The Nineteenth Byte

Anonymous

12:02 PM

@Qwerp-Derp That's not true. PHP wasn't designed.

7

Anonymous

It was an unfortunate accident involving tequila and a cow

Jan Dvorak

PHP is a templating language for MySQL

user165474

12:14 PM

Why does everyone seem to hate PHP? (I've never used it before)

Anonymous

@HyperNeutrino Give it a try

Anonymous

You'll understand

Anonymous

Here's a challenge for you: write a web page that takes a string and outputs all rows in a MySQL table where the column "term" contains that string

Anonymous

Without introducing SQLi exploits

Lembik

people have very different views about programming languages

many people love python and many other people think it's insane to have a programming language with dynamic typing

user165474

12:17 PM

@Mego Okay, I'll try that. (the challenge is to use PHP right?)

ASCII-only

@HyperNeutrino Yeah

user165474

Okay, I'll try that.

ETHproductions

But hey, this proves that JS isn't at the very bottom of the "worst languages" list :P

betseg

Default PHP session lasts 24 minutes. Why you would ask. Because a day is 1440 minutes. It was supposed to be a day, but someday minutes got converted to seconds, and 1440 seconds equal to 24 minutes.

7

user165474

Oh um... that sounds sketchy...

ETHproductions

12:23 PM

If there's a thedailywtf.com for language design, that sure belongs on it....

user165474

hah That would be a funny site.

Jan Dvorak

There's this site: phpsadness.com

5

ETHproductions

That's great

user165474

Perfect. Now I know to never use PHP (Except for Mego's challenge)

ETHproductions

> Sort functions operate in-place

Sadly, that is also a JS feature

Anonymous

12:27 PM

A lot of languages have in-place sorting

Anonymous

But the sane ones also have copy-sorting

AdmBorkBork

PowerShell gives you options

Anonymous

e.g. list.sort() vs sorted(list)

AdmBorkBork

[array]::Sort($a) vs $a|sort-object

Yay, PowerShell is sane

ETHproductions

In JS that'd be list.slice().sort()

A terrible workaround, but at least it doesn't take a lot of code

user165474

12:28 PM

In Java there is no function to sort arrays... lol

user165474

Wait, actually Arrays.sort

user165474

Nevermind :P

Anonymous

@AdmBorkBork That's affirming the consequent :P

betseg

b=sort(a) and a=sort(a) are enough imo :p

Anonymous

Having copy-sorting is a necessary but not sufficient condition for saneness

ETHproductions

12:30 PM

And consistency, consistency, consistency...

ais523

what about a language in which all operations mutate by default, but cloning is easy?

I'm not sure there are any languages like that, but I wouldn't necessarily call it insane

user165474

oh god someone's playing music at about 80-100 dB outside my school right now and the beat is really disrupting my work

Anonymous

It would be seriously difficult to use

ASCII-only

Well, it depends IMO

Destructible Lemon

@HyperNeutrino at least its not someone inside the school playing a mosquito tone

ETHproductions

12:33 PM

It might be easier in a language where no operations mutate by default, but you have a .= operator or something that mutates it

Fatalize

"The ternary operator is left-associative" haha

ETHproductions

Or like Ruby's ! functions

user165474

@Destructible Ok, yeah, true.

Fatalize

One more argument to laugh at php

ais523

was going to ask if you were talking about PHP

ETHproductions

12:33 PM

Wow, references are converted back to values if you wrap them in parentheses...

ais523

does that site even mention PHP arrays? IIRC those were fairly insane

ETHproductions

> INF < []

Um, no. Just no.

AdmBorkBork

Infinity is smaller than empty-array?

ASCII-only

@ETHproductions Why not

ETHproductions

Well, INF < [], [] < [1], [1] < INF

11

Fatalize

12:36 PM

They're not the same types of things so comparing them doesn't really make sense anyway

^^ that's a problem though

ais523

I like Perl's comparison operators

it has two sets of them, one which coerces both sides to numbers before comparing

and one which coerces both sides to strings

ASCII-only

@ETHproductions At least JS does it the sane way

ais523

(this is because values don't have types in Perl, there's no real distinction between 12 and `"12`` in the language, rather operators and builtins specify the type that should be used for operating on the vlaues)

ETHproductions

@ASCII-only By converting everything to a string before comparison? :P

Oh my bad, it just gives false if you compare anything with an array

But with == it converts them both to strings: 1/0 == ["Infinity"] gives true

ais523

what's the most useful comparison behaviour between arrays and numbers, for golfing?

