There is a contradiction. I changed the code to not push the first finish molecule forward, and instead to let go of it sooner. Even though it doesn't push it forward, it is still within the goal area with no manipulator holding it. The infringing molecule has a manipulator holding it at that moment.
Mention that in the freaking post then. I'm not psychic. EDIT: It's still not a contradiction. The game sees the finished molecule after the last turn of it but before dropping, so while it is correct at that moment, the molecule can't be taken away because it's still held. Then it's dropped and the next molecule's salt also gets into the goal area. The game sees an incorrect solution after that cycle, just like we can see, thus it does not remove the correct molecule. Like I said, it's a dumb check. It just looks at the elements that are in the goal area and compares them to the correct order, and apparently it doesn't ignore held molecules. Thanks for finding that one out.
However, the check for victory does not happen until the end of the loop.
Yes, that's what I've been saying for several posts already. It checks on cycle change.
I think a good way to remedy this would be pretty simple. Since a drop action must precede a score, then checks for score should be made at the very beginning of a drop action, rather than a the end of it. This makes the graphics of the game match the reality, and also allow for more efficient and clever designs that bring the next molecule right up the goal area as the previous molecule is disappearing.
It would require a change in the rules and limitations of the game. The developer could've done that, but instead he added the limitation that victory checks are only done on a cycle change. Doing it otherwise would add more complexity to the game on all sides.
I like the way I got Litharge to come out: b,10,10;b,3,10;m,7,10,90,2,wcrw22rror;m,1,10,90,2,cdlow lllu;m,9,6,90,3,crroll;m,13,10,90,2,wwwcdroluw;
But my Oil of Vitriol is a lot clunkier than the others I've seen on here. Oh well, here it is anyways: b,7,8;t,5,8;c,1,11;t,3,11;m,3,8,0,1,wcuuodd;m,11,13,0,1,wcluodrwww;m,1,9,90,3,cddouuw ww;m,11,8,90,1,wwwwcuwrodl;m,8,6,90,1,wwwcllorrww;m,5,11,180,3,jwwcddruolu;
Philosopher's Stone with 104 symbols in 100 cycles: c,1,9;b,3,6;c,13,9;b,10,11;b,3,11;b,11,6;b,8,6;m,1,11,90,1,crrrorclorcrrror; m,13,11,90,1,clllolcrolclllol;m,6,11,180,1,oucd;m,8,11,0,1,oucd;m,12,4,90,1,orcl; m,6,8,90,2,crol;m,8,8,90,2,clor;m,10,1,90,2,crow;m,11,8,270,1,wwwc2lor; m,6,6,0,1,rrcl11lo;m,8,10,270,1,u11uoddc;m,7,8,270,3,1oddcu1u;m,1,5,90,3,cddouurr; m,1,6,270,2,rorc;m,3,2,0,2,dcuo;
The number of cycles could really be cut down if I calcified from the Air and Earth supplies too, but I'm just happy to have a design that works tonight.
EDIT: Holy FSM, has anyone else seen XV: ????? That's a nightmare and a half.
FINISHED! And it's awesome, 72 symbols, 40 cycles, Brimstone: ... Anyone out there who wants to try and do it faster or with less symbols?
How about both, 48 symbols, 37 cycles c,1,11;t,3,9;t,5,11;b,7,9;m,8,11,180,1,rolucd;m,11,9,90,3,wwwwwjcrdolu;m,3,11,90,1,crrror;m,1,13,0,1,wwclor;m,1,9, 0,1,orwwcl;m,5,9,180,1,clloll;m,8,6,270,2,orrcll;
[And a slightly more efficient version of nickname change's. t,9,11;c,7,13;c,11,9;t,7,9;b,3,6;m,7,11,270,1,orrcll;m,9,13,0,1,cllorr;m,11,11,270,1,llcrro;m,9,9,0,1,rrollc;m,4,9,0,2, clwowr;m,7,6,180,2,olcwwr;m,1,6,0,1,owwc11;m,3,2,90,3,cl1owr;m,3,10,180,2,cruodl;m,0,13,270,2,orclww;]
Finally finished Aqua Fortis: t,5,9;c,11,11;t,7,7;b,10,9;c,3,11;b,9,7;m,9,5,90,1,wwwjcllorr;m,13,9,90,3, cdroluww;m,10,5,90,3,wcddouuw;m,11,13,0,1,wjcluuoddr;m,7,9,180,1,crolwww; m,5,11,180,1,wjcrolwww;m,3,13,180,1,crolwww;m,5,7,0,1,wcuuoddw;
67 symbols and 50 cycles. It's got fewer cycles than Nickname Change's solution, but the problem I keep having is that all my designs look like theyr'e destined to fail eventually, where as the others that have been posted look like they could run forever.
