The Nineteenth Byte

Alex A.

03:00

h1 = #, h2 = ##

Doorknob

oh, h2s is the plural of <h2> :P

@Doorknob Firstly, we should discourage the use of single-# in answers. It's bad practice to have more than one <h1> in an HTML page.

Alex A.

Bad practice according to Peter?

I disagree entirely. I think h1 looks much nicer.

What's worse is mixed though.

Doorknob

bad practice according to every HTML standard ever

(I just had never thought of it that way, and when Peter brought it up I realized h2 was more proper)

Alex A.

Why does it matter?

Doorknob

54 secs ago, by Doorknob

bad practice according to every HTML standard ever

Alex A.

03:04

1 min ago, by Alex A.

Why does it matter?

We're not writing HTML pages.

Doorknob

uh, yes we are

view-source:http://codegolf.stackexchange.com/questions/60152/standard-loophole‌s-only

@quartata haha, beat me by 0.1 of a byte

quartata

yasss

Alex A.

Well yeah, we are writing HTML pages in that sense, but I mean that isn't the point of what we're doing.

Doorknob

sure it is

Alex A.

If we were working on one big collaborative HTML page, then I would agree.

Doorknob

03:06

How else are people supposed to view your posts? >.>

quartata

I didn't know CJam could grab web pages.

Alex A.

CJam can do a few things. Pyth can do more.

Doorknob

@AlexA. well, we are, aren't we?

Alex A.

I realized that was poor wording on my part right after I said it.

Because yeah, we are.

@AlexA. Who starred you?

Sp3000

@mbomb007 Try something like self.end = 0 in Tape.__init__, self.end = max(self.end, index+1) in Tape.__setitem__ and replace the while in Interpreter.call with while pc < self.tape.end: maybe?

(might have an off by one error, haven't tested thoroughly)

El'endia Starman

03:09

Hey, I see double underscores on both sides! Must be Python!

Alex A.

@Sp3000 There are three difficult concepts in computer programming: naming variables and off by one errors.

9

quartata

0

A: Standard loopholes only

GNU sed, 0 bytes Even better..

The real winner

Maltysen

@quartata but you didn't get the bonus ;P

quartata

Still won.

First code golf contest I've won.

Tastes like victory.

Doorknob

:D

Maltysen

03:13

would've won tumbleweed if not for the -3...

Doorknob

or the answers

or the comments

or the views

or the anything

El'endia Starman

@Maltysen Err, Tumbleweed is no comments, no votes, no answers, and low views for a week or two weeks.

Maltysen

derp

El'endia Starman

Which is probably effectively impossible on PPCG, incidentally.

xsot

Hello, anyone knows how I can insert tabs in my code submission?

Doorknob

03:15

you can't

Dennis

@El'endiaStarman Not quite.

codegolf.stackexchange.com/help/badges/21/tumbleweed

Alex A.

@El'endiaStarman codegolf.stackexchange.com/help/badges/21/tumbleweed

Doorknob

@xsot The Stack Exchange software automatically removes them

El'endia Starman

It's happened four times though. All in 2011, 2012, 2013.

Alex A.

@Dennis Damn, ninja'd. :D

El'endia Starman

03:16

Huh, zero score. I guess +1/-1 is possible.

xsot

Oh, okay. How do people usually work around it then?

El'endia Starman

I wonder if deleted answers count.

@xsot Four spaces.

Doorknob

@xsot What's the context? Why are tabs necessary?

El'endia Starman

Also, really, tabs are converted automatically, from what I've heard.

Doorknob

they are

Alex A.

03:17

Note that tabs are rarely necessary. Even when writing code, one should use spaces to indent rather than tabs.

xsot

I have a python submission that has 3 levels of indentation.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

CUDDLE calculation code-golf math According to the Wikipedia page on the number 69, it is of note that 692 = 4,761 and 693 = 328,509 together use all decimal digits. The number 69 is in fact the lowest number that satisfies this property. For a similar reason, 32,043 is remarkable: 32,0432 = 1...

El'endia Starman

> LogarithEm

Okay, that's pushing it.

Dennis

The E had to come from somewhere...

Alex A.

Why were you casually browsing the Wikipedia page for 69?

