« first day (1498 days earlier)      last day (3355 days later) » 

3:08 AM
@Sp3000 I need a catchy name for that regex substitution challenge type ;)
I'm only coming up with silly things like Regex-Mania and Regex Revenge... :/
would be nice if it was also remotely indicative of the challenge structure
 
Regex regret
 
Regex repeater?
Not sure :/
 
or Regex remorse
 
@feersum there is no such thing.
 
your challenge will be first to induce it!
 
3:11 AM
:(
Regex Juggling?
 
0
A: Sandbox for Proposed Challenges

xnorIs this matrix rank-one? code-golf Your goal is to check whether a given integer matrix is rank-one, using as few bytes as possible. A matrix is rank-one if every row is a multiple of the same vector. Example: 2 4 20 10 -3 -6 -30 15 0 0 0 0 The definition also holds true ...

 
Atomic regex golf?
 
meh :D
 
Hey! Just trying to help :P
 
I know, thanks ;)
 
3:18 AM
(in another news, my language looks really nice on paper but now for the hell part of implementing it...)
 
How about make the title in the form of a regex?
 
could be appalling :D
 
@Sp3000 Your 2D language?
@MartinBüttner As long as it's pronounceable, it'll be fine.
 
also, the reason I'm looking for a name for the challenge type is that I hope to use this challenge format again in the future (because classic regex golf is only interesting to a point...) and then it would be nice to identify this challenge type by a name so people don't have to double check the format each time again
 
Yeah :) but the parser's not letting me use square brackets for my char classes for some reason and it's annoying me :(
 
3:20 AM
regex rigamarole
 
what's that?
 
I dunno, it's your challenge
 
@Sp3000 Did you post a spec of your language anywhere?
 
Not yet - as I said it's all on paper :P
 
3:26 AM
My spec is too rambling and self-contradictory, so I 'm not posting it
 
I want to make sure it actually does what I want to it before I post it, but the diamond one should look something like ^1(X/*)<(_(1)X'*)<(_(1)X/*)<(_(1)X*)` (^1 is set direction to diagonal up right, backtick is escape, < is rotate left, (_(1)...) is match length of group 1)
 
I think the most difficult challenge is the diamond one.
 
it would be hilarious if anyone (including me) tried to read my regex traversing code later
 
@Sp3000 maybe just plainly Regex Coding
 
meh :D
 
3:34 AM
:D
 
(nah just kidding, it's nice and straighforward)
 
It doesn't describe the challenge (beyond regex) at all
 
sure, it implies that you're actually programming with regex... in the sense of performing computations
 
only if you can read minds
 
lol
How about something like: "Programming with Regex: Adding Numbers"
 
3:45 AM
I'm now at 1100 measures of music in my sketchpad. 100 measures in a day, not bad.
Programming with Regex is probably good enough.
 
alternatively, I'll just call this one "Adding Numbers with Regex" and ask for challenge name suggestions in the comments
 
Regex Abacus: How to add
 
that seems very specific to arithmetic
 
4:02 AM
I found a long word and want to make a challenge out of it: en.m.wikipedia.org/wiki/Nonfirstorderizability
 
'regex abacus' conveys well for this challenge the idea of performing arithmetic in a tedious manual way
 
yeah, my point ;)
 
@PhiNotPi use Hungarian: megszentségteleníthetetlenségeskedéseitekért :)
 
pfff
Rinderkennzeichnungs- und Rindfleischetikettierungsüberwachungsaufgabenübertragungsgesetz listen (RkReÜAÜG) (literally, Cattle marking and beef labeling supervision duties delegation law) was a law of the German state of Mecklenburg-Vorpommern of 1999, repealed in 2013. It dealt with the supervision of the labeling of beef. The name is an example of the virtually unlimited compounding of nouns that is possible in many Germanic languages. German orthography uses “closed” compounds, concatenating nouns to form one long word. This is unlike most English compounds, which are separated using spaces...
(and now someone bring on that Welsh town name...)
 