(designing a golfing language atm and am interested in that one)

AdmBorkBork

12:40 PM

@ais523 I like PowerShell's where it acts kinda like a filter

ASCII-only

@ETHproductions Except for 0==[]?

ETHproductions

Comparing its length might be somewhat useful, i.e. [9,8,7] < 4

ais523

right, filtering or vectorizing makes sense for < and >

AdmBorkBork

(1,2,3,4,5) -gt 2 will return (3,4,5)

ais523

but what about sorting?

e.g. what do you get if you sort [2, [1,3]]?

AdmBorkBork

12:41 PM

It compares against the first element, so you get ((1,3), 2)

I'm not sure if that's useful, but it's what happens

ETHproductions

@ASCII-only I don't know how that one works... maybe it converts them both to numbers

Fatalize

In Prolog, Variables < Numbers < Strings < Lists. That's pretty useful because you can group elements of the same type when sorting

AdmBorkBork

Oh, that's kinda handy

Kritixi Lithos

This is where Java comes handy, it just errors out

ais523

@Fatalize OCaml likewise has a universal sort order that sorts everything

via that sort of rule

Fatalize

12:43 PM

e.g. when you find an int in a sorted list you know that all previous elements are variables

ais523

however, treating variables as a type for sorting is rather nondeclarative, IMO

because what if they retroactively turn into integers later?

Conor O'Brien

@ETHproductions except for arrays

arrays cast to string >.>

Fatalize

You would in that case need to implement a sorting predicate which only really sorts once everything in the list is grounded

ETHproductions

@ConorO'Brien Um, wow

Conor O'Brien

which actually gives a "handy" way to do deep equality: a <= b && b <= a

for arrays a, b

ETHproductions

12:45 PM

Huh, interesting

Conor O'Brien

because they are both cast to string, and <= checks for lexical sorting, which is basically the same if two objects (and their toString's) are the same

ETHproductions

IMHO though a <= b && b <= a should equal a == b in any sane language :P

ais523

is a.toString().equals(b.toString()) always necessarily equivalent to "a deeply equals b"?

Conor O'Brien

usually, but if that's java, idk

ASCII-only

Wait why does [1,2]<=[1,2] and [1,2]>=[1,2], but not <, >, or ==?

Conor O'Brien

12:47 PM

@ASCII-only [1,2].toString() < [1,2].toString() is false because "1,2" < "1,2" is false

ASCII-only

@ConorO'Brien Yeah I know

ETHproductions

>> a = [1,2,3]
>> b = [[1,[[[2,[3]]]]]]
>> a <= b && b <= a
true

ASCII-only

but what about ==

Conor O'Brien

that, I have no idea

ais523

how do [1,2,3,4] and [1,"2,3",4] compare via <=?

Conor O'Brien

12:48 PM

@ETHproductions oh. jolf has a bug then >_> forgot about that

ETHproductions

@ais523 That would give true as well

ais523

because they stringify the same way?

Conor O'Brien

tho Array.prototype.toString = function(){ return "[" + this.join(", ") + "]" } fixes that

ETHproductions

Perfect, though maybe you don't want to mess with the default toString()

@ais523 Right, Array.prototype.toString() is about the same as function(){ return this.join(",") }

Conor O'Brien

class BetterArray extends Array {
    toString(){
        return "[" + this.join(", ") + "]"
    }
}

Learn How To Be Transparent

12:50 PM

@ConorO'Brien ES6 extend? IDK that, let me try it.

