Bay 12 Games Forum

Please login or register.

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

Author Topic: [CHALLENGE]Dwarven Combination Locks  (Read 13541 times)

Pie Maker

  • Bay Watcher
  • [DOES_NOT_MAKE_PIES]
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #30 on: July 20, 2013, 08:29:56 am »

PTW; these designs might prove to be very useful sometime.
Logged
"W-w-w-w-what do you want?"
"...MORTAL. I COMMAND THEE. BRING ME CHEESE."

Larix

  • Bay Watcher
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #31 on: July 20, 2013, 09:48:37 am »

Posting this separately - i forgot to mention the hatch links that make my minecart lock work:

Lever one - 'true' hatch of cell one
Lever two - 'true' hatch of cell two, 'false' hatch of cell one
Lever three - 'true' hatch of cell three, 'false' hatches of cells one and two
Lever four - 'true' hatch of cell four, 'false' hatches of cells one to three

If levers get linked to 'false' hatches later in the sequence, the 'solution' also includes returning those levers to their original positions. It would also be possible to use lever returns as direct input by building lock cells which return 'correct' when a specific lever is _closed_ and 'false' when another lever is moved to open.

The idea with a lock which 'evaluates' an array of lever-operated machinery _after_ the levers have been 'set' sounds interesting. Could you provide a concept/example? One limitation would be the behaviour of triggers - you might need to 'cycle' a lever to make sure the linked building is actually in the position corresponding to the lever. It would also, i think, be limited to 2^n combinations, where n is the number of levers. A sequence of 'pull these levers in the correct order' allows n! (factorial) combinations, which is bigger than 2^n for n>3. A sequence of 'operate these n levers m times, in the correct order (opening and closing both applied)' would allow n^m combinations with a theoretically unlimited m - basically, it'd be the number of lock 'cells' applied. I'd expect such designs to get hideously complicated very fast.

EDIT: expanding on that thought, basic design for a perpetual-motion minecart 'cylinder' reacting to the first 'on' signal:

Cart arrives from the west, into the 'upper' opening of the southern pit. Both pits with straight ramps, outlet covered with a hatch. First one to open lets the cart out, either to the north or south.

Basic design for a cylinder reacting to the first 'off' signal:

Straight N-S ramps again. All hatches open when ready to accept input, of course. As long as the hatches are open, the cart circulates through the array, the first closed hatch it encounters sends it back out of the ramp - to the south from the eastern pit, to the north from the western pit, the emerging cart follows the track corner and leaves the array either to the NW or SE.

One step advanced: Cylinder reacting to the first of three levers to trigger, one originally open, two closed:
       
To the left with all hatches visible, to the right when awaiting input - the left-hand one of the southern pits is originally open.
Straight N-S ramps. The cart arrives from the SE, underground (actually, it's cycled in from straight south, across the southern pressure plate - the input speed must be limited for the circuit to work). Cycles through the array as long as no input changes, the first input to change - either a closed hatch opening or the open hatch closing - sends the cart to one of the three pressure plates.

For laughs, i glued it all together and built this abomination:

Edit: revised debugged version, with minimal wiggle room.
Since all 'false' results before the final three-argument cell are simply cycled back to the beginning, you're practically guaranteed to get _a_ result eventually(*). The canonical way to the solution is 1. microcline on - 2. marble on - 3. cobaltite on - 4. cobaltite off - 5. marble off - 6. microcline off - 7. cobaltite on - 8. CHOICE.
Basically switch levers on in the correct order (six possibilities), then switch all levers off in the correct order (six possibilities again), open one lever (three options), make the final choice to get one of three results. Each of the three pressure plates opens one of the schist doors, providing access to, as the designer may choose, water, oncoming minecarts, or treasure.
As you start, the olivine hatch cover is open and the pitchblende one closed. The olivine hatch cover should not become visible before the third lever is pulled - if it does, you did something wrong and need to close a lever or two before proceeding. When the pitchblende hatch cover opens, you're getting close. If it re-appears and the olivine hatch cover doesn't open again, you're ready to make the final choice.


(*) Of course, it'd also be possible to 'punish' the user for false combinations, and there are five distinctive 'failure' paths you can play with if different results are desired.
« Last Edit: July 20, 2013, 05:15:51 pm by Larix »
Logged

Atomic Chicken

  • Bay Watcher
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #32 on: July 23, 2013, 09:50:25 am »

That is a very fine piece of dwarven machinery you have there Larix. Took me ages to grasp, 5 stars for dwarven ingenuity and over-complicatedness! Added to OP.
Logged
As mentioned in the previous turn, the most exciting field of battle this year will be in the Arstotzkan capitol, with plenty of close-quarter fighting and siege warfare.  Arstotzka, accordingly, spent their design phase developing a high-altitude tactical bomber. 

