The Nineteenth Byte

 Place  Rank                 Name  Points  Penalty
   1st   264               Jakube     100  1:16:04
   2nd   756               Sp3000     100  2:06:19
   3rd  1393               aditsu      71  2:07:48
   4th  1425             PhiNotPi      71  2:14:29
   5th  1481       DennisMitchell      71  2:28:33
   6th  2282           BrainSteel      55    49:25
   7th  2544                 xsot      55    58:20
   8th  3900             kennylau      55  2:07:32
   9th  4744          TheOnlyMego      20    05:46

Kenny Lau

@TheNumberOne Which makes their count... odd....

Nathan Merrill

wooo last :)

TheNumberOne

@MarsUltor Where's me?

aditsu

3:32 AM

well, good news, I have a chance to use even more languages

bad news, I kinda ran out of languages I know

orlp

we can talk now?

Maltysen

ya

aditsu

I just have java left now

El'endia Starman

So, Jakube and Sp3000 are definitely moving on. We still have two more rounds for the others to make it.

Maltysen

@Jakube can you post your C code?

Mars Ultor

3:33 AM

@TheNumberOne What's your GCJ username?

Nathan Merrill

woah, 52% acceptance rate

orlp

I didn't compete, but since for C we know n = 1000, which is small, you can just use repeated brent's cycle detection

TheNumberOne

@KennyLau I thought of that and then "proved" to myself that that was wrong :/

PhiNotPi

For those unaware, the trick to B was to find unpaired numbers.

TheNumberOne

@MarsUltor TheNumberOne

aditsu

3:33 AM

I golfed A and B small

Jakube

@Maltysen gist.github.com/jakobkogler/c7c09b58d1825a881a649029ded19c78

Mars Ultor

  14th  7615         TheNumberOne      20    34:01

Maltysen

@Jakube ty

Mars Ultor

Forgot usernames were case sensitive

Nathan Merrill

Ok, back to my original request: what bugs did you guys run into while coding?

TheNumberOne

3:34 AM

@PhiNotPi C was basically a bowties with the center as a 2-cycle.

Nathan Merrill

(whether the bugs were related to the challenge or not)

orlp

maybe I should've competed :(

PhiNotPi

@NathanMerrill shallow copying, putting stuff inside loops that should be outside, etc.

orlp

I think I would've gotten sub 30, sub 45 for sure

aditsu

I had one bug where I reused a counter variable

Maltysen

3:35 AM

@Jakube so possibility 2 was basically that everyone's bff works out with out wrapping around?

orlp

I just don't like stress coding

TheNumberOne

I had a bug where the list of kids that didn't have any best friends was somehow empty always.

Nathan Merrill

@PhiNotPi @aditsu can you expand?

Jakube

@NathanMerrill In B I wanted to read the N (the length of the matrix) with int N; cin >> N; but forgot the cin >> n part. Wondered for a few minutes why I didn't get any output.

orlp

@Jakube why code in C++ for speedcoding though?

Mars Ultor

3:36 AM

@orlp It's the most popular language for GCJ

Jakube

@orlp Because the solution runs faster.

aditsu

@NathanMerrill I had one bug ... kidding, I had a for(i...) and then reused i in another for loop inside it

Mars Ultor

Twice as popular as Java/Python

Nathan Merrill

oooh, that brings up a good language design question

Jakube

@Maltysen Yes.

orlp

3:37 AM

@Jakube isn't programmer time what costs you, not runtime?

did you write a single program that ran for more than 1 second?

Mars Ultor

@orlp Some brute forces may be too slow

Nathan Merrill

should uninitialized variables be allowed

why?

orlp

@NathanMerrill what do you mean with that

Mars Ultor

@orlp Python-style foo=1

Nathan Merrill

aka, is var a; a code smell?

Jakube

3:38 AM

@orlp No, C-large was the slowest with 0.7 seconds.

aditsu

anyway, I noticed and fixed my bug between downloading the code and submitting the solution

orlp

@MarsUltor that's not uninitialized though

Nathan Merrill

@QPaysTaxes python does it best with n = b if a else c

New Main Posts

0

Q: How to find several rectangles with minimum area to cover the red cells?