67 symbols and 50 cycles. It's got fewer cycles than Nickname Change's solution, but the problem I keep having is that all my designs look like theyr'e destined to fail eventually, where as the others that have been posted look like they could run forever.
I wonder about that too. The way I solve the problem is I make a machine that creates infinite copies of the required compound by keeping all the parts in sync. However, it's very possible to make a machine that goes out of sync, and will eventually break, but will still produce enough of the compound to beat the level.
I think in this case it is best to quote Ferdinand Porsche. I think the quote goes something like this.
"The best built race car would fall apart immediately after crossing the finish line."
Black Powder 50 symbols and 63 cycles b,4,11;t,7,9;t,4,6;m,6,6,0,2,l2orrwwwcl;m,4,9,90,1,clll11rror;m,1,11,90,1,cluodrwwww;m,7,11,180,1,olwcr;m,5,13,0,1, wclor;m,4,13,270,1,owwc2w2ww2;
67 symbols and 50 cycles. It's got fewer cycles than Nickname Change's solution, but the problem I keep having is that all my designs look like theyr'e destined to fail eventually, where as the others that have been posted look like they could run forever.
I wonder about that too. The way I solve the problem is I make a machine that creates infinite copies of the required compound by keeping all the parts in sync. However, it's very possible to make a machine that goes out of sync, and will eventually break, but will still produce enough of the compound to beat the level.
I think in this case it is best to quote Ferdinand Porsche. I think the quote goes something like this.
"The best built race car would fall apart immediately after crossing the finish line."
I have a feeling that the best way to optimize the later levels is to make a machine that goes out of sync after some point.
Oh my god, I thought it to not be possible in less than 40 cycles. My design is still more awesome! Yay for awesome, quick, symmetrical(ish), dual feeder. I knew less symbols would be easy to do, the main point was the cycles, good job Zarlack, wonderful job.
67 symbols and 50 cycles. It's got fewer cycles than Nickname Change's solution, but the problem I keep having is that all my designs look like theyr'e destined to fail eventually, where as the others that have been posted look like they could run forever.
I start designing with phases in my head, 3 steps, 4 steps, and then reset in the same phase length. Then it's just a lot of optimizing of symbol order. All these sort of games are the same for me. I'm a sucker for nice looking designs. An example, in the case of that Bureau of Steam Engineering game. I prefer to use a 3 length pipe and 2 length pipe instead of a 4 and 1. 1 length pipes are ugly. And what the heck does that walking system do there in my way?! Makes awesome dual-shooting, unbeatable (if you can time ducks), simple design look ugly because I need to put the control valve one square lower.
Philosopher's Stone 88 symbols, 80 cycles: b,3,11;b,10,11;b,11,7;b,11,5;b,3,9;c,3,5;c,3,3;m,1,11,90,1,clor;m,6,11,180,1,oucd;m,8,7,90,3,lorc;m,8,11,0,1,ucdo; m,13,11,90,1,rolc;m,6,9,90,1,crol;m,7,7,0,2,ollcrr;m,11,1,0,1,ruuolddwwwwc;m,11,9,270,1,ouucdd; m,8,5,270,2,rdwwwwcullor;m,1,9,0,1,lcro;m,1,4,90,2,rrclwwwwwlwo;m,7,5,180,1,ruulrw1olddc; (It works, but sometimes the collision detector thinks that the square hits the upper left 'Spawn' on the swing back.)