> The German language has lost its longest word thanks to a change in the law to conform with EU regulations. Rindfleischetikettierungsueberwachungsau
lol
 
4:09 AM
the antiindefiniteridiculouswordconcatenationwithincomprehensibleresultsregulations?
 
seems like a normal German expression just the spacebar didn't work
 
it's actually fairly comprehensible
 
parsed as 'within comprehensible results'
 
should have said "ambiguous" :p
 
the Hungarian one is one single 5-letter word + a ton of suffixes
 
4:12 AM
"This is the official short title of the law"
 
maybe short = no unnecessary whitespace
 
0
A: Sandbox for Proposed Challenges

FunctinoMake a Space Heater I've been out shoveling snow all day, and my hands are freezing! Heat them up with my computer. Here's an (ungolfed) Linux C solution: #include <stdlib.h> #include <stdbool.h> #include <unistd.h> #include <sys/prctl.h> #include <sys/wait.h> #include <sys/sysinfo.h> int ...

 
@randomra yeah, you don't get that in German... you just get compound nouns
and that one is quite extreme
 
@MartinBüttner I hope so
 
Isten áldja meg
 
4:14 AM
@aditsu ?
 
@randomra that said, I've seen official bodies in the UK government whose names aren't any shorter... they just put spaces inside their compound nouns...
 
a common greeting in my grandma's village
 
it looks almost like Icelandic
 
and golfers don't need spaces
 
Java standard library beats out German for longest compound: InternalFrameInternalFrameTitlePaneInternalFrameTitlePane
MaximizeButtonWindowNotFocusedState
 
4:17 AM
that name-generator stuck in a loop
 
I'm hearing birds outside my window
I should probably sleep
 
I'm already up :)
 
 
2 hours later…
6:43 AM
cool, when I had a compiler error gcc told me exactly what to write in order to fix it
I had tried putting the offending line in gcc because the approach of VS was to spew hundreds syntax errors inside the STL source
 
 
6 hours later…
1:02 PM
hm, I totally forgot about the next fortnightly challenge
will set that up after breakfast
 
please don't make the theme Pathfinding, I have a busy couple of busy weeks ahead of me
 
it'll be lego
I was thinking about generalising De Bruijn sequences to two dimensions this morning
Of course, De Bruijn was a clever guy and thought of that himself
In combinatorial mathematics, a De Bruijn torus, named after Nicolaas Govert de Bruijn, is an array of symbols from an alphabet (often just 0 and 1) that contains every m-by-n matrix exactly once. It is a torus because the edges are considered wraparound for the purpose of finding matrices. Its name comes from the De Bruijn sequence, which can be considered a special case where n is 1 (one dimension). One of the main open questions regarding De Bruijn tori is whether a De Bruijn torus for a particular alphabet size can be constructed for a given m and n. It is known that these always exist when...
I was afraid these would grow way too fast to make an interesting challenge
(actually, maybe De Bruijn didn't think of this, but someone did)
 
 
2 hours later…
That blew my mind into a million droplets.
 
4:13 PM
@Sp3000 "starts at the top right". did you mean top left?
 
... yes top left, oops
Thanks :D
Man, prelude's going to be a fun one to do
 
Yay! My parser can handle directions now :D
 
:D yay
 
for a good laugh compare the source code of python entries with my c++ one
 
5:22 PM
@Sp3000 Congrats on getting 10k!
 
Thanks :) - I think a lot of people are approaching that number now
 
omg, writing a backtick inside a code-formatted bit is so hard
 
double backticks
 
A couple of days I looked up how do it, but I was still unable
 
(the surrounding ones)
but if you're writing an answer (as opposed to a comment or chat message), you could also simply do <code>`</code>
 
5:38 PM
it seems I have found the limits of codegolf.se with my two challenges...
I suppose there had to be one
hi :)
hi @feersum
 
hi
 
there should be a fortnightly challenge tag I think
 
that would be the definition of a meta tag
if you want to find all fortnightly challenges, go to the meta post (which is linked from all fortnightly challegnes)
 
ok thanks
 
also searching for "is:question fortnightly challenge" is probably gonna do the trick as well
or maybe not
the first few used to be "weekly challenges"
 
