The Nineteenth Byte

Neil

00:00

@user here you go: ^(((<|>|$(?3){3}$)(?=(?>(?4))?>)(?3))?(\((?1))?)(.) - result is in $5.

user

00:14

@Neil That's brilliant!

Redwolf Programs

00:52

CMC: Given a number, zero or one indexed, return the number with that index in [0, 1, 2, 4].

So 0 -> 0, 1 -> 1, 2 -> 2, 3 -> 4

Neil

@RedwolfPrograms Python 2, 15 bytes: lambda n:2**n/2

user

01:25

@RedwolfPrograms Scala, 16 bytes: n=>((n+5)%4+4)%5 (there's probably a better solution)

@RedwolfPrograms Scala, 6 bytes: _*10/7. I was being dumb before

Neil

@user note that the leading ^ was only necessary for your /gm flags in your linked version, I submitted my version as an answer without it

user

@Neil Ah, I was wondering why that was there

New Main Posts

01:56

0

Q: Golf a number larger than "Bitetremar"

Definition Bitetretar is a large number equivalent to the number in the Fast Growing Hierarchy, so use that as the number to surpass. Rules You may not accept STDIN input. You may only use an esoteric programming language. You cannot use any functions. Scoring This is , so the program with the...

code-golf

Jonah

02:51

Can anyone with experience in both speak to how Wolfram/Mathematica compares to Sage? I want to learn a high-level math-oriented language, and like the idea of Sage, but wouldn't paying for Wolfram if the language is significantly better.

Razetime

03:12

@Jonah Sagemath is an extension of Python which is made specifically to be math oriented

2x-1

@Jonah Mathematica is a unique functional language on its own

Razetime

If you're looking for math only then I'd use SageMath, it's got enough in that department

Mathematica is just.. insane

Jonah

Meaning... insanely fully featured?

2x-1

I love the M-expressions in Mathematica.

Razetime

It's got so many quirks and updates with random functions nobody will ever use

It's got hundreds of custom datatypes

and functions to use those with

2x-1

03:14

But sometimes, they can help you win challenges, if you find the right built-ins.

You can do a lot in a few bytes.

(Really, I hate Python.)

Razetime

hopefully Jonah doesn't lol

I think the wolfram engine is free for use

so if you want, try that out

mathematica has excellent documentation

Jonah

I like Python fine. It's not really a negative or positive. I'm more interested in knowing.. is SageMath mature enough that, say, many professional mathemeticians would use it over Wolfram?

2x-1

Obviously, Mathematica is a lot more advanced than SageMath.

Razetime

Mathematica is a separate language meaning it has its own stuff for displaying images, visualizations, and stuff in house

any mathematician would prefer mathematica unless they had a soft spot for python

sagemath also has visuals, just not equally customizable or varied

Jonah

Thanks.

Redwolf Programs

04:22

Opinions on VTRing this question?

Lyxal

Wow, closing the question actually brought the OP back to edit it

Also, I'm unable to think of a new music meme for this weeks chat bio

Redwolf Programs

Pretty impressive for an 8 year old question, esp. a day or so after it was closed

Lyxal

nvm i got it

Redwolf Programs

I'm excited

Lyxal

okay updated

I change both profile and chat bio weekly

usually with verbose memes

CG profile gets a regular meme

Chat bio gets a music meme

Redwolf Programs

04:27

I liked the one a few weeks ago

With the gist

Lyxal

hehe

I put effort into it

Redwolf Programs

Do you still have the link to it

Lyxal

one second

you reposted in the wrong Lyxal

also, i just remembered i need to reapply for gh pro

actually nvm

that's tomorrow's job

mainly because it hasn't expired yet lol

Redwolf Programs

Gonna put it in a markov chain

Lyxal

@RedwolfPrograms can you put this conversation into the chain next?

Redwolf Programs

04:33

gist.github.com/RedwolfPrograms/…

I'll copy the conversation next

gist.github.com/RedwolfPrograms/…

Ignore the misspelled file name lol

gist.github.com/RedwolfPrograms/…

Lyxal

@RedwolfPrograms that is extremely hard to rap

Redwolf Programs

gist.github.com/RedwolfPrograms/…

@Lyxal Bet :p

Lyxal

@RedwolfPrograms razetime country.