Tevish Szat

  • Bay Watcher
  • [PREFSTRING:diabolical schemes]
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #33 on: July 23, 2013, 02:25:08 pm »

This is an interesting one, especially for me when I'm not great at advanced logic with minecarts and the like.

I'm sort of working on designing a multistage combination lock based on levers alone as a logic puzzle

Essentially, you'd have a sequence of N doors, each with 2+ levers between them: one is an "escape" lever which does nothing but open/close the door you had to get through to get to this room.  The other levers control some combination of doors and other devices (pumps, floodgates, hatches, and the like) that, if you try to simply make forward progress through the doors, will flood the vault with your fluid of choice.  The solution would have to involve using your ever-expanding lever access to open the next door AND maintain the trap in a "safe" state.

Ideally, some doors in the sequence would have to advance the trap's progress to be opened, reducing your margin for error as you got deeper into the lock.

The design behind this is actually somewhat complicated to do in my head.  I'm going to have to bust out the graph paper of just strike the earth and blunder forward at building it.
Logged
A medium-sized humanoid fond of fantasy and science-fiction.

Tevish Szat likes books, computers, board games, and cats for their aloofness. When possible, he prefers to consume hamburgers and macaroni and cheese. He needs caffeine to get through the working day.

wierd

  • Bay Watcher
  • I like to eat small children.
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #34 on: July 23, 2013, 03:30:20 pm »

I dunno.. I can usually defeat that kind of puzzle in 2 steps when #toggled items <9.

To pose a significant challenge, it needs to have about 16 compound triggered obstacles, with no ordering between the order of presented levers, and the kinds of connection they have.

Let's say we have 16 doors in a hallway:

O==============================================O
_D__D__D__D__D__D__D__D__D__D__D__D__D__D__D__D__D_
O==============================================O

And 12 levers:

/_/_/_/_/_/_/_/_/_/_/_/

Lever 1 triggers doors 1 and 7
Lever 2 triggers doors 5 and 6
Lever 3 triggers doors 2 and 3
Lever 4 triggers doors 4 and 9
...etc

With some of the levers triggering a door that is triggered by a different lever, but only one of them. (Each lever triggers 2 doors to toggle state)

There is no logical sequence between the levers and the doors. This makes it considerably more difficult to solve for intuitively, and also makes multiple possible sequences for solution, some much longer than others.

The "penalty" could be introduced by having the room with the levers slowly begin to fill with a liquid, and if not solved in time, the adventurer drowns in 7/7 water. Succeeding in opening the hallway allows the adventurer to walk the hallway, and step on the (ahem) "reset" pressure plate, which opens the drains in the entry hall, closes the flood system, and slams all the doors.

The toggle states of the levers will be all mismatched, having gone through the solution once before, giving no useful clues as to the solution.

A "reset on critical failure" should also be implemented with a pressure plate sensitive to 7/7 water, that resets the puzzle as well. That way when the adventurer drowns, the puzzle will still reset.

I might see if I can implement this.

Logged

Tevish Szat

  • Bay Watcher
  • [PREFSTRING:diabolical schemes]
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #35 on: July 23, 2013, 03:42:57 pm »

I dunno.. I can usually defeat that kind of puzzle in 2 steps when #toggled items <9.

To pose a significant challenge, it needs to have about 16 compound triggered obstacles, with no ordering between the order of presented levers, and the kinds of connection they have.

Let's say we have 16 doors in a hallway:

O==============================================O
_D__D__D__D__D__D__D__D__D__D__D__D__D__D__D__D__D_
O==============================================O

And 12 levers:

/_/_/_/_/_/_/_/_/_/_/_/

Lever 1 triggers doors 1 and 7
Lever 2 triggers doors 5 and 6
Lever 3 triggers doors 2 and 3
Lever 4 triggers doors 4 and 9
...etc

With some of the levers triggering a door that is triggered by a different lever, but only one of them. (Each lever triggers 2 doors to toggle state)

There is no logical sequence between the levers and the doors. This makes it considerably more difficult to solve for intuitively, and also makes multiple possible sequences for solution, some much longer than others.

