« first day (1411 days earlier)      last day (3434 days later) » 

2:07 AM
@MartinBüttner I don't have anything concrete for it. If you have something that uses the concept, feel free to post.
 
 
7 hours later…
8:38 AM
@Geobits no, nothing in particular, I was just wondering
 
9:25 AM
@MartinBüttner I am deleting my comments.
(on your answer)
 
I had the 14 token solution all along, but Joe was not replying to my question :|
 
oh lol, your string exactly the same, down to the XX :D
 
what ?
 
your lookup string
is the same as mine
 
9:28 AM
Yeah, The one I was using had ZZ in the blank spaces, but its on my other computer
 
ah okay
 
and the code in the answer is not running (due to weird characters)
 
I can't check the other one
yeah exactly
 
I should have given a mother... link but that link is also in my home computer now
 
9:50 AM
@Doorknob冰 are you gonna post the call for themes for the weekly challenge challenge today?
 
or are we going to post the meta challenge challenge challenge
@MartinBüttner Hey, builtins were banned :P
 
that's why I included it just fyi at the bottom
 
and WTH is there a built in for Isotopes :|
 
because Mathematica
 
how much does a license cost ?
 
9:56 AM
I don't know
 
how do u have it ? student license ?
 
yup
wolfram language comes with the raspberry pi though, so that's a cheaper alternative
 
but on ur personal computer too ?
 
yeah sure
 
10:10 AM
You know what
JS beat Mathematica
 
post it then
 
0
A: Sandbox for Proposed Challenges

Martin BüttnerCaesar-Cypher-Mania code-golfsource-layoutstring The Caesar cypher is a very simple substitution cypher where each letter is shifted by a fixed offset (looping around Z to A). We can also define a Caesar cypher for the set of printable ASCII characters. These are the 95 characters from code poi...

 
Hmm I'd almost suggest allowed the programs to differ by whitespace... but that'd just make it too similar
(Time to Caesar all of Python's builtins and see what happens)
 
@Sp3000 how is the smoothing for PbN coming along?
 
I started my day doing this. Then I was doing this, but my program returned no solutions (maybe a bug?)
But the plan tonight is to implement this :)
(Also in my plans is colour correction, but I'm not sure how well I'll be able to manage that one)
 
10:30 AM
 
does it have the same bug as your CJam submission?
 
should be. You are not having the same bug ?
 
yeah I think JS does
no I don't
also, your JS function doesn't use the specified input format
I can easily shave off 5 tokens, if I use separate arguments
(in fact, it would shorten my submission by 8 tokens)
 
You mean using only 1 argument ?
I can fix that.
 
yeah a space-separated string
 
10:35 AM
that will reduce my tokens in general
 
and I think the question asks for a program
 
that will reduce even further :3
 
then why didn't you do that in the first place :P
 
I was not thinking
but this B and I bug .. is a little tricky
 
I was annoying for me... because of that I had to switch to InputString
because 123 I is a complex number
 
10:39 AM
0
Q: How do we thank people for outstanding work (beyond +1)?

CarpetPythonAs I understand it, comments should be used to ask questions about the puzzle or the solutions, but not for a "Thank You" note. Sometimes people are very responsive and provide extensive information. There is no information or question I could add, but I would like to thank that person for their ...

 
Ninja'd by Martin :P
(Was going to comment bounty as well)
 
Ninja'd ?
 
@Optimizer on that meta post
 
11:16 AM
0
Q: Count big-ints/symbols/variables as N tokens in atomic-code-golf

KennyTMThe goal of atomic-code-golf was to "level the playing field". However, languages with built-in bigints, symbols/atoms, regex literals etc can easily encode arbitrary programs into a few tokens. For example, in Python 3 any program can be encoded in 14 tokens using big-int + exec exec(int.to_...

 
@MartinBüttner Ahaha I was thinking of making a paint by numbers template which you can fill in, but first I'd need to get rid of the fuzziness :P
(my only concern is that some parts are only 1 or 2 pixels wide, and would be obscured by any cell borders)
 
you'd probably have to scale it up to fit in the numbers
 
Hmm yeah that too. I hope PIL has support for text...
 
i like how jeff atwood edited the sigsegv question
 
which one?
 
11:31 AM
14
Q: Shortest code that return SIGSEGV

AryaShortest code that returns Segmentation Fault (SIGSEGV) in any programming language.

he doesn't even have enough rep to edit questions
 
Ughh, 42 tokens :(
 
so, which language beats which now? ;)
btw, I've got a 7 token submission if digits aren't counted separately
 
i've been trying to make a Python "atomic code golf" with very few tokens with something like import binascii;exec binascii(n) for appropriate n
 
@Optimizer but I think the CJam equivalent is 6 tokens
 
11:38 AM
What do you mean by digits aren't counted separately ?
 
a number like 23495804980353 is a single token
 
that is given, he's talking about something else
 
@Optimizer as separate tokens
@Optimizer if that's given, why do you still have a 112 token version :P
mine is 9 tokens though
 
oh, you mean similar to by 13 token , you have a 9 token version ..
Damn, only if I was not in office to have time ..
 