5:45 PM
is the idea that fortnightly challenges will involve more work than a typical challenge?
i.e. be a larger task
 
not necessarily
the idea is that it's a tricky challenge to work out which requires collaborative effort
the resulting task might well be simple
(although I guess there is some correlation between difficulty of writing the challenge and writing an answer)
 
ok so would my snake bending challenge be suitable for example?
11
Q: The number of reachable snake orientations

LembikThis challenge is not about the game Snake. Imagine a 2d snake formed by drawing a horizontal line of length n. At integer points along its body, this snake can rotate its body by 90 degree. If we define the front of the snake to be on the far left to start with, the rotation will move the back...

or codegolf.stackexchange.com/questions/47163/… which seems to have foxed everyone :)
 
I don't know
spec-wise these are fairly straight forward, aren't they?
 
oh you mean it should be hard to pose the question in the first place?
 
well if it isn't, then why ask the community for help with writing it?
 
5:58 PM
Did kuroi neko leave PPCG?
 
something happened ?
 
huh looks like it
 
oh, that guy who just kept on posting invalid answers and then replying rudely ?
 
he posted amazing answers on the lab rat race
I don't remember about other challenges
 
He posted answers on a few other code challenges.
Lembik's question up there for an example.
 
6:01 PM
hm I can't seem to search for his answers
that explains the upvote that disappeared on lab rat race
 
That explains the 80 rep I lost a month ago.
He really liked Carpet Python's challenges.
 
6:21 PM
there's a new (so far private) beta site
Music Fans
@Sp3000 I really like your 2D regex submission btw!
(?_()...) would be a nice addition to normal regex flavours
 
Ahaha yeah :P .NET can kinda do that I think
 
yes, but it's tedious, especially if you want to refer back to the same length more than once
 
I wish my code was neater though - took me hours to figure out how to implement concatenation because of backtracking issues ((ab)*ababc match ababababc)
Ah, yes more than once...
 
@Sp3000 your laughter starting with an "a" sounds like a really tsundere laugh.
 
"tsundre"?
 
6:26 PM
Does it?
(Although, yes manga might have something to do with it アハハ...)
Hmm I think double chests and Prelude are probably the two hardest ones in the list
 
prelude should be easy
and of two regex possible ?
 
Not sure how to do the bracket matching for Prelude hmm
 
I'm constructing a 2D lang based on conjunctive grammars, Prelude should be easy with that. Not sure when it'll be ready for posting though.
 
@Sp3000 back referencing ?
 
Hmm... backreferencing might be something else I need
 
6:34 PM
@Sp3000 BTW I also find your 2D regex thing awesome. :P
 
any other question incoming Zgarb ?
 
Conjunctive grammars sound cool too :P Not sure what that entails though
 
@Optimizer backreferencing can't match balanced strings
the only constructs in normal regex flavours which are capable to matching balanced strings are recursion and, well, balancing groups.
 
@Optimizer I have a randomness challenge in mind, but it requires some work. Basically, "guess from which random distribution these numbers are drawn from".
 
that'd be difficult to spec.
 
6:37 PM
True. :D
 
If it's one of those classification types I'd be interested. I tried doing something like that with shuffles
 
It would probably have a handful of predefined distributions, then M lists of N numbers drawn from each, and you'd have to guess at least X% of the lists correctly.
 
(On a side note, I think if I have anchors then slip left/slip right might be able to act like a stack...)
 
13
A: Convert a repeated decimal to a fraction

xfixPerl 6 (93 101 100 80 68 66 bytes) $/=split ".",get;say ($0+($1+$2/(9 x$2.comb||1))/10**$1.comb).nude The size was increased to handle nothing, instead of just failing. Mouq proposed to use $/, so it's now being used, and the code is 20 bytes shorter. Ayiko proposed replacing / with , so the c...

Shortening old code golf solutions is so fun.
These days symbolic languages like GolfScript pretty much always win.
 