Philosopher's Stone 88 symbols, 80 cycles: b,3,11;b,10,11;b,11,7;b,11,5;b,3,9;c,3,5;c,3,3;m,1,11,90,1,clor;m,6,11,180,1,oucd;m,8,7,90,3,lorc;m,8,11,0,1,ucdo; m,13,11,90,1,rolc;m,6,9,90,1,crol;m,7,7,0,2,ollcrr;m,11,1,0,1,ruuolddwwwwc;m,11,9,270,1,ouucdd; m,8,5,270,2,rdwwwwcullor;m,1,9,0,1,lcro;m,1,4,90,2,rrclwwwwwlwo;m,7,5,180,1,ruulrw1olddc; (It works, but sometimes the collision detector thinks that the square hits the upper left 'Spawn' on the swing back.)
I've gotten collisions every time I've run your solution.
On slower machines the app will not run entirely in sync. I ran into a situation where a specific part of my construction would work just fine but after adding a few more components it would start to collide. I had to switch from a six cycle to a nine cycle to get rid of the problem.
I fixed my Philosopher's Stone (and its has better numbers). Symbols 76, Cycles 73 b,3,11;b,10,11;b,11,7;b,11,5;b,3,9;c,3,5;c,3,3;c,6,11;c,4,7;m,1,11,90,1,clor;m,8,7,90,3,lorc;m,8,11,0,1,ucdo; m,13,11,90,1,rolc;m,6,9,90,1,crol;m,7,7,0,2,ollcrr;m,11,1,0,1,ruuolddwwwwc;m,11,9,270,1,ouucdd;m,1,9,0,1,lcro; m,1,4,90,2,rrclwwwwwlwo;m,0,11,0,3,odcu;m,8,5,270,3,odrwwwcullr1;
The other thing to think about is number of components. The game does not give you a count, but I would think ideally a solution should have as few parts (arms and glyphs) as possible.
Result: 64 symbols, 35 cycles. So, I used an extremely long pipeline, which now uses far more symbols, but is only 2 cycles faster.
EDIT: Incidentally, on that basis, I think the best benchmark is therefore symbols * cycles. Why? because this equates to the amount of work done per unit energy/cost, and hence efficiency. Go ahead and try to make a more efficient design (by those criteria) than any of mine. EDIT2: This doesn't make sense anymore; no. of manipulators makes more sense than symbols.
Incidentally, on that basis, I think the best benchmark is thereforesymbols * cycles.
I propose that components * (cycles/unit assembled) is the best. It eliminates cycles used for initial setup and doesn't regard "programming" as a resource.
That might be a good idea Timo. So the benchmark is, nominally speaking, "long-term production efficiency". i.e. setup time costs and initial component costs are neglected, and so the only costs are the operating costs of the manipulators.
On that basis, here are the best scorers so far (mine, except where I say otherwise): I) Aqua Vitae: 6 (1 manipulator, 6 cycles/unit)
Very nice. I would include the non-manipulator components in the component count tough. It gets fairly interesting when you try to make do with one calcinator on Philosophers Stone ;-).
OK, I'm not really fond of this rather wasteful design but since no one else has posted anything I'll assume the ???? broke other peoples' brains as well.
I am a little proud of the 12 cycles/unit as 10cycles/unit is the theoretical maximum and that would require 5 and 10 cycles for sub-parts of the machinery instead of the 6 and 12 cycles I used. The shorter cycles would increase the number of manipulators and the faster pace would make race conditions even worse than they are in this design.
Comments
EDIT: It's still not a contradiction. The game sees the finished molecule after the last turn of it but before dropping, so while it is correct at that moment, the molecule can't be taken away because it's still held. Then it's dropped and the next molecule's salt also gets into the goal area. The game sees an incorrect solution after that cycle, just like we can see, thus it does not remove the correct molecule. Like I said, it's a dumb check. It just looks at the elements that are in the goal area and compares them to the correct order, and apparently it doesn't ignore held molecules. Thanks for finding that one out. Yes, that's what I've been saying for several posts already. It checks on cycle change. It would require a change in the rules and limitations of the game. The developer could've done that, but instead he added the limitation that victory checks are only done on a cycle change. Doing it otherwise would add more complexity to the game on all sides. NICE! 38 symbols.
