The Nineteenth Byte

4:11 AM

@fomin happens on weekends (also, it's holiday season)

7:24 AM

Hi all

anyone up?

Redwolf Programs

I am :p

@Lyxal Just invent a language that only has two operators and usually has really long answers but can sometimes be short because the control flow is compact and the operators are weirdly overloaded and then use it a few times and forget about it every few months then sometimes realize someone else is better at it than you and try to make more answers in it and sometimes update the interpreter then realize it isn't turing complete so change a major aspect of the language so it is but

3

then decide you prefered it when it was probably a finite state automaton because it stuck way closer to your original vision and there was so much more room for creativity within the limited possibilities so just add an option to the interpreter to use the old version but never actually do because you have more important things to do than spend hours working on an answer in a language you made a long time ago and need a lot of practice in

3

or is that just me

@fomin I am

:) what time zone?

IST

need help with reopening the question?

oh it''s already open

Thanks!

Just needs answers now

7:40 AM

@RedwolfPrograms no, I think what you missed is the part where you make an entire language based upon a concept you vaguely remember from a dream that you woke up from at 7am, spend about an hour making it only to have it turn out to be just another BF clone that uses curly brackets and desmos math logic, and have control structures that you aren't even sure work.

Then name it with a cringey name, put effort into making an esolangs page, forget about it for several months finally asking Dennis to place it on TIO.

And then laugh when you see your CG friend use it in an answer and ping you in the comments.

but that's prolly just me

(note that I actually asked Desmos why the expression {} evaluates as 1 and they said it was because it allows for people to type empty brackets without breaking their equation)

8:20 AM

@fomin you'll find that non-CG challenges are a whole lot less popular here.

for the "i don't like curry" question, wonder if rust macros are allowed

i feel like that'll probs be the only way to do it at all

or generics i guess? but still not sure if that counts as a function

Anush

9:48 AM

Morning all

@Adám that's a bit sad. Are they more popular somewhere else?

10:01 AM

@Anush I don't think so. In fact, I don't know of any programming sites where that's the main focus. I think that most people just find shortening code inherently more exciting and enjoyable than speeding it up.

Rahul Verma

I guess this ruby code isn't valid for I don't like curry

... but I can't find any valid reason

Actually in ruby, curried proc supports both syntax: f[1][2][3] and f[1,2,3]

Bubbler

10:18 AM

@RahulVerma It's worth asking in comment

Rahul Verma

@Bubbler Oh nice. I was expecting it a silly comment. But, I have to wait till the user wakes up :p

Anush

10:40 AM

@Adám Each to their own, said the woman as she kissed the cow

I love fast code when the speed comes from a clever insight

I would ask on topanswers but it seems almost empty, except for tex

@Anush no, it's: everyone has a taste of their own, said the dog as it licked its

3

Q: Generate 100 Java Tuple classes

11:05 AM

@Adám and less subjective/easier to judge, of course

@RahulVerma you're correct, it's not valid; i replied why

Newly Featured Questions

37

Write a script that writes to standard output, or your language's equivalent, 100 lines of (valid) Java code that begins with: class Tuple1 {public Object _0;} class Tuple2 {public Object _0,_1;} class Tuple3 {public Object _0,_1,_2;} class Tuple4 {public Object _0,_1,_2,_3;} class Tuple5 {publi...

11:21 AM

@ngn :)

I have so many things I want to discuss about my challlenge :)

@fomin this? are you Anush under a different username? :)

11:41 AM

@ngn iirc they're a friend of Anush

@Lyxal ah.. cool

And please don't check the two above message's edit history... It's late for me and I accidentally forgot how to grammar for a very short moment

grammar don't matter

ok so what lang next for i dont like curry

@Lyxal late as in 9pm

@ASCII-only 10 you Queenslander

Smh

Y'all refusing to believe in daylight saving assuming that the rest of us don't suffer having clocks an extra hour forward

Disappointing

\s

11:53 AM

@ngn not the same person but we are friends

12:12 PM

@ngn any interests in my challenge? I can think of a number of ways of attacking it which is what interested me about it

