2016 Spacechem Summer Challange

[etammate95]

Welcome to this small event!

I started this event, to ease the lack of tournament this year. I hope i can set a new tradition, and we will see more small tournament, or even a long one.

The event starts on Friday(07.08), I will post 2 similar, but different puzzle to solve. It will last for 11 days (so you got two weekends). It ends on monday(07.18) midnight (CET).

Send your solutions here: [e-mail]
You have to include the number of atoms you used, the solution, and i would be glad if you explained how it works. UPDATE: send me wich puzzle is the base of your solution, if you altered the puzzle send me the puzzle too. (use export)

I accept screenshots, but Savechem [www.coranac.com] would be better. (its much faster so please send me the code). I will make a video with the solutions, and write notes about the interesting ones.

The rules, puzzles, solutions will be posted under this discussion.

Have fun!



Ps/disclaimer:
I will update the placeholder comments later.
If turns out my puzzles have well know solutions, or they have been in older tournament, i have a spare puzzle. But that can mean the cancellation of the event.
Im not a native speaker, so if something is not understandable, ask me politely about it.
My puzzles arent like in the 2015 Tournament, the optimisation is not the main goal. But i hope, you will like it.

Change log:07/07: Updated rules and added some inroductions.
07/08: puzzles uploaded. Good luck.
07/09: Updated infos after reading the questions. Tripler and quadrupler added.
07/12: added "AS 42" with 42 atom, i hope its enough.
07/14: update: please send me the puzzle too.
07/16: new Q/A. Make sure you have read the rules!
07/21: tables uploaded with the scores.
07/28: solutions are avaible, commentar in the last placeholder comment

[etammate95]

RULES:
-General forum etiquette
-Your solution has to be your own work, and you have to send it me while the event lasts.
-The discussion of the solutions are prohibited. You can make suggestion like in the previous tournament about the possible scores, but use the spoiler tags like this: its possible to go under 1000 cycle, but its hard


RULES FOR PUZZLES:
moved to puzzles. lol

[etammate95]

Introduction, random thoughts:

About the puzzle: It will involve a random inputs like my puzzle for 2014 tournament. The goal will be to use the space and the sensor to the max. I hope new players will find interesting to experiment with new things, and veterans will do their best to push the limits. The two puzzles will be similar, but with different goals. I made two very different solutions, so i hope you will see the difference and make awesome solutions.

About the randomness: The solution has to be independent from the input order. I will test it, but i wont be able to test all of them. Also this means there is no need to optimise for speed.

If everything go as planned i will uploed them on Friday at 20:00 CET. I would like to ask the veterans to chek them out, and message me immideitly if the puzzles has known solutions.

[etammate95]

***Read the whole post! Rules after description!***


Your job is to make the max out of a single sensor. How many atoms can you identify with it?

I coded different number of atoms into the sandbox. The number in the puzzles name indicates, how many atoms will it input. The first sequence is reversed, then an ascending, after that randoms are repeated. Obviously i couldnt put every possible sequence in the sandbox, but your solutions has to work with all of them.
There is a plus atom on input A. Dont use it! Its for more input. Description below this comment.

Copy the puzzle and add it as a custom assignment.
Here are the puzles: on my drive [drive.google.com]
(you dont need to solve all, just the best you can make)

What do i mean by identifying?

Puzzle #1 Atom Sequencer
Your job is to arrange atoms in ascending order.
The sandbox has a random input starting from hydrogen and going up. You can use only 1 reactor! You have to put the ascending sequence into a output printer. So the printer goes like: H, He, Li, Be, B, C, N, O, H, He…
Scoring: The more atoms your sequencer can manage, the better. Then the fewer symbols win.

Puzzle #2 Atom Organizer
Its similar, but you have to output them in one single block! The atoms have to be ALWAYS at the same place! Make them in a row or spiral, ideas on my drive [drive.google.com]
For this i programmed the atoms from scandium upwards. The transition metals have a lot of bond, so there is no problem bonding them together.
Scoring: The more atoms your sequencer can manage, the better. Then the fewer symbols win. ( so its the same)

Rules: (READ THEM, IMPORTANT)
-It has to fulfill the criteria
-only one reactor
-you CANT use fusion/fission!
-it has to run infinitely without crashing
-you cant use the recycler
-you cant use the pseudo-random input, it has to work with every input-orders. I will test if the solution is independent from the input order.( the ratio is fixed btw)
-no telekinesis, bugs, and other hard glitches. Like last year
-no branch-control!

But you are free to:
-Make your own shape for #2. But it has to be always the same.
-Choose the input place! If its better for you to input the atoms in the corner you can do it. (but then you have to change all atoms!)

[etammate95]

Thoughts and tips:

You can solve the puzzle with more then 8 atoms? Good. Do the following: Have a second reactor(this wont count in the symbols). Just transfer the first sequence, then fuse one atom from input A to the second sequence. You doubled your atoms! Like this: pic on my drive [drive.google.com]
Delete some of the syncs to make it work with fewer numbers.
(There is now a tripler for 8, revire the flip flops to get less. Also a quadrupler for 7, add or delete the syncs to get different numbers.)
Instead this use AS 42 to get 42 atoms. I put the atoms in a big chunk, then you have to break it down in a reactor. ( there was a tourament puzzle with this idea.)

Every input has 7 atoms, two will be fused to get noble gasses.
It makes now 42, you can reduce it by deleting atoms from the last input. Watch out for the fusion, if it gets you a non-noble gas, may atoms gets stick together.



