Talk:/ˈæmbiːɛf/

From Esolang
Jump to navigation Jump to search

About [+]

Given that +- randomly increase or decrease the cell, is it theoretically valid (and why?) to assume [+] will return to zero every time instead of going more towards positive or negative infinity (with occasionally "backtracking" (I don't know a right term) -- that is, while increasing, occasionally decreasing, then increasing again, always more than decreasing)? I assume that it is, but this seems to be a good opportunity to ask the question, as I have a couple of nondeterministic language plans myself. A nice language, by the way. --Keymaker (talk) 16:46, 15 October 2012 (UTC)

Fair question... When we were talking about that on #esoteric I didn't pay attention to the "bignum" part and thought cells were bounded. In that case it would be very easy, using very loose lower bounds, to prove that [+] eventually reaches 0. If cells are unbounded it's more complicated... Here is the shadow of the sketch of the beginning of a proof.
let p(n, k) = probability to have reached 0 starting from n after k iterations in [+]
p(0, k) = 1 for all k
p(1, 0) = 0, p(1, 1) = 1/2
p(n, k) = 0 for all k < n
p(n, k) > (1/2)^n for all k >= n > 0
p(n, k) = p(n-1, k-1)/2 + p(n+1, k-1)/2
p(n, k) > p(n-1, k-1)/2
p(n, k) > p(n-2, k-2)/4 + p(n, k-2)/4
p(n, k) > p(n-3, k-3)/8 + p(n-1, k-3)/8 + p(n-1, k-3)/8
p(n, k) > p(n-4, k-4)/16 + p(n-2, k-4)/16 + p(n-2, k-4)/16 + p(n, k-4)/16 + p(n-2, k-4)/16 + p(n, k-4)/16
(etc.)
all of those terms p(n', k') such that k' >= n' respect p(n', k') > (1/2)^n (that's why I only kept terms with n' <= n)
so for any given n, when k -> +inf, the number of such terms is high enough that it should work? or maybe not.
In my defense, I haven't slept much recently! --Koen (talk) 20:06, 15 October 2012 (UTC)
Also, according to Wikipedia:Random walk, the answer is yes: "How many times will a random walk cross a boundary line if permitted to continue walking forever? A simple random walk on Z will cross every point an infinite number of times. This result has many names: the level-crossing phenomenon, recurrence or the Wikipedia:gambler's ruin." --Koen (talk) 21:12, 15 October 2012 (UTC)
As I recall and as that wikipedia article probably mentions, a random walk on a one- or two-dimensional grid will always return to any point (with probability 1), while a three-dimensional grid has a non-zero probability of never doing so. --Ørjan (talk) 10:26, 16 October 2012 (UTC)
For nondeterministic languages, you're mostly trying to prove that it completes with probability 1. I already knew that random walks on a 1D line reached every point with probability 1, so I didn't really think about the issue further. --ais523 21:14, 16 October 2012 (UTC)

The name, oh my word, the name

The introductory paragraph seems rife with... conceptual problems regarding the name.

  • "Its name exists only when pronounced" -- Well, this is sort of a philosophical point, isn't it? Tree falling in the forest and all that lot.
  • "/ˈæmbiːɛf/ is how its name is rendered in IPA [...] It officially cannot be spelt" -- What is the difference between "rendering" and "spelling"? Clearly the name can be spelled, even if it must be spelled in a particular alphabet (the International Phonetic Alphabet). But in the same breath, there is the claim that it officially cannot be spelt (which I'll assume refers to arrangement of symbols, rather than something about being wheat.) This contradicts the fact that it has just been spelled, in the previous sentence, in IPA.
  • "it is also impossible to determine whether or not its name starts with a capital letter" -- whether a name starts with a capital letter or not is a matter of the choice of the author so it seems odd to think of this as something people can attempt to "determine" (that is, it seems to imply there is some objective method to this aside from asking the author what their wish is.) Also, since the name either (1) is spelled in IPA, which does not have capital letters, or (2) cannot be spelled at all (due to the official statement to the effect), this claim about capital letters seems doubly futile.

Of course, if these conceptual problems were intentional, well, then, well done! Chris Pressey (talk) 07:33, 18 October 2012 (UTC)

Actually, the name thing was decided specifically to pay homage to the way you named languages. What I was trying to get at was that the name was defined in terms of its pronounciation, and doesn't have a separate existence beside that. (And I didn't really consider IPA to be a spelling, more a notation to express a pronounciation in text.) I guess a better way to express it is that it can only be spelled in phonetic languages. (As for the capital letter thing, that was taking a jab at brainfuck's non-proper-noun-ness.) --ais523 13:58, 22 October 2012 (UTC)
Ah, well, I do like a pathological name! (Names are their own particular aesthetic hell, and names of programming languages, for whatever reason, inhabit its lowest rung...) It's not that I couldn't follow what you were [most probably] saying in the first paragraph, it's that the alternate interpretations were blowing my mind. I guess you could define "spelling" as the way one (prescriptively or descriptively) arranges symbols to represent sounds in an alphabet that is not wholly phonetic; in which case, words aren't "spelled" at all in IPA.
I also guess I also have several other thoughts on the matter, but they are too wide-ranging and not directly relevant, so I will save them for The Manifesto. Chris Pressey (talk) 14:16, 29 October 2012 (UTC)