yeah
but you can very easily turn your 13 tokens into 6 tokens
 
11:41 AM
oh wait
 
the ideal thing would be a language with both eval/exec and a built-in function that turns a number into a string
 
LOLOLOL
 
0
A: Sandbox for Proposed Challenges

LembikTell the computer to do the math This challenge is partly an algorithms challenge, involves some math and is partly simply a fastest code challenge. For some positive integer n, consider a uniformly random string of 1s and 0s of length n and call it A. Now also consider a second uniformly cho...

in case anyone is interested
 
@user2179021 patience ;) ... we have a bot that posts these every 10 minutes
 
@MartinBüttner I have no patience :)
but I did like the painting by numbers :)
 
11:45 AM
I have no money :(
 
I just want to use the code :)
 
@user2179021 you should participate then :P
 
@cjfaure contract coding¬
@MartinBüttner your answer is too good already!
 
@MartinBüttner Yo dawg, I heard you had 9 tokens!
 
11:46 AM
@user2179021 it's ironic that you put a grave after that
 
:)
@MartinBüttner oops.. I mean David Carraher
 
@Optimizer I wonder how you got the idea you could shorten yours to 6 :P
 
;)
I wonder that too
 
your code is still invalid though :P
 
doesn't matter now
it will still be 6 tokens after the fix
which I am doing right now
 
11:48 AM
this is like a conversation between a movie hero/villain
 
@Optimizer I know, but 6 invalid tokens are still invalid :P
 
who is the hero here ?
 
@Optimizer it works both ways
depending on whether tokens are bad/good
since they're bad, martin is the hero
 
@xnor Nice job on Pseudoku! Now I feel really sad because my program probably had a bug :(
 
or maybe he's the villain
 
11:49 AM
tokens are bad ?
 
i dunno D:
 
tokens are the worst, man
 
0
A: Sandbox for Proposed Challenges

LembikTell the computer to do the math This challenge is partly an algorithms challenge, involves some math and is partly simply a fastest code challenge. For some positive integer n, consider a uniformly random string of 1s and 0s of length n and call it A. Now also consider a second uniformly cho...

 
hooray ! :)
 
"what makes you think that you can take it down to 6 tokens?"
"oh, but i already have!"
"haha, but it didn't work, remember?"
"but i can fix it without adding any more tokens! YOU'RE TOO LATE!"
dramatic fight ensues
 
11:51 AM
I look forward to cutting comments and suggestions :)
 
the steam winter sale reminds me of youtube.com/watch?v=tRxcSNaVCPA
 
Fixed
still 6 tokens
 
"MUAHAHAHA! you'll never win!"
 
how is the other version still 112?
 
I don't have the string for that ..
let me create a new one
 
11:56 AM
ah, and the intermediate version would be 15 now I suppose
 
I think so
 
it's been minutes and no one has commented on my question yet! What happened to "Internet time" ? :)
it would have been closed 6 times already on SO :)
 
:D
even if you post it normally as a question
it would have been
(I did not look at the question, just telling about the rate of closure)
@MartinBüttner fixed!
 
@user2179021 it's too early
(in the day)
 
@MartinBüttner ah.. the americans
@Optimizer :)
 
12:07 PM
it's like 4am in the us, right?
 
depends on the part of the US, but yes
 
4am sounds like PPCG time :)
do you have a feeling for how many regular users PPCG has now?
I don't know if there is an easy way to measure this
 
Hmm... The golden rule of upvoting seems to be "that answer you spent not-so-long on will get more upvotes than that other answer that took you days"
 
@Sp3000 maybe the latter answer is harder to understand?
 
@Sp3000 Exactly
 
12:13 PM
More like the difference between HNQ and not HNQ
 
@user2179021 you can check how many users got more then X rep in the last week/month/quarter
@Sp3000 yeah... your PbN submission really deserves more attention... I shared it on reddit two days ago... that got the question a lot of views, but no upvotes for the answers unfortunately :/
 
Oh you shared it? But yeah I wouldn't expect everyone on Reddit to have an SE account
(In other news I just weighted k-means based on cell size and I think that makes the colours look better. Still testing to make sure.)
 
@MartinBüttner interesting
 
@Sp3000 I made the same experience on Puzzling, btw
 
Ahaha did you?
Is this the dungeon one?
 
oye. I will also promote my answers then :P
and I will directly go ahead : codegolf.stackexchange.com/questions/32359/…
 
I'd totally upvote...
if I hadn't already...
 
I know. I just need 1 more..
 
There you go
 
woot woot. Necromancer!
 
12:30 PM
congrats :)
 
Beating CJam by one char, eh
I mean GS
 
:D
I don't know if peter tries today, he might make it 33
 
:P
 
my goal was the badge, not accepted answer.
 
@Sp3000 wanna show some examples for comparison?
 
