The Nineteenth Byte

« first day (3483 days earlier) ← previous day next day → last day (1343 days later) »

2:18 AM

Find the Smallest number with exactly n divisors code-golf math Task Write a program/function that when given a number \$n\$ as input outputs the smallest number with exactly \$n\$ divisors. A reference implementation can be found here here. Scoring This is code-golf so shortest bytes wins. Sampl...

HighlyRadioactive

3:04 AM

@Szewczyk Szewczyk's daily routine

HighlyRadioactive

3:21 AM

This polyglot outputs 2 in 1+, 3 in Python 2, 4 in Foo, 5 in COW, 6 in Starry, and 7 in Prelude:

print 2|1##11+"4":[MoO MoO MoO MoO MoO OOM .9-999993 + + + + +!

(In case you wonder why 1 is missing it's supposed to be a hypothetical language based on "recursive definitions" and can only output 0 and 1)

(For example a parenthesis check program is **\n(*)\n[*]\n{*})

(And the polyglot given here just matches the program here since there are no * which means "any matching string")

(But it implicitly matches the empty string too so it outputs 1)

Hi Razetime

@Razetime

I can't stop Starry from erroring out without making the polyglot long however sadly

Razetime

hey radioactive

HighlyRadioactive

~~Apparently Parcly Taxel thinks Radioactive and Razetime are similar because both of them begins with "Ra".~~

Guess the source!

Razetime

I have no clue

HighlyRadioactive

Of course

A: Tips for golfing in 1+

Use comments for conditionals Found this while exploiting interpreter features. Apparently Parcly Taxel thinks comments and conditionals are similar because both of them begins with "co". The comment indicators [] are actually instructions because there's no such thing called parsers in 1+. This ...

I commented "Apparently Parcly Taxel thinks comments and conditionals are similar because both of them begins with 'co'." When I discovered that you can skip the comment character with gotos in 1+

Razetime

interesting

HighlyRadioactive

3:37 AM

Fix polyglot:

print 2|1##11+"4":[MoO MoO MoO MoO MoO OOM .9-9999991 + + + + + +!

@Razetime not that interesting

but it makes the painful conditionals in 1+ a tiny little bit easier

Razetime

so you need to fix the starry code to make it work

HighlyRadioactive

@Razetime what

Razetime

nevemind

HighlyRadioactive

The thing I want to point out is, even though it might not be the shortest, the executed "useful" commands overlaps a lot

For example the 1 in the 2|1 that gives 3 in Python makes 1+ jump over the #

The 11+ in 1+ and 4 in Foo is used in the Prelude code

2 hours later…

Lyxal

5:27 AM

Today I remembered a period of time when I tried to write challenges no one could answer. Heh. Example

I said to myself "for sure that can't be done in non-esolangs"

but of course, it was done in Python/C/R/Javascript/etc

I learnt that we sure are a determined bunch around here.

1 hour later…

Razetime

6:34 AM

many people sure are

those answers are crazy good

3 hours later…

Szewczyk

9:15 AM

@HighlyRadioactive lol

HighlyRadioactive

9:55 AM

@Szewczyk You crack languages!

So when I post my new very trivial 1+ derivative

under the "appears to be unusable" challenge

be sure to take a look at it!

Szewczyk

sure thing

HighlyRadioactive

It was created several months ago (privately) but I'm struggling to solve it

user202729

10:46 AM

@Neil Golfing languages could make a built-in for indirect self-reference as well, and it's just as short.

ATaco

Languages like Javascript's quine often use a mix of self-reference and self-duplicating data. One cannot submit the quine function f(){return `${f}`}, As that is direct reading of source. However, f=()=>`f=${f}` must also attach data.

One could also theoretically submit function f(){`${f}//`}// as the comment would be included in the source, but not produced by ${f}.

My RProgN2 quine utilizes this particular method by appending skipped whitespace manually to the function's definition.