ETHproductions

@ConorO'Brien Better

ais523

@ETHproductions oh, for some reason I thought we were talking about PowerShell…

ETHproductions

JSON.stringify I think would work too: JSON.stringify([1,"2,3"]) != JSON.stringify([1,2,3])

@ais523 Oh, haha

ais523

@ConorO'Brien what language is that? it looks a lot like Java but toString() is missing a return type and would need to be marked public there

ETHproductions

It's actually JavaScript ES6

Conor O'Brien

12:52 PM

oh, that's better, loses conciseness tho

@ais523 yeah, JS ES6

Learn How To Be Transparent

Hey where can I learn ES6?
I see folks do it everyday.

Conor O'Brien

JS is getting beautiful

ASCII-only

@LearnHowToBeTransparent The internet

Conor O'Brien

@LearnHowToBeTransparent developer.mozilla.org/en-US/docs/Web/JavaScript

ASCII-only

@ConorO'Brien Node will too when ES7 comes out for V8 with async/await replacing callback hell

Learn How To Be Transparent

12:53 PM

36256730 is this place good enough sorry fat finger?

Conor O'Brien

@ASCII-only TBH I hate how node etc. stuff async down your throat.

ASCII-only

@ConorO'Brien Not me

Node is a serverside framework

ETHproductions

@LearnHowToBeTransparent I learned by doing; I'd suggest reading through that article and playing around with the different features

ASCII-only

Non-async servers are terribly inefficient

@LearnHowToBeTransparent That's more for learning ES6 (probably better to learn basic JS first)

But if you know a C-derived language (especially Java) you should be fine

Learn How To Be Transparent

i know ES5 only

ETHproductions

12:55 PM

You should be fine then

phpsadness.com/sad/39 Moral of the story: NEVER HAVE BUILT-INS RELY ON SOMETHING THAT CAN BE OVERWRITTEN. EVER.

AdmBorkBork

@ETHproductions So, Java and reflection is ... bad?

ETHproductions

Well, maybe my statement was a little strong... I guess what I mean to say is "never have built-ins rely on something that can be accidentally overwritten"

Conor O'Brien

deletes scathing comment

ETHproductions

You can't accidentally use reflection. You can accidentally create a function called __lambda_func() (without knowing what that will do)

ais523

C has a rule that if you use identifiers starting __ (except when the standard specifically tells you to), you deserve what you get

(i.e. no guarantees are made about how the compiler will react)

if PHP has a similar rule, then defining a function __lambda_func() would be something you aren't allowed to do

ETHproductions

1:03 PM

PHP doesn't have any rules like that AFAICT

user165474

@ETH What does __lambda_func() do? (Also what language is it)

ETHproductions

PHP: phpsadness.com/sad/39

Conor O'Brien

PHP

user165474

hm. Strange.

ETHproductions

The only perfect language is the one you design yourself...

Poke

1:09 PM

and Java

Kritixi Lithos

^ True story

Conor O'Brien

@ETHproductions none of my languages are perfect >_>

ETHproductions

The languages you design yourself are often the ones you are best at using though

Then again, every language idea is perfect until it's implemented

Conor O'Brien

this is true

user165474

I've designed an esoteric golfing language and I suck at using it ;_;

ETHproductions

1:17 PM

Poll: which octal literals are still useful today?

user165474

Quick question: Is answering a challenge you just posted allowed/encouraged?

Fatalize

I don't think it's disallowed but I would never do it

New Main Posts

0

Q: How Many Holes?

Challenge Given a graphical input of a shape, determine how many holes there are in it. Input Input will be given as some 2D form of input, either a multiline string, an array of strings, or an array of character arrays. This represents the shape. The shape is guaranteed to be in only one piec...

code-golf

user165474

ok. I was thinking of trying to answer my own challenge, I'll do it but I probably won't post it

Fatalize

You can but just don't post it right away

user165474

1:21 PM

Yeah, after the submission deadline I probably will post it.

Fatalize

otherwise it discourages people that intended to use the same algorithm / language

user165474

Yeah. I'm going to use Python, which is a fairly common language around here, so I won't post it until I've accepted an answer as the winner.