3

Dennis

03:19

Found a link in this challenge:

4

Q: Find a numeric coincidence when representing numbers in various bases

Recently, when doing some code-golf challenge, I came up with with two solutions, in 69 and 105 bytes. It's a remarkable coincidence, because: 69 (decimal) = 105 (octal) 69 (hexadecimal) = 105 (decimal) What other numbers have this property (using different bases)? You should help me answer t...

Alex A.

Likely story. ;)

Maltysen

@Dennis maybe work "exponent" in their somehow

"CUmulative Decimal Digit Least Exponent", maybe?

Sp3000

@xsot Can you do something like for j in range(128*(len(set('%o0'%i))>7))?

Dennis

@Maltysen Neat, I'll use that.

El'endia Starman

Seems to me like all CUDDLE-less numbers are some power of 10 (and 0).

xsot

03:27

@Sp3000 for j in len(set('%o0'%i))/8*range(128) works. Thanks!

Alex A.

@Doorknob Any idea who this guy is?

Doorknob

totally not me

Sp3000

Ah of course, i is at most length 8 :P

Alex A.

Doorknob

Compare.

Alex A.

03:29

Illuminati confirmed

> ARCTIC SCIENCE PRIZE [NORWAY, GERMANY, USA, CANADA]: Eigil Reimers and Sindre Eftestøl, for testing how reindeer react to seeing humans who are disguised as polar bears.

PhiNotPi

I just went and saw The Martian

so much space

Doorknob

I just finished reading that!

It's awesome.

Alex A.

I also just finished reading that

Can confirm: Is awesome

Was the movie theater underwater? Did you have to tread water during the movie?

PhiNotPi

no

Alex A.

> PROBABILITY PRIZE: [...], for making two related discoveries: First, that the longer a cow has been lying down, the more likely that cow will soon stand up; and Second, that once a cow stands up, you cannot easily predict how soon that cow will lie down again.

Sp3000

03:38

Special announcement: anarchy golf now supports ><> (and it has replaced A+ as being the first language in the list :D)

PhiNotPi

So... the amount of time a cow stays standing follows a geometric distribution?

Alex A.

No, then it would be easier to predict when it will lie down again.

PhiNotPi

(I should have actually said "exponential distribution" which is shaped like geometric but is for real numbers not integers)

Alex A.

Yes

But that would apply to the probability that a lying cow will stand.

PhiNotPi

An exponential distribution describes things like half-lives. The probability of decay is completely independent of how long the particle has gone without decaying.

Which is similar to how the probability of the standing cow sitting is independent of how long it's been standing.

Alex A.

03:46

Yep. Because the exponential distribution is "memory-less"

PhiNotPi

Which would means that the time it stays standing follows an exponential distribution.

Alex A.

Well, we don't know that it's independent of how long it's been standing. Presumably it isn't; one could imagine that the cow will want to lie down after having been standing a while.

El'endia Starman

@AlexA. What year? (i.e., link?)

PhiNotPi

> The probability of cows standing up within the next 15 min increased (P < 0.001) with lying time in all experiments, which was consistent with the first hypothesis. The probability of cows lying down within the next 15 min did not significantly increase with standing time.

Alex A.

@El'endiaStarman I don't know how to link to specific spots on that site but it's 2013.

PhiNotPi

03:49

From the paper abstract.

El'endia Starman

Ah, PhiNotPi is doing what I wanted to anyway. Read the paper. :P

PhiNotPi

appliedanimalbehaviour.com/article/S0168-1591(10)00054-7/pdf

Alex A.

Haha I didn't bother to read it

PhiNotPi

lol @ "$31.50 USD (24 hour online access)"

Alex A.

That's a deal!

I'd totally spend $31.50 to read cow math.

PhiNotPi

03:51

That's probably one of the worst deals I've ever seen.

El'endia Starman

> The probability of cows lying down within the next 15 min did not significantly increase with standing time.

That sounds like it's a uniform distribution.

Dennis

How hard is it to find the integer partitions of a number?

El'endia Starman

Hard-ish.

I don't remember if it's in NP, but I think it is.

DankMemes

@PhiNotPi such exciting research!

Alex A.

@Dennis 7/10

Calvin's Hobbies

03:53

We need a cow-spots drawing challenge

Alex A.

