« first day (1534 days earlier)      last day (3310 days later) » 
00:00 - 19:0019:00 - 00:00

12:02 AM
I thought google code was dead.
 
Google Code as in their source hosting? Or Google Code Jam as in the coding challenges?
 
Why aren't y'all participating?
 
@AlexA. no
 
@BrainSteel: Looked at it, didn't have the attention span at the time to understand it, didn't bother to look at it again.
 
12:18 AM
Eh, can't fault you for it. Not entirely sure why I'm doing it :P
 
12:36 AM
You guys are doing it for the same reason you participate on this website
 
@NathanMerrill I'm being forced to by my brain?
I can't stop
 
I just tryed to post the small solution to D
and it says it's wrong
but I'm quite sure it's right, I checked them just in case
 
D is the kind of problem where there seems to be lots of edge cases
 
but there aren't
and as I say
I checked them all by hand
the small test is only for up to 4minoes
 
you can try again
 
12:40 AM
so it's pretty easy to check
 
well, they don't give you the correct output
 
I know, but I made them in my brain
and they looked right
ooooh right
just found the prooooblem
 
it's the freaking
_X_
XXX
I have a love-hate relationship with that tetris piece
 
oh, I had no idea that max X was 4
that makes things much simpler
I didn't try it because I thought that max X was much larger
 
12:48 AM
no, max X for the small test case is 4
and for X > 6 richard always wins
because he can always create a piece with a hole
 
so the only problematic inputs are for x=5 and x=6
 
well, not always true
if the width or height = 1
 
when not?
well
but then he wins automatically
 
that's also not true
 
12:49 AM
because he just chooses an L shape
 
wait, his piece doesn't have to fit?
 
richard is the one trying to choose a piece so that it can't be filled
gabriel is the one trying to fill the space
 
I thought that the piece richard picked had to at least fit in the grid
 
no, it doesn't
in the test cases it's shown
moreover, in the small test case there is one that is a grid of 1x1 and has to be filled with 3-ominoes
so it's right out impossible
it's pretty easy all in all
compared with the others at least
All right!!
Let's try the large input
 
I'm glad I'm not attempting problem D. :-P
 
12:59 AM
@ChrisJester-Young are you attempting any of them?
 
@NathanMerrill I solved A hours ago. I'm currently working on C.
 
1:20 AM
This problem D solution has got to be my most efficient Code Jam solve ever (hardcoding) :P
 
@Sp3000 @rcrmn I'm curious how you guys are solving D Are you simply testing for a bunch of different cases, or are you iterating through the different pieces?
 
I'm just hardcoding all the cases :P e.g. if X == 3, then (R*C) % 3 != 0 is an immediate fail
 
1:41 AM
@Sp3000 Exactly this: I check for the area to be multiple of x and then check that r and c meet the constraints, mainly that they are greater than some number
 
Plus an edge case or two :P
I hope I've got them right
 
I haven't found any edge case that I could not reduce to a size constraint
but I haven't look too deep into it
 
I think I've got one, but I might be wrong
 
I'm going to sleep, but tomorrow I'll read this if you explain it
good night!
 
:P k night
 
1:56 AM
T-minus 3 minutes...
 
Everyone qualified? :P
 
I'm qualified
 
Everything except C large which I didn't attempt, woo
 
I did not qualify, lol. That's what I get for leaving things to the last minute.
 
@rcrmn The one case I had which wasn't dependent on min(R,C) = some number was 3 x 5 for X = 5
 
2:06 AM
man, the top guy finished all of the problems in 1 hour
 
I made it :D
 
I'm taking my 24 points and being happy :)
 
If I find a part of a solution to a challenge (not a full program or implemented code, just an idea for what could be possible), is it OK to post that in the comments of the challenge?
 
@ASCIIThenANSI I don't understand.
 
2:56 AM
28 points :P
 
25 here. To my credit, I only started with 12 hours left on the clock :P
 
unfortunately, we can't discuss the problems for the next round
 
right.
 