b,10,10;b,3,10;m,7,10,90,2,wcrw22rror;m,1,10,90,2,cdlow lllu;m,9,6,90,3,crroll;m,13,10,90,2,wwwcdroluw;
But my Oil of Vitriol is a lot clunkier than the others I've seen on here. Oh well, here it is anyways:
b,7,8;t,5,8;c,1,11;t,3,11;m,3,8,0,1,wcuuodd;m,11,13,0,1,wcluodrwww;m,1,9,90,3,cddouuw ww;m,11,8,90,1,wwwwcuwrodl;m,8,6,90,1,wwwcllorrww;m,5,11,180,3,jwwcddruolu;
c,1,9;b,3,6;c,13,9;b,10,11;b,3,11;b,11,6;b,8,6;m,1,11,90,1,crrrorclorcrrror;
m,13,11,90,1,clllolcrolclllol;m,6,11,180,1,oucd;m,8,11,0,1,oucd;m,12,4,90,1,orcl;
m,6,8,90,2,crol;m,8,8,90,2,clor;m,10,1,90,2,crow;m,11,8,270,1,wwwc2lor;
m,6,6,0,1,rrcl11lo;m,8,10,270,1,u11uoddc;m,7,8,270,3,1oddcu1u;m,1,5,90,3,cddouurr;
m,1,6,270,2,rorc;m,3,2,0,2,dcuo;
The number of cycles could really be cut down if I calcified from the Air and Earth supplies too, but I'm just happy to have a design that works tonight.
EDIT: Holy FSM, has anyone else seen XV: ????? That's a nightmare and a half.
48 symbols, 37 cycles
c,1,11;t,3,9;t,5,11;b,7,9;m,8,11,180,1,rolucd;m,11,9,90,3,wwwwwjcrdolu;m,3,11,90,1,crrror;m,1,13,0,1,wwclor;m,1,9,
0,1,orwwcl;m,5,9,180,1,clloll;m,8,6,270,2,orrcll;
[And a slightly more efficient version of nickname change's.
t,9,11;c,7,13;c,11,9;t,7,9;b,3,6;m,7,11,270,1,orrcll;m,9,13,0,1,cllorr;m,11,11,270,1,llcrro;m,9,9,0,1,rrollc;m,4,9,0,2,
clwowr;m,7,6,180,2,olcwwr;m,1,6,0,1,owwc11;m,3,2,90,3,cl1owr;m,3,10,180,2,cruodl;m,0,13,270,2,orclww;]
the dude that made this game also made a steampunk themed game:
The Bureau of Steam Engineering
t,5,9;c,11,11;t,7,7;b,10,9;c,3,11;b,9,7;m,9,5,90,1,wwwjcllorr;m,13,9,90,3,
cdroluww;m,10,5,90,3,wcddouuw;m,11,13,0,1,wjcluuoddr;m,7,9,180,1,crolwww;
m,5,11,180,1,wjcrolwww;m,3,13,180,1,crolwww;m,5,7,0,1,wcuuoddw;
67 symbols and 50 cycles. It's got fewer cycles than Nickname Change's solution, but the problem I keep having is that all my designs look like theyr'e destined to fail eventually, where as the others that have been posted look like they could run forever.
I think in this case it is best to quote Ferdinand Porsche. I think the quote goes something like this.
"The best built race car would fall apart immediately after crossing the finish line."
50 symbols and 63 cycles
b,4,11;t,7,9;t,4,6;m,6,6,0,2,l2orrwwwcl;m,4,9,90,1,clll11rror;m,1,11,90,1,cluodrwwww;m,7,11,180,1,olwcr;m,5,13,0,1,
wclor;m,4,13,270,1,owwc2w2ww2;
b,1,5;c,1,7;c,1,9;b,3,9;b,3,11;b,12,5;b,6,7;b,6,9;c,12,7;c,10,7;
m,4,5,0,2,lljwcdouw;m,1,3,0,2,dljclloll;m,10,5,0,2,dolcur;m,10,1,0,2,crdoul;
m,10,13,0,2,wwwcrroww;m,4,13,0,2,wwcluordw;m,1,11,0,2,wwwwwjowwwldclu;
m,3,7,0,2,djlcl1lw1lo;m,10,9,0,2,ljddluoruwc;m,7,5,0,2,rc1rr11or;
88 symbols, 80 cycles:
b,3,11;b,10,11;b,11,7;b,11,5;b,3,9;c,3,5;c,3,3;m,1,11,90,1,clor;m,6,11,180,1,oucd;m,8,7,90,3,lorc;m,8,11,0,1,ucdo;
m,13,11,90,1,rolc;m,6,9,90,1,crol;m,7,7,0,2,ollcrr;m,11,1,0,1,ruuolddwwwwc;m,11,9,270,1,ouucdd;
m,8,5,270,2,rdwwwwcullor;m,1,9,0,1,lcro;m,1,4,90,2,rrclwwwwwlwo;m,7,5,180,1,ruulrw1olddc;
(It works, but sometimes the collision detector thinks that the square hits the upper left 'Spawn' on the swing back.)