Redwolf Programs

Uhh

> Tasty are dennis same person.

Lyxal

> Offended God p.

Redwolf Programs

04:39

> Everyone 22 can't do anything about being.

Facts

Lyxal

"Dad, why do we make people fight wars?"

> For your country, Because Razetime.

5

Another favourite

> Meanwhile always is no.

Razetime

@Lyxal what

oh

markov chains

Lyxal

05:54

@Razetime no questions

Comprehend?

Razetime

06:30

comprehend indeed

Lyxal

good

you will not receive any punishments tonight for your compliance

scurry along now

Unrelated String

07:03

huh, it's been exactly 2 years since i joined cg&cc

Lyxal

happy cake day!

oh wait this isn't reddit.

Unrelated String

@Lyxal ~~damn, i was about to repost the left or right challenge again~~

Lyxal

@UnrelatedString ~~do it~~

~~you wont~~

user41805

08:12

200 rep to the best dc answer to codegolf.stackexchange.com/q/11336

Newly Featured Questions

08:33

29

Q: Print a string in as few distinct characters as possible

The task is simple: print The quick brown fox jumps over the lazy dog using as few distinct characters in your source code as possible. print "The quick brown fox jumps over the lazy dog" has a score of 29 characters, as all 26 lowercase letters appear at least once, as well as an uppercase T, sp...

code-challenge

user202729

08:46

Any remaining problem with codegolf.meta.stackexchange.com/a/19310/69850 ? going to post it soon.

seshoumara

10:49

@user41805 your bounty brought me back to make new bash, sed and dc answers and I was shocked no one thought of the trivial dc answer to that question. I posted it at codegolf.stackexchange.com/a/218683/59010

@user41805 In fact I suspect there are multiple 3 char solutions possible, besides the trivial ones of replacing 1 in my answer with any digit from 1 to 9. For example, one can use multiple v operations to do successive square roots from an incredibly large unary number until you get to the ASCII code.

user41805

11:35

hey @seshoumara, nice seeing you again

user202729

It's obvious that there must be one digit, one operator and P, right?

seshoumara

hey, thank you, nice to be back.

user202729

If it's not Z, it must be an arithmetic operator.

However it will throw various errors on stderr (which is allowed)

seshoumara

@user202729 Yes, that's how 3 char solutions will work exactly. With Z though, no error is given.

user202729

@user202729 Why would anyone go upvote the other comment there. :/

seshoumara

11:41

There are a few operators in dc that work with a single number, but besides Z, perhaps v (square root) can be used, but I highly doubt it one can be that lucky. (as explained above).

user41805

i'm still thinking if there is yet a shorter approach (but some bytes can be shaven off of your solutionto beat the jelly solution)

user202729

Obviously there's.

Not with Z though.

seshoumara

wait, but wasn't your bounty specifically to the best dc answer?

user41805

yes, so 'best' first by the count of unique chars, then by bytecount if there is a tying dc submission

user202729

Guess I have something to do now.

seshoumara

11:45

Well currently there is only one 3 char dc answer, so I can relax for the moment :)

user41805

hang on, i think you can have 2 unique bytes

user202729

I can't figure out how.

12 mins ago, by user202729

It's obvious that there must be one digit, one operator and P, right?

...

user41805

oh i remember now, i didn't generate that dc solution because i couldn't (be bothered to) figure out how to split an array in j

user202729

You can do zzzzP but that...

Wait that might actually work.

user41805

:)

user202729

11:49

Expect the resulting program to be not runnable, however.

user41805

i mean, you can still golf it to probably being able to run it

user202729

Is it fair if I try to beat you to posting it? You came up with the idea.

user41805

that's why i put the bounty, to encourage others to it

so go ahead

seshoumara

@user41805 nice one!

user202729

Oops. The program takes 271 TiB.

And it's the best one.

seshoumara

12:03

hmm, how do you decrease the number after some succession of z?

user202729

You can't decrease it, right?

Only z and P can be used.

Since you need to store that many elements in the stack...

It's impossible to run the program without swap, and thus it will definitely take years to run.

Well anyway. It's valid right?