3:20 AM
@ChrisJester-Young If I find what seems like a good solution (or part of one) for a challenge, but not any actual code, can I post it in the challenge's comments?
 
Why is Martin suddenly anon here and here? I could swear that it said his name there yesterday.
 
@Calvin'sHobbies Chris disassociated it so that he could accept the answer without waiting two days. :P
 
I see. Mods have strange powers :P
 
SE devs*
 
@ASCIIThenANSI I'd say giving suggestions for how to complete a challenge is fine in comments, especially if the challenge is particularly hard or does not have many answers. If you want to post an entire block of code then you should probably link it, e.g. "Here's a pastebin of my attempt.", or even discuss it in chat instead.
@Doorknob I think of Flatland when I think of mods. Normal users see the site exist only in 1 or 2 "dimensions". Mods see 3. Devs must see 4...
2
 
3:40 AM
@Calvin'sHobbies OK, thanks for the info.
 
4:36 AM
@Doorknob I reassociated so that I'd stop getting these strange questions, but unfortunately, the system now treats it as an OP accept (since that's what it is) and so it's no longer floated to the top. D'oh.
 
0
A: Sandbox for Proposed Challenges

mbomb007Identify an "Identity" BrainF*** Program Your goal is to write a program or function that takes a string representing a BrainF*** program, and prints or returns a boolean value (or your language's equivalent), identifying whether the input program was equivalent to an identity operation or a NOP...

 
Maybe I should associate it with Community so it'll float back to the top.
 
@ChrisJester-Young Now the answer is owned by Community and the comment is still anon... O_o
 
@Doorknob Yeah, the comment can't be reassociated, sadly.
I should nuke the comment to avoid confusion, since Martin can indeed upvote it now.
@Doorknob Curious, when you look at the post history, does it show all my crazy shenanigans? I know it shows it for dev and doesn't show it when I view it in an incognito window.
and I'm sure as hell not going to downgrade my account just to test it out. ;-)
 
@ChrisJester-Young Nope, it looks exactly the same as it does in incognito.
 
4:41 AM
Bummer.
@Doorknob (Also, see my comment in the other chat room. :-P)
 
Second weekend closed well. 5 more shows of Drood for this droog.
 
4:57 AM
Uh, okay.
 
I'm playing the upright bass in a local musical.
Somehow, I'm the only one who can tell how awful I am. :)
I bought the thing a month ago, not having playing one in 10 years. So, I'm riding on talent rather than skill, per se.
Apropos of nothing. That's what's going on with me.
More programmatically, I taught inca3 to do currying like J does.
` (2&+)⍳9
9 #
2 3 4 5 6 7 8 9 10
`
So, I'm pretty thrilled about that. Nowhere near being a complete language, yet. But I'm very pleased with the code that's written. 1400 lines of dense C.
It's probably closer to 10,000 if written in a normal C style. But it's easier to navigate a small file.
 
5:19 AM
@orlp thanks for the answer.. although it doesn't actually fit the spec :)
 
 
3 hours later…
8:36 AM
@luserdroog So, what's your inca3 about?
 
9:03 AM
@ChrisJester-Young I actually found D easier than B and C... but looking at the results, apparently loads of people failed the large set.
@NathanMerrill @rcrmn @Sp3000 if you're interested, I also just hardcoded all special cases for D:
n, w, h = line.split.map &:to_i
min, max = [w, h].minmax
winner = 'GABRIEL'
if (w*h) % n != 0 ||
   (n+1)/2 > min ||
   n > max ||
   n > 6 ||
   n == 6 && min == 3 ||
   n == 5 && min == 3 && max == 5 ||
   n == 4 && min == 2
    winner = 'RICHARD'
end
 
python?
 
Ruby
the first one is simply if the grid can be partitioned into N-ominoes. the second is if Richard can fold the shape into an L that is too wide on either side to fit along the shorter dimension. the third is if an straight polyomino can fit along longer dimension. the fourth is if Richard can choose a polyomino with a hole. the other three are special cases, in which Richard can choose a polyomino which separates the grid into two regions which are not multiples of N.
 