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Verify Cyclic Difference Sets code-golf A cyclic difference set is a set of non-negative integers associated with a particular integer divisor n, strictly greater than any member of the set, with a unique property: for any remainder 0 < r <= n/2, the equation (b - a) % n = r has the same number...

user165474

Question: Why is the PPCG graduation script still working even though I turned it off in TamperMonkey?

user165474

And when I turn it on, the added details are duplicated???

betseg

1:27 PM

@HyperNeutrino maybe we got the graduation design? :O

user165474

Maybe! :O

user165474

(Verifying by opening in Chrome without the possibility of cache)

user165474

nope :(

user165474

well I was hopeful :P

ais523

I often answer my own challenges but I tend to not start writing the answer until after the challenge is posted

to make the FGITW race fair

ETHproductions

1:30 PM

I tend to write an answer before the challenge is posted so I can verify the test cases, and make sure that the problem isn't too easy or hard

I don't post it though, at least not until I've got several other answers

AdmBorkBork

I very rarely write an answer for my challenges.

I think maybe two or three times out of the 40+ I've done

ais523

I often write an ungolfed answer before posting the challenge

Lembik

hmm.. so many downvotes for my harmless question :(

Lembik

1:49 PM

the other detail is that a lot of the answers don't work if there is only one integer

Hi everyone..

HI HAUNTED

Hi Poke..

hi..

i am looking at a question which is kind of unrelated to this chat room

can i post it?

ETHproductions

1:52 PM

Sure, go ahead

Haunted

its more of an algorithm kind of thing.

Poke

do what you feel, guy

Haunted

Thank you..:)

I was looking at this questions

Given a number line from -infinity to +infinity. You start at 0 and can go either to the left or to the right. The condition is that in i’th move, you take i steps.
a) Find if you can reach a given number x
b) Find the most optimal way to reach a given number x, if we can indeed reach it. For example, 3 can be reached om 2 steps, (0, 1) (1, 3) and 4 can be reached in 3 steps (0, -1), (-1, 1) (1, 4).

This was asked in 1 of the interviews.

Source: geeksforgeeks.org/minimum-steps-to-reach-a-destination

ETHproductions

a) Definitely. By alternating between left and right you can move in steps of 1 in either direction

Haunted

What do you think is the optimum solution for this in terms of time complexity?

All i can think of is brute forcing.

ais523

1:54 PM

at a complete guess, O(sqrt n)

oh, for the time taken

Haunted

O(sqrt n)?

ais523

wouldn't surprise me if the time needed to work out the position was O(sqrt n) too, though

ETHproductions

O(2n) is a worst case, which I think cancels out to O(n)

ais523

@Haunted big-O notation, it basically means that doubling the time you take lets you calculate 4 times as many numbers

Haunted

got it..:)

ais523

1:56 PM

@ETHproductions O(n) means "doubling the number doubles the length of time it takes to calculate for that number"; O(2n) would be the exact same pattern so they're equal

ooh, I think there might be an O(log n) solution (and O(1) if you assume that arithmetic takes constant time)

ETHproductions

To do e.g. 17 though, you can move R, R, R, R, R to get to 15, and then L, R, L, R to add the extra 2

ais523

here's how you do it: first you keep on going in the direction of the number until you're beyond it and have the same parity (odd or even)

then you change some of the moves to the other direction based on how far you overshot by

ETHproductions

Heh, I just thought of that as well

ais523

so in the case of 17, we start off with six R moves to get 21, then reverse the second move from right 2 to left 2

ETHproductions

17 would be R, L, R, R, R, R

Poke

1:57 PM

CMC do what haunted proposed

ais523

RLRRRR

ETHproductions

ninja'd

ais523

I think this is provably optimal; obviously, if you don't reach the number going towards it at every opportunity, you won't reach it if you go away sometimes, so you need at least enough moves to overshoot

and you can't change the resulting parity for a given number of moves

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Calculate Wind Chill math arithmetic code-golf number The Australian Apparent Temperature (aka, wind chill) in °C AT is given by this algorithm from the Australian Bureau of Meterology (wp, source): AT = Ta + (0.33 * e) - (.7 * ws) - 4.0 Where: Ta = Dry bulb temperature (°C) e = ...