@Lyxal although we are all friends here of course :)

@ASCII-only rust

@Razetime there's a bounty for that, and i'm ineligible (since i already know rust) so i was leaving that for someone else

oh you're not supposed to know rust?

not to mention it may be a bit complicated

alright then

12:21 PM

> Since Rust is the language of the month, I will be offering a 50 rep bounty on Rust answer to this question if you haven't used Rust before.

emphasis mine

also the shortest (only?) way would be via a macro but i have no idea whether that's as valid as a function is

looks too non trivial for me to do in rust

has julia been the language of the month yet?

I am not sure how to find the list of languages of the month

Q: Nominations for Language of the Month, 2020 edition

16

We've decided to revive the Language of the Month event. We've also decided to start a new post for nominations. (You can see the old nominations post here.) So let's get to nominating! Procedure A language can be nominated for Language of the Month by posting an answer to this thread. We all vot...

discussion language-of-the-month

cool

julia is not on the thread

so you can add a post for it

12:31 PM

When I add a bounty to my challange, is it ok to say I will award it to the fastest solution for a language and choose which one on the basis of how impressed I am by it?

absolutely

it's your choice who you award your rep to

@Razetime I don't have time to write a lovely list of reasons yet but I will do if no one else does

@Razetime cool

@fomin i'm fine with it (and i'd imagine basically everyone would) as long as you do what you say in the bounty text

@ASCII-only cool. Thanks

@Razetime any other suggestions?

12:33 PM

umm

nobody has done haskell yet

how do you get the large font but with a link as in codegolf.meta.stackexchange.com/a/20473/91489 ?

tl;dr click edit to see how

I can't click edit it seems

but I worked out a method

thanks

2:42 PM

@ASCII-only Jelly could be a fun challenge hastily shoves 2 hours of failed Jelly golfing behind me

question

does jelly even have functions

Kinda

Each line is a "function" I guess

ok, does it have curried functions

For sure not in the conventional manner, but I'm sure there's some work around there :P

I tried going down the "generate code then eval" route

:|

@cairdcoinheringaahing "fun"

more "fun" than aoc day 20 part 2

2:46 PM

4 hours ago, by ngn

@Anush no, it's: everyone has a taste of their own, said the dog as it licked its

:P

@cairdcoinheringaahing lmao

I'm pretty sure this is ready to go, but I'd appreciate confirmation.

A: Sandbox for Proposed Challenges

@Adám Aside from the tags (I'd use ascii-art instead of code-generation, and string rather than generation), it looks good to me

New Sandboxed Posts

0

Help Stack Exchange transpose tables Stack Exchange doesn't know how to transpose tables. Let's help. Given a markdown table, transpose it. Input assumptions: All cells have equal either no leading space or exactly one leading space (you must handle both) If the cells have a leading space, then ...

3:09 PM

@cairdcoinheringaahing Hm, but technically it isn't ASCII art at all, but rather generating markdown code/generating a table.

4:57 PM

Any feedback?

@cairdcoinheringaahing Assuming you've triple-checked your test cases, it looks good. Two changes I'd make: I'd use i instead of k (for obvious reasons), and (also) provide test cases in a format that is easier to copy and paste.

@Adám Updated to have both a full table as well as some easy-to-copy test cases

5:16 PM

easy to copy test cases are everything

Neil

@ASCII-only I was thinking of adding a bit more support for dict objects in Charcoal

the bounty is up!

6:06 PM

@fomin yes, as i'm generally interested in fastest-code challenges, but i solved a very similar one recently, so i need some time to forget

6:26 PM

@ngn oh what was it?

oh, there's a bounty!

@fomin what was what? :)

"but i solved a very similar one recently" <--

@fomin as you can see ^^ i've forgotten already :)

:)

hooray!

what language do you like to use for fastest-code?

@fomin c, of course

@fomin i also like vector languages like k or apl, but they don't cope well with scalar operations and non-uniform data

7:04 PM

@fomin now that i read it more carefully, this challenge seems a little too trivial. you just have to find the right buffer size.