 
1 hour later…
10:18 AM
Huzzah for hardcoding
@MartinBüttner btw B was actually really easy once you thought about it the right way
Rather than trying to split stacks all the time by introducing more plates, it's easier to think of it as adding an extra diner to help finish the same plate
 
I'm not entirely sure how that helps
 
Say you have 10 15, and both plates start with 1 diner. That's 15 minutes.
To improve on that, you add a diner to the one which will finish last, so the 15 pile. That gives 1 minute + 10 minutes, since 10 will now finish last
So add a diner to the 10 pile, giving 2 diners on each. Now you have 2 minutes + 8 minutes, which is 15/2 rounded up
Add a diner to the 15 pile, giving 3 minutes + 5 minutes
 
oh right, that makes it easier to split the plates evenly
 
^^ yeah that
 
how do you know when to stop?
 
10:27 AM
I just did it for M iterations, where M is the max number of pancakes on one plate
Since that's a clear upper bound
 
oh that makes sense
did you do C as well?
 
I did C small by constructing the string and traversing in one passing, checking whether at some point I get i, at a later point I get k (= ij) and at the end I get -1 (= ijk)
I was too lazy to think about how to handle the repetition :P
 
yes, that's what I did
oh right
you can simply cap the repetitions at 12 + (X%4)
 
Was that still fast enough? I wasn't sure
 
yes
it solved the large set in one or two seconds
 
10:30 AM
Damn :/ should have tried that then
The numbers looked big :P
 
nah
that gives strings of length 150000
(at most)
the base-string length was limited to 10k
 
Ah right, I forgot about that part
Hmm where did the 12 + (X % 4) come from? I thought the repetition could be capped, but I didn't work out an exact bound
 
I first tried 12, but that failed. basically, 4 repetitions always give unity. so if i took more than 4 repetitions you could also find an i by removing four repetitions. same for j and k. so that gives you 12 as an upper bound.
however, when you omit repetitions, you also need to make sure that those are unity, otherwise they'll mess up your k
 
Oh... that makes sense
 
so you use 12 + (X%4) to ensure that a) you have at least 12 repetitions and b) you omit a multiple of 4 repetitions.
of course, if your base string comes out as -1, you could use 12 + X%2 and if it's 1, you can just use 12
but that seems like a very minor optimisation
I also exited early if the whole thing didn't evaluate to -1
 
10:37 AM
:) nice
 
I'm amazed by how well they anticipated the difficulty of the small vs large test set for D (looking at % correct and the points it gave)
 
I'm more surprised at just the raw percentage that passed D large. I wonder if there were some common mistakes floating about
 
I'm pretty sure a lot of people just missed those three special cases where Richard has to separate the grid into two regions
it's actually quite funny that there are less than 100 distinct inputs for the small test set :D
plus, half of them are equivalent
 
:P
 
11:02 AM
yingluck
 
what is the removed post ?
woot! we are in the 90s!
 
yeah
I'm still contemplating whether I should attempt this Unlambda
but I don't think I can be bothered
1998 has Malbolge...
 
I dare someone to do JavaScript for 1995... but in JSFuck :D
 
oh, 1998 also has C++
 
@Sp3000 JS is 96
95 was with some other name and out only as public beta ..
C++ was made after JS ? :D
that is why JS did not have classes from beginning ..
 
11:09 AM
Java did (obviously) exist though
actually C++ was only standardised in 98
it's been around since 83
 
contrary to popular beliefs, JS was not derived/influenced from Java
 
it didn't come up with the name independently though :P
 
I can ask Brendan about it, but I bet some of my friends have already asked that, so I can ask them
 
With 3x5 you have this:
__XX_
_XX__
_X___
Which Gabriel won't be able to fill
 
11:22 AM
Hah, apparently my system works quite well as a Javascript module. i.imgur.com/7ddB2Zk.png -> janijohannes.kapsi.fi/frp/test.html
 
11:42 AM
Should I implement a random generator/checker for the jigsaw puzzle challenge?
 