Haunted

how did you come up with RLRRRR for 17?

can this be implemented in a program?

user165474

2:00 PM

Create a file called cmd.bat. Type this into it:

:top
start cmd.bat
goto top

Save and run. mwahahaha

user165474

(don't do it please, especially if you're on windows)

ETHproductions

@Haunted Start off with six R moves to get 21, which is greater than 17 and has the same parity (even/odd). Then switch the 2nd move from R to L to subtract 2+2 = 4

Haunted

"here's how you do it: first you keep on going in the direction of the number until you're beyond it and have the same parity (odd or even)
then you change some of the moves to the other direction based on how far you overshot by" - I thought of this but do you think this would give the minimum steps taken?

I am sorry guys this is taking some time for me to full understand..

@ETH

ETHproductions

I do think it's optimal. If T(n) gives the index of the first triangular number (1, 1+2, 1+2+3, etc.) greater than or equal to n, which would be the absolute minimum, this algorithm always results in T(n) or T(n)+1 moves

Haunted

@ETHproductions - Can you please explain what this means? "Then switch the 2nd move from R to L to subtract 2+2 = 4"

ETHproductions

2:04 PM

RRRRRR = 1+2+3+4+5+6 = 21. Since 21 - 17 = 4, we just need to subtract 4, which we can do by changing +2 to -2, or RRRRRR to RLRRRR.

ais523

so the only remaining problem is to find an additive partition of half the difference which has no repeated elements

but I think you can do that using the greedy algorithm

ETHproductions

Isn't half the difference always representable by a single member of the numbers you've added?

Even in e.g. 20, 1+2+3+4+5+6+7 contains 4, which is half the difference from 28 to 20

Haunted

wow i took some time for me to get it but i understand now..

Awesome..

New Main Posts

0

Q: Count the number of permutations of 1,2,...,n such that no 123 sub arrays are there

Suppose there is a set of {1,2,...,n} and we have to find the number of permutations of this set such that there is no increasing sub array of length 3. For example, if n=4, then 1,4,2,3 is not a permissible permutation as 1,2,3 is an increasing sub sequence. I know the answer is Catalan number b...

combinatorics permutations

Haunted

Thank you @ETHproductions and @ais523..

ais523

2:08 PM

@ETHproductions the worst case is n=2, I believe

but even then, half the difference is 2 which is still in the set

so yes, should always be representable by a single number

ETHproductions

@Haunted No problem. Thanks for the interesting math problem :-)

Haunted

just one more question, please. What if the number is 14

?

ais523

for larger worst-cases, something like n=46 overshoots to 66, but half the difference there is 10

which still just about fits

Haunted

1+2+3+4+5=15

ais523

for 14, we overshoot to 1+2+3+4+5 (which has the wrong parity) + 6 (still wrong parity) + 7 (now it's correct)

Haunted

2:09 PM

should i add 6 to this?

ais523

that's 28

28 - 14 is 14, halve it and we get 7

huh, so what about n = 12?

ETHproductions

Oh darn

ais523

I think I was wrong about what the worst cae is

*case

for 12, we still have to overshoot to 28

but now we have to subtract half of 16, i.e. 8

ETHproductions

I forgot the sequence doesn't change parity each iteration

ais523

we can still do that trivially, i.e. by reversing the 7 and the 1

but we need two reversals

Haunted

2:11 PM

:):):)

ais523

I knew it changed parity every iteration

ETHproductions

So the solution is always either T(n), T(n)+1, or T(n)+2

ais523

just didn't realise immediately the worst case was if adding 1…even was too small (I thought it was if adding 1…odd was too small)

but yes, you can't possibly need more than 2 additions to get the parity correct

because that adds an odd and an even numbre

Haunted

that link has a program which i couldnt fully understand.

can i paste it here?

ais523

don't paste the program directly unless it's a one-liner

you can paste a link to the program if you have comments about it

Haunted

2:13 PM

sure..

