The Nineteenth Byte

Sp3000

12:00 AM

Ah, the only voter I see

VTCAKAVSMoACE

ahem Misread.

feersum

@VTCAKAVSMoACE I don't understand.

VTCAKAVSMoACE

I had misread the post. Apologies.

feersum

I'm not sure how complicated the task should be.

I want it to demonstrate programming capability somewhat, but not be too tedious in a very low-level tarpit type of language.

Maybe I'll remove length restrictions for the solution program.

Sp3000

btw is anyone else confused why the sys recursion one is apparently valid? I thought of that but it always errored out immediately for me, and I assumed that wasn't allowed since it never got to the following code...

feersum

12:06 AM

@Sp3000 Seems invalid to me.

Anonymous

It was within the (albeit ill-defined) rules of the contest, as far as I could tell. It literally made it unusable.

Sp3000

I don't see how that's any different from the invalid loop rule, or even just calling exit() on the first line o_O (the fact that jimmy said it was valid confuses me the most)

feersum

Never mind, it actually does seem valid.

I can reach statements after setting the limit.

In Python 2, I can even print something.

It behaved differently in teh REPL.

Sp3000

Ah... I was trying in REPL :/

feersum

I can also do math and print the result. I guess function calls for those are optimized out of the bytecode.

Sp3000

12:13 AM

Hmm weird, mine still errors

Anonymous

It errors in REPL or with -c for me in Python 3, didn't try with 2

Sp3000

Had to command line it for it to work

In 2 that is

feersum

You need to make a program in an actual file.

Sp3000

Well this is weird. I can actually set the recursion limit back to some high number o_O

feersum

yep me too, guess it's invalid then, lol

Sp3000

12:27 AM

It even works on ideone - would you like to post a comment feersum?

Anonymous

I think it works on ideone because it doesn't actually let you do most of the sys stuff

feersum

Why me? You found it.

Anonymous

We're a bit past that point, the question has been closed

Sp3000

k, then - just thought it'd be faster cos phone

You can set recursionlimit on ideone btw

Anonymous

I'd like to see someone try to beat Dennis's CJam answer on my question - 59 bytes doesn't seem like it's the shortest possible answer

El'endia Starman

12:34 AM

@PhiNotPi Oooh...intriguing...

Sp3000

If it's Dennis getting 59 on CJam, Pyth is probably the only choice left

btw anyone tried breaking muddyfish's Python one in a more standard way? I'm trying to build strings but I'm missing a few letters

I have aefilnorstuFLNT0123456789{}[](),:\."'

VTCAKAVSMoACE

In Java, "*" is referring to the literal *, not regex, right?

Anonymous

@VTCAKAVSMoACE ?

Anonymous

@Sp3000 Link for the lazy?

feersum

@Sp3000 Why would it matter what letters you use?

VTCAKAVSMoACE

12:44 AM

So, say I did case "*":, would that accept the regex * or would it be the literal *?

Sp3000

4

A: Make your language unusable

Python 2 import types def function(): #INSERT CODE HERE pass c = function.__code__ code = types.CodeType(0, c.co_nlocals, c.co_stacksize, c.co_flags, c.co_code, (None,)*len(c.co...

undergroundmonorail

Have you guys played Human Resource Machine?

Sp3000

Because I wanted to try exec - seems like I can do a few things there I can't do outside the exec scope

feersum

@Sp3000 I still don't understand. What does it do that would depend on specific characters?

Anonymous

@VTCAKAVSMoACE Literal *

VTCAKAVSMoACE

12:46 AM

Alright, good. Would've been bad otherwise.

PhiNotPi

@El'endiaStarman Nothing really came from it, though.

Sp3000

Almost everything else, eg strings, are replaced with None and builtins are gone

But you can still use some things like keywords and dict/list literals

Anonymous

The builtin len(iterable) is gone, but what about iterable.__len__()?

feersum

You say you have backslash and digits. That's all you need to write all strings.

Sp3000

You can't dot syntax outside exec, but it seems you can inside, which is what i wanted to try

Oh, that's a nice idea...

