The Nineteenth Byte

6:00 AM

Yeah. Only integers for any base greater than 96. That sounds good.

Anonymous

6:22 AM

r'([+-]?\d+\.?\d*)([+-]\d+\.?\d*[ij])?' will work for ints, floats, and complex. For ints and floats, you can eval(match.group()) to get the value, and for complex, you can eval(match.group().replace('i','j')) will get the value.

feersum

25 mins ago, by feersum

Doesn't work for .1

Sp3000

ninja'd

feersum

Personally I find regexes unuseful for number parsing.

Anonymous

I don't see a problem with that :P Just require a leading 0

feersum

That's so verbose.

Anonymous

6:24 AM

You're so verbose

El'endia Starman

@feersum I won't be using them to parse, but rather as a "is this a number?" check.

feersum

Regexes are only really helpful if you can outsource to eval, which won't work if you're trying to allow different bases.

@El'endiaStarman How is that not a part of parsing?

El'endia Starman

I won't be doing the actual work of parsing with regexes...?

I'll be stepping through the given string one character at a time.

feersum

Exactly.

When you see a digit, you know you have to parse a number.

Anonymous

He means interpreting/evaluating vs recognizing

feersum

6:29 AM

And keep taking chars one at a time until you hit a non-number char.

So using regexes there is redundant.

Alex A.

ψ

Anonymous

@AlexA. !π

Alex A.

Psi not pi?

Anonymous

@AlexA. Yeah, that's Phi's sock

Alex A.

Hm. Okay.

I meant ψ to be like sigh

Because it amuses me

Way to ruin my dreams, Mego.

El'endia Starman

6:32 AM

New challenge idea! "Calculate Psi(not phi or pi)" where "or" is bitwise or. :P

feersum

Poll: do you pronounce the p in 'psi'?

El'endia Starman

Silent.

Alex A.

No, because you aren't supposed to

feersum

How do you bitwise-or non-integers?

I do!

Alex A.

@feersum Don't, it's wrong

Like it actually is

El'endia Starman

6:33 AM

@feersum Well, considering my last challenge was about phi(x) and pi(x)...

feersum

If @AlexA. says it's wrong, that means it's right. Yay.

Sp3000

I pronounce it like "sai"

El'endia Starman

^ That is correct.

Alex A.

^

Sp3000

@feersum Bitwise or the binary representations :P

feersum

6:34 AM

That could be a good challenge for a non-golfing language to win.

Very easy in C, very hard in some language where you can't cast memory.

Sp3000

Hmm not necessary...

Alex A.

What isn't necessary?

Sp3000

Well except the psi part, I dunno what that's meant to be

Anonymous

@AlexA. Always glad to help

El'endia Starman

@Sp3000 I don't know either.

Sp3000

6:35 AM

*necessarily

Anonymous

@AlexA. You obv /s

feersum

psi has got to be the input.

Sp3000

I keep dropping the "il" from "rily" endings for some reason

feersum

Since we do not want constant-output challenges.

Anonymous

psi is the wave function

feersum

6:36 AM

Of what?

Alex A.

👋

^ wave function

👋(x)

Anonymous

@feersum I dunno. A photon?

Alex A.

Alternatively, 🌊(x)

Anonymous

The problem is, most interesting wavefunctions require 4D coords

Anonymous

(3+1D technically)

Alex A.

6:37 AM

Yes, 3+1=4. Good job, little buddy.

Anonymous

@AlexA. Except in physics

Alex A.

3+1 overflows in physics because physicists can't use computers?

Anonymous

Where ds² = dt² - (dx² + dy² + dz²)

Sp3000

Except in physics, where 3 + 1 = 5.124

Anonymous

(Minkowski space if I remember my terminology)

El'endia Starman

6:40 AM

@Mego Correct!

Anonymous

So it's not quite simple addition of the 4 dimensions

Anonymous

@El'endiaStarman yay

Anonymous

Someone else who can help explain why alex-is-wrong

Alex A.

@Mego Did you study physics?

Anonymous

@AlexA. No it just appeared in my brain one night

Alex A.

6:41 AM

ಠ_ಠ

I mean was it your major

Anonymous

The 2 years of physics classes in college had nothing to do with it

Anonymous

No, though it was almost my minor

Alex A.

Almost?

