Timeline for Sandbox for Proposed Challenges
Current License: CC BY-SA 3.0
24 events
when toggle format | what | by | license | comment | |
---|---|---|---|---|---|
May 25, 2018 at 0:31 | comment | added | N. Virgo | @wastl not sure if I'll ever post this, but the preprocessor would ignore all whitespace, so that would be fine | |
May 24, 2018 at 14:05 | comment | added | wastl |
In the non-preprocessed program, what whitespace is necessary? Is a+a+a+b+b+a{a-b+} valid?
|
|
May 29, 2017 at 21:59 | comment | added | Esolanging Fruit |
A potential loophole for cops: The cops' program searches for a string that has a certain hash. If the string's length is more than 400, it outputs true , otherwise it outputs false . It would be very difficult for the cops to prove that the program generates an incorrect result.
|
|
May 6, 2015 at 21:15 | history | wiki removed | Martin Ender | ||
May 3, 2015 at 15:37 | history | made wiki | Post Made Community Wiki by Martin Ender | ||
Oct 30, 2014 at 2:02 | history | edited | N. Virgo | CC BY-SA 3.0 |
added 81 characters in body
|
Oct 29, 2014 at 10:36 | comment | added | N. Virgo | @PeterTaylor yes, indeed, that's what I based it on. | |
Oct 29, 2014 at 10:35 | comment | added | Peter Taylor | Looks like a restricted form of Minsky register machine. As a fan of MRMs, I approve. | |
Oct 29, 2014 at 10:27 | comment | added | N. Virgo | @PeterTaylor in case you're interested, I invented my own Turing-complete language (imperative but much more minimal and easier to parse than IMP) and added its spec to the draft. | |
Oct 29, 2014 at 10:10 | history | edited | N. Virgo | CC BY-SA 3.0 |
added 2800 characters in body
|
Oct 29, 2014 at 8:57 | history | edited | N. Virgo | CC BY-SA 3.0 |
added 608 characters in body
|
Oct 28, 2014 at 3:44 | history | edited | N. Virgo | CC BY-SA 3.0 |
added 50 characters in body
|
Oct 28, 2014 at 3:33 | history | edited | N. Virgo | CC BY-SA 3.0 |
added 1468 characters in body
|
Oct 27, 2014 at 12:16 | comment | added | Peter Taylor | If you want to take a functional approach then pick a universal set of combinators (e.g. SK or Iota). If you want an imperative language, either a well-specified BF or a "toy" language like IMP (as in the one described in Winskel's book on semantics, not the one described in the Wikipedia page - which might confuse people too much). Or pick something interesting from esolangs.org. | |
Oct 27, 2014 at 0:55 | comment | added | N. Virgo |
@PeterTaylor I'd be quite happy to pick an L for everyone - that would also solve the problem feersum pointed out - but the question is, what should it be? It needs to be Turing complete and simple enough to analyse programmatically, and I guess it would help a lot if it was also fairly human-readable and intuitive. (That pretty much rules out lambda calculus, which is the obvious choice, though it's actually already a bit fiddly to deal with lambda expressions programmatically because of having to track free variables and do alpha-conversions all the time.) Do you have any thoughts?
|
|
Oct 27, 2014 at 0:36 | comment | added | N. Virgo | @feersum ah, you're right, this challenge can be gamed in that way. Even worse, you could use something like the twin primes conjecture instead of prime factorisation, and nobody could crack your entry without collecting a Fields medal. I'm not sure if there's a way to fix that :( | |
Oct 26, 2014 at 17:40 | comment | added | Peter Taylor |
Not really. You could pick a semantic model and require people to use it, but then you exclude people who don't know it and don't want to learn it just for this challenge. You could pick an L and require everyone to use it, but I expect you to say that that would take away a lot of the fun.
|
|
Oct 26, 2014 at 15:30 | comment | added | feersum | I see this as much too easy for the cops, who can simply use a tedious integer factorization or similar task as always. Define L as follows: after performing trial division on a hard-coded integer, execute a Brainfuck program if it has a 200-digit prime factor, terminate without executing those instructions if it has a 201-digit prime factor, or otherwise go into an infinite loop. P simply prints 1. | |
Oct 26, 2014 at 13:40 | comment | added | N. Virgo | @MartinBüttner I like hard challenges, and by posting them I hope to attract other users who feel the same way. I think "this challenge is too hard" is not a thing that should be said here; it might seem hard to some people (especially before a solution has been posted) but one shouldn't underestimate how clever people can be. I'm not sure but I think being hard for the cops rather than the robbers is a good thing in this type of challenge. | |
Oct 26, 2014 at 13:38 | comment | added | N. Virgo | @PeterTaylor I've fixed the first of those issues - do you have an intuition on how to deal with the second? | |
Oct 26, 2014 at 13:37 | history | edited | N. Virgo | CC BY-SA 3.0 |
added 108 characters in body
|
Oct 26, 2014 at 10:53 | comment | added | Martin Ender | I think this will be way too hard on the side of the cops to be feasible for PPCG. | |
Oct 26, 2014 at 8:16 | comment | added | Peter Taylor |
As written this seems to allow P to fail to halt if the supplied program fails to halt. "Completely define its syntax and semantics" could lead to disputes, especially if someone chooses to use an obscure semantic model or to define their L in terms of an existing language which doesn't have formally defined semantics.
|
|
Oct 26, 2014 at 5:45 | history | answered | N. Virgo | CC BY-SA 3.0 |