The "penalty" could be introduced by having the room with the levers slowly begin to fill with a liquid, and if not solved in time, the adventurer drowns in 7/7 water. Succeeding in opening the hallway allows the adventurer to walk the hallway, and step on the (ahem) "reset" pressure plate, which opens the drains in the entry hall, closes the flood system, and slams all the doors.

The toggle states of the levers will be all mismatched, having gone through the solution once before, giving no useful clues as to the solution.

A "reset on critical failure" should also be implemented with a pressure plate sensitive to 7/7 water, that resets the puzzle as well. That way when the adventurer drowns, the puzzle will still reset.

I might see if I can implement this.

An interesting take, but I had thought to allow the adventurer to enter the lock as it comes unlocked, finding more (mission critical) levers within... so if you'd pass through a dozen doors, each one would add 2-4 levers to the problem, some of which will help you defuse the trap or make progress, while others are red herrings or actively harmful (but might become useful later when you get another lever that opens a door but also toggles a trap that an old one toggled)

I like the idea of flooding the lever room (with automatic reset) rather than the treasure, though.
Logged
A medium-sized humanoid fond of fantasy and science-fiction.

Tevish Szat likes books, computers, board games, and cats for their aloofness. When possible, he prefers to consume hamburgers and macaroni and cheese. He needs caffeine to get through the working day.

Gentlefish

  • Bay Watcher
  • [PREFSTRING: balloon-like qualities]
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #36 on: July 23, 2013, 10:10:07 pm »

Did any of you guys play those games where you could "toggle" the color of a series of objects in a square and had to make a certain patter to win?

You can toggle gears, right?

Why not have an array like so?

Code: [Select]
{L = Lever G = Gear}
L L L L L     G G G G G

L L L L L     G G G G G

L L L L L     G G G G G

L L L L L     G G G G G

L L L L L     G G G G G

Where each "L" Corresponds to the center of a 3x3 (nxn) area on the grid, or picks its x line and y line. There'd be axles underneath the gears that all run into a master gear once all the proper gears are fitted to drive a water-activated pressure plate.

Depending on how large you make it, you could easily starve an adventurer deep underground with this and never require a reset switch so long you can remember the pattern, and you could make a "clear" switch to reset all the gears. Maybe. At least flip the pattern.

Button

  • Bay Watcher
  • Plants Specialist
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #37 on: July 31, 2013, 05:49:05 pm »

OK, after a few rounds of revision, I have finished my lock. It is minecart-based, so any challenge using this design will need to be marked as a lair to prevent item scatter.

Features:
  • Resettable.
  • Result is unknown until you pull the 'submit' lever.
  • Improper use of/attempts to hack the 'primer' and 'reset' levers may create false negatives, but will not create false positives. False positives are theoretically possible, but you'd have to have correctly entered all but one of the levers already, have a speed of 1, and also be very lucky.
  • After each lever pull (except for the 'reset' and 'primer' levers), the lever pulled is disabled until the system is reset.
  • Expandable to an arbitrary number of levers, limited only by the amount of space you have.
  • Logic components can be arbitrarily far away from the input terminal
  • No fluid logic=no need for pumps, power, or any geographical feature. (Unless you want to use them as part of the punishment for failure.)

Drawbacks:
  • Requires dfhack to export to adventure mode (minecart-based system)
  • Reset operation may cause lag
  • Large numbers of activated levers may cause lag
  • Will break if Toady fixes impulse ramps

The correct combination for this example is 23615. (Lever 4 will cause failure if pulled.)

Screenshots are using Phoebus. If you're not familiar with Phoebus, please note that the purpley-magenta bear-traps are pressure plates. All other graphics should be self-explanatory.

My solution was to model the lever combination as a finite-state automaton.

Spoiler: Logic (click to show/hide)

Since pressure plates send an 'off' signal after 99 seconds without re-activation, I needed a way to keep the consequence of success or failure active for longer than that. So I made a separate, mini-automaton for tracking only success and failure.


Allowing the user to close failure doors after opening them was a huge security hole in my original design. So I redesigned using more minecart automation!


Logged
I used to work on Modest Mod and Plant Fixes.

Always assume I'm not seriously back

Larix

  • Bay Watcher
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #38 on: July 31, 2013, 06:53:41 pm »

Oh, some nice details in this one. I particularly like the single-use (resettable) levers and the 'held' states.