Closed as too broad

PhiNotPi

(I assume Dennis is talking about golfing. Time complexity is of no concern.)

Alex A.

Haha like Dennis' groups solution that was O(n!^n)

Dennis

I'm writing a Pyth program. Input 28 has been going for 5 minutes without results...

Alex A.

CJam is going to miss you once you've completely turned to the dark side.

Dennis

03:54

There seem to be only 3718 partitions...

El'endia Starman

The algorithm on Wikipedia seems like it would be in P with just a quick glance.

Sp3000

If you want it to be fast, you could pentagonal number theorem it I guess

Calvin's Hobbies

@AlexA. >:|

Alex A.

:3

Calvin's Hobbies

There are like 6 versions of starred alex-is-wrong tags now

Dennis

03:57

@Sp3000 No, Pyth has the built-in ./. I'm trying to use it in a challenge.

Alex A.

@Calvin'sHobbies I know. :/

Calvin's Hobbies

No you don't

Sp3000

@Dennis Oh, I thought you were trying to test your abelian question without ./. nvm

PhiNotPi

I just thought of a random question about algorithms and time complexity. Let's say that you have a problem, and there is a set of algorithms that can solve the problem. Is the following situation possible? For any E>1, there exists an algorithm to solve the problem in O(n^E), but there does not exist an O(n) algorithm?

El'endia Starman

Sorting?

You have to make at least n-1 comparisons.

PhiNotPi

04:02

Isn't sorting like O(n log n)? Whether or not n*logn < n^E depends on the size of the list.

Alex A.

@xnor Thank you for that meta post. I think that's an important discussion.

PhiNotPi

I was thinking that the algorithms themselves have time complexity of the form O(n^value)

Sp3000

Ditto Alex's comment

I think someone mentioned taking us off HNQ at some point, but I forgot how that discussion went

PhiNotPi

Maybe I can phrase it this way... can you have a set of algorithms O(n^2) O(n^1.5) O(n^1.25) O(n^1.125) etc.

without actually having an O(n) algorithm?

Sp3000

The closest example I think think of is matrix multiplication, which is O(n^3) naive, O(n^2.373) exists but we don't know whether we can get O(n^2)

PhiNotPi

04:07

That's actually what inspired the question.

Sp3000

But I can't imagine how you can get arbitrarily close but not reach it though

... ah :P

Alex A.

@Sp3000 Our site traffic would plummet.

New Meta Posts

3

Q: Easy challenges and Hot Network Questions

I've noticed many challenges getting to the top of Hot Network Questions because they are really easy and get many answers that feed the activity-rewarding hotness formula. In turn, the visibility attracts more answers and more votes and keeps the cycle going. Can we do something about this? Som...

discussion challenge-quality

El'endia Starman

It's probably worth noting that Christianity.SE actually has its questions penalized, so it's harder for them to hit the HNQ list. This was a good decision in the past because otherwise there would have been many, many low-quality questions that got upvotes and answers. However, this is now under review because we've tightened our scope and we have better quality control.

8

Q: What is the reason for the lack of "hot network questions"?

In my two and a half months here, I don't believe Christianity.SE has ever had a hot network question, whereas the average site should get at least one every few days. I see several possible reasons for this: Our site has a much lower vote-to-view ratio than average Our questions attract signi...

discussion status-review answers questions voting

@PhiNotPi Perhaps you can treat this limit-style. For every n, there exists an E where n^E - n < 1, so O(n^E) ~= O(n). In other words, a true O(n) algorithm may not exist, but you can get close enough that it doesn't matter.

Sp3000

04:27

@xnor How much more expensive is it to convert to Python 3?

xnor

I think only one char for the print

lemme make sure

Sp3000

Because 3.5 {*range(1,8)} should save 2

So -2+1?

xnor

haha right

or a bytes object

{*b'1234567'}

Sp3000

Ahaha, nice :P

PhiNotPi

I'm making some KOTH server progress.

Alex A.

04:30

Yay

xnor

@Sp3000 same # chars though

Sp3000

btw if you Python 3, you can error out with -print(l)

xnor

probably worse if something needs escaping

PhiNotPi

My server has now been able to download and run this "KOTH" from GitHub.

xnor

@Sp3000 that's great

Alex A.