Unless dc has some weird limits like "there can't be more than 2**32 elements in the stack"

seshoumara

wait, I have another idea :D

Razetime

hey seshoumara

user202729

I can't find any instruction that pushes without an existing number except digits and z, and z alone makes the program that large.

seshoumara

using z and P, to make it drastically shorter than terrabytes

user202729

12:06

I did do that.

That's the optimal one.

seshoumara

@Razetime hey

user202729

#!/bin/python
from math import isqrt
import math
import subprocess
from typing import List, Tuple, Union
import functools
print()
s="The quick brown fox jumps over the lazy dog".encode('u8')
assert int.from_bytes(b"AB", "big")==(65<<8|66)
@functools.lru_cache(maxsize=None)
def value(i: int, j: int)->int: return int.from_bytes(s[i:j], "big")
@functools.lru_cache(maxsize=None)
def work(i: int, j: int)->Tuple[Union[int, float], List[int]]:
	assert i<=j<=len(s)
	v=value(i, j)
	if j==len(s): return (v+1, [v])

Razetime

dc answer i see

user202729

That's the generator program.

Razetime

@user202729 what

seshoumara

12:08

my python is rusty at best, so just post your answer and then mine, using 2 chars, then see, unless we think of the same thing

user202729

Wait, P does pop after printing.

It needs to be changed a little but...

You can't get much shorter than that.

#!/bin/python
from math import isqrt
import math
import subprocess
from typing import List, Tuple, Union
import functools
print()
s="abca".encode('u8')
assert int.from_bytes(b"AB", "big")==(65<<8|66)
@functools.lru_cache(maxsize=None)
def value(i: int, j: int)->int: return int.from_bytes(s[i:j], "big")
@functools.lru_cache(maxsize=None)
def work(i: int, j: int)->Tuple[Union[int, float], List[int]]:
	assert i<=j<=len(s)
	v=value(i, j)
	if j==len(s): return (v+1, [v])
	best: Tuple[Union[int, float], List[int]]=(math.inf, [])

user41805

@user202729 how is this run? py2 gives invalid syntax at the i: int

Neil

12:48

CMC: Given a list of finite floating-point numbers and sentinel values, return the average of the finite numbers, but return the same sentinel value if there are no finite numbers. In SQL this is just the AVG function, where the sentinel value is NULL.

user202729

13:12

Correction: 122.47315051766873 TiB

@user41805 There, I posted mine.

With only z and P I doubt it's possible to do anything else.

New Sandboxed Posts

13:40

0

A: Sandbox for Proposed Challenges

Hello World! - Generation 3392 Your job is to create a genetic algorithm that evolves and mutates its string until the output is equivalent to "Hello World!". The genetic algorithm should start with a random string generated, and continue for generations, making random mutations to its code, like...

user202729

14:46

@NewSandboxedPosts Someone help me to explain to the op please.

Any final comment for codegolf.meta.stackexchange.com/questions/2140/… ?

New Main Posts

15:35

0

Q: BF memory layout optimizer

Given a BF program consisting of only the characters +-[]<>., with the property that there's an equal number of < and > between every matching pair of []. Then, we can tell exactly where the memory pointer is for each instruction executed. The memory can be rearranged so the program is shorter, w...

code-golf brainfuck

user202729

16:11

We should just not use the sandbox anymore and tell people to just live with the down votes on main.

Razetime

@user202729 why?

caird coinheringaahing

@user202729 The sandbox isn't perfect, but it's better than not having it

user202729

Currently its sole purpose is to avoid people initially downvoting/closevoting the challenge on main, no?

caird coinheringaahing

@user202729 Yes, but indirectly. It's purpose is to provide a place where people can point out issues with challenge ideas (and allow the author to fix those issues) before posting to main

user202729

Hm. So the important point is that people must not be able to answer before it's reviewed, right?

caird coinheringaahing

16:16

For your latest challenge (which is what I suspect caused this), I get what Jonathan Allan is saying. The formal definition is (like most "formal definitions" out there) difficult to understand for the lay person. I personally disagree with Jonathan that every challenge must be easily accessible, but the challenge could be better if someone who did understand it could phrase it more accessibly

user202729

The informal part is pretty readable if you already know what BF is though.

we don't get to design the site anyway, but

caird coinheringaahing

@user202729 No, the important point is that people can fix holes, make things clear and get basic feedback before a challenge could be ruined on main

user202729

@cairdcoinheringaahing That's the same as what I'm saying, no?

caird coinheringaahing

@user202729 Similar. Some challenges are posted to the sandbox with no issues and immediately posted over to main. The sandbox allows for fixes, it doesn't require fixes

user202729

"no issue" is a type of review.

