Live data from Hacker News

2048 Numberwang

louhuang.com

51–60 of 116 posts

Re: 2048 Numberwang

#51
Found a bug you might want to look into:

After 8 steps, I "won" after reaching the following configuration (all on the edges, from top-left clockwise):

16, 584, 26, 6, 18, and 7

That's not numberwang, according to the official definition [1]

[1] D. Mitchell, B. Russell, A. Turing, and R. Webb. Numberwang determination and the Entscheidungsproblem. Principles of Mathematics and Computation, 1944. Cambridge Press 14(2).

Re: 2048 Numberwang

#53

Found a bug you might want to look into: After 8 steps, I "won" after reaching the following configuration (all on the edges, from top-left clockwise): 16, 584, 26, 6, 18, and 7 That's not numberwang, according to the official definition [1] [1] D. Mitchell, B. Russell, A. Turing, and R. Webb. Numberwang determination and the Entscheidungsproblem. Principles of Mathematics and Computation, 1944. Cambridge Press 14(2)…

No, "That's numberwang" does not indicate winning nor the end of the game. It just pops up at random throughout the game.

Re: 2048 Numberwang

#54

Found a bug you might want to look into: After 8 steps, I "won" after reaching the following configuration (all on the edges, from top-left clockwise): 16, 584, 26, 6, 18, and 7 That's not numberwang, according to the official definition [1] [1] D. Mitchell, B. Russell, A. Turing, and R. Webb. Numberwang determination and the Entscheidungsproblem. Principles of Mathematics and Computation, 1944. Cambridge Press 14(2)…

No, "That's numberwang" does not indicate winning nor the end of the game. It just pops up at random throughout the game.

Ah yes - that's true in the BBC 4 version of the rules, introduced in 1968. I think most countries use that now. In Australia we play according to the original rule set, hence the misunderstanding.

Re: 2048 Numberwang

#56

    C:\> TWENTY48
    Loading...
    Ready.

    You are in a darkened room. On the floor in front of
    you are sixteen tiles, arranged in a grid.  The tiles
    are labelled A1, A2, A3, A4, B1 and so on up to D4.

    There is a two here, on tile A1.
    There is a two here, on tile D3.

    > GET ALL
    The numbers appear too strongly glued to the tiles.

    > INVENTORY
    You have a copy of On Lisp by Paul Graham, a thing your 
    Aunt gave you that you don't know what it is, and a 
    small leaflet.

    > READ ON LISP
    Time passes. You have a profound enlightenment experience.  
    Sadly, this does not help you win the game.
    
    > READ LEAFLET
    "WELCOME TO TWENTYFORTYEIGHT!

    TWENTYFORTYEIGHT is a game of numbers, addition and high 
    bits. In it you will explore some of the most tedious 
    territory ever seen by Hacker News readers. No computer 
    will be without one!"

    > GO LEFT
    Some numbers move, and a new number appears!

    There is a two here, on tile A1.
    There is a two here, on tile C1.
    There is a two here, on tile D1. [Footnote 6]

    > FOOTNOTE 6
    There are no footnotes.

    > GO UP
    Some numbers move, some change, and some new numbers 
    appear!

    There is a four here, on tile A1.
    There is a two here, on tile B1.
    There is a two here, on tile D4.

    >GO DOWN
    Some numbers move, and a new number appears!
    
    There is a four here, on tile C1.
    There is a two here, on tile C4.
    There is a two here, on tile D4.
    There is a two here, on tile D1.

    >PANIC
    Not surprised.

    >QUIT
    Amazed you survived this long.

    Your score is 4, out of a possible umpty squillion and six.
    Good bye!

    C:\>
    C:\>
    C:\> DEL TWENTY48.EXE
    C:\>

Re: 2048 Numberwang

#57
post #48

I think I should have won by now: https://www.dropbox.com/s/2hokoh94hujyf2j/Screenshot%202014-... I thought it played normally, and it does for a while, but it always eventually seems to randomly revert your high tiles (I verified that I had at least 1024 by checking the debugger). [edit:] Ah, I found the code. It's possible to win but you have to survive the small chance that it will perform a random merge and destr…

Hm, the comment is wrong, it's 0.5% chance.

Re: 2048 Numberwang

#58
Brilliant! Even better because Numberwang is our internal term for "any random number in some code that isn't declared as a constant". Love it.

Re: 2048 Numberwang

#59
1724 first go. NFI as to what I was doing, but that seems pretty good to me?

EDIT: I just followed the arrow keys around in circle and got 7056. That's pretty, pretty, pretty good.

Post reply on HN