I thought Golfscript was a traditional, old guard language ?
 
6:51 PM
I'm not saying GolfScript. I mean, all of those GolfScript-like languages (or alternatively APL-like).
Like Pyth and CJam.
 
what's symbolic about them?
 
One character maps to one instruction.
(at least for common things)
 
It's fun when something like Python wins (or even Perl or Ruby, for that matter).
 
Perl/Ruby winning probably indicates a regex question, although Perl doesn't seem to do too badly for a lot of things...
 
7:18 PM
@Sp3000: Actually, I linked an answer a while ago written in Perl 6. It's not a regex question. But stuff like that occuring is fairly rare.
 
That one was a while ago though, i.e. near birth of CJam while :P
 
8:08 PM
@MartinBüttner I see. I thought the challenges were meant to be more work to solve and that the community wrote them together because they had to be right and interesting as a result
hi @Sp3000
 
Hello
 
how things?
 
Hi
 
I'm still coding and it's 7am
 
so... good :)
coding for ppcg or for other stuff?
 
8:09 PM
that's actually the best time to code
 
@Lembik Language design challenge
 
@TheBestOne sorry what is that in reply to?
oh.. what sp3000 is doing?
 
yes
 
sad no one puts so much effort into my recent challenges :(
@Optimizer did you say you might try one of them?
 
no
:P
I am a very bad coder
 
8:12 PM
@Optimizer oh.. must have been someone else :)
now I want to know who it was! :)
 
oh..
i read half
yeah, I said I'd try your subsequence one
 
substring...
cool!
 
but it has been a long week. with 3 movies today ..
 
you made 3 movies already ? :)
 
LOL
IMMA ROCKSTAR
 
8:14 PM
@Optimizer you are not really a bad coder are you? You have lots of rep :)
 
compare it with Martin
 
Rep on this site doesn't measure coding ability.
5
 
@Optimizer well.. that doesn't make you bad.. it just makes you not completely awesome :)
 
See... Peter knows that I am bad
 
@PeterTaylor Good point.. did you get a chance to check out codegolf.stackexchange.com/questions/47163/… It seems to have foxed people
@Optimizer again.. peter and martin are just completely awesome :)It's not a fair comparison
 
8:27 PM
I haven't thought much about it. It looks like it should be a trivial variant on longest common substring, probably tackled by dynamic programming, but it doesn't interest me enough to work out the details.
 
@PeterTaylor OK. If you had the time to give a hint to others that would be great.
@PeterTaylor I think it can be solved in O(n log n) time
which would make it very fast on 1 or 2 million length input
I need to work out how to pose a challenge that does interest you :)
 
@Lembik Since longest common substring can be done in O(n), that's not very good ;) Adapting the suffix tree approach to also have a parallel suffix-with-up-to-one-error tree should be O(n), although the size of the alphabet will be relevant.
 
 
1 hour later…
9:49 PM
@Lembik naming Peter and me in the same sentence (in that context) isn't really fair to Peter :P
@Lembik I think you managed that before with the drunken walk and the property X matrices ;)
 
10:02 PM
I think it would be nice if this new user's proposal wouldn't be ignored:
0
A: Sandbox for Proposed Challenges

FunctinoMake a Space Heater I've been out shoveling snow all day, and my hands are freezing! Heat them up with my computer. Here's an (ungolfed) Linux C solution: #include <stdlib.h> #include <stdbool.h> #include <unistd.h> #include <sys/prctl.h> #include <sys/wait.h> #include <sys/sysinfo.h> int ...

something tells me way too vague, but I don't have any constructive feedback for how to improve it
 
10:59 PM
time to get some work done on Retina. :)
 
11:13 PM
I think if I ever implement all the features I've got in mind for Retina, I need to stop calling it a regex interpreter... it would actually be more if a regex-based (esoteric) programming language then.
 
11:53 PM
Looks like Lego is the next fortnightly challenge.
 
I don't really have any ideas for it yet.
 

« first day (1498 days earlier)      last day (3355 days later) »