Anonymous

At my uni, to get any sort of BS, you had to take 4 science credits

Anonymous

I chose to take physics for the majority of them

Alex A.

6:43 AM

What was your major? CS?

Anonymous

Yep

Anonymous

CS major, Math minor

New Main Posts

toot toot i am a roobat

Anonymous

And I was like 2 or 3 physics classes away from being a physics minor

Anonymous

@NewMainPosts wut

Anonymous

6:44 AM

On the front page, it shows it as being asked by Chipperyman

Alex A.

It isn't on the front page.

That shit was deleted, yo.

Anonymous

Anonymous

And now it's gone

Anonymous

I guess the cache strikes again

Sp3000

If I had a dollar every time someone asked a calculator-with-strings question... I'd have about $5 right now, I think?

xnor

6:46 AM

you mean FIVE dollars

Alex A.

Canadian dollars

Anonymous

FIVE loonies

GamrCorps

100 rep away from 2000! Sooooo close...

Anonymous

@GamrCorps it's not worth it turn back while you can

GamrCorps

lol

Alex A.

6:50 AM

100 rep away from 13120! Sooooo close...

El'endia Starman

False, it's totally worth it.

@AlexA. Wait, did you just pass 13k?

Alex A.

Yep

El'endia Starman

Ah! Indeed!

Anonymous

@El'endiaStarman Go to the Sandbox when you have 2000, you'll see how not-worth-it it is

El'endia Starman

oh, right, that...

Sp3000

6:51 AM

@xnor btw xnor, out of curiosity - have there been any must-pass-these-test-cases-only type questions on PPCG?

I feel like it might be interesting to see some anagol-like exploits for once

xnor

fixed output ones? :-)

i don't know, i've considered posting ones

well, just saying "as long as you pass the test cases, that's fine"

Alex A.

So you describe how it should work and it's okay if it doesn't work like that so long as it works in like 3 specific cases?

Anonymous

Whenever I provide test cases, I make sure to specify that it should theoretically work with any input, and those test cases are just there to help people out

Sp3000

Well the scenario would be that this is the problem where the test cases came from (to provide some basic intuition), but that the test case system is incomplete and you only need to pass the given ones (or as xnor puts it, "as long as you pass the test cases, that's fine")

Alex A.

if input == 1
    output test case 1
else if input == 2
    output test case 2
...

Sp3000

6:56 AM

You could do that, but that generally wouldn't be the golfiest way

Alex A.

Idk, just seems less interesting to me.

El'endia Starman

This is a good example: has22 on CodingBat in Python.

def has22(nums): return '2, 2' in str(nums) passes all test cases but it's wrong.

nums = [12, 2] is one case where it would give a false positive.

feersum

What kind of Bat is it supposed to be?

Alex A.

Batman

Anonymous

@feersum Baseball

Anonymous

7:01 AM

Battery

xnor

@xsot I forgot about m=240. Minutes should still work though?

xsot

How would it work for minutes?

xnor

wait, never mind, that doesn't make sense

that's what i get for trying only 3 values

xsot

The m/2%30 is such an eyesore.

Alex A.

The OP changed the input lower bound from 60 to 0 after 28 answers had already been posted. ಠ_ಠ

xnor

7:07 AM

@xsot Is m%60/2 the same?

no that it changes anything

xsot

@AlexA. Oh what, that breaks both my submissions.

Anonymous

@AlexA. Yeah that's bad on him

xsot

@xnor I think it's equivalent.

Alex A.

@xsot I left a comment telling him to change it back to the original. It's pretty bad form to make changes that invalidate answers after they've been posted.

Anonymous

He also changed the acceptable format for round hours, removing x hours as an output

Alex A.

7:13 AM

@Mego Ugh, seriously? Alright. I left another comment addressing that too.

Anonymous

@AlexA. ಠ_ಠ

Alex A.

Why give me the disapproval face? I'm the one trying to keep answers from being invalidated. :P

Anonymous

Because seriously

Anonymous

That's my standard reaction to anyone who uses that word in a message addressed to me :P

Alex A.

Oh right, I forgot.

Anonymous

7:21 AM

sure..

Anonymous

I believe you...

Anonymous

~~evil mod~~

feersum

Language name idea: the

Anonymous

@feersum Language name idea: no

feersum

That's even better actually.