geeksforgeeks.org/minimum-steps-to-reach-a-destination

this is the link to teh question and it has a program written in C++

ais523

oh, it's basically a nondeterministic search program implemented using backtracking, which is in turn implemented using recursion

basically brute force

or, well, not backtracking, although it has the same evaluation order

it just always evaluates both probabilities and takes the better one

it runs in O(2^output), which is terrible performance for this problem

Haunted

i did not fully understand but i will look up the terms that you used. Thank you @ais523..you are kind and very smart. :)

ais523

so the basic idea is that when the program isn't sure what to do

it just does both (this is called "nondeterminism")

and the way it does both is to do one, then go back to the state before it did it ("backtracking"), then do the other (although it's not quite a typical backtracking-style program as it remembers the answer it got from the first case)

user165474

This might freeze your browser

Jan Dvorak

I'd just use BFS. I think it's O(n^3).

ais523

2:22 PM

the algorithm we came up with earlier is definitely O(n) and probably O(1)

@JanDvorak that's trivially beatable

if you assume that things like addition, square root, etc. are O(1)

Jan Dvorak

hint?

ais523

the hint is: find a solution for a nearby number, then change that solution so it works for the number you want

(that isn't a full solution, just a hint as to how ours works)

Jan Dvorak

gotcha.

Erik the Outgolfer

@HyperNeutrino How many popups will be blocked?

Poke

@HyperNeutrino probably don't want to post that as a clickable link... or at all

Jan Dvorak

2:26 PM

I clicked it accidentally trying to click the link I thought it was

Erik the Outgolfer

And that's why popup blockers are useful.

user165474

2:45 PM

@Poke ok sorry, can't delete it now :(

user165474

@Erik Well it just opens a copy of itself and then reloads, so if you have popups blocked, it will just keep reloading and telling you "popup blocked".

Erik the Outgolfer

Well, at least you do get warned that it might freeze your browser.

user165474

Yeah lol I posted it to this one annoying guy in my class who kept asking for cool programming "hax" (i.e. stuff like color a on batch) so I gave him the link saying it had cool programs on it.

user165474

He didn't click it, sadly.

user165474

Because people in my class know not to trust links that I post.

ETHproductions

2:47 PM

I don't think it's a good idea to troll anyone like that

...especially since you'll get a reputation for posting untrustworthy stuff :P

Down Christopher

I was tweeted

Yay

AdmBorkBork

root@terminal:~# love
-bash: love: not found
root@terminal:~# happiness
-bash: happiness: not found
root@terminal:~# peace
-bash: peace: not found
root@terminal:~# kill
-bash: you need to specify whom to kill

2

Down Christopher

kill *

or drop table people --

Getting stars in TNB 101. To get stars post a very context comment that when taken out of context has no good meaning. Then if no stars come say "I hope [this](your message url here) does not get starred or I will have the out of context question in a day"

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Playing Darts: become an amateur with the most efficient prioritizing code-golfarithmeticintegernumbergame Introduction: I think most people will know how darts work. But since some might not, here are the rules of playing Darts in a nutshell: Amount of players: 2 Starting points: 501 Goal: ...

AdmBorkBork

> become an amateur with the most efficient prioritizing code-golfarithmeticintegernumbergame

Lembik

3:26 PM

hi

does anyone understand why I am getting all these downvotes for codegolf.stackexchange.com/questions/113674/… ?

Down Christopher

Trivial?

Idk

Jan Dvorak

You're still deeply positive, though

Lembik

8 upvotes net

and a million answers

it's an odd mix

@DownChristopher It's certainly trivial that's true

AdmBorkBork

I downvoted the challenge due to the restrictive input requirements.

Lembik

@AdmBorkBork aha.. thanks for saying

Fatalize

3:30 PM

I didn't downvote but didn't upvote either. This challenge is basically "do X with cumbersome IO"

AdmBorkBork

Yeah, thought I had commented that. Sorry. I usually try to comment when I downvote something.

Lembik

that's interesting.. I don't see a text file with numbers in it as cumbersome

@AdmBorkBork Yes it would have been good if you had

@Fatalize What makes it cumbersome?

Fatalize

Either the challenge is about reading lines from a file, or it's about summation

Lembik

I mean there are 2 byte solutions!