The puzzles may look very similar, but i came up with very different solution.
I hope newer player will experiment with new ideas, and think outside the box. And the veterans will use their knowledge to push their limits. Good luck!



You stuck and you have no idea? Check out my old puzzle for the 2014 tournament. http://steamcommunity.com/app/92800/discussions/0/522730700071141331/?ctp=52#c617319460828372087

[etammate95]

Q and A
I will copy here my answers so you dont have read the whole discussion to get easily some answers:

Originally posted by ToughThought:

A few questions:

1. May we build a solution that handles input streams containing N different input atoms, without having a solution for fewer than N?

2. Why is argon included as the A input in the organizer 5 through 8 puzzles? (Fusing argon with scandium, for example, does not seem to produce a usable atom if N is less than 18.)

3. Am I correct to think that branch-control symbols are not allowed?

4. I notice that using a second reactor in the form of your image "how to double the input for 8 atoms.png" would affect not only which atoms will arrive, but also how fast they can appear. In other words, it would serve to slow the input stream. Are we permitted to modify the second reactor to fashion any desired input speed? Are we permitted to use a second reactor solely to alter the input speed, even without affecting the input atoms?

1. Yes. I would like to get your best solutions. I wont have time to validate all, if you send me a dozen solutions. Also most of the solution can be downgraded.

2. If you fuse argon into the stream, you get the second row of the transition metals. This way you can make nice structures, no noble gasses.

3. Yes

4. Stupid limits of the programmed input....
I didnt thought this would be a problem. I think this would make some nasty solution so no. I will make official versions. THe identifying should be done in one reactor only.

Originally posted by 12345ieee:

Queston: puzzle 2, the output positions must be always the same: it's also true for the number of bonds between the atoms?

Yeah, its acceptible, just make sure the atoms are always at the same place, and its in one piece.

[etammate95]

Well the event wasnt so polular, but you sent me very good solutions. The goal was to think outside the box: use a pipe that feeds back to its own reactor, and the communication between red and blue with the sensors help. I think it was succesful, congratulation to everyone!

Solutions #1

1. Toughtought: 40-229
   
2. 12345ieee 39-204
   
3. gggol 36-185
   
4. rexkix 4-78

(not on the list: Toughtought: 41-237 arrived a bit late, G Wen: 8-71 disqualified for using fusion)

Solutions#2

1. Toughtought: 16-69
   
2. gggol 16-102
   
3. 12345ieee 16-121

[etammate95]

Savefile is avaible for download on my drive [drive.google.com]. (name: 2016SummerChallange.user the second file from top. To view it, you have to change its name so spacechem can recognise it)
It contains 7 tournament solution, 2 bonus solution from ToughTought for 41 and 42, and my very first solution for 8 atoms (its good for understanding the basic method for the new comers)


Commentar for the sollutions:
My main goal was to achive 'communication' between the blue and red. Every solution had this so i was happy with the results. The main mechanism was the following: red inputs N atom and puts them on a pipe that feeds its own reactor. Then grabs one atom and puts on a sensor at the output zone. The blue is sitting on a sensor- branch, if the said atom is good, it outputs it, and goes a bit forward. If it doesnt happen red puts it back to the pipe. There was some variations:
-gggol made the best blue maze, with reusing the outputsign it can manage 34 atoms. The red outputs 2 more by going around until everything has been output even in the worst case.
-12345ieee has made a unic solution what i did not expect. In every solutios blue outputs the sensed atom if its the next one. But here its the opposite: if its the next one, blue leaves it there so red can grab it. This allowed him to reduce the repaeting subunit of the blue path down to 2 sign.
-ToughTought: Its similar to gggols solution but red starts with the outputing, with the worst-case-senario method. The blue is sitting on a sync, while red works. This allowed him to make the best solution.

The Organizer solutions are similar, the blue bonds the atoms instead outputting them. ToughTought made an interesting trick with the flipflops.

[12345ieee]

Just a note, since you are using a sandbox production, please, hardcode the whole random sequence, so that we have the same puzzle across all the OSes.

[etammate95]

Originally posted by 12345ieee:

Just a note, since you are using a sandbox production, please, hardcode the whole random sequence, so that we have the same puzzle across all the OSes.

I added some explanations in rules and in the introductions. Dont worry i will code them, but you still have to make sure its independent from the actuall order.

[etammate95]

Yes, im late with the puzzles... Sorry

[etammate95]

Its complete i hope, good luck!
Please write a comment if you plan on taking part in the challange, so i know how many solutions i should expect.
Also if something wrong message me immidietly.

[rezoons]

I will try to participate. I have a bad experience with puzzle with random inputs but hey, still worth a try.

[12345ieee]

I'm trying, too.

If I am solving the N atom problem, can I assume that all N atoms will be output before any duplicate (e.g. I'll get a 2nd copy of each atom before I get my 3rd He)?

[etammate95]

Originally posted by 12345ieee:

I'm trying, too.

If I am solving the N atom problem, can I assume that all N atoms will be output before any duplicate (e.g. I'll get a 2nd copy of each atom before I get my 3rd He)?

Yes. All atoms appears once, before you get the second. (except if I made a mistake, but I hope I didn't)

[12345ieee]

No, you made no mistakes, I just wondered what the limits of "you cant use the pseudo-random input, it has to work with every input-orders. I will test if the solution is independent from the input order.( the ratio is fixed btw)" were.

Thanks.