Alex A.

7:22 AM

registers it on github

There's a language called FALSE.

That's not as good as No though.

Someone should make a golfed version of PL/I and call it GL/I.

xnor

Would it be crazy to disallow accepting answers as a matter of policy?

with exceptions for questions that are a single language or have a single language-independent metric like algorithmic complexity

Alex A.

I think so, yes.

xnor

why do you think so?

Alien G

Crapppppppppppppppppppp

I just realized that I've been drinking coffee since 10PM and it's now 11:30PM

We all know who isn't getting sleep.

Sherlock9

@xnor @AlexA. There is a guy username markshancock that is suggesting the edits

Alex A.

7:30 AM

@xnor One of the fundamental parts of the Stack Exchange network is accepting answers. It's not possible to disable that on a per-site basis. If we were to make it our policy to disallow accepting answers entirely, we'd have a lot of explaining to do every time we got a new user from another SE site.

@Sherlock9 What edits

Sherlock9

The ones that break the hours minutes question

xnor

@AlexA. so should i take this to be a practical objection rather than a principled one?

say, if magically SE devs would turn off accepting, what would you think?

Alex A.

I would still disagree on principle, but I figured the practicality argument is stronger.

:P

xnor

i see

Alex A.

Why do you think it would be best never to accept answers?

xnor

7:34 AM

it implies that some languages don't really matter

Alex A.

How so?

xnor

since realistically a pyth answer will always be shorter than a python one, a python one will pretty much never be accepted

or a C answer, and so on

Alex A.

Why does not being the accepted answer imply that the language doesn't matter?

xnor

"doesn't matter" is stronger than i meant

more like, are playing a side game

whereas the real game is between those coding in Pyth and CJam and etc

feersum

Suppose one language solves it with a 2 byte builtin and another works hard to anwer.

It's stupid that the 2 byte answer should be considered better.

Sp3000

7:38 AM

I guess the other thing about accepts is that they're pinned to the top, which doesn't really mean much for golfing...

Alex A.

It's not inherently better because it's accepted, it's just the shortest answer that complies with the challenge spec. I feel like as long as someone enjoyed putting together an answer in their language of choice, be it Pyth or Java or what have you, then it doesn't really matter whether they win.

Interesting, well thought out answers get upvotes.

feersum

@AlexA. Can you name any positive aspects to accepting?

Alex A.

Declaring a winner is a nice privilege to have

feersum

That doesn't even make sense.

xnor

for who?

Alex A.

7:40 AM

For the person who can click the button and say, "Hooray you to win," and the person who wins.

No, not at all.

xnor

they're not choosing then, right?

Alex A.

"Choose" was a bad choice of words on my part

Edited above message

xnor

so are you saying it's a perk for askers?

Alex A.

I think so.

Perhaps not a particularly strong one, but I enjoy being able to click the check mark. :P

xnor

i don't like it as an asker

feersum

7:43 AM

Is the perk to decide whether to accept the shortets answer, if you like it, but ignore it if you don't like it?

xnor

i feel that that i'm unjustly ignoring most of the answers

so i've stopped doing it

Alex A.

No one says you have to accept an answer.

xsot

I feel that it is meaningless to compare languages, but having an accepted answer seems to validate certain languages.

Alex A.

@feersum No, not at all. The perk, at least for me, is being able to congratulate the person who found the absolute shortest means of accomplishing the task you've laid out with +15 rep.

xnor

as opposed to just the 10 rep of upvoting?

Alex A.

7:46 AM

Yep

For me, upvote = Nice work! and accept = Hooray you win

xnor

i'm never going to get congratulated though, no matter how good my python golf is, and that's a downer

and, sometimes a person who just ports my python code into pyth will

Sp3000

(irrelevant to current topic, but we all think your golfs are cool xnor, nod nod)

Alex A.

^

xnor

thanks!

Alex A.

He ninja'd what I was about to say. :P

xnor

7:48 AM

but your compliment isn't green enough

or check-marky

i just need something to fill the checkmark shaped hole in my heart

WE LOVE XNOR

aww thanks!

Is it a putter?

Apparently so

xnor wins the green putter of justice

xsot

@xnor You're pretty dang good at recursion.

Alex A.

7:50 AM

He's pretty dang good at every dang thing I've seen him do on this site.