Anonymous

12:50 AM

And since exec is a statement, you can't break it by disabling builtins

Sp3000

Which is the nice part :)

Anonymous

@feersum str is a builtin, and so would be disabled

Sp3000

Don't need it though :)

feersum

That didn't make sense anyway, lol.

I called str on a string.

Anonymous

Are backticks (shorthand for repr) disabled?

Sp3000

12:53 AM

Nope

That's how I'm getting strings

Anonymous

Then you could do \btlambda n:n\bt for a few more chars

Anonymous

\bt meaning backtick, I don't know how to escape them in code blocks

Sp3000

I'll try feersum's thing in a bit

Anonymous

<function <lambda> at 0xffe195dc>

Anonymous

Which puts you up to abcdefilmnorstuxFLNT0123456789{}[](),:\."<>'

Sp3000

12:59 AM

Lambda doesn't work for some reason. Ditto generators

Anonymous

You should also be able to use the same trick on a nested named function for even more characters

Sp3000

1:12 AM

@feersum The backslashes are literal backslashes, and I can't evaluate them to get octal code points :/

And yeah def/lambda seem to hang things...

undergroundmonorail

why does every programming game make you implement a sort

i know how to do it it's just tedious

feersum

@Sp3000 Can you call member functions of strings?

Sp3000

Only in exec, I think

Also do you reckon file output is okay? That seems to still work despite Maltysen's update

feersum

It was closed as unclear for a reason, lol

Anonymous

The obstacle to overcome is constructing a string from what we have to worth with that does something useful in exec

Sp3000

1:20 AM

True :P

Well let's just keep in mind that input/open seem to still work where print doesn't, in case it ever gets reopened

But yeah all the good ideas I have for exec I'm missing a letter or two

Anonymous

What are you missing?

feersum

Can you assign a variable inside exec?

Sp3000

Yes I can, with [x for x in [...]] (thank god Python 2 is broken)

I have no = or ; though

Anonymous

@feersum If he had a way to get a = he could

feersum

It shouldn't be that hard then, if we can call string functions and then assign strings.

Then exec a second time.

Sp3000

1:23 AM

The problem is that variables I define with exec don't seem to be usable outside, and I don't have an x/c to do exec inside :/

Anonymous

Can you call str.__doc__() on one of your literal strings?

feersum

Oh, I thought you could use the variable outside.

Sp3000

Dot syntax errors out

Anonymous

Oh right, crap

feersum

I should stop being so lazy and set this up myself, wow

Anonymous

1:24 AM

Variables in exec only update outside variables, they can't create new ones

feersum

In the REPL, you can create new vars with exec

Sp3000

I can't seem to update outside variables either :/

feersum

This is a good programming puzzle.

Anonymous

The evil that was done to wipe the builtins may be messing with that

Sp3000

Yeah, I was up all night trying to get letters :P

But the nice thing is I can exec, say, ''.title and it doesn't error. I can't actually tell what happens though because I can't print it, but lack of error is a good sign

feersum

1:27 AM

I don't get why print 3 prints None, but print 3>2 prints True.

Anonymous

Can you import anything inside of the exec?

feersum

Oh, it prints False, derp.

Sp3000

Yeah False makes more sense :P

Can't import, no __import__

Anonymous

Oh wait, you don't have the letters for it

Sp3000

That too :P

Anonymous

1:28 AM

Phooey

Sp3000

The other confusing part is I don't seem to be able to use outside vars from the inside

But building them again shouldn't be too hard

Anonymous

Hmm... I guess builtin classes are wiped, too

Sp3000

If only I had _ - it's basically the one char I really need :P

feersum

I found a good test harness is to give the function an argument and pass in a string. Then I don't have to bother with actually constructing what I want to exec.

Anonymous

Wait, what happens if you try to exec syntactially-invalid code?

Sp3000

1:31 AM

You get the appropriate syntax error

@feersum I should... really do that :P

Anonymous

So it raises SyntaxException

Anonymous

Wrap it in a try/catch

Sp3000

Yeah, but you can't catch it cos Exceptions are gone

feersum

All you need to do is change the first arg of CodeType to 1.