12:32 PM
 
I really like the pastel-y look of the colours at the moment
 
I retried a few times - this is one of the times I got all the shadows in
(due to the randomness in the algorithm, sometimes you get better results than other attempts)
 
ah I see
 
user image
4
 
wow
 
12:34 PM
 
(fyi that's without blurring)
 
in small res. that doesnt matter
 
f=lambda i:print["cat",["win","lose"]["ooo"in i]]["xxx"in i] would technically work, if i is a list of strings representing every line
@Sp3000 that's beautiful ;u;
 
@cjfaure diagonals
and columns
 
@Optimizer diagonals are included in i
by "line" i mean every winning line
 
12:37 PM
give an example input
 
k
for the example board in the post
 
you were not supposed to transform the input
1 answer does it, and I think that is also wrong.
 
["xxo","xox","oox","xxo","xoo","oxx","xox","ooo"]
 
yeah, that is I think an invalid input
the question states "you decide the input structure"
 
and my program prefers lose over win anyway
 
12:39 PM
@MartinBüttner Sure, I'll post that some time today.
 
@cjfaure so you can just decide on the structure of input, not completely process the input and do everything in the input itself
 
@Doorknob冰 cool :)
 
I can very well have the input as 0 1 or 2 and print win lose or cat based on that value.
ughh, why am I not getting the necromaner yet ..
 
relax
the silver cron jobs don't run that often
some only daily
 
@MartinBüttner PbN probably won't hit HNQ, but I think it'll get upvotes if we do things the Calvin way - bump it up every so often with updates :P
(not on purpose of course, this is so fun I think I'll just keep adding stuff)
 
12:47 PM
lol yeah
we could try hacker news, but I think with only 3 answers it's less likely to get much attention there. and even then it would be mostly views I think
 
these are cron jobs ?
 
I see
got it :D
 
1:15 PM
:/ I hate hindsight. Currently it's telling me "I should have made a class for colours so I can overload +"
 
Good thing I don't write production code (yet)
Apparently the last half hour of messages are between the three people with most rep gain this quarter
(I blame Martin)
 
:D
you're welcome :P
 
I need to start asking (good) questions too :(
I was ~300 rep ahead of Martin for this month a couple of days back.
 
no you need to start asking good questions because we're down to 2.4 questions per day :/
 
1:24 PM
Collaborative question writing, anyone? :P
 
well, we'll get that soon
 
Ahaha meta post, right
I had an idea which was "print a suffix tree" but wasn't sure how that'd affect golfing languages like CJam
 
what was the "print a suffix tree" idea? I mean was that the whole question?
how about represent a suffix tree in the minimum number of bits
 
Basically you print each node, with subtrees represented by indenting
 
@Sp3000 I think coding a suffix tree to store in small space is an interesting programming challenge
 
1:31 PM
How would representing with minimum number of bits work?
 
@Sp3000 A suffix tree supports various operations. Mostly "step down by following a character" and "report matching substring or mismatch"
 
I guess you'd have to clarify whether you want to store the substring at the node or just a pair of indices
 
@Sp3000 but a naive representation of a suffix tree takes about 30 bytes per input char I think
@Sp3000 always a pair of indices!
:)
@Sp3000 I think a cleverer one can take much less space
 
(When I wrote mine I didn't have to worry about memory so I was lazy :P)
 
:)
@Sp3000 you would have to specific which operations are constant tmie
 
1:34 PM
Hmm I guess that's heading into the too-hard-for-me realm :P
 
@Sp3000 if you can implement a suffix tree it doesn't sound too hard does it?
 
Well that was just with your typical representation of data and list of children
Maybe it'd be fun to do something that allowed suffix trees in it... fastest code for 1 million substring queries? :P
 
@xnor has so many good questions waiting in sandbox ...
 
Ahaha yeah bowling
 
and minesweeper too
I like that one.
 
2:04 PM
Hmm wish polyominoes goes more submissions :/ (not that I'll be updating mine any time soon)
 
its a hard one
 
2:27 PM
it's raining badges
 
for you :|
 
@Optimizer lol why are you comparing my 32 token version with your 6 token version? :D ... double standards!
 
:D anyways, doesn't matter. The message behind the statement is conveyed
@MartinBüttner there
 
@Sp3000 lol, nice bounty message
 
:D
Basically because I'm sure that if Pypy can get up to 18, then feersum or any other C-like implementation should be able to do well
I might just be a bit naive though
 
2:35 PM
also, no one has tried subset sum yet
 
Ahaha I still haven't figured out how to put that in my algorithm, but I don't think I'll be getting around to polyominoes any time soon
 
I think it would be a completely different approach
but yeah, I'd rather see you spend your time on PbN :D
 
:P graphical output
 
2:50 PM
I just used one of ReSharper's solution wide cleanup features, and ended up with a file that looks like var var var var v for (int i = ar 0; i < va 100; i+var var+) r var var var var var
It was supposed to replace explicit types with var, but instead it replaced everything with var.
 

« first day (1411 days earlier)      last day (3434 days later) »