About the challenge... I could describe what BF is, but I think that the effort required to understand that is more than that to understand the formal description (and to actually understand the challenge you need to formalize it anyway)
So what's the point?

caird coinheringaahing

16:21

@user202729 Agreed. I think the main issue with the challenge is that the informal bit simply goes "The memory can be rearranged so the program is shorter. Write a program to do this", which is far too undersimplified for the site, meaning that if you don't understand the formal definition, the challenge is unclear

Unfortunately, while I understand what the challenge is, I don't understand the formal definition enough to help you write it in a more accessible way

It's a weird blend of being both too unclear and not unclear at all, depending on how much you know about the topic

user202729

16:35

@cairdcoinheringaahing Is the informal one unclear? What's the issue?

caird coinheringaahing

@user202729 It's not so much "unclear" but more "underspecified" in that it's missing the standard rules that go along with most code-golf challenges (I/O, "shortest code wins", etc.). One key thing that sticks out to me is that, as written, there's no need to manipulate the input, so long as the output's length is minimised and the memory the same. That may be correct, but it doesn't say that anywhere in the challenge

A worked explanation is also never a bad thing. The fact that these are minor things that are easily corrected is why I haven't cast a VTC

user202729

@cairdcoinheringaahing Manipulate? The input? What?

caird coinheringaahing

@user202729 By the input, I mean the inputted bf program. "Manipulate" as in the program doesn't have to be related to the output aside from the memory

The swapped < to > in the second example is what's prompting this. It seems as though the output can simply change arbitrary things in the input, so long as the memory is the same

seshoumara

16:54

@user202729 your dc answer can't run on any architecture, because of the RAM first of all, then the limitation to the maximum file size. Still, I liked your answer.

user202729

@seshoumara Can it? There's swap and NTFS/ext4.

I think 64-bit systems can handle it just fine?

seshoumara

Well I googled and on Windows you can't have files larger than 16 Tb I think it was.

I guess you can split your code into multiple such 16 Tb files, but then the OS needs to give to dc probably 122*2 Tb (source and stack). Surely no current PC can do that.

user202729

> The ext4 filesystem can support volumes with sizes up to 1 exbibyte (EiB) and single files with sizes up to 16 tebibytes (TiB) with the standard 4 KiB block size.

Although I guess the swap size is much smaller than that.

Besides just make the block size larger.

Redwolf Programs

Can't NTFS handle files up to an EiB? Or is that also volume size

user202729

@cairdcoinheringaahing Now it seems like that you don't bother to read the formal description and nitpicking me. -- Anyway, I added a FAQ.

(don't take offense...)

@seshoumara Finding the shortest 3-distinct-character program is more interesting.

caird coinheringaahing

17:02

@user202729 Don't worry, I know how frustrating it can be if you don't see what's unclear in a question that people are insisting is unclear. Plus, I'm a difficult person to offend :)

seshoumara

@user202729 I agree. I already found a shorter 3 char version with 1ZP. My intuition tells me there are other 3 char dc solutions out there, much shorter than 3.6Kb.

Razetime

Try it online! Any mathematica experts know how to make this work?

user41805

17:48

@user202729 agreed too, even though the bounty will likely go to yours, i'm interested in seeing the shortest dc program with 3 unique chars

@seshoumara "your bounty brought me back to make new bash, sed and dc answers", i'm not sure if i told you already, but here's 400 rep for a sed fizzbuzz solution shorter than mine (currently 129 bytes) codegolf.meta.stackexchange.com/a/18428 . my rationale is mitchs's 99b sed (no extra flags, so no -r) solution on anagol golf.shinh.org/p.rb?FizzBuzz#sed

yeah i'm really awful at arithmetic in sed

seshoumara