04:32

@PhiNotPi Nice!

Sp3000

Too bad {n}|s=={0}and-print(l) is the same length :/

xnor

it might be worth it to get everything on one line

PhiNotPi

The results file is stored here: koth-phinotpi.rhcloud.com/servertest/report.txt

xnor

replacing the if with another short circuit

and the loop with the comprehension

Sp3000

Actually, I don't think f(1234) works with the current one

(or f(29) for that matter)

xnor

04:34

uhoh

oh, does it not check that p==0 at the end?

I think changing {n,p}|s=={0} should do it

maybe i should just check directly that sum(l)==n

Sp3000

I don't think that works

Since n changes

xnor

yes, this would be with not updating n

Sp3000

Ah, right

In any case, stupid way to save bytes maybe? def f(n,s={*range(8)},l=[],p=0):{n,p}|s=={0}and-print(l);p and f(n-p,s,l+[p]);[f(n,s-{x},l,p*10+x)for x in s]

xnor

you know, I think that could be a lambda :-)

Sp3000

... actually yeah it could :P

xnor

04:42

oh, wait, that probably doesn't actually save chars given that we have to print

Sp3000

Ah :P

xnor

how are you dealing with s having 0?

Alex A.

He's been in therapy for it

xnor

you could do p*10+x, but you're not doing that

Alex A.

05:58

@PhiNotPi Relevant to your interests

Beta Decay

06:39

@PhiNotPi Also relevant to your interests

Lembik

07:52

hi all

Martin Büttner

Brace yourselves, binary matrix challenge imminent.

Lembik

@MartinBüttner :)

or python bashing...

Lembik

08:16

I notice a sad lack of fastest-code challenges since my absence :(

New Sandboxed Posts

08:40

0

A: Sandbox for Proposed Challenges

Repeatable combinations List all of the repeatable combinations (I'm not sure how it is usually called) of size k from n items. A repeatable combination is an unordered multiset that every item in it are also in the n items. Note that: It is unordered. So a previously printed set with a diffe...

undergroundmonorail

@Sp3000 i was actually really disappointed that they didn't have it the last time i went there

nice

Lembik

09:07

@MartinBüttner I have a slightly controversial challenge idea :)

by which I mean.. very...

fancy hearing about it?

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Make your language unusable Try to write some code in your language and make it not satisfying our criteria of being a programming language any more. Examples like making it not able to output, or disabling all the loop constructs so it won't be able to do primality test and making sure the use...

New Sandboxed Posts

09:35

0

A: Sandbox for Proposed Challenges

"Standard compliant" web browser Your task is to write a standard compliant web browser that: Support HTTP protocol: (link to be added) Support TLS 1.2 for HTTPS: https://tools.ietf.org/html/rfc5246 Support rendering HTML5 documents: http://www.w3.org/TR/html5/ Support rendering XHTML1.0 docum...

Sp3000

Jimmy's on a roll

Lembik

hi @Sp3000

undergroundmonorail

hi @Lembik

Lembik

hi

Sp3000

Hey Lembik

VTCAKAVSMoACE

09:52

Is this question's title a little unclear?

Sp3000

Wrong link?

VTCAKAVSMoACE

No...?

Sums of digits 1 through 7, yeah?

Sp3000

Oh question title, right

A little

VTCAKAVSMoACE

It's more like "Sums of Numbers Composed of Only Digits 1 through 7"?

Sp3000

That's probably a bit too long IMO - it's kind of a tradeoff between brevity and clarity. The current one isn't too bad, and you find out the exact intention pretty quickly upon opening the question

VTCAKAVSMoACE

10:01

Yeah - when I read it, though, I thought it meant 1+2+...+7, and it confused me reading the question.

Sp3000

Same - it's the subtle difference between "sum" and "sums" :P

VTCAKAVSMoACE

Oh, and, since it would take longer than necessary to post this to sandbox, would just printing a number to console (as code golf) be a good question? Super simple, just define a number and print it.

Sp3000

How is it determined which number to print?

VTCAKAVSMoACE

Like, given the number "-1221", (as hard coded) print the number.

Sp3000

How is the hardcoding done?

VTCAKAVSMoACE

10:09

Totally up to the programmer.

