The Nineteenth Byte

AZERTY have µ ._.

@Shebang Of course.

@Fatalize It's a GERMAN PENNY SIGN.

Well yeah it's useless

since germany now uses the euro

Wow, I lived in Germany as a kid and I don't think I ever saw that symbol o_O

Even when there were Marks

2:01 PM

What if the EU collapses? It will be useful then.

That makes it even more useless than I thought, thanks @Geobits

@mınxomaτ Oh, is that just the Build Tools install?

Why do people always ask "does this line of code count towards the code byte count?".

@TimmyD Yes. I'm setting up Rider+.NET w/o VS.

betseg

@Fatalize mobile ;_;

@EriktheGolfer mobile ;_;

Don't you have è in the list of e variants when holding on e?

Q: How to specify dialects of C?

2:03 PM

I've seen similar things happen when add-on modules get installed -- the option is controlled by the parent application. Likely it's looking for a regkey and not finding it.

New Meta Posts

0

When writing an answer, people here usually start with stating in which language it's implemented. When C is concerned, there is the "implicit int" rule that was, as far as I know, removed in 2011. This rule is very important for golfing. For example, a function calculating a square of a number: ...

@TimmyD It shouldn't depend on any regkey at all, or the regkey is mi$$ing.

Shebang

@mınxomaτ Because there are two questions in Meta with conflicting answers

That doesn't make it much better, since it requires agreement to an otherwise optional "privacy" policy.

@mınxomaτ Or, should I say... GOVERNMENTAL policy??

2:05 PM

Uh, no.

@mınxomaτ ¥ou n€v€r kno₩... brainwa$hing.

ℓ°ℓ ωντ

2

Q: Print the digital root

To find the digital root, take all of the digits of a number, add them, and repeat until you get a one-digit number. For example, if the number was 12345, you would add 1, 2, 3, 4, and 5, getting 15. You would then add 1 and 5, giving you 6. Your task Given a number N, print the digital root of...

code-golf number

@NewMainPosts We get it!

I just saw that Pyke answer go from 0 upvotes to 5 immediately

2:17 PM

Yep, standard practice when a language happens to have a one byte builtin that solves it exactly :/

If you're going to make a beeline for the builtin answer at least take the time to look at some of other answers too

Downgoat

@NewMainPosts pls. Y u all upvoet trivial built in instead I'd interesting approach

3

/me Just spent a half hour traversing URLs for repl.it because when I saved my program yesterday it didn't put the URL in my browser history. I finally found it.

Normally in the case of a trivial challenge I wouldn't complain but this isn't a completely trivial challenge and the builtin doesn't seem like it would be useful in any other situation

Oh nice JS answer.... n=>n-9*(--n/9|0)

2:21 PM

Not really muddyfish's fault though. Can't control the upvotes

@NewMainPosts Huh Pyke outgolfed Jelly

Something is wrong in the universe rn

Because it has a useless over-specific built-in

@Downgoat hey I spent a good minute making that builtin

xD

Downgoat

@NewMainPosts was this sandboxed?

2:22 PM

no

feature request: make @NewMainPosts answer correctly (or @NewSandboxedPosts)

@muddyfish The digital root builtin is newer than the challenge

At least the last rev

Fix digital root sum in map - didn't change arguments in time

muddyfish committed on 10 Jun

@muddyfish github.com/muddyfish/PYKE/commit/…

> 9 minutes ago

2:25 PM

@TuxCopter github.com/muddyfish/PYKE/blob/…

(My entire commit base broke about half an hour ago)

Yeah he changed it to mod 9 after the challenge and then changed it back

xnor

@MitchSchwartz twas me

anyone know how to fix my commit base so github says the commits were on the day they were on?

A: How can we help users who are put off by the use of golfing languages?

What happened to

64

Stop upvoting trivial solutions It's easy to imagine why a new user would be frustrated when the highest-voted answer is a 3-byte solution in a golfing language with a built-in that nearly solves the problem. It feels like no matter hard they golf in a conventional language, their solution will ...

