The Nineteenth Byte

4:03 PM

@MartinBüttner Well I guess I will. There are so many variants that it probably makes sense anyway.

Oh, I started to write my explanation and realized need a wrapping/left-infinite tape. That probably explains why it doesn't work with BrainSteel's interpreter.

Lembik

hi all

Κριτικσι Λίθος

@Lembik Hi

Lembik

any low level coders able to tell me if checking for NaN is the same speed as checking equality of floats?

Optimizer

4:18 PM

I assume NaN should be faster. Actually, I call dibs on it.

Manage your digital lifestyle with Windows 95. http://t.co/WvxaaM8LRS

Tweeted by Windows95Tips on October 12, 2015 at 2:04 PM

Κριτικσι Λίθος

?

Lembik

@Optimizer call dibs?

Optimizer

¯\_(ツ)_/¯

Martin Büttner

@Lembik I can tell you that NaN doesn't have a unique binary representation. Also, the bits that are the same for all NaNs are also the same for infinities, so I wouldn't be surprised if it takes more processor instructions to check for NaN. (IANALowlevelcoder though)

Lembik

@MartinBüttner hmm.. thanks. I can't think of a sensible way to benchmark it sadly

Optimizer

4:22 PM

why not ?

Martin Büttner

compile some code and disassemble it?

Lembik

@MartinBüttner that would work ..

@Optimizer because of -O3 just removing my code

Optimizer

@MartinBüttner but float equality means comparing all bits, NaN for a float variable will be faster if you start checking bit by bit

anOKsquirrel

@Mego how did you get extended ASCII characters to work in Seriously? I can't get it to work :/

Martin Büttner

@Optimizer what processor are you using that checks for equality bit by bit...

Optimizer

4:24 PM

@MartinBüttner that was just a guess. its not bit by bit ?

Martin Büttner

if I remember my hardware lecture correctly, processors operate on entire words at a time (that is, 32 or 64 bits these days)

Optimizer

shouldn't 2 == 2 check that both LHS's and RHS's bits are 00000010 (for 1 byte data type)

Martin Büttner

yes, but the hardware checks all of those bits at the same time.

Optimizer

so in 1 cpu cycle, they can compare more than 1 bit ?

oh

well, if we are really talking low level, nothing it really in parallel.. right?

Martin Büttner

at the very lowest level, I'm pretty sure it is.

Optimizer

4:27 PM

also, even if we consider all bits are checked in parallel, then it might simply mean that for NaN, hardware has to check lesser number of bits

so eventually effecting throughput.

Martin Büttner

no, because the bits that identify a NaN would also match in an Infinity. you have to check that one chunk of bits is 1 and the other chunk is anything except all zeroes.

I don't know if any single processor instruction can do that.

Optimizer

this representation you are talking about is processor dependent or language?

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

The Language Relay! A typical relay race only has four competitors per team, but where's the fun in that? Instead, let's see how many teammates you can cram onto the bus to the stadium and still finish the race. (This analogy is starting to break down, so I'll cut to the chase.) Your task is to...

Dennis

@Optimizer Float comparison is more than just checking if the representation is the same. For example, positive and negative zero are equal.

Martin Büttner

@Optimizer neither. it's the IEEE 754 standard for floating point numbers

Optimizer

4:30 PM

@Dennis umm, wouldn't that already be covered if you compare all bits ? so all 32 bits for 4 byte float data type

Dennis

No, because positive and negative zero are two different floats.

quartata

0

A: Implement a Truth-Machine

SPLEMIT21 1 Byte T SPLEMIT21 (short for Shortest Programming Language Ever Made In The 21st [