Sp3000

What does that do?

feersum

1:32 AM

Also you need to declare your function with 1 argument.

It's the number of arguments.

Sp3000

Ah

feersum

Good news: lambdas work fine inside exec.

And you have all the chars.

Sp3000

I'm missing m and b I believe

feersum

Oops, accidentally copied Mego's string.

PhiNotPi

Anonymous

1:40 AM

Yeah, that was when I thought lambdas would work

Sp3000

If you guys can come up with ways to get any char not in the list, I'd be interested to know :P

Anonymous

What about repr values of binary/hex int literals?

feersum

Literals are replaced with None.

Sp3000

^^ that

PhiNotPi

my battery life doesn't make any sense

Sp3000

1:41 AM

Don't worry Phi, at least it's honest

Anonymous

asjhksahfkjhf

Anonymous

If we could just import sys, all our problems would go away

feersum

@Sp3000 how do you get curly braces?

Anonymous

...wait, dot syntax doesn't work. sys.something would error out

Anonymous

@feersum dict/set comprehension I believe

Sp3000

1:44 AM

Dict and list literals still work it seems

Tuples don't though

PhiNotPi

I'm trying to figure out how to break Perl. (Un)fortunately there's the CORE:: namespace.

Sp3000

Or at least pure tuples with immutable contents

feersum

([],) works

Anonymous

@PhiNotPi exec 'rm -rf /'

Sp3000

;)

I tried for a long time trying to get a complex number for parens, but Python 2 didn't let me do (-1)**.5 :/ (Python 3 does)

Anonymous

1:46 AM

In Python 2, complex numbers are hidden away in the cmath module

VTCAKAVSMoACE

sigh I need to learn Python...

3

Seriously though. I tire quickly of using Java for interpreting.

feersum

@Sp3000 You don't seem to have - on your list.

Sp3000

Hmm -+ should be on there oops

feersum

How do you make +?

Sp3000

1e+100

feersum

1:49 AM

oh e+

Sp3000

:)

Anonymous

aefilnorstuFLNT0123456789{}[](),:\."'+-

Anonymous

You know, I don't think Brainfuck or Whitespace could be broken in that challenge

feersum

I still don't see how to break the sandbox, but I think we can be Turing-complete at least.

We can make an infinite loop by using for on a list and then extending the list indefinitely.

Sp3000

Yeah we're probably Turing complete and can take "some" form of input (see comments), but I was wondering if builtins was recoverable

Anonymous

2:00 AM

If we could get <> we could build a brainfuck interpreter

Sp3000

We could do CaneCode instead I guess :P

Anonymous

Perfect

Anonymous

The only issue is input/output for the ,. commands

Sp3000

Well that doesn't really affect TC so that's fine

VTCAKAVSMoACE

Hmm... My code is returning only zero for sine and I have no clue why... ;-;

Anonymous

2:04 AM

@VTCAKAVSMoACE sin(x) where x = a*pi, a in ints?

VTCAKAVSMoACE

No... But I figured out what it was...

feersum

I have an idea for how to communicate variables between inside and outside of exec

Make a list outside, and then modify the contents inside.

Sp3000

Oh?

feersum

We should be able to get some more letters from inside of exec somehow.

Not working somehow.

Sp3000

:(

feersum

2:09 AM

It seems exec can't read the variable from outside.

In a normal function it works to go [x for x in [1]];exec'print x' but not in here.

Sp3000

Yeah... damn :/

New Main Posts

0

Q: Anagram Bubble Sorter

The Challenge A bubble sort takes a list of non-repeating integers and sorts them in ascending order, with only the use of one operation: switching two numbers which are next to eachother. However for this puzzle, you must take in two strings, which are both anagrams, and bubble sort the letter...

Anonymous

I think the issue is the fact that everything gets set to None, and None is a value-type, not a reference-type

Sp3000

The frustrating part is that if we just had, say, c/d/v/mb we might have a chance

But we're slightly off in each case

VTCAKAVSMoACE

3:12 AM

Welp. My brain officially hurts.