It is a trivial problem but I didn't expect/want so many upvotes for that.

2:29 PM

I don't think the overlap of people who read our meta and the people who drive-by from HNQ is that large ;)

It hasn't had time to hit HNQ yet

@Geobits Here, I drew a Venn diagram --> O O

@DJMcMayhem the clearest answer is to upvote my questions more :)

@DJMcMayhem You sure? It's had a bunch of answers quick, and that usually throws it over there almost immediately.

It is trivial; though this is yet another instance of a 1-2 byte built-in which is completely useless in general and which can be reimplemented in just a few bytes anyway

2:30 PM

stackexchange.com/questions?tab=hot

I will never understand the point of these

I'd much prefer this answer to get more upvotes as it took so much effort

It's the third hottest network question right now.

Is there like a "software" RAID library? I.e. input is a file and the RAID type, and it outputs n files containing the parts.

@Geobits it is... how?

2:31 PM

Lots of fast answers.

Because beta sites get on HQN more easily

^^ That's the number one contributor IIRC

(lawl)

Our trivial/simple challenges like this almost always hit HNQ really fast.

@mınxomaτ Zip can do that, but only for RAID-0 ;-)

2:33 PM

Most other sites never have questions that get 20 answers in 30 minutes, so I doubt the algorithm was designed with that in mind.

Huh, catalog no longer exists?

I do think there is an oddity where questions are downvoted if they are hard for esolangs and upvoted if they are particularly easy for them

it's inconsistent in my view. The line goes "Don't ask that as esolang don't support that sort of thing well" but no one says "don't ask that because non-esolangs don't support that sort of thing well"

Questions that are hard for esolangs are generally hard for normal langs

@Fatalize for example complaining about the use of floating point numbers

or I/O

Questions that take lots of effort to do tend to get fewer upvotes

2:34 PM

I see complaints about floats much less now then I used to. Now it's usually a debate about how precise it needs to be instead.

perfectly normal coding challenges are downvoted for those reasons

A: Should there be a [catalog] tag (e.g. shortest in every language)?

6

No there shouldn't be a tag. They aren't really different from a normal challenge, except that they focus on simple/standard programming exercises and explicitly ask for answers in non-competitive languages. That makes the tag seem very much like a rather arbitrary meta tag. I would be in favour...

ok

Downvoting for that is obviously stupid

it happens to me sometimes too

2:35 PM

But I can understand people not upvoting challenges that are fairly restrictive and very difficult

Well, you think that's why people are downvoting. Unless they say it explicitly, it's not useful to assume that.

6

there are all sorts of rules in the meta for what make bad questions which seem aimed at those who love esolangs

I do think that esolang challenges and answers are great here but they should be only part of the story

I'm going to say it now: Don't let this argument get heated. (I've fallen into that trap before...)

I thought they were aimed at being inclusive, but not to the point of banning challenges that don't suit all languages

to my eye, once you have seen 10 answers in Jelly you have seen them all

ok let's take codegolf.stackexchange.com/questions/96825/…

2:38 PM

@Lembik They are? On the digital root for example, half are eso/golf and half are "normal" languages. Even with the trivial builtin, people are still posting in other languages.

"Those challenges include operations on prime numbers because they actually add something to the challenge (or are the core of the challenge). Requiring complex support in this challenge is adding unnecessary fluff by requiring the use of unnecessarily complicated number types, which causes it to be a chameleon challenge. Three hits on the "don't do these things" list is not good. "

there are three rules quoted there

I can't copy the links sadly

The vast majority of languages don't support complex numbers

0

Q: Convert Chinese numbers

In Chinese, numbers are written as follows: 1 一 2 二 3 三 4 四 5 五 6 六 7 七 8 八 9 九 10 十 For numbers above 10, it is expressed as the number of tens and the number of ones. If there is only one ten, you do not need to explicitly say one, and if there are no ones, you don't need to put anything afte...

code-golf

@Fatalize Esolangs

@Fatalize let's say that is true.. so what?

2:39 PM

Even outside of esolangs

Most languages now support complex numbers

I am not even sure it's true for popular languages

I mean C does

which popular languages don't?