Sp3000

I'm trying to see how this adds to a standard Hello World

VTCAKAVSMoACE

So, like, you could do -3*11*37 or just -1221.

Sp3000

I'm a bit confused what the task is. Is it just you pick any number you want, hardcode it into the program and all you do is print the hardcoded number (and there's no input)? If so, I don't see why you would choose to factorise it like that, say

VTCAKAVSMoACE

No - I specify the number.

Sp3000

So... you pick a number, and all submissions need to output that same number? Is this going to be a big number?

VTCAKAVSMoACE

10:14

Larger than Java's Integer.MAX_VALUE.

Or less than it's negative component.

Sp3000

Depending on how easy the number is to compress, I'd imagine that for some languages it'd just be easier to print the number as a string

VTCAKAVSMoACE

Exactly. That's why it'd be a code golf.

Sp3000

Well that's what I meant about Hello World - if you pick a number that's hard to compress, then it'd be no different from a HW

(basically the whole challenge depends on the number you pick - some numbers are easier to compress using arithmetic, others by printing char-by-char, and some just can't be compressed well at all)

VTCAKAVSMoACE

It wouldn't be too difficult to compress. Different languages would be more efficient for different methods.

(it probably should be a popularity contest with a focus on brevity.

Sp3000

10:30

Well, I can't really sandbox-check it without the number and I'm assuming you're not going to post the number until you make the question, so we'll see

VTCAKAVSMoACE

I'm gonna come up with a fairly difficult number to arithmetize. I'm still thinking about it.

Here's my number: -184642425887198622.

Sp3000

Doesn't look easy to use chars, so I guess you'd need an expression that beats 19 bytes

VTCAKAVSMoACE

Correction : -184642425887198623*

The 2 makes it easy to divide.

Sp3000

I don't see how the 2 changes anything

VTCAKAVSMoACE

-19×101×140663×684032359 = -184642425887198623

Sp3000

10:40

Yeah, but the factorisation of the previous one wasn't any easier to use either

VTCAKAVSMoACE

It's slightly easier - same distinct primes, smaller chars total.

-2×3×7×4396248235409491 = -184642425887198622

Sp3000

The thing is, unless you have a large prime power in there somewhere, it's not any more efficient to use the factorisation

VTCAKAVSMoACE

Unless you calculate that specific prime somehow (i.e. a small sieve).

For example, 140663 is the 13067th prime.

Sp3000

That'd only potentially help for golfing languages

VTCAKAVSMoACE

But, nonetheless, it's still another solution - there are hundreds of ways to find this. Therefore, the question.

Sp3000

10:50

Well let's put it this way: even if a golfing language had a one-char operator to get the nth prime, you wouldn't save any bytes doing P13067 compared to 140663

VTCAKAVSMoACE

Mm, true. But there are certainly more (way more clever than my generation technique) ways of finding this.

Maybe a restriction should be that you cannot use the number itself?

Sp3000

People'd probably circumvent by adding 1 or something

VTCAKAVSMoACE

At least three non-reductive steps*

Sp3000

So far the only language that I can think of that'd be able to compress anything is CJam

VTCAKAVSMoACE

Simply because it loads to stack then empties?

Sp3000

10:54

Maybe Bubblegum too. Should check that actually

Because W = -1 and J = 19

VTCAKAVSMoACE

You could also take it in chunks, with that method. Or hexadecimal, in general.

-28ffb4edd790d9f is shorter.

Sp3000

True

I dunno, feel free to post it if you want? I just don't think there'll be that many interesting answers, that's all (maybe someone might find something and everyone else copies, for all I know)

VTCAKAVSMoACE

Or base 36: -1ei29yg72ra7.

Yeah... it should probably be a popularity contest with the most interesting method, then.

Sp3000

That's not what I meant, but okay...

VTCAKAVSMoACE

Well, if someone copies, it'll be obvious then.

Sp3000

11:01

Don't forget that base36 might look shorter like that, but you're neglecting the code to do the conversion

VTCAKAVSMoACE

I thought that mathematica could deal with it...? (looking up now)

Yeah, you're right. :\

Wait - is compiled code typically shorter than source? It must be, right?

(totally different subject)

minxomat

Not really.