A few thoughts: as i understand it, the 'submit' lever basically 'reveals' the already-determined solution (or forces failure), right? I'm aware that'd take a very different design and would likely restrict to 2^n (n=number of levers) choices, but the idea of only running the evaluation after all levers have been set up in a specific way still has a certain appeal.

What's the rationale behind putting failure doors for the next following correct lever on the path between one cell and the next? Is this just to punish too fast operation?

Couldn't the design be condensed quite a lot by linking all 'failure' levers for each passage to a single door, if you're collecting them all into a single failure pool anyway? Couldn't you replace the 'Primer' lever, too, since a reset doesn't do anything to carts on the 'primed' hatches? And couldn't you link the door latch plates of the 'lock' carts to their respective hatches as well, so that they already initialise the closing of the hatch by the time the reset is triggered?
Logged

Atomic Chicken

  • Bay Watcher
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #39 on: August 01, 2013, 06:58:33 am »

OK, after a few rounds of revision, I have finished my lock. It is minecart-based, so any challenge using this design will need to be marked as a lair to prevent item scatter.

Features:
  • Resettable.
  • Result is unknown until you pull the 'submit' lever.
  • Improper use of/attempts to hack the 'primer' and 'reset' levers may create false negatives, but will not create false positives. False positives are theoretically possible, but you'd have to have correctly entered all but one of the levers already, have a speed of 1, and also be very lucky.
  • After each lever pull (except for the 'reset' and 'primer' levers), the lever pulled is disabled until the system is reset.
  • Expandable to an arbitrary number of levers, limited only by the amount of space you have.
  • Logic components can be arbitrarily far away from the input terminal
  • No fluid logic=no need for pumps, power, or any geographical feature. (Unless you want to use them as part of the punishment for failure.)

Drawbacks:
  • Requires dfhack to export to adventure mode (minecart-based system)
  • Reset operation may cause lag
  • Large numbers of activated levers may cause lag
  • Will break if Toady fixes impulse ramps

The correct combination for this example is 23615. (Lever 4 will cause failure if pulled.)

Screenshots are using Phoebus. If you're not familiar with Phoebus, please note that the purpley-magenta bear-traps are pressure plates. All other graphics should be self-explanatory.

My solution was to model the lever combination as a finite-state automaton.

Spoiler: Logic (click to show/hide)

Since pressure plates send an 'off' signal after 99 seconds without re-activation, I needed a way to keep the consequence of success or failure active for longer than that. So I made a separate, mini-automaton for tracking only success and failure.


Allowing the user to close failure doors after opening them was a huge security hole in my original design. So I redesigned using more minecart automation!


Oh, some nice details in this one. I particularly like the single-use (resettable) levers and the 'held' states.
I agree with Larix, very nice design Button! Trying to understand all the logic made my head hurt, but it was worth it. :P Added to OP.

Spoiler (click to show/hide)
Spoiler (click to show/hide)
Should I add these concepts to the OP or wait until working designs are submitted?
« Last Edit: August 01, 2013, 07:09:04 am by Atomic Chicken »
Logged
As mentioned in the previous turn, the most exciting field of battle this year will be in the Arstotzkan capitol, with plenty of close-quarter fighting and siege warfare.  Arstotzka, accordingly, spent their design phase developing a high-altitude tactical bomber. 

Button

  • Bay Watcher
  • Plants Specialist
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #40 on: August 01, 2013, 09:22:14 pm »

A few thoughts: as i understand it, the 'submit' lever basically 'reveals' the already-determined solution (or forces failure), right? I'm aware that'd take a very different design and would likely restrict to 2^n (n=number of levers) choices, but the idea of only running the evaluation after all levers have been set up in a specific way still has a certain appeal.

Using minecarts for that kind of design would be quite simple. You'd set up a straight track with each lever intended to be in the 'on' state linked to a door, and each lever intended to be in the 'off' state linked to a raising bridge. So that if every lever was in the correct position, the cart could zoom on through; but if one of the levers was in the correct state, the straight track would be blocked and the minecart diverted into a failout.

You'd need to link each door switch up to a mechanism to delay its opening and closing, though, otherwise you could hack the system by turning each lever on, waiting a bit, and then switching it off and on very quickly. That would leave all doors open and all bridges down.

Quote
What's the rationale behind putting failure doors for the next following correct lever on the path between one cell and the next? Is this just to punish too fast operation?