@Lembik ~~TI-BASIC.~~ Wat, even TI-BASIC supports?

they do with libraries

@Fatalize complex numbers are in the C standard.

2:40 PM

Even Java doesn't without something like Commons (yuck)

and you still have to include a library

The meta list of things to avoid when writing challenges has a lot of very useful advice in it, but none of them are rules. They let you know what is difficult to get right, but you can still go against these if you are confident you found a way that works

Yeah, that's what I mean. Not even a standard library.

@Fatalize In C you can't do anything without a library[33160824].

@wizzwizz4 true!

2:42 PM

There's a diff imo between a standard library and an arbitrary third-party one.

@Lembik Are you sure? What are they called?

That's like saying OpenGL is "part of C" :P

@DJMcMayhem #include <complex.h>

The point is that by including complex numbers in your challenge, the only thing it does is make it much more obnoxious to solve with languages that don't support complex numbers easily, and has nothing to do with the challenge you are really asking about

@Fatalize I just don't get it. It's makes it a programming challenge for languages that don't support it!

which is the whole point of this site

2:43 PM

By including complex numbers you change absolutely nothing to e.g. possible J answers but you make it so that the majority of a Java answer will be about dealing with complex numbers

it's a fantasy, in my view, to pretend that the parts that are hard are going to be the same in every language

@Lembik Example: Add 2 numbers together. You must allow support for complex numbers

@Lembik But the challenge isn't to implement complex numbers! It's to do something else.

@wizzwizz4 that's not right.. the challenge was to do something else using complex numbers

This could have been avoided if you'd named it differently and introduced complex numbers in the setup. "To make the challenge a little more interesting, the matrix may include complex numbers" is the death sentence here, because it's all about presentation.

2:44 PM

you might be able to find a particularly efficient way of doing this

@Geobits ah

I think in some cases the complex case requires a sufficiently different approach that for some languages the challenge wouldn't be a duplicate if posted in both forms (with and without complex numbers). However, it isn't clear whether duplicates should be judged on whether some, all, most or one language can be trivially ported and remain competitive

@Geobits "To make the challenge a little more boring" better?

:)

@Geobits I know it sounds dumb and arbitrary, but it's true.

@Lembik Print the contents of this array. You must support the array being in Small Basic's internal array format.

@wizzwizz4 ?

2:45 PM

So all in all my point is: by including complex numbers, this doesn't pose any additional constraints to language that natively support them, whereas for languages that don't this shifts the majority of the work from the actual challenge problem to supporting complex numbers

@Lembik It's simple in Small Basic; you just set a variable to the input then loop through the array. In other languages, it's far from trivial.

@Fatalize right.. but I think that is vitally true for a huge range of challenges. See the 5 byte Jelly answers

those questions which permit a 5 byte Jelly answer are clearly introducing a constraint for all the other languages that don't have some super short way of computing whatever the function is

@Lembik Jelly just makes it short, not necessarily easy.

@Fatalize I don't have much empathy for that viewpoint. As a non-esolanger, There are lots of challenges where other languages have builtins for things which eliminate what turns out to be a bit of work for mine.

No they don't, because the Java answer for that would only be about solving the question posed in the challenge

2:47 PM

@wizzwizz4 look at +/ .*

it just turns out Jelly has exactly the right functions built in

@Lembik For some things, but not others.

@Geobits But that work is directly related to the problem posed; dealing with complex numbers is not needed to implement the algorithm that computes the permanent of a matrix

right.. but my point is just that it's almost always the case that in a code golf challenge, some esolang has a very useful function which other languages have to work harder to do

that's the whole point of the golfing languages!

no one votes to close those questions

How about this: you are only allowed to use six different non-alphanumeric characters in the source code. Instantly Jelly finds it harder.

LOL the most upvoted answer for the primality check challenge is hello, world!

2:49 PM

@wizzwizz4 don't get me wrong.. I love jelly (maybe jello to you)

My point is that dealing with complex numbers has nothing to do with the implementation of the permanent computation

@wizzwizz4 easy in brain-flak. :P

@Lembik Nope, still Jelly to me. (You do mean the wobbly thing, right?)