My Oil of Vitriol is nothing short of epic: 24 symbols, 35 cycles
Symbols 76, Cycles 73
b,3,11;b,10,11;b,11,7;b,11,5;b,3,9;c,3,5;c,3,3;c,6,11;c,4,7;m,1,11,90,1,clor;m,8,7,90,3,lorc;m,8,11,0,1,ucdo;
m,13,11,90,1,rolc;m,6,9,90,1,crol;m,7,7,0,2,ollcrr;m,11,1,0,1,ruuolddwwwwc;m,11,9,270,1,ouucdd;m,1,9,0,1,lcro;
m,1,4,90,2,rrclwwwwwlwo;m,0,11,0,3,odcu;m,8,5,270,3,odrwwwcullr1;
Aqua Fortis
42 symbols, 38 cycles Haematite
44 symbols, 46 cycles
My first solution - 42 symbols, 37 cycles (the best so far) Then I tried it Pentium 4-style - a long long pipeline: Result: 64 symbols, 35 cycles.
So, I used an extremely long pipeline, which now uses far more symbols, but is only 2 cycles faster.
EDIT:
Incidentally, on that basis, I think the best benchmark is therefore symbols * cycles.
Why? because this equates to the amount of work done per unit energy/cost, and hence efficiency.
Go ahead and try to make a more efficient design (by those criteria) than any of mine.
EDIT2:
This doesn't make sense anymore; no. of manipulators makes more sense than symbols.
i.e. setup time costs and initial component costs are neglected, and so the only costs are the operating costs of the manipulators.
On that basis, here are the best scorers so far (mine, except where I say otherwise):
I) Aqua Vitae: 6 (1 manipulator, 6 cycles/unit) II) Sal Ammoniac: 18 (3 manipulators, 6 cycles/unit) (by nickname change) III) Cinnabar: 18 (3 manipulators, 6 cycles/unit) IV) Litharge: 24 (3 manipulators, 8 cycles/unit) V) Oil of Vitriol: 24 (4 manipulators, 6 cycles/unit) VI) Aqua Fortis: 42 (7 manipulators, 6 cycles/unit) VII) Haematite: 48 (6 manipulators, 8 cycles/unit) VIII) Brimstone: 42 (7 manipulators, 6 cycles/unit)
p,3,10;p,10,11;b,8,7;t,1,7;t,13,7;b,5,7;t,3,5;b,12,5;m,8,5,90,3,cddouu;m,9,9,0,1,lorwwc;m,8,11,270,1,owrclw;
m,10,13,270,1,lucdro;m,3,13,0,3,cldoul;m,13,9,180,1,rwwclo;m,13,11,180,1,owwlcr;m,11,13,0,1,wwclor;
m,13,10,90,2,jllollc;m,6,10,270,2,wowrcr;m,1,10,90,2,clor;m,5,5,180,1,lorc;m,1,5,0,1,rclo;
m,1,11,90,1,ddcrruuollww;m,3,7,0,2,wcd1ouucduod;m,11,7,180,1,rolc2uuoddcw;m,7,10,270,2,uoddcu;
m,10,5,180,3,wwcrddroluul;m,9,5,0,2,dcuuodwwwwww;m,13,1,180,3,dwwlc21wr2uo;
151 symbols, 84 cycles
or
20 manipulators * 12 cycles/unit = 240.
I am a little proud of the 12 cycles/unit as 10cycles/unit is the theoretical maximum and that would require 5 and 10 cycles for sub-parts of the machinery instead of the 6 and 12 cycles I used. The shorter cycles would increase the number of manipulators and the faster pace would make race conditions even worse than they are in this design.