I've got another puzzle for you

[davidn]

BrickAt( "BoardObjects", (X( "Robot Sprite" )/CELL_SIZE)+1, (Y( "Robot Sprite" )/CELL_SIZE)) = CELLTYPE_BARREL and I'm having one of those moments when I wonder why I spend my weekends like this. Somehow, though, I decided I wanted to try making another game, so here's yet another concept sketch thing that I've put together while avoiding my main project.


Play prototype level

It's called "Sokodeban" as a throwaway title (that will accidentally stick firmly if it's anything like my other projects). The general idea is that of Sokoban, but without having direct control and instead having to program robots to work together to push the barrels around into the empty sockets.

My idea is to eventually have each different robot coded with a slightly different method, and each having their own advantages and tactics, but for now, each of them are directed with simple NSEW characters in edit boxes to get them to move north, south, east and west (parsing that was enough work for a day). With some experimentation, you should be able to complete this level that fell out of my testing within the character limits given beside each box, but they're not enforced as yet so you can experiment as you like. And it doesn't check for the level being complete, either, so you'll have to just award yourself a biscuit instead if you make it.

Good luck. And I hope that this doesn't end up on the Indiegames feeds like the last time as a demonstration of what my finished work's like.

Comments

[tamakun.livejournal.com]

I really like this! I've always been a fan of programming puzzles. I'm trying to remember what games I would play ad infinitum, the only one being a Dr. Brain puzzle: Motor Programming (http://www.youtube.com/watch?v=xqDGPYMcWGk).

I appreciate the work that goes into something like this: "Programming programming". Similarly, I appreciate tools that allow a person to build their own levels, it's a UI to simplify the actual creation of the levels.

I wasn't able to determine the "right" way of solving the puzzle in the given character limits. It gives me something to work towards :)

[davidn]

The "robot with subroutines" puzzle! I've gone through something like that, as kjorteo ruined my day by providing a Flash version once, though I hadn't been aware that the same template had existed as far back as the Dr. Brain games.

I really liked how mad the solutions had to get in the particularly variety I played, though at the same time I'm hoping that I can make this different enough not to rip it off. In the underlying engine just now I've given provision for code to be "compiled" before it's run, but the two robots there just run the NSEW move commands directly and just ignore anything they don't understand - I was thinking of having one type of robot that had LOGO-like commands, one with limited memory but provision for subroutines... I'm not quite sure yet about how to strike a balance between the puzzle of telling the robots what to do within the "memory" available, and that of getting those programs to synchronize with each other to solve the greater puzzle - but with the two layers of it, it sounds like it could work quite well if I get it right.

And, well, every puzzle has an answer ;) If it were in a full game, though, this particular one would definitely be placed a few levels in after some explanatory tutorials, so it's possibly harder than it appears. Like I said the last time I showed off a puzzle game prototype, puzzle games are completely impossible to sense the difficulty of, seeing as you already know the answer by the time you've built a level.

[dr-dos.livejournal.com]

That took me far longer than it should have to complete.

The idea is pretty neat. I recall one time playing a board game where each player loaded cards with instructions for robots to move and turn and fire lasers in a fight to the death.

And man I know all about that "hey I know what I should do, instead of working on my game, start making a new one!" feeling. Somehow I got it in my head that it would be amusing to make a short Atari 2600 style version of Team Fortress 2.

[kibet.livejournal.com]

I quite enjoyed this and it is a good starter level and straight forward. I think i had to activate 3 times. Once when I activated thinking that we switched between the bots. The second time was after only moving Bertha 6 moves and Jack none to realise that you have to tell them both at the same time what to do and noticing the holes become solid and The third to actually complete the level.

[davidn]

Excellent :) Did that idea go anywhere more than is shown in the screenshot? (Or is it a project that's going on just now?)

[davidn]

You definitely seem to have a good sense for... playing things out in your head, if you only had to use two tries for working out the rules and then immediately solved it! But I would have expected that from you :) Thanks for trying it out - perhaps something will grow out of it.

[dr-dos.livejournal.com]

The exciting adventures of being a red soldier and capturing the enemy intelligence in 2Fort.

I justify it by saying that it makes me learn concepts for games that aren't tiles like Tyger.

[kibet.livejournal.com]

I have played a lot of Sokoban and other pushing games. I prefer games that don't give you the rules as working out how to play can be half the fun. CryptRaider was a bit like that as it reminds me back in the day of games for the Apple II and Commodore 4⁺-64, where they never gave hints to start only straightforward levels to introduce concepts.

[davidn]

Every concept needs a small project to get a feel for it :) For this one, I was surprised how much I needed to consider for what looks like a simple "pushing" game on the surface, because I thought that the program interpretation would be the hardest part... can a robot move? If it's next to a barrel, can it push it, or is there a wall or another robot in the way? Or - less obviously - is there a space to which another robot or barrel is going to move in this turn?

Each object now declares the space it's in and the space that it thinks it's going to move into, when it's examined... so only the object that happens to be examined first is allowed to get into a contested square. I'll do a very long and dull post about this at some point.