I do :)

@Fatalize Sometimes. It depends on if you consider "changing an integer to a string" directly related. In some cases that's a quarter of the code :/

2:50 PM

@Lembik Jelly == jam, jello == dessert

@Fatalize I get that you can split challenges into parts. You could split the permanent function into parts too

@DJMcMayhem only in the US I think

I still think that if the challenge had been posed as "complex permanents" instead of having complex tacked on, it wouldn't have received the same comments. Even being the same exact challenge.

That's all I know

@Fatalize also it's boring to just ask "Give functions that do arithmetic with complex numbers" Yawn!

It's much more interesting to ask some to actually compute something worthwhile with them

@Geobits a very interesting point!

@Lembik Let us found the CAMRJ here and now, and forever more have real jelly! :-)

2:51 PM

A different example -- my Totally Invertible Submatrices challenge. I was originally considering R, but rejected it as being too complicated, so I went with Q instead. However, Peter (rightly) pointed out that Z is equivalent because you can just scalar-multiply the matrix to get from any Q to an equivalent matrix in Z. So, rather than focusing on Q, I went with Z, because the number type wasn't the interesting part of the challenge.

@wizzwizz4 :)

@Lembik My point is that what's interesting in a challenge about "computing the permanent" is the way you implemented that. If Java suddenly supported complex numbers the golfed answer that would have been posted would be the same

Sometimes golfing gains can be made which only apply when two tasks are combined. To insist on only ever having a single atomic task per challenge would remove this possibility. I like seeing a range of challenges, some simple, others involved.

4

Q: Calculate the permanent as quickly as possible

@Fatalize It's also not clear to me that you can never use the particular function you want to compute to golf complex arithmetic differently

@trichoplax snap :)

here is another example where number type is important

22

The challenge is to write the fastest code possible for computing the permanent of a matrix. The permanent of an n-by-n matrix A = (ai,j) is defined as Here S_n represents the set of all permutations of [1, n]. As an example (from the wiki): In this question matrices are all square and w...

math fastest-code matrix

@Lembik Sometimes you need to rewrite the entire implementation of multiplication to work with complex numbers. Sometimes you don't.

2:53 PM

turns out I didn't realise this would become a big Int challenge

it would have been better with floats

@wizzwizz4 right

in fact I will make another float only challenge ...I would like it to be one where all the answers are ranked per language

I think there is a code snippet for that

@Lembik There is.

I think it's interesting to give the fastest bash code :)

@Lembik You're going to do another challenge just like that except floats? :/

@Geobits yes.. it's very controversial but the aim of the other one is speed per language

I am happy to accept it might be a terrible idea

(and not about big Ints )

@Geobits I did ask about this earlier on chat

and no one shouted at me ;)

You could offer a bounty for the fastest code in the slowest language: Take the fastest answer in each language and award the bounty to the slowest of them. It sounds counter to the winning criterion but still pushes for the fastest possible per language

2:56 PM

@trichoplax nice!

I really like that idea in general

@trichoplax I think you might just end up with ports of the fastest algorithm into some slow scripting language.

let's do more of that!

It sounds like the challenge would end up finding the slowest language rather than improving the speed.

@Geobits hard to port codegolf.stackexchange.com/a/97123/9206 :)

that's an answer which doesn't exist in most languages

@trichoplax I can see a few ways that could go wrong. What if I'm the only one who posts in the vbscript emulation of that obscure processor where the emulation introduces new behaviour?

2:57 PM

@Geobits I wonder how many tasks that would be true of (certainly some). I can imagine there being some languages for which the asymptotically fastest algorithm wouldn't be fastest on the test cases

@Lembik You don't even have to port. Just find some really obscure, slow language and you win the bounty because nobody else will bother with it :P

@Geobits what do you have in mind?:)

@wizzwizz4 Then you've advertised the existence of it, and the fact that it's the slowest and therefore the bounty target. If no one picks up that information to compete with you, then you win...