Fatalize

Some languages can't read from files

or not easily

Lembik

3:31 PM

oh you really like it broken down into tiny little pieces

@Fatalize why does that matter? I mean every challenge you can imagine is terrible for some languages

look at the number of answers!

it's clearly not that restrictive

Fatalize

Because the challenge appears to be about summation, so IO should be irrelevant

Lembik

to my taste, a challenge that you can actually run and that does something you might plausibly want to do is more interesting

I get that not everyone agrees

but I wouldn't downvote a question that was totally useless

Fatalize

Well I didn't downvote, I just don't find it particularly interesting for the reasons mentionned

AdmBorkBork

@Lembik I've now commented. Like I said, sorry I didn't do that earlier; I thought I had.

fəˈnɛtɪk

Cumbersome IO as in a large number of languages can not read a variable number of lines

Jan Dvorak

3:37 PM

O_o

flawr

@MartinEnder It's a pity Mathematica's built in names are so long=)

AdmBorkBork

Uh-oh, you summoned him and he arrived.

Martin Ender

@Lembik it's interesting for those languages that can actually participate. ;)

a lot of people here probably view challenges more from the perspective of esolangs. in many esolangs, your challenge isn't about summing integers at all. it's about parsing input.

look at the Jellyfish answer for example. summing a list of integers is 4 bytes: p/+i. my answer is 24 bytes, because I need to parse the input.

and Jellyfish is fairly high-level as far as esolangs go.

many other languages have to read the input character by character, reconstructing the numbers from decimal digits, before they can do the sum, and then they have to decompose the number into decimal digits again to print the output.

so yes, there are many languages (which people like to use here), where a linefeed-separated list of integers is a cumbersome I/O format. but the problem isn't really the specific format, but allowing only a single format.

Jan Dvorak

but it would be too trivial without the cumbersome IO

Martin Ender

I have no problem with "trivial" challenges. No challenge is so trivial that you can't pick a language where it's a fun challenge.

AdmBorkBork

3:46 PM

I have a problem with trivial challenges, as I think they drag down the overall site quality. Simple challenges are OK, trivial challenges are not.

IMO

Lembik

@MartinEnder Thanks Martin. You are right that a lot of people here probably view challenges more from the perspective of esolangs

I have noticed that

Martin Ender

@AdmBorkBork I think the quality problem arises from people thinking that simple challenges require less care when writing the specification, not from the challenges themselves.

AdmBorkBork

Mmm, that's a fair point.

Poke

that right there is a reason i liked popcons

but i'm not going down that rabbithole right now

it's almost lunch time :]

Lembik

@MartinEnder I approach this from a work perspective. In my normal work I come across tasks that we really want to do and I like to make them into challenges

like summing the integers in a file

I just personally find it more fun if a question is something you might want to do but then taken to an extreme

of course the solutions are largely useless but that's part of the fun :)

fəˈnɛtɪk

3:51 PM

One of the issues with summing the integers in a file is that it is a strict IO format which disallows a lot of interesting languages from participating. It also is made up of two boring single part tasks for most non esolangs.

Lembik

@fəˈnɛtɪk but but.... is the number of answers not evidence that it does allow a lot of interesting languages

I mean a fastest-code challenge excludes almost every language

Martin Ender

@Lembik allowing a lot of interesting languages is not contradictory to disallowing a lot of interesting languages

Lembik

true in a literal sense

I don't really see why testing two aspects of a language is terrible I have to say

that seems normal

I mean any coding you do typically involves more than one aspect of a language. Loops AND assignment etc.

@fəˈnɛtɪk I challenge the boringness claim :) Just look at the bash solution

Martin Ender

golfing challenges aren't typical coding. ;) I think challenges work better if they focus on one thing. if your challenge is made up of two parts then you end up comparing the shortest summation algorithm of one language with the shortest input parsing algorithm of another.

Jan Dvorak

Aren't we supposed to compete within languages, not across?

Lembik

3:57 PM

@JanDvorak a good point

Martin Ender

ah well, shouldn't have said "comparing". my point is, for some languages it's a summation algorithm challenge and for others it's an input parsing challenge. I'd like to figure out a summation algorithm in Brian & Chuck for instance, but it would be completely dominated by a massive I/O parsing/stringifying code, that I'm not interested in writing (and golfing).

Transcript for