VTCAKAVSMoACE

Oh. Never mind then.

minxomat

Let's say you invoke MessageBoxA to display something. That's going to be a few bytes. But the compiled executable must have a valid DOS stub and PE32 header ;)

VTCAKAVSMoACE

Ohhh, okay, that does make sense.

But for long code, would it be shorter?

minxomat

11:11

Depends. If it's native code, yes.

If it has a lot of compiled resources no :)

You can generally decrease the executable size using a compressing linker (like crinkler) or PE compressor such as MEW11SE as it's commonly done in sizecoding.

VTCAKAVSMoACE

Huh. Okay. Reason why I ask is mainly for force-compiled languages, like Java. Why do we count source bytes, not compiled bytes?

Compiled is the actually executed code, after all.

minxomat

Because this is codegolf, not sizecoding.

Java is another story, where the "compiled" code is for an abstract machine, not actual hardware. Can't really compare it to native code (PE/ELF), where headers would create an overhead. That's why, in traditional sizecoding, Java/.NET and the like are disallowed because they require an external runtime.

> Compiled is the actually executed code, after all.

And this is slightly inaccurate, since compiled native code is machine code, but what's really executed is Microcode.

Sp3000

11:29

@Dennis Damnit Dennis, you made me have to reverse eng Bubblegum

(Is there an encoder somewhere that I missed?)

Jakube

11:44

@Sok I would really like to see a better Pyth-solution for my Rubiks Cube question. Try again? 38 bytes are possible (though I used a feature, that wasn't in Pyth when the question first was published)

Sp3000

12:04

@Dennis btw I get 129 bytes for Python's zlib for FizzBuzz

New Main Posts

13:01

-3

Q: Java program code challenge?

Write a program that asks the user to enter certain number, then asks the user to enter another 20 numbers, after entering them all, it prints out the number of occurrences of the first number. See example below (text shown in boldface is supposed to be user input). Enter number to search for: 2...

random

aditsu

compare allfileformat.com/info/unicode/char/1f66d/index.htm and allfileformat.com/info/unicode/char/1f66e/index.htm

seems to be a 🐛 in the symbola font

Dennis

@Sp3000 I use zopfli for DEFLATE. That seems to have been a mistake for FizzBuzz. LZMA isn't even worth trying for such a short output. I haven't published an encoder yet.

Geobits

13:28

Thought about sandbox-sniping CUDDLE, but it just gets too unwieldy with BigInteger :(

Martin Büttner

I've been wondering about how to determine when there is no CUDDLE. or can you show that that only happens for 10^n and 0?

Sp3000

@Dennis Ah k, well I've basically set mine up over here to try a bunch and take the best - Python beats zopfli some of the time it seems

Geobits

@MartinBüttner It would make intuitive sense that it only happens for those, but I can't think of a good way to prove it beyond an exhaustive search up to the limits of the input data type ;)

quartata

I really dislike it when people downvote challenges without saying why they didn't like it...

Rainbolt

Link?

Lembik

13:38

dear combinatorics experts... If I have a vector n elements. How many different ways can I set a of the elements to 1 and b of them to -1? No LaTeX :)

quartata

My RTTTL Obfuscation challenge.

11

Q: RTTTL Obfuscation

RTTTL, or Ring Tone Text Transfer Language is a music format that was invented by Nokia in the dawn of time when dinosaurs roamed the land. It allows for fairly crude pieces of music (no chords or anything), and the format is pretty simple which is why I think it's pretty cool. Take a look at thi...

Rainbolt

You know what I really dislike

Geobits

You're at +12/-1, I wouldn't lose any sleep over it.

Rainbolt

People who receive a downvote and then feel entitled to an explanation

quartata

@Geobits I know but I want the feedback so I can make better challenges.

randomra

13:40

(n) * (n-a)
(a) * ( b )

Geobits

@quartata Sometimes there's just no constructive feedback to be had.

Rainbolt

Well, let's start the questionnaire then.
1. Have you argued with anyone in the past week?

quartata

No.

Lembik

@randomra oh! Is that the same as {n \choose a+b}{a+b \choose a} ?

Rainbolt

You passed the questionnaire.

quartata

13:40

Ah I see.