Holy hell, it's 4AM. >.> Did not see that coming.

El'endia Starman

3:23 AM

LOL.

PhiNotPi

PPCG Level: Literally Monday

Sp3000

3:45 AM

@feersum I can get variables out of exec, and I think I've got more chars!

feersum

@Sp3000 how to got variables?

Sp3000

exec ... in {}

feersum

Very nice.

Sp3000

4:02 AM

Almost there :)

DankMemes

@PhiNotPi 10 minutes into monday

Sp3000

4:22 AM

The Python jail has been broken

6

DankMemes

4:49 AM

holy crap what is that O_o

El'endia Starman

5:01 AM

@DankMemes He rebuilt a few select lines of code and evaluated them. Starting from None as the only value available.

> code = lb+a+a+sp+f+o+r+sp+a+a+sp+i+n+sp+lb+qu+qu+dot+t+i+t+l+e+rb+rb

== code = [aa for aa in ["".title]]

It's insane and incredible.

2

Sp3000

^^ that. Technically I get dicts, sets and lists as well, but tuples, strings, ints, bools get turned into None. Keywords still work though, which is good for exec

I'm trying to see how much cleaner I can get the code, since I probably did more than necessary :P

orlp

I had an idea for a metagolf challenge

where the goal is to minify CSS

DankMemes

Dang, if only that "Make your language unusable" question was a cops n robbers

waste of a really good idea on a closed question

well, "on hold" but effectively closed

Sp3000

I don't think it'll necessarily stay closed

El'endia Starman

@Sp3000 If I understand correctly, after the last line of code before pass, you've executed code that loads built-ins, right? So strings, ints, etc should work after that point?

feersum

5:15 AM

If you have an idea, you could post it with an exec in your language as the "interpreter" for my version, hehe

There has to be a solution though.

Sp3000

I think it works inside the exec, but I'm not sure about outside

It wouldn't be like loading __builtins__ makes ints work after that point though, since the literals are actually replaced before running the function (I think that's how that works?)

El'endia Starman

hmm

But the int function itself ought to work, yes?

feersum

I don't think the builtins thing is really important. What matters is to get the whole alphabet available for use in exec.

Then it's easy to use builtins or do whatever.

Sp3000

^^ that

El'endia Starman

6:37 AM

@feersum import string; string.letters

If you have those letters, you can get all of them.

feersum

I'm not sure whether import is working, but anyway there was no m before the full jailbreak.

Sp3000

Apart from the m thing, I think import needs __import__ from builtins

feersum

6:52 AM

How do I set permissions of a directory so that I don't need sudo/root privileges to write in it?

El'endia Starman

@Sp3000 Don't you have built-ins? You'd have to call it differently, but still...

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Permute into No Double Letters You should write a program or function which receives a string of letters as input and outputs or returns a permuted version of this string where no adjacent letters are the same. If the goal is not possible you should remove the minimum number of characters from ...

Sp3000

@El'endiaStarman I think what feersum means is that by the time you get builtins it doesn't really matter any more, since you can do whatever

import string could be one way to get the remaining chars, chr could be another yeah

New Main Posts

0

Q: Make my pseudocode real

I've got some Java pseudocode that uses whitespace instead of curly braces, and I want you to convert it. I/O Your program should take an input file along with a number designating how many spaces are used to indent a block. Here's an example: $ convert.lang input.p 4 // Convert using 4 space...

code-golf string

El'endia Starman

7:17 AM

@NewMainPosts ...so...the task is to take Python and turn it into Java... :P

2012rcampion

@orlp Here

orlp

hey

did you just think of the 42 by chance?