12:40 PM
0
A: Sandbox for Proposed Challenges

randomraConnecting letters code-golfascii-artconnected-figure(weird tag but exists) You should write a program r function which receives a block of chars represented as a string and outputs or returns a similar block as a string in which the letters adjacent in the alphabet are connected. A visual exa...

 
@randomra I like that :)
 
@MartinBüttner I was surprised it hasn't been asked yet
 
you should include a test case that goes up to z
(you could just expand the second one)
can we assume the input is padded with spaces to a rectangle?
 
@MartinBüttner aeti sqjh k p u cfm vb owgzyx rnd l => aeti--sqjh-k--p--u--cfm-vb-owgzyx-rnd-l ? (in the name of laziness)
 
lol, sure
 
12:52 PM
@MartinBüttner yes, will add that
 
I fixed your formatting btw... block-formatted code in bullet lists needs a full 8 spaces of indentation
 
@MartinBüttner great, thanks. There are some weird magic stuff with markdown.
 
1:44 PM
@Lembik what do you mean?
 
@orlp it was meant to output answers until I kill it
@orlp yours takes a value of N as input
it should output 2,12,230,12076,1446672,...
and take no input
also if you make it python 2 friendly I can try it with pypy :)
I think "yield from" doesn't work in python 2
 
ok give me a sec
 
there is pypy3 I just don't have it installed
 
yay, first Strunk & White on our meta. :)
 
:)
what do you get that for?
 
1:49 PM
click the link
 
80 edits
 
nice
@orlp clicking the link didn't do much for me
but hovering over it did...
 
it says "Edited 80 posts" next to the badge
 
it does.. I wasn't sure if that was a description of the badge
or just a statement about you
but now I know
 
Such strunk
 
1:54 PM
Fuck!
why doesn't Ubuntu come with a good memkiller by default >.<
everytime I try something memory intensive on Ubuntu I end up having to forcibly restart
can't even ctrl+c or ctrl+alt+f1 to kill
 
So that's why my Ubuntu froze all the time, huh.
 
@MartinBüttner Speaking of which, I wish more people followed rule 1 in Strunk & White.
 
@ChrisJester-Young I've intentionally attempted to stray away from that rule
 
@ChrisJester-Young In dutch we always use 's for possessives, but "The chicken is rich, it's wealth is massive." is incorrect.
 
2:02 PM
That's not what rule 1 is about.
 
oh
 
Rule 1 is about saying "Chris's code" vs "Chris' code".
The former is compliant. The latter is not.
 
I'm all for "Chris's"
 
@MartinBüttner Thank you!
 
that's the possessive, no?
> Form the possessive singular of nouns with 's.
"it" is a possessive singular noun, no?
 
2:03 PM
@ChrisJester-Young Not a fan of rule 2 though ;)
 
@orlp Its, his, hers, ours, theirs, whose, etc. are special exceptions.
 
@orlp "it" is a personal pronoun
 
so outside of those exceptions, 's is always correct?
 
@MartinBüttner I'm a stickler for Oxford commas.
@orlp Outside of exceptions (not just the ones I listed), 's is correct when applied to singular nouns.
 
@orlp "Exceptions are the possessives of ancient proper names in -es and -is, the possessive Jesus', and such forms as for conscience' sake, for righteousness' sake. But such forms as Achilles' heel, Moses' laws, Isis' temple are commonly replaced by "the heel of Achilles", "the laws of Moses", "the temple of Isis""
 
2:05 PM
Thus "a bus's tyres" or "those buses' tyres".
 
@MartinBüttner why is Jesus an exception?
 
@orlp Historical usage.
 
@ChrisJester-Young I think the main reason I don't like it is because it's plain wrong in German.
 
@MartinBüttner Interesting, fair enough.
 