Geobits

The problem I have is that many times if I DV something and do explain it, the OP will get defensive and start a mini-argument in the comments. It's just not worth it sometimes.

Rainbolt

I am getting very suspicious of Geobits right now

randomra

@Lembik should be

Geobits

If I can think of something constructive to say that won't be taken as "hostile", I normally do.

Martin Büttner

@quartata Also, some people will just downvote any pop-con on sight.

quartata

13:41

I think he did say that I should use the beautiful PLAY format instead of RTTTL

Lembik

@randomra I am not 100% sure what you wrote.. is it {n \choose a}{n-a \choose b} ?

quartata

@MartinBüttner That's kinda what I figured but I was holding out for it to be something sane

Rainbolt

Maybe I should consider adding "Is it a popularity contest?" to the questionnaire. Some people just don't like them.

Geobits

True, but I didn't downvote it despite your use of a terrible format :D

randomra

@Lembik yes

@Lembik if you write out the factorials, they are the same

(your and my version)

New Main Posts

13:43

-1

Q: fraction of positive numbers in a matrix (Mathlab)

Which command will return the fraction of positive numbers in a n-by-n matrix A in Mathlab?

math array-manipulation

quartata

A lot of off topic main posts today

Geobits

It's probably still September in some places.

Lembik

@randomra great..thanks

@randomra it feels like it should simplify

Martin Büttner

@Lembik so what's your challenge idea?

quartata

Ooh someone's making a challenge?

Geobits

13:52

It's Lembik, so expect it to be something to do with matrices or strings filled with small numbers :P

Lembik

I think it's too controversial actually

quartata

Alright one second while I pull out Octave

Lembik

but here it is (was).. on old unix network systems "ypcat passwd" would give you the password file

which you could then use to hack people's passwords by bruteforce

quartata

What does ypcat do?

Is it just like regular cat?

Geobits

It looks like a butchering of copycat.

Sp3000

13:53

I thought it was y(our) p(et) cat

Lembik

@quartata the passwords are managed by a network server .. yp is short of "yellow pages" strangely

quartata

Ah I see.

Lembik

eventually..in the 90s, people realised this was a bad idea

as fast computers could find a lot of passwords

but.. these systems still persist

Geobits

Well, it probably wasn't all that bad until processing speed caught up with pw complexity anyway.

Lembik

really bad sysadmins just remove the ypcat binary so users can't use it

@Geobits password complexity was fixed at 8 characters on these unix systems

Rainbolt

13:55

Hmmm... we are going to have a real problem in a few years at my work

Geobits

@Lembik Ah, right, we're talking old ones. Rainbow tables FTW :)

Rainbolt

We started with X products and a single code branch

Lembik

good sysadmins moved to shadow passwords

Rainbolt

Now we have X products and X branches

Lembik

@Geobits old and still existing systems :)

Rainbolt

13:55

My team got to keep the original main branch

Now Visual Studio says I have incoming changes to nearly every file, and the numbers keep increasing

Geobits

That sounds about right. Nobody wants the main branch.

Lembik

so.. the challenge was going to be to implement ypcat passwd in the smallest number of chars, assuming the sysadmin has deleted the binary

Rainbolt

We were about to make a drop when we split, so they gave us the main branch to minimize the impact to our team

Lembik

a golf challenge!

who would have thought it :)

Geobits

That doesn't sound like you at all, Lembik :P

Sp3000

13:57

So... how do we test it works?

Lembik

@Sp3000 I can run it :)

Sp3000

I can't :/

Lembik

I have an old system to try it on

@Sp3000 you could compare what it does to a real ypcat passwd ?

the code is freely available

do you have linux?

randomra

@Lembik it is n! / (a! b! n-a-b!) which is a 3-way choose

Sp3000

Not currently :P (guess I'll sit this one out)

Lembik

13:58

but for roughly these reasons I cancelled the idea

@Sp3000 I am not going to do it

trichoplax

When chess gets too easy, some people move to blind chess. This seems to be an example of blind golf.

Lembik

@randomra thanks!

Sp3000

Ah k

Lembik

@trichoplax :)

trichoplax

:)

Lembik

13:59

I do like the idea of golfing security hacks

Sp3000

I was doing cyber security all of last week, which is where the Python pickle came from

Transcript for