I just chose it as a 'fun' constant

2012rcampion

I originally had it as 2, but the performance on repeats was not too good

Since you get 2 => 4+ 1 => 5

So if you have to repeat three times, one is a duplicate of a previous run

orlp

also, did you read my comment in my linked code?

it's a pretty cool observation that I haven't seen elsewhere

2012rcampion

About the modular multiplication?

orlp

7:22 AM

yes, that you never have to convert to montgomery form

2012rcampion

Yeah, I read that in one other place, I think it might have been Hacker's Delight

orlp

fairly sure it's not in hacker's delight

but it might be an one-off note

2012rcampion

I remember that (wherever it was) it was mentioned offhand while they were talking about operations on Montgomery forms (addition and subtraction)

orlp

I've also tinkered a lot with the RNG for the pollard-brent

in the end I just settled for +1

because I didn't find measurable improvements even with CSPRNGs

2012rcampion

Yeah, I found that even on the largest numbers I only got ~1% failures, so anything affecting the repeats has a tiny effect on performance

Since none of the test cases failed, I didn't have to worry about it =)

orlp

7:26 AM

my is_prime code is different than yours though

I use strong (lucas) probable prime tests

@2012rcampion you could save a handful of cycles in the trial division btw

github.com/orlp/libop/blob/master/bits/math.h#L550

2012rcampion

I saw that that one depended on knowing the factorization of n-1 and passed over it

orlp

hrm?

not that I'm aware of

it's a general method

I described it here: math.stackexchange.com/questions/1251327/…

2012rcampion

Ah, on Wikipedia it says: "In computational number theory, the Lucas test is a primality test for a natural number n; it requires that the prime factors of n − 1 be already known."

orlp

@2012rcampion en.wikipedia.org/wiki/Baillie%E2%80%93PSW_primality_test

"The Baillie-PSW test is a combination of a strong Fermat probable prime test to base 2 and a strong Lucas probable prime test."

that's what I do

it has no known counterexamples

2012rcampion

I'll have to give that one a look

orlp

7:32 AM

the divisibility trick is mentioned in hacker's delight btw, but I don't think it contains a general method as I linked above

2012rcampion

Yeah, it looks like the fast division would save me around 250 cycles on average.

orlp

@2012rcampion in the context of the challenge, catpiss

but when using as a general method, the vast majority of numbers will be composite, and the vast majority of numbers will be small

2012rcampion

Very true

orlp

just a small trial 'division' table of 168 elements will completely factorize any number smaller than 1 mil

2012rcampion

If I were planning to actually use this code I would definitely work on optimizations like that, but as it stands I'm ahead of the second-place by something like 100 times, so I think I've leave the code as-is for now

orlp

7:39 AM

@2012rcampion yeah, this is more out of interest than necessity :)

either way, I enjoy your answer, and I hope you enjoyed the GOLF (and didn't curse it every second while enduring it)

2012rcampion

I actually quite like GOLF, partly because it's a nice change-of-pace from my usual, but also because it's pretty different from most architectures I know of

10 cycle division but no add-with-carry means you make a lot of interesting choices

I'm planning on writing my own assembler that will let me do macros and such, but I have to learn how to write an assembler first =)

orlp

@2012rcampion it's an architecture I designed to teach the basics and ballpark estimate costs of CPU instruction sets without complicating the architecture with tons of instructions or the mess that is memory (caching, load ordering, threading, etc)

it's a mix of realism and simplicity - not going to either extreme

Fatalize

Just got my ticket for Star Wars 7, WOOOOO

2012rcampion

I would say it goes towards the extreme end of simplicity... I really miss that status register

Jacoby Yarrow

hi guys

Fatalize

7:44 AM

and we get to see it 2 days before muricans here in France, suck it

orlp

@2012rcampion it definitely leans more on simplicity, yes

@2012rcampion the main thing that sets it apart from other architectures from what I'm aware is the call/return mechanic

2012rcampion

What I got from the spec was that there's no access to the call stack?

orlp

@2012rcampion not entirely

you are free to maintain your own call stack

it's just that the CPU itself has an internal stack to maintain the stack of register files

@2012rcampion the z register holds the ztack pointer, you should use that if you want to store stuff on the stack

2012rcampion

Yeah, I saw that as the intended use

You don't have access to the return address though either.