No joke.

xnor

@xsot it's funny how much different IO standard lead to different solutions

in my stint on anarchy golf, dealing with lines of input from STDIN made me realize there's a whole new bag of tricks

and it trickles down surprisingly much into the code, not just changing boilerplate

xsot

I agree, I like how both sites are unique in their challenges this way.

xnor

would it actually be crazy to let the asker accept what they want?

i do imagine there might be complaints of unfairness

Sp3000

2

xnor

haha

oh no, it's being negated!

Alex A.

7:53 AM

26 mins ago, by Alex A.

I think so, yes.

feersum

Isn't that the output?

xnor

oh right

i'd find it cool to give a nice checkmark to answer i though was best, as long as i got to choose

Alex A.

Bounties are an expensive way of doing that

(expensive in terms of reputation, that is)

feersum

When you're asking, you could make a different rule for accepted answer.

xnor

i think there's a particular appreciate you get from coming up with a spec of the solution strategies someone might take

feersum

7:56 AM

Like lowest length / length of shortest same-language answer on a certain question was mentioned before, I think

xnor

@feersum would people be annoyed if i just put "I'm going to accept the answer i like most"?

feersum

Probably.

Alex A.

^

It negates the purpose of objective winning criteria, IMO.

xnor

"i'm going to accept the answer that i think is shortest relative to its language, based on a total eyeballing of it and not any statistics"?

Sp3000

With our current site situation, most likely. We've even gotten some people to revert when they accept "the wrong answer"

xsot

7:57 AM

Well, the objective winning criteria isn't exactly a fair one.

Alex A.

In a broad sense, objective == fair

The point of objectivity is to be fair, is what I mean

xnor

ok, but surely if i gave an objective but totally out there criterion, like "I'll accept the answer that has the least punctuation", people would complain?

how about, if I said, "I'm accepting the shortest Python answer"?

feersum

Major butthurting from golfing language users no doubt

Alex A.

I would consider that unfair. I have a feeling that at least some of the community would agree with me, though I'm sure some wouldn't.

xsot

I think removing the need to accept answers might encourage people to return to non-golfing languages.

xnor

8:00 AM

i'm not really proposing it, just trying to juxtapose it with the current de-facto "I'm going to accept the shortest answer in a golfing language"

i feel like people would be annoyed if you said it much more than if it was basically the case, but not stated

Alex A.

Some challenges require tasks that golfing languages can't do. For example, I won a challenge that involved the SE API with a ~200 byte Julia solution once. So golfing languages don't always win, they just usually do.

feersum

Aren't there golfing languages that can download from URL?

xnor

yes, of course, but i think that tiny minority doesn't justify the overwhelming trend

feersum

Probably it was only that none of them bothered to answer.

xnor

and even among that minority, a lot of it seems to be "Mathematica has a built-in", and few people are in the running for that either

Sp3000

8:07 AM

tbh, at this rate, even Mathematica's advantage might be dwindling...

Alex A.

Seriously is shaping up to be the Mathematica of golfing languages, it seems.

(in terms of standard library capabilities)

xsot

Code golf is turning into a game of language design.

Alex A.

In some ways I think that's true. But I've also learned a lot about language design by seeing golfing language implementations, so that's fun.

xnor

i think it's a nice competition to have

TanMath

@AlexA. ?

xnor

8:12 AM

it's impressive what type of things have come out of pushing golfing languages to their limits

i just want to make sure it doesn't create the feeling that these languages are the real competition and everyone else is doing some side-contest

Alex A.

In general my view of the site is to just come up with a solution, shorten it, post it, and enjoy the fact that I've made a working solution in my favorite language. I don't feel that competitive. (Could be because I don't know any competitively brief languages :P)

Sp3000

On the one hand I've liked seeing how new languages try to come up with golfy ideas (eg prefix notation, blocks), but I'm not too sure wherethe line is between that and "builtins for everything"

Alex A.

Yeah.

I think Arcyóu is a neat idea. It's a Lisp-like golfing language (so it's almost guaranteed not to win :P) and it doesn't rely all that heavily on built-ins.

The user's name who made it is bkul

Sp3000

@xnor Hmm tbh I'd still golf in Python since it's what I like, but I'm not sure how other users feel about it

Alex A.

