While taking PSAT practice tests, I invented a sort of game to be played on the answer grid. For those who don't know what the PSAT answer sheet looks like, it's basically a 4 by X grid of bubbles, with each row representing a different problem, and each bubble in the row representing an answer choice. I would pretend that four balls began falling down the top of the grid, and the filled bubbles would block their path and cause them to split. The goal of this contest is to display the columns where balls will exit.

Only one bubble will be filled per row, and no row can be. blank. A ball will continue down a column if there is no bubble in that column, and if there is, it will split and move to the two columns on either side only if there is no wall and no filled bubble above it.

( )(*)( )( )
( )( )(*)( )
( )( )(*)( )
(*)( )( )( )
( )(*)( )( )
( )( )( )(*)
( )(*)( )( )
(*)( )( )( )
( )( )(*)( )
( )(*)( )( )

The output for the example "map" above would be `A, D`. The pattern for the falling balls would look something like this:

(|)(*)(\)(|)
(|)( )(*)(|)
(\)( )(*)(|)
(*)(\)( )(|)
( )(*)(|)(/)
( )( )(|)(*)
( )(*)(|)( )
(*)( )(^)( )
( )(/)(*)(\)
(|)(*)( )(|)

All entries must be complete, no code snippets allowed. The input will be in the form of `yxyy<newline>yyxy` with `y` representing an empty space, `x` a filled bubble, and <newline> representing a newine. `x` and `y` may be replaced with any character you want. The program may receive input through STDIN or read a file. The program output may be any legible form, eg. `{1,0,0,1}`, `1001`, `AD`, `A, D`, `{true, false, false, true}`. All BASIC and Axe entries will be scored on source size, minus the VAT header, asm (calc or computer) entries will be scored on binary size (minus VAT header if on calc), and computer entries will be scored on source size. If there are any questions, please ask. You can submit your entry by PM'ing it to me here, or on IRC on freenode or efnet.

Scores:

Juju: Ruby, 101 bytes
gudenau: Java, 616 bytes

Looks nice! It reminds me of this: http://www.ticalc.org/archives/files/fileinfo/139/13958.html

Quote from: Vogtinator on June 22, 2015, 08:05:41 am

WOW, this is unexpected. Well done! How does it run? I assume it's not really at a usable speed.

I do wonder why "Minor compilation changes for older GCC versions" was necessary, binary literals have been supported since ages.
You should definitely update your SDK and toolchain!

Boot up is slow, taking ~3-5 minutes (Although I believe I can speed it up), but after most of the core services are starting the system is actually pretty usable (imagine an HTC Dream running at half-speed), making this a feasible alternative OS to run on the calculator.

This is awesome! Where is the rootfs stored? I assume it's on an external flashdrive or something.

Here's my collection:


The 83+ SE isn't mine, it's my sister's, but I can use it whenever I want practically, I got the 84+ CSE, 85, and one 84+ SE all for free And yes, the 73 Explorer is running the TI-83+ operating system, so I can run 83+ programs on it.

Thank you! I actually like it that it looks so self-crafted, though :p

Yeah, even in its current form it looks pretty neat.

Nice job! The end result looks pretty great, but could be prettied up a bit. If you have some thinish black plastic (I've found that ground coffee lids work great) then you could cut a square and a hole in that to put over the hot glue.

11 bytes in TI-BASIC :p

Aye (for reasons I explained above)
Aye aye (it's nearly like writing your code in hex at this point.)
Aye aye aye (If assembly is gonna get counted in binary, then the same goes with BASIC. That's pretty much what we did before IIRC.)
Nay nay nay nay (I'd like if each contest had the same standarized, agreed-upon counting method.)

I vote exactly the same as Juju