orlp

nope

Jacoby Yarrow

7:51 AM

codeshare.io/T7ZcF

go there

type something

2012rcampion

I kind of missed that, since there's one point where I use call to access a subroutine, but then jmp out of it, meaning that if I call ret again I pop back to the original call location

call/ret is super-convenient as a way to store data though

orlp

@2012rcampion yeah, jmp wasn't really intended to be used across basic blocks

or well

it doesn't mix/match well with call/ret

2012rcampion

Being able to access to 25 quadwords in one cycle is a nice feature, and I didn't want to give that up to have nicer code flow

orlp

@2012rcampion but if you are interested in interesting architectures you should definitely take a look at the Mill CPU

the talks are long, but man are they worth it: millcomputing.com/docs

(except maybe #1, unless you are interested in instruction encoding/bandwidth stuff)

2012rcampion

Looks pretty interesting, I'll take a look later

Right now it's about 3am here and I need to get up for class tomorrow

So I'm signing off

Thanks for the challenge =)

Sp3000

10:08 AM

Updated jail code to look a lot neater if anyone's interested in an explanation

New Main Posts

11:40 AM

0

Q: Array Output Problems

I need the user input to be saved into my array and then output the array before the user inputs the next time. I have moving things around different ways but cannot seem to get them to perform properly. I tried to cut down the code to the two functions I am having issues with. void PlayGame() ...

c++

2

Q: Program the Cup-Stacking Robot

I'm sure everyone has seen before that cups can be stacked into pyramids (and other shapes): A A A A A A A A A A A A A A A A A Yes, A is definitely an adequate character to represent a cup. New cups could be added either on the ground, to the right of the stru...

metagolf hexagonal-grid

randomra

12:12 PM

@MartinBüttner Have you seen this 2-operand cup-stack building somewhere or is it your original idea? It is nice, that you can make any legal configuration.

Martin Büttner

@randomra I don't know if this has been studied before. I came up with it when trying to design an esolang where the memory model works like such a stack of cups.

in Esoteric Programming Languages, Oct 15 at 7:33, by Martin Büttner

@Sp3000 what's the simplest set of operations that can build any shape of cup pyramid? I've been considering "add a cup on the ground to the right", "stack a cup on the left-most adjacent pair of cups" and "remove the left most cup with no other cup on top", but I don't think it's sufficient

in Esoteric Programming Languages, Oct 15 at 7:41, by Martin Büttner

Maybe, if I change "left-most" to "first in reading order"

randomra

nice, I immediately tried to imagine an esolang with only these 2 operands

and now I learned about the Esolangs room :)

Sp3000

12:42 PM

Issue #25274: sys.setrecursionlimit() now raises a RecursionError if the new
recursion limit is too low depending at the current recursion depth.(https://hg.python.org/cpython/file/default/Misc/NEWS). Now we have an excuse to not upvote if another one of these gets posted for a popcon :P

Anonymous

@Sp3000 You are either a Python god or a raving lunatic. Nice job on cracking it

Sp3000

Probably the latter :P

(but yeah, if they repost in Python 3 then we're pretty much screwed)

New Main Posts

1

Q: *Overwrit*labels

If you've ever tried adding labels to a really dense plot, then you'll realise that sometimes labels will overlap one another, making them hard to read. We're going to do something similar but in 1D. Input will be a sequence of (label, x-coordinate) pairs, and output will be the result of drawin...

code-golf string

Anonymous

Well, first the question will have to be reopened

Anonymous

I wonder just how many bugs on interpreter issue trackers are from PCCG users finding extreme edge cases in languages

Sp3000

12:54 PM

Ahaha maybe a few, but I don't think that many :P

It's probably more likely that someone on PPCG saw something in an issue tracker then used it in an answer

Martin Büttner

:( I was going to solve this in (.NET-flavoured) regex, but I think I found a bug in Regex.Replace which I can't work around... do I have two problems now? — Martin Büttner ♦ Apr 20 '14 at 12:18

Anonymous

Lol Martin

Martin Büttner

I actually reported that, but I can't seem to find the bug report any more...