In Figure 1, (a) is the input mesh, we want to find several rectangles to cover the red cells in (a), at the same time, the sum area of these rectangles should be as small as possible. Figure 1(b) and Figure 1(c) are two ways. Obviously, (c) is a better way due to the smaller sum of area the gr...

code-generation

Doorknob

let n: i32 = if a { b } else { c }; // Rust :)

Sp3000

3:38 AM

@Jakube Congrats :) Also, nice work on getting C so fast :P

orlp

IMO uninitialized variables should be allowed

Nathan Merrill

then its a function.

orlp

but only for writing to them

not for reading them

Doorknob

let n: i32 = if a {
    doSomeStuff();
    someMoreStuff();
    andMoreStuff()
} else {
    otherStuff()
};

Nathan Merrill

@Doorknob I really like that

Dennis

3:40 AM

@aditsu I have a very unholy mess of CJam for C small. :P

Doorknob

let n: i32 = functionThatReturnsAnOption().unwrap_or(default_value);

Nathan Merrill

the only other time I can think that people use uninitialized variables is in try/catch blocks

which I'm trying to think of another way of doing

because I can't stand try/catch blocks

Doorknob

@QPaysTaxes nil is broken :(

orlp

all right

I'm gna try to do C now

let's see how long it takes me

(5:41 now)

Doorknob

let n: i32 = try!(someFuncThatReturnsResult()); // expands to return Err if func returns Err, Ok(x) otherwise

Mars Ultor

3:42 AM

Dennis' mess:

li{"Case #"V):V": "]o

li:L;l~]:(:K;L,{a{_W=K=|}L*}%{)K=\1>W<#W=},:X{X\f#:e>1<},_{_)K=#!1$W%)K=#!|\,2>&},_@@-
2m*{~:B\:A&{AB&BW%2<={ABW%|}A?}{AB+}?}%:$Q|_,{[2m*{~:B\:A&A{AB|}?}/]:$Q|}*
+:,:e>

N]o}*

Jakube

@Sp3000 Congrats to you too. I also messed up C a few times. But the examples cases pushed me towards the correct solution a little bit each time.

TheNumberOne

@Doorknob var n = if (a) b else c in kotlin :P

Nathan Merrill

@Doorknob I can't do "return Err", because I'm enforcing types

Sherlock9

Well I overslept through most of the first of two classes and still missed 1A

TheNumberOne

@QPaysTaxes same with kotlin

Mars Ultor

3:42 AM

@QPaysTaxes He used CJam because it's less to type, and he's really familiar with it

Sherlock9

Yup

Doorknob

struct Result { Ok(T), Err(V) };

or:

Sherlock9

I only opened 1A 15 minutes ago :D

Almost exactly two minutes before the end

aditsu

I used 3 more languages this round, total so far should be 11

Sherlock9

Didn't even try A-small though I had a solution :P

Sp3000

3:44 AM

@Jakube The examples don't have a case of multiple bowties, so that's what took me 2 hours (i.e. after I finally decided to try random small cases) :P

Doorknob

too many messages got posted in between so I gave up because there would be no context :(

I was going to mention result.unwrap_or_else(|_| computeDefaultValue())

zyabin101

Well done, guys. Unofficial results: Winners are Jakube and Sp3000.

Nathan Merrill

blech. I'm still not a fan of returning errors, even if it is typesafe

Maltysen

@Jakube what are bowties?

Nathan Merrill

@QPaysTaxes I know

Jakube

3:47 AM

@Sp3000 This piece I figured out quite early. Although at the beginning I though only of multiple Pairs like A<=>B and forgot the chains on both ends.

Downgoat

El'endia Starman

I could listen to this guy talk all day.

Doorknob

@QPaysTaxes it just calls the lambda variable _, which is Rust's "black hole" variable meaning "ignore this"

Jakube

@Maltysen Something like this: A->B<=>C<-D

Doorknob

@QPaysTaxes In Ruby it's just another variable name

In Rust it has a special meaning

Maltysen

3:49 AM

@Jakube ohhhh I see

Doorknob

@QPaysTaxes is it? how?

oh yeah, Rust does the same

aditsu

zyabin101

   1st   264               Jakube     100  1:16:04
   2nd   756               Sp3000     100  2:06:19
   3rd  1393               aditsu      71  2:07:48
   4th  1425             PhiNotPi      71  2:14:29
   5th  1481       DennisMitchell      71  2:28:33
   6th  2282           BrainSteel      55    49:25
   7th  2544                 xsot      55    58:20
   8th  3900             kennylau      55  2:07:32
   9th  4744          TheOnlyMego      20    05:46
  10th  4792         KeyboardFire      20    07:20

Downgoat, RikerW, molarmanful, winny3.14 and zachgates7 are excluded from the contest.

Downgoat

@zyabin101 D: there is no Downgoat

@zyabin101 why?

zyabin101

@Downgoat They didn't qualify in the Qualification Round.

Downgoat

3:51 AM

:(

aditsu

s/They/you/

Downgoat