Same in dutch
A, B en C
 
2:08 PM
And Finnish, but we probably copied it from the Germans.
 
1
A: Oxford Comma Conventions

Adam KatzThere are situations where use of the Oxford comma will make or break a sentence. Choose a style and be consistent. When you run into a situation in which your choice suggests a misinterpretation of the sentence, rewrite it in another manner to avoid the confusion. Consider these two pairs whe...

 
unlikely? isn't finnish etymologically unrelated to german?
 
Yes, but it didn't really exist in standardized written form until relatively recently.
 
@ChrisJester-Young Yes, I'm aware of that example. That's why I meant, I don't think the Oxford comma is a bad idea, I just don't like it because of habit. ;)
 
2:11 PM
oh another confusion with dutch
in dutch we create plurality using -en
so "boek" (book) -> "boeken" (books)
but lent words from other languages can't use this
for example "taxi"
 
Boxen and Unixen. ;-)
 
@ChrisJester-Young oh, I didn't read it properly. I was not aware of the counter example in that answer.
 
those words we create the plurality by appending 's, so it becomes "taxi's"
so in general I'm always very confused about when to use 's in english, because in dutch we use it for entirely different reasons :P
 
god, how I hate that double clicking the upper left corner of a window closes it in Windows...
it's such an obscure UI feature, and sometimes I just misclick the leftmost tab in my browser twice in a row...
 
@MartinBüttner didn't know that, never actually double clicked on the left corner of a window
 
2:13 PM
@MartinBüttner That is for historical reasons.
 
you can hate M$ all you want, they do very long backwards-compatability
 
Finnish was first written using the typography and spelling rules of Swedish, German, Latin and who knows what else. At some point it was decided that creating consistent spelling rules from scratch would be a better idea.
 
Back in the Windows 3.0 (or perhaps even earlier days), the upper-left corner brings up the "system" menu. A shortcut was to close the window by double-clicking it, because it predated the Windows 95 convention of the top-right close button.
 
@Zgarb you should add a test case where u == v.reverse
@ChrisJester-Young that menu still exists on many applications
 
Well, of course.
 
2:15 PM
and even though the "close" option is bolded in that menu, associating a double click with closing it, just seems weird to me
 
How else would you close a window, before Windows 95? Other than Alt-F4 or Ctrl-F4 (for closing an MDI window), that is?
 
@orlp I don't hate MS... there are just a few really quirky "features" in windows :D
@ChrisJester-Young I don't remember using a mouse before Windows 95
 
looking at past codejam problems...
how would you guys get the three digits before the decimal point in (3 + √5)^n
 
then again, I was very young before Windows 95
 
when n can be as large as 2000000000
 
2:17 PM
@NathanMerrill by using Mathematica.
 
@MartinBüttner I remember playing Minesweeper before Windows 95, so obviously I remember using a mouse. ;-)
 
@MartinBüttner Done.
 
in c++
or java
or python
 
@NathanMerrill you implement a tiny tiny CAS
 
2:18 PM
@ChrisJester-Young well I think I may have used a mouse in games and stuff as well, but nothing that wasn't fullscreen as far as I remember
 
using gnump
 
@NathanMerrill (a+b*sqrt(N))^k = c+d*sqrt(N)
 
what is c and d?
 
you loop
to find that out
just work out a few steps on paper
 
@NathanMerrill with iterated squaring (or what that's called) it's not hard to find out in O(log k)
 
2:23 PM
@NathanMerrill c is oeis.org/A098648
@NathanMerrill d is oeis.org/A084326
when solving computational problems
OEIS is your biggest friend
just compute a few terms
dump it into OEIS
and you've usually got a result to implement instantly, or at least a name to look up more references
@NathanMerrill in particular, the expansion of (3 + √5)^n is very closely related to the fibonacci series
 
what is the set of a+b*sqrt(N) called where a,b integers?
I thought Gauss integers but that's apparently just for N=-1
 
2:51 PM
it has been done... there's an answer to the Fish challenge in ><>
 
bounty it
 
@orlp testing it now
@orlp gets to n = 5 even with pypy
 
ok
it's a slow solution :)
 
:)
but a winning one!
(so far)
stackoverflow.com/questions/29484864/… has some new ideas which might be interesting
or not :)
 
3:20 PM
0
Q: How to write good tag wikis?