"«` . " (Qoutes added for readability)

John Dvorak

11:06 AM

Methinks the rules for allowable quines are way too strict

ATaco

11:59 AM

The intention is to make quines require more of a restriction in challenges based around them.

Razetime

I'm trying to solve codegolf.stackexchange.com/questions/207404/the-cantor-function/…

in ruby

but my function returns nil for many cases tio.run/…

HighlyRadioactive

12:15 PM

Esowiki says:

Since April 2005, 2,947 articles have been created by 76,002 edits, including 2,258 esoteric languages. There are 6,699 registered users, but most of them are spambots.

Nice

2 hours later…

user202729

2:19 PM

@NewMainPosts It's entirely possible that this number is generated with Python built-in random module with some small seed value.

my pronoun is monicareinstate

I have already brute forced c++'s mt19937 and mt19937_64 with UNIX timestamp seeds for a few months back, didn't find anything (I assumed the digits were obtained by rng() % 3) (I think this sentence turned out to be very confusing)

user202729

Some languages used by OP in last answers: Trigger, Octave, Python 3.

@mypronounismonicareinstate I think C++'s mt19937 is not initialized with the timestamp by default.

I don't know how Python 3 (or the 2 other languages) seeds the RNG, but that may be worth trying.

my pronoun is monicareinstate

I researched that a while back, and it involves this (quoting: "suitable for cryptographic use").

@user202729 I like this approach!

user202729

Is that used by default in Python?

Okay, it is.

Perhaps some Gaussian elimination can be used. It's just mt19937.

So there are 19937 bits of state...

While that number is 2816 bits long.

Isn't worth it, unless the seed is short.

my pronoun is monicareinstate

let me just google how is Gaussian elimination used in inverting PRNGs...

user202729

2:31 PM

Generating short seed from the initial state is known to be hard.

It also depends on how Python uses urandom (if it takes more than ~2816 bits then there's no point).

And whether urandom is really safe on OP's machine.

>.< where did you read that by default Python random module seeds from os.urandom?

It seeds from the current time.

github.com/python/cpython/blob/3.8/Lib/random.py#L780

my pronoun is monicareinstate

Octave seems to use either /dev/urandom, or, if not available, combines the current time, clock() and the fractional part of the current second, so I guess it isn't really feasible to brute force. (sending this message so that I can start typing the next one)

I read that somewhere on stackoverflow.com, and this link seems to confirm that (the way I read it, if nothing is provided as the seed, it invokes random_seed_urandom and only if it fails it falls back to something bad)

user202729

Guess that comment is wrong, then.

my pronoun is monicareinstate

Might be simply outdated

user202729

This requires digging incredibly deep into stuff's source code.

... messages can be easily lost if the tab is closed before they are sent.

(the whole 624 dwords are obtained from urandom.)

New Main Posts

2:49 PM

-2

Q: I was writing a c program it shows error like process.h can't be complied and more.someone can help me

#include<stdio.h> #include<process.h> #include<stdlib.h> struct node { int info; struct node *link; }; struct node *start=NULL; void create(); void traverse(); int main() { int choice; create(); while(1) { printf("\n\nMENU\n\n"); printf("\n1.Traverse"); printf("\n2.Insert At Beg"); printf("\n3.In...

user202729

@NewMainPosts Unrelated: op already tried to post on stack overflow. stackoverflow.com/a/63378545

I hope there's an easier way to delete such questions.

Also perhaps it's better if the "You may be able to get help on Stack Overflow." part in the close reason is removed.

1 hour later…

Razetime

4:17 PM

codegolf.stackexchange.com/a/209476/80214 @HighlyRadioactive

Try doing this in 1+

it might actually have similar length lol

1 hour later…

Szewczyk

5:36 PM

https://cdn.discordapp.com/attachments/733242619567800383/743142414352515142/unknown.png
A small update on the lisp interpreter, I got it to execute a simple factorial program

suggestions on what to run using it are more than welcome lol

also: don't use MT19937 to generate safe random numbers ;)

« first day (3483 days earlier) ← previous day next day → last day (1343 days later) »

all rooms