Rainbolt

@Calvin'sHobbies Do you have proof? All I can find is that they are "related" and share the same grandmother. They could be brothers.

Anonymous

1:10 PM

lotr.wikia.com/wiki/D%C3%A9agol

Anonymous

They're cousins

Martin Büttner

@Dennis I've got 23 for Overwrit.

Rainbolt

@Mego Refresh the page. It clearly says brothers.

Martin Büttner

I started with the same approach as yours (but I used null bytes instead of empty arrays). It turns out to be a byte shorter if you don't use .e| but do it manually.

@Rainbolt You go really far to prove your point. ;)

Rainbolt

I was really just disproving his

Anonymous

1:12 PM

Tricksy hobbitses

Sp3000

@Calvin'sHobbies Maybe something like *zT and Al*? shrugs

Anonymous

fooooood

Geobits

Hmm. Even for Java, 242 seems pretty bad for Overwrit. Must rethink this.

Martin Büttner

@Dennis If you want to include this approach in your post as well (or if you're able to golf it to 22: Q~{~S*'*+\+}%:.{_S&@@?}

Using your input format, that's Lr{iS*'*r++.{_S&@@?}r}h (still 23)

Dennis

1:44 PM

@MartinBüttner Thanks! If only CJam had a stack-based ternary (<select if true><select if false><Boolean>)...

New Main Posts

0

Q: Check if a given 8-bit Number is sparse

The task is the following: Program a fully functional program that checks, if an 8-bit value contains a sequence of 2 or more ones in its binary form. If this is the case, the output shall be 1, if not: 1. Examples: Input: 1 Output: 1 Input: 3 Output: 0 Input: 72 Output: 1 Input: 73 Output...

number bitwise

Martin Büttner

@Dennis yeah :/

Dennis

I've been thinking about your cup code. If I understand it correctly, it stores the the state as lines. I've had two ideas so far, but no time to implement them.

Martin Büttner

duplicate hunt, go: codegolf.stackexchange.com/q/61157/8478

Dennis

(Not sure how good they are. Fastest code isn't really my thing...)

Martin Büttner

1:46 PM

@Dennis yes it does

TreFox

codegolf.stackexchange.com/questions/47870/… maybe?

Sp3000

Man, I left two comments and missed the fact that it had no winning criteria

Martin Büttner

@Sp3000 haha, I didn't notice either of the things you commented on

Geobits

:(

class S{public static void main(String[]a){System.out.print(Integer.valueOf(a[0]).toString(2).contains("11")?0:1);}}

Assuming the 73 test case is wrong, of course ;)

New Meta Posts

0

Q: Does PPCG fulfil a role as a catalogue for golfed solution of standard programming exercises?

Rosetta Code collects solutions to standard programming exercises in hundreds of programming languages. However, it focuses on "good" or idiomatic solutions in those languages. I'm not aware of a similar catalogue from a golfing perspective. Is PPCG capable of fulfilling that role? Do we have cha...

discussion code-golf

Dennis

1:52 PM

@MartinBüttner One is to store the state in a 1-dimensional reading order array, using three separate symbols for cups, empty slots and available cup positions. Each command now boils down to a #, a t and two conditional ts to change available positions. Indexing and modifying large arrays seems very fast in the Java interpreter, but I'll have to try online.

Martin Büttner

ah

22

Q: Find the next 1-sparse binary number

A positive integer N is K-sparse if there are at least K 0s between any two consecutive 1s in its binary representation. So, the number 1010101 is 1-sparse whereas 101101 is not. Your task is to find the next 1-sparse number for the given input number. For example, if the input is 12 (0b1100) o...

code-golf arithmetic

Sp3000

That answers that then?

Dennis

@MartinBüttner The other is to store the state as two arrays of coordinates: one of cups, another one of available locations. It might require less memory, but searching for the first cup/slot would be more expensive.

Rainbolt

If enough people say welcome to PPCG maybe he won't be sore about his closed question

Lol Geobits I was just doing that

Geobits

:D

Transcript for