Keep in mind that by the time a dwarf gets down to the combination room, the minecart is traveling at high speed. The time it would take the dwarf to step from one lever to another is less than the time it takes for the minecart to travel from one state to another. So the only way the failure door will be open, is if it was pulled before the correct lever was.

Quote
Couldn't the design be condensed quite a lot by linking all 'failure' levers for each passage to a single door, if you're collecting them all into a single failure pool anyway?

Yep! I think I mentioned it in my post, actually. If I were to do it again, I would definitely link all the failure levers for each transition to a single door. The one-door-per-failure-lever design was left over from an earlier version of the design, and I didn't want to completely re-write the automaton just because it was a little more complex than it had to be. (I already had a few failed early versions scattered around the fortress!)

Quote
Couldn't you replace the 'Primer' lever, too, since a reset doesn't do anything to carts on the 'primed' hatches? And couldn't you link the door latch plates of the 'lock' carts to their respective hatches as well, so that they already initialise the closing of the hatch by the time the reset is triggered?

I did originally plan to have the lever reset batteries not use primer hatches, but it turned out to be a lot of work. The problem is that it takes a pressure plate 99 ticks before it sends its 'off' signal, in which time a minecart moving at max speed can move 267 tiles. Then there was the problem of minecarts moving at different speeds based on how long they had been active and whether or not their hatches were open at the time of reset, which made derail-based looping unpredictable. Ultimately I decided it was easier to stop the carts completely, and have the user be responsible for getting them started again.

It's possible, but you'd have to be better at minecarts than I am to do it. :)
Logged
I used to work on Modest Mod and Plant Fixes.

Always assume I'm not seriously back

Larix

  • Bay Watcher
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #41 on: August 02, 2013, 04:25:34 am »

A few thoughts: as i understand it, the 'submit' lever basically 'reveals' the already-determined solution (or forces failure), right? I'm aware that'd take a very different design and would likely restrict to 2^n (n=number of levers) choices, but the idea of only running the evaluation after all levers have been set up in a specific way still has a certain appeal.

Using minecarts for that kind of design would be quite simple. You'd set up a straight track with each lever intended to be in the 'on' state linked to a door, and each lever intended to be in the 'off' state linked to a raising bridge. So that if every lever was in the correct position, the cart could zoom on through; but if one of the levers was in the correct state, the straight track would be blocked and the minecart diverted into a failout.

You'd need to link each door switch up to a mechanism to delay its opening and closing, though, otherwise you could hack the system by turning each lever on, waiting a bit, and then switching it off and on very quickly. That would leave all doors open and all bridges down.

With hatches and doors, 'jamming' the system is not possible, since they react instantly to any signal. A lock can be built with only such devices, but you wouldn't use a straight path for the correct solution, you would need to use one where the cart's supposed to go straight in some places but around corners in others. Unless the lock combination was 11111111, of course, but we're not trying to emulate Minuteman missile locks, do we?

Quote
Quote
What's the rationale behind putting failure doors for the next following correct lever on the path between one cell and the next? Is this just to punish too fast operation?

I looked at the basic logic again and realised i had mentally put it upside-down. I hadn't properly taken into account that all the possible failures for a sequence step are only consulted _after_ the correct lever for the step is pulled. Of course, with this mode of operation, a failure must be defined, say, for lever three between cells two and three, else you could get away with pulling lever three before two.

Quote
Quote
Couldn't the design be condensed quite a lot by linking all 'failure' levers for each passage to a single door, if you're collecting them all into a single failure pool anyway?

Yep! I think I mentioned it in my post, actually.

Oops, yes you did. I must've had lost the thread of your argument there for a moment. Sorry.

Quote
I didn't want to completely re-write the automaton just because it was a little more complex than it had to be. (I already had a few failed early versions scattered around the fortress!)

Oh yes, the botched designs and the testing grounds. I have three z-levels of one fort filled with nothing but minecart logic testing rigs. Not for the combination locks, but i spent quite a lot of dwarftime carving, engraving, installing and testing devices for my dwarven precision clock there.

Hmm, i'm actually even more impressed now, considering the entire lock is based on derail-speed minecarts. My ramp-based designs are by definition at less than derail speed, which removes many of the problems outright that you had to deal with - if a 'slow' minecart has a fitting and unique track exit from a tile available, it will follow the track without issue. Getting a derailing minecart around a corner takes a _bit_ more effort and space.