^

xsot

8:19 AM

Golfing languages have their place, but I see a trend where they have become the focus of the game, to the extent where poorly golfed answers in a golfing language can draw more attention and awe than well golfed answers in regular languages.

Sp3000

^^ "Wow that's short! Can you please provide an explanation?" - first timer seeing a golfing language

Explaining good golfs helps a bit for regular langs, but people still vote weirdly I've given up figuring out how they tick

xnor

it's hard to overcome the fact that 14 is a lot less that 54

it's the thing people see

feersum

I can't say for sure as this is a counterfactual, but I suspect I would have been much less likely to join PPCG if there were as many golfing language answers as there are now.

To me it would give the impression "these guys have their heads up their own a****"

Alex A.

# Java, 14 bytes
Plus another 83 more because I lied in the header.

2

xsot

@Sp3000 I actually hesitate to explain my submissions unless someone asks as I find that part of the fun is in figuring out how someone else's code works on your own :)

xnor

8:24 AM

# Python 2, 80 bytes * 0.2 (language conversion adjustment) = 16 bytes

Alex A.

@feersum Why so?

feersum

Dude. That's a great excuse for not providing any explanations.

I'm going to use that all the time now.

Anonymous

@AlexA. <3

feersum

;)

Alex A.

ಠ_ಠ <-- Me every time I see a submission without an explanation, regardless of the language

5

Sp3000

8:25 AM

@xsot I've never thought about it that way, tbh - interesting...

xnor

i'm too lazy for that ...

Anonymous

@AlexA. That's why I built an explanation generator into the online interpreter for Seriously :P

Alex A.

If I see something in J I immediately know that I won't be able to decipher the code. I'd rather see an explanation that just say "Oh it's short and they posted it so it must work as expected."

feersum

@AlexA. When there are are those gibberish languages that only 5 people can read it makes it harder for outsiders to participate.

randomra

@Zgarb you can use 0's and 1's in J for masks effectively: codegolf.stackexchange.com/a/65201/7311

Anonymous

8:26 AM

Though I usually try to provide a higher-level explanation of what's going on

Stefnotch

@Mego Seriously?

Anonymous

@Stefnotch ಠ_ಠ

Sp3000

@feersum Unfortunately, ditto with your counterfactual, which is why I've kinda refrained posting solutions lately (and kept them on my wiki) :/

Alex A.

@Sp3000 What is your wiki?

Anonymous

I think golfing in Python is much more rewarding than using a golfing language, because there are so many neat tricks to use

Sp3000

8:27 AM

Maybe I'll post if it's really interesting...

Stefnotch

Does your online interpreter really have that?

Anonymous

I love golfing in Python and learning new tricks

Sp3000

@AlexA. Just the Github wiki for Gol><>, example programs page

Anonymous

However, I also love golfing in Seriously and getting to know my own language better

Anonymous

@Stefnotch Yep, though it only provides definitions for individual operations, not an abstract overview

Stefnotch

8:29 AM

That's pretty awesome...

Anonymous

The way I see PPCG is, there are multiple layers of competition going on simultaneously

Anonymous

The most obvious one is competing for the shortest overall answer on each challenge

feersum

And the most important one is chat.

The question is, who's winning?

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Bernoulli Numbers The Bernoulli numbers (specifically, the second Bernoulli numbers) are defined by the following recursive definition: Given a nonnegative integer m as input, output the decimal representation of the mth second Bernoulli number, rounded to 6 decimal places. Here is the input...

Anonymous

Then you have the competition for the shortest answer per language on each challenge

Stefnotch

8:30 AM

Was Mego writing up that challenge while chatting? XD

Anonymous

Then there's the subtler competition between language authors, who are competing to get the shortest answers across all challenges

Anonymous

@Stefnotch No, I was writing it up while I was conspicuously absent from the conversation

Anonymous

Bets on how long it takes Peter to point out something I missed?

Alex A.

@feersum Dennis

feersum

@AlexA. At chat?

Alex A.

8:31 AM

Yes

Anonymous

Well in the chat category, there are multiple competitions

Alex A.

No

Anonymous

Stars, trolling, helpfulness, etc

Alex A.

Dennis wins all

Anonymous

Doorknob is winning at stars obviously, but he cheated

3

Anonymous

8:32 AM

Alex and I are the main contenders for trolling

Alex A.