Turbo C Shell time

the main real problem is who is going time it?

it's hard enough when the code is in Rust

2:59 PM

@trichoplax But then the challenge is "find the slowest language", not so much "do this task quickly".

@Geobits QFTASM comes to mind?

@trichoplax But it's implemented in an old, proprietary "fork (of specification)" of vbscript that isn't available for purchase any more...

All you have to do is wait for the bounty timer to be (almost) up and snipe in something ridiculously slow.

find the slowest language that the OP can actually run on his/her computer...

that sounds more challenging

@Geobits I agree it wouldn't make a good winning criterion, as it discourages fast languages in the same way the fastest-code discourages slow languages. I just thought it might be an interesting bounty to offer after the fast languages have finished competing

3:00 PM

or you could insist it in the top 50 list

@muddyfish What if I wrote a language that would compile to a QFTASM implementation of the Game of Life running a QFTASM program that completes the task, but only if the hash of the program is X...

tiobe.com/tiobe-index

@Geobits Good point. Maybe it would work better as an open ended bounty with a 7 day period after the last entry before awarding

That's it. I'm creating DecafJava. It's just like regular Java except it inserts a Thread.sleep(1000); between every statement.

@wizzwizz4 If you can write this tailored language before the challenge is posted then hats off to you ;)

3:04 PM

@Geobits Yeah? Well, I'm creating SleepyJava. It's just like regular SleepJava except the sleep time is the same as (the rolling character count of the program)^(the total character count of the program).

Oh yeah, well ComaJava does that, but also adds an additional for(int i=0;i<1000;Thread.sleep(i++)); all over the place.

What is happening here

Not much. Mostly sleeping.

Sleeping while golfing sounds dangerous.

I'm not golfing right now though, so it's all good :)

3:11 PM

Nah. You fall asleep, your head hits the keyboard randomly, and you've banged out the answer in a new esolang.

If you hit $ enough while that happens, it's probably also a powershell answer.

A: Sandbox for Proposed Challenges

True

Of course, I do love how $$|?{$_} is valid PowerShell (and if you understand the syntax, even makes a little bit of sense).

New Sandboxed Posts

0

Algorithm Analysis Background: Recently, at least one of our fellow golfers landed a job at Google. From what I've heard, Google's interviewers often ask about various algorithms, so it is recommended that an interviewee learn as many as possible before attending one. However, it's been awhil...

Anyone here have experience with Python's ctypes (specifically how it works internally)?

How does it know the return type of a function without the header?

3:28 PM

@quartata magic

Thanks. That really helped.

@quartata it looks like you use ctypes to call other C functions?

Yeah.

you only need to know the type of the function being called

so, at runtime, if your variable doesn't match the parameters the function needs, you throw an error

that seems like the logical choice

Yeah, but how does it know the return type?

dlsym doesn't give you that information.

3:31 PM

why do you need that info?

To know what I'm getting back

It's just meaningless garbage otherwise

ctypes does it somehow because it automatically coerces to a Python object, but I don't know how it does it.

I should probably just read _ctypes.c

so, it just gets a stream of bytes back?

Well yeah. C doesn't keep type information around at run time

dlsym gives you a (void *)() function handle. You have to cast the pointer to something

right, I guess I assumed it would return a string back

popularity-contest obfuscation

-1

Q: Write a Burning Rope

Here is another common interview question: You have two ropes. Each rope takes exactly one hour to burn. The ropes can be lit at either end at any time, but do not necessarily burn at a uniform rate. Can you use these ropes to measure 45 minutes? If so, how? If not, why not? The solution to...

3:33 PM

and you could detect the type from the string

but that makes sense

Ugh, Python extension modules are so ugly.

Oh no. It better not be doing what I think it's doing.

I need to make one myself!

having never done it before

Oh no. It is.

casting it to an int everytime?

It doesn't cast the pointer at all. It assumes it's numeric until you do something otherwise

No wonder the example in ctypes is printf and not sprintf. It wouldn't even work at all

3:39 PM

so, if I do for c in returned, would it assume its a string?