Martin BüttnerI occasionally find that important tags are missing tag wikis or have only very rudimentary information in them. At the same time, I often find it hard to write good tag wikis. The challenge-type tag wikis don't seem to have that problem - I think they are all fairly decent: code-golfcode-chall...

 
@NathanMerrill I have made an logarithmic time solver for this, in Python
@NathanMerrill using Lucas numbers: gist.github.com/orlp/7df4f8db3e0f03c3329b
@Lembik you might find it interesting as well ^
solves for a, b in (3 + sqrt(5))^n = a + b*sqrt(5)
despite being a logarithmic algorithm it's still quite slow because the numbers get huge
 
3:47 PM
@ASCIIThenANSI you might want to slow down with the tag wiki edits a bit until the above meta post has reached some sort of consensus
 
@MartinBüttner how long did it take in mathematica to find (3 + sqrt(5))^n for n = 2000000000?
 
I didn't actually try
 
oh?
 
hm it doesn't expand it by default
 
why not? what was the original problem?
 
3:49 PM
I don't know
2e5 already takes a while
 
@MartinBüttner 2e5 is instant for my Python solution
the solution has 143801 decimal digits
 
yeah I don't think Mathematica does anything clever there
 
with gmpy I might be able to calculate for n = 2000000000
hope I don't run out of memory
these numbers are massive :P
phew
my laptop barely had enough memory to compute the numbers
they're 1438010552/1438010553 decimal digits long
 
4:31 PM
ugh, I've got one of those days where I just can't bring myself to do anything productive...
 
I had one of those yesterday
but I managed to snap out of it at the end
still did some good work
 
still at 1999 only ? :D
I thought that I'll miss 96 while i was away
 
what do you expect on a sunday...
also you can't really "miss" any year
 
4:47 PM
I can miss the language though
 
5:04 PM
@JoeZ. yes, I've created a new rational numbers tag. yes, you have quite a few questions this tag is applicable, too. no, you don't need to retag all of them now. ;) (I've got a list of all applicable questions I could find, and will go through them myself slowly.)
 
5:50 PM
@orlp any thoughts for a faster solution ? :)
 
@JoeZ. Out of curiosity, is it "Joe Zee" or "Joe Zet"?
 
@MartinBüttner is it possible to see which tags no one uses?
 
it's "Joe 'zed'"
although
my last name is Zeng
also, @MartinBüttner Noted.
A "rational numbers" tag?
which questions would those be...?
 
I think I saw at least two about resistors
 
oh
yeah, those two
most of my problems deal with integers only, TBH
or, well
I guess also the repeated decimal one
 
5:57 PM
@Lembik the tags page is sorted by number of questions using the tag
@Joe, yes that one too
 
6:44 PM
How do you search for something only in question titles/bodies (no answers)?
 
@Calvin'sHobbies is:question
 
Thank you! I should have known that by now
 
pro-tip: hit enter in an empty search bar to get to the search page. then expand the Advanced Search Tips on the right
@Calvin'sHobbies are you cooking something up? ;)
 
hey, I'm actually wanting to do an origami-based challenge
 
@Calvin'sHobbies yes you should have :P
@NathanMerrill fold to form a shape ?
 
6:59 PM
If I provide a square paper and a list of folds (and whether they are concave or convex)
are those folds guaranteed to provide the same shape?
 
list of fold is via 4 coordinates ?
 
@Nathan that sounds vaguely familiar. You might want to search meta if it's already in the sandbox somewhere.
 
00:00 - 19:0019:00 - 00:00

« first day (1534 days earlier)      last day (3310 days later) »