7:20 PM

Sometimes I wish that i could just get pure numbers from input instead of having to convert them when using FALSE.

because instead of doing something simple, I have to use 1_0[^$1_=~][0\[$32=~][48-\10*+^]#%]#% to get a list of integers from stdin

@ReedsShorts Have you had a look through the accepted I/O methods? There might be something that's nicer/shorter allowed (e.g. functions, ARGV etc.)

7:38 PM

well that saves a bunch of bytes off my code

@ngn I don't think that's right. How are you going to do the iteration efficiently? Too much duplication will kill you for example

@fomin iteration is nothing compared to i/o

i will try to the the closest strings in FALSE

however the problem is that the output might exceed 128k

@ngn I will pipe it to dev/null for speed testing

also should i assume that string data is already pushed to stack or should i read it from stdin?

7:53 PM

@fomin how much ram do you have? :D

@ngn 16gb. The output size is less than 20mb! :)

@fomin oh, so it can be a single write(), nice

Yes

Neil

... so many secret hats...

I am intrigued how fast your code is going to be!

7:57 PM

@fomin without algorithmic optimisations: ~200ms but this laptop is old and slow

That's pretty cool. For what length input string?

@fomin obenquillo - that's 10

Got you. Try it with a length 16 string with all letters different

That should be the slowest

@fomin should i assume string is already on stack or should it be input from stdin?

@Neil Those are the fun ones :P

7:59 PM

Inputted from stdin makes my life easier

will do

@fomin 560ms

Nice... But surely can be made faster :)

just got all three string operations implemented

and already 128ms, great

based on tio

@ReedsShorts ah, good idea :) i forgot tio exists

8:05 PM

Try or don't try, there is no time.

@ReedsShorts how do you redirect to /dev/null on tio?

the problem with tio is that output might be > 128k

i dont think you can redirect to /dev/null

@ReedsShorts Why not?

But how would you do that with other languages?

You'd run the language from the command line

Try it online!

8:09 PM

@cairdcoinheringaahing , i.e. run under TIO's bash, redirecting output to /dev/null

Try it online!

aha! and the .c file can be cat-ed from a heredoc

thanks @cairdcoinheringaahing @Adám

@ngn I would love to know what your algorithmic approach is?

@fomin the dumbest possible :)

tio.run/…

CMQ: What's the best way to word "Here are some inputs that aren't known if their corresponding outputs exist. You don't have to worry about them for now, but if the outputs are shown to exist, your program should work for those inputs"?

aight i kind of understand

8:17 PM

Your code is impossible to read. The dumbest approach is to iterate over all strings and check the edit distance so it can't be that

@ngn ^^^

i would have to figure out how to run FALSE code redirecting to /dev/null though

@fomin no, that would have been impossibly dumb :)

it applies all possible edits in two levels of recursive calls

my approach currently is to iterate over methods and positions

@ngn do you make all strings at distance 1 and then make all strings at distance 1 from those?

@fomin i make all edits in place, and fill a big buffer with them, and at the end write() it

8:19 PM

@ngn :)

this is gonna be hard and my goal is to just make a working program for now

@ngn how do you iterate over all the different ways of doing 2 edits?

@ReedsShorts Try it online!

@ReedsShorts This should work. Put your FALSE code between the DOCs (or the .s for Adam's)

@fomin recursion

8:21 PM

Sounds not so dumb :)

yep works

How much duplication do you have?

@fomin possibly a lot.. i think it depends on the structure of the input

Yes it will. Try all a's

@fomin i believe you without trying :)

but dupes are ok, right?

8:23 PM

@fomin my approach is to basically have 3 nested while loops

@ngn yes. They might just slow you down

I look forward to your answer. Please do make it read from stdin

@fomin i'll make an offering to the gods of randomness so they won't give you such input :)

@fomin but that costs an entire syscall! :)

ok, i'll make it read from stdin..

omg im gonna have 4 nested while loops why

HyperNeutrino

lol