>

By default functions are assumed to return the C int type. Other return types can be specified by setting the restype attribute of the function object.

@NathanMerrill Array. But yes.

I suppose I should have read the Python docs a little more carefully

OK, so I need the header to do it the way I want to

Does anyone know where I can find the specification for GCC's precompiled headers?

0

Q: Write a metaquine polyglot

Your task is to write a program, that when run in two different languages, produce quines for that language it is being run on. When run in two different language, it should produce a quine for the language it is run on. Restrictions: The quines your program produces cannot be the same. The t...

code-golf quine polyglot

^ Whoops... didn't realize I had earned dupehammer powers

@ETHproductions It's a dupe, though.

@TimmyD I read it again and it's slightly different; it needs to output a quine in each language, but it doesn't need to be a quine.

Don't know if that's enough distinction though

El'endia Starman

3:53 PM

@ETHproductions: I just noticed that your avatar is inverted in two different ways. Clever. :P

@El'endiaStarman ha, thanks

@NewMainPosts I'm not sure why this received so much negativity

@ETHproductions Wait, I'm not following -- you're saying I take the program, run it in Python, and it produces Python code that is a quine, but is not the original code? Is that what's being asked?

I mean, its nearly a duplicate, but sufficiently different

its code that is a polyglot that produces a quine

its not a quine itself

OK, if that's the case, that's completely unclear to me from reading the challenge.

3:58 PM

the first sentence is pretty clear:

Your task is to write a program, that when run in two different languages, produce quines for that language it is being run on.

@TimmyD Same here, which is why I voted to close (and accidentally closed it myself; I wanted other people to vote on it)

The quines your program produces cannot be the same.

I don't think you guys should close it because the first time you read it, you misunderstood it

Me neither. I'm going to vote to reopen, if I can

Isn't that the literal definition of closing as unclear?

the language is clear

4:00 PM

Does it just need some emphasis added in places?

Oh look, I have un-dupehammer as well

literally what he wrote is clear, its just that "metaquine" makes you think that you are going to be writing a quine

@trichoplax that would be helpful, I think

Maybe something like "Polyglot distinct quine generator" would be clearer, but there's only so much you can expect to be understood from just a title

right :)

what a mouthful

What about "Write a quine generator polyglot"?

That is, s/metaquine/quine generator/

4:03 PM

That sinks in more readily, I think.

can we really change the title of a post, though?

like, I'm all for suggesting that edit, but I think that's going too far

We can certainly suggest a new title in a comment, and heavily upvote it

that sounds better

@trichoplax Oops, I just changed it. I'll change it back

I need to stop jumping ahead of everybody :P

More than just "does not have to be a quine itself" it cannot be a quine, if it is required to give distinct quines for the two languages

4:08 PM

Theoretically, it could be a quine in one language but not the other...

"Your task is to write a program, that when run in two different languages, produces a quine for the language it is run on"

oh, I understand

you're saying quine 1 could simply be the same output

@ETHproductions Oh good point...

I've edited my comment to hopefully reflect that

I'm curious: there are at least three well defined "terms" we can use in our challenges. Aka, we can refer to the "output", the "input", and the "source code". We can put restrictions on any of these and its well defined.

is there anything else that is well defined?

So, what "welding mirror" should I use :D

4:25 PM

@NathanMerrill If you treat the program as a black box and don't score based on internal workings (which seems advisable) then the only other thing I can think of is running time

yeah, that's true. otherwise fastest-code isn't well defined

@xnor If you want to extend your Python tip on combining multiple loops into a single one, I made a tool that can do it for any number of loops: repl.it/EHwa

I think the cost outweighs the gain pretty quickly for golfing purposes.

except that each inner loop adds more indentation

so I think it depends on how big the code you're looping is

@NathanMerrill It's still only approximately measurable, depending on using a defined machine and running multiple times, but it seems well defined

that said, its usually not more than 1 or 2 lines

4:28 PM

@NathanMerrill Yeah, but when it gets to a large number of loops, it's probably shorter to use exec and have a format string for the for part.

