, I make special "team" from 5-15 dwarves who have only stone hauling and push vehicle jobs enabled, and only farmers harvest. Also, all recruits dig stone/ores before earning their military status so the stone is always available. Most often, I dig something like 21x21xA where A is amount of z-levels of rock which i need/like, plus any nearby ores [automine ores designation]. When stone runs out, I set hauling team to wood/food/refuse etc., dig another block, set up tracks and repeat.
I play on medium-deep maps, with magma sea in last example that I remember being at ~160 z-lvls from surface, and I love surface castles, citadels etc. On that map there were ~5 z-lvls of usable, non-cavern granite -- and another few at the bottom of the map, I think, so the distance was quite long.
Making food/drink stockpiles and small quarters with dining hall and universal temple zone in 1st dug out area should take away most of distracting things.

@Insert_Gnome_Here
The only problem is that even with burrows you can not dig same rock type at 2 places, thay will try to collect distant stone outside and cancel job, leading to cancelation spam and wasting time. For 160 tiles, roller setting speed to 50k should be able to go all the way, IIRC friction is 160x10=1600 speed loss. Another few thousand could be lost at corners, so lets make two of them [few tiles from stop1--push and after track stop dumpzone] for sure, 3-5 tiles long to be absolutely sure -- the real thing is that unless you are using pepretual flow generator or bring aquifer nearby by dropping piece of if in cave-in, you have, like 200-400 power loss from machinery alone [from surface river]. My very crude semi-automatic system used 800 units of power, to run at full speed at flat track and do ~20z-lvls. Mostly, power was used for loads and loads of gears to power rollers every z-level and for shaft, which was ~100 units long from river on map's border.

I dont take impulse ramp [thats the name, right?] into my calculations.

@NESgamer190

I guess that whole series of experiment should be done [no time now, will try in few next weeks :/]. Probably, all should be done with same dwarves on the same map to reduce randomness in attributes.
Something like:
**dwarf-powered and machinery powered
*bring stone 50/100/150z up from 21x21x5 block
_|__dig
_|__dig
_|__stockpile
____dig
____dig
*bring stone 50/100/150z from distant 21x21x5 block [as above but 60-80-more x/y from main stockpile]
*bring ore from 1/4th [48x48x5?] map, both direct and x/y distant? There will be only small amount of stone + moderate ore, but paths wil be longer and annoying for haulers

Also, the major advantage of the system is that building cost [time] is diminishing the more sectors you dig out and bring to surface/magma forges at bottom. And the major [personal] disadvantage is that tracks+ramps are annoying to designate