And what i failed to mention before - sending an extra 'open' to the hatch to force it closed via the paired nature of signals is a pretty clever trick. I'd been aware of the principle but managed to overlook a good many of its possible applications.
Logged

fricy

  • Bay Watcher
  • [DFHACK:ZEALOT]
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #42 on: August 02, 2013, 04:40:55 am »

Quote from: Larix
With hatches and doors, 'jamming' the system is not possible, since they react instantly to any signal.

Maybe it's not what you'd want, but it's possible to delay the signal until another lever is pulled by using power-to-signal converters.
The "puzzle-levers" switch  the power-in gears of these signal converters, an "evaluation lever" provides power to these gears, and the power-to-signal converters switch the maze path. These can be water, or minecart based, and also there's a compact solution in DFhack called power-meter, which is a modified pressure plate that detects power in the gears built next to it.

Ravendarksky

  • Bay Watcher
    • View Profile
    • DFMon.exe - Get rid of DF SPAM
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #43 on: August 02, 2013, 05:43:42 am »

I don't suppose it helps anyone but I came up with the following piece of machinery which may be helpful for this (and no cheaty impulse ramps). I discovered that minecarts will reliably push another cart, plus themselves if left on top of turned on rollers!

Square 1: EW track, Max speed rollers, minecart set to forbid
Square 2: EW track, Pressure plate, minecart set to forbid
Square 3: Vertical bars/floodgate linked to whatever you want
THEN a U bend (don't make square 4 part of the U bend or the rollers will cause derail)
Square 4: EW track, max speed rollers
Square 5: EW track
Square 6: Vertical bars/floodgate linked to whatever you want
THEN a U bend back to square 1.

This gives you a switchable permanent ON/OFF. Basically a Pressure plate lever, where toggling square 3 sends an off signal and toggling square 6 sends an on signal. You can easily make this two opposing switches by simply adding a pressure plate to square 5.

Also a quite aesthetically pleasing (and simple) signal generator if you link up loads of these in a row and have the pressure plates set to open the gate just in front of them.

Here is a screenshot showing both:
 
Logged

Larix

  • Bay Watcher
    • View Profile
Re: [CHALLENGE]Dwarven Combination Locks
« Reply #44 on: August 08, 2013, 02:59:04 pm »

I fiddled with 'evaluation' concepts, but it got completely out of hand:

     

Spoiler (click to show/hide)

Only three 'lock levers' are used, but three separate evaluations must be run, and for each of them, a correct three-lever combination must be set. Failure will only become apparent after the second or third evaluation, granting bountiful gifts of water, magma or both. The system can be reset. There are three correct 'keys':
000-010-111
011-010-111
001-001-100
Two different 'solution paths' were installed specifically to allow different correct keys.

EDIT: this is not an official entry, since it went way beyond the specifications and is so obscenely byzanthine that i can't properly explain it. It works exactly as intended, although at the cost of quite a lot of debugging.
Self-imposed design targets were:
-use more than one cart, making the processing partially sequential and partially simultaneous, using weight-sensitive pressure plates, all that while keeping them apart and without the need for additional 'regulator' carts.
-keep the number of levers and direct links from the levers themselves low. The controls consist of three levers for the lock itself and two 'evaluation' levers, necessary because i wanted to run _three_ evaluations. The first lock lever has only one real connection, the other two have three each; i added two dummy links to the first lever. All other processing is done through signals created within the lock itself.
-allow more than one correct combination, if possible with meaningful differences in keys for every solution triplet. Effectively, there are only two solutions, because the processing can't differentiate between '000' and '011' on the first and second evaluation, which would mean either excluding them or allowing them as a freely interchangeable pair of correct combinations.
-all combinations must result either in success or failure and every finish combination must be returnable to the normal start configuration through reset. The reset lever is consequentially linked to eight hatches. In the current configuration, a magma or obsidian failure would destroy the control area (marble levers aren't magmaproof and you'd have to disable the pumps _behind_ the lock or their power supply upstream from the magma door before getting anywhere near them anyway), so the reset ability is ...limited. Disconnecting the power train before trying the lock would limit but not remove the impact of failure, all the water already in the pipeline would still get in and a good helping of magma would slowly seep through the door, quick enough to cut off the exit.
-no hacking through fast operation or the like - enforced through separation between evaluation levers and built-in delay between execution of the second and third evaluation.
-any amount of lever-pulling allowed between evaluation actions without harm; that was easy, it just required keeping the lock-operated and evaluation-operated events separate.
« Last Edit: August 11, 2013, 04:28:43 am by Larix »
Logged
Pages: 1 2 [3] 4