I guess memory used is also well defined, for a given implementation and for a deterministic program, but again awkward to measure

@trichoplax its well defined in the fact that "I will run your code, and your score will be the time it took"

Or to use product(*map(range,v)) where v is a list of the range values for each loop.

@NathanMerrill But your computer might slow down suddenly for some reason not related to the running code.

There's no way for you to judge that.

you're right

but its still well defined

even if it isn't fair

I can objectively say "It took 3.42 seconds on my computer"

@NathanMerrill I guess I should be using the word "reproducible". Input, output and source length are all the same measurement for anyone, whereas running time depends on being able to use the same computer, and memory used depends on being able to use the same implementation (and possibly the same machine?)

4:31 PM

It's not a deterministic scoring method

Even if the code is.

@NathanMerrill Yes, but it might not take 3.42 seconds next time you run it on the same computer

that's actually a good distinction. We require challenges to be well defined, but the best challenges are reproducible (which is why fastest-code challenges aren't the best solution for challenges looking to measure runtimes)

Ideally you measure the run time a large number of times, and then take the minimum as the score, for a deterministic program. For a stochastic program the minimum is misleading, so I guess the best you could do would be taking an average, unless you take the minimum with a fixed seed, and then test a wide range of seeds

@xnor Here's a related program when I was working on it: repl.it/EHtK/2

@trichoplax true, but even then, its not reproducible, even if everybody was testing the submissions on the same computer.

fastest-code, technically speaking, isn't reproducible. We just try to minimize the distribution

4:36 PM

This is why I quite like competitions to find the best solution rather than measuring run time directly. It's a bit vague, as people can run their code for as long as they like on as many computers as they like, but it avoids having to run everything endlessly on one machine

right :)

I don't see huge numbers of answers for such challenges, but I like seeing the few answers I do get, so I'm inclined to try and think up more of them

code-golf graphical-output

0

Q: Display an xkcd

xkcd is everyone's favorite webcomic, and you will be writing a program that will bring a little bit more humor to us all. Your objective in this challenge is to write a program which will take a number as input and display that xkcd and its alt-text (mousover text). Input Your program will ta...

@NewMainPosts Since graphical-output requires "drawing pictures on a screen or generating image files", it doesn't seem to apply to a challenge that only requires fetching pre-existing images. Does anyone see a way it would apply? I'm removing it for now

4:52 PM

Intriguing: Intel Atom and ATMega on one SoC

@Geobits @NathanMerrill Could you take a look at the comments I made for my latest sandbox post and see if they would help?

Hrm, looks like from webbrowser import* is shorter than __import__ :/

@trichoplax does graphical-output care about how the the image is obtained?

That's what I'm trying to find out. The wiki wording to me suggests generating images rather than accessing them from elsewhere

@NewMainPosts Huh, this is a dupe of "open PPCG in the def. browser" per the answers. Not the question though.

4:55 PM

@Yodle as I said: "there are an infinite number of possible O-runtimes, and fairly easy to come up with bigger ones"

@NathanMerrill I would guess most people following the tag would be interested in methods of generating images rather than internet access, but that's only a guess so I wanted to check what others use it for

A: Sandbox for Proposed Challenges

that's a fair point

are we tagging for semantics or tagging for users

New Sandboxed Posts

0

Solve the Laplace equation Sandbox note: I will replace the formulas with images of nice LaTeX outputs This is the "Hello, World!" of PDEs (Partial Differential Equations). The Laplace or Diffusion Equation appears often in Physics, for example Heat Equation, Deforming, Fluid Dynamics, etc... ...

@NathanMerrill Yeah i guess it still doesn't address that, do you think there's a way to incorporate an implementation's big O into the score to stop people from just implementing slower and slower versions of the same thing?

@JonathanAllan Don't quite get your comment

4:58 PM

well, if you put a cap, then there is a finite number fo algorithms possible

@mınxomaτ I think that's a consequence of the challenge being wordy and requirements being overlooked. Simply opening the page in a browser is invalid

so, it'll basically be code-golf with multiple algorithms