Haha

Anonymous

(which means we're in last place for helpfulness)

Alex A.

@Mego Note that Doorknob wasn't the one who pinged that message

@Mego I try to be helpful but I rarely have the knowledge to help people who have questions in here

Anonymous

@AlexA. Someone else cheating on your behalf is just as bad

Anonymous

@AlexA. Same

Alex A.

8:33 AM

If you can't help 'em, troll.

Anonymous

Martin is probably the most helpful in chat, though Sp and xnor are pretty awesome too

Sp3000

One day I swear I'll make a codegolf bingo challenge - must use one LISP-like language, one of [Ruby, Python, Perl, Julia], ...

Anonymous

@Sp3000 Bingo multiple-holes?

Stefnotch

About the Bernoulli Numbers:
Mind adding an explanation of the sequence that an idiot could understand? (I hate seeing those fancy formulas without any other explanation..)

Sp3000

Yeah

Anonymous

8:34 AM

@Stefnotch Unfortunately that formula is probably the simplest explanation

Alex A.

@Mego Bingo, you found all of the multiple holes

@Sp3000 One of Ruby, Python, Perl, and Julia?

How darest thou lump Julia in with the common filth!

Sp3000

Be glad I didn't include JS :P

Alex A.

ಠ_ಠ

Stefnotch

@Mego You could try giving an example program/function for the formula...nobody around here can't read code!

Alex A.

Aug 4 at 4:10, by Dennis

Hey, if you can't win, troll.

Alien G

8:37 AM

@AlexA. Like purple?

Alex A.

What?

Alien G

@AlexA. You know, the color? Do you not know what purple is?

Alex A.

No what is purple I've never heard of it

Alien G

Well, it's impossible to explain. So, that's what purple is, you're welcome.

Maltysen

its 790–666 THz

Alien G

8:38 AM

I can't win, so I'm trolling (Or attempting to)

Alex A.

@Maltysen 666 so metal \m/

Maltysen

nah i'm not really that metal its wikipedia:

Violet (color)

Violet is the color of amethyst, lavender and beautyberries. It takes its name from the violet flower. On the visible spectrum of light, violet light is at the end, with the lowest wavelength of 380-450 nanometers (in experiments under special conditions, people have so far seen to 310 nm). Light with a shorter wavelength than violet but longer than X-rays and gamma rays is called ultraviolet. In the color wheel historically used by painters, it is located between blue and purple. On the screens of computer monitors and television sets, a color which looks similar to violet is made, with the RGB...

Sp3000

@xnor Hmm not accepting from now on seems like a decent idea... Don't think I'll mention it in the post though

xnor

yes, silently not accepting is pretty convenient

Anonymous

@Stefnotch Done

Anonymous

8:45 AM

Not golfed, horribly optimized, but done :P

Sp3000

I "forgot" to accept, even after you reminded me :o

orlp

@Calvin'sHobbies you haven't wrote an answer since october!

Stefnotch

@Mego Great! :D

orlp

how can you survive?!

Alien G

It think I just trolled @AlexA out of this chat.

xnor

8:45 AM

@Sp3000 "i'm waiting for more submissions" :-P

Alien G

Win?

Stefnotch

I can't read Python, but since it is so verbose, it is almost like reading a book! :P

Sp3000

@xnor Touche

orlp

@Stefnotch anyone can read Python

it's like the simplest language out there

Stefnotch

XD

Alex A.

8:46 AM

@Sp3000 Touch

Anonymous

The reference implementation will take forever around m=13 because of the horribly-inefficient way of computing factorials and combinations, but hey, it exists :P

Anonymous

The Akiyama-Tanigawa algorithm is probably the way to go for that challenge, but it'll be interesting to see what people come up with

orlp

@Mego which changenge

Anonymous

1

A: Sandbox for Proposed Challenges

Bernoulli Numbers The Bernoulli numbers (specifically, the second Bernoulli numbers) are defined by the following recursive definition: Where denotes a combination. Given a nonnegative integer m as input, output the decimal representation of the mth second Bernoulli number, rounded to 6 dec...

orlp

the algorithm description on wiki is weird

for j from m by -1 to 1 do

Anonymous

8:59 AM

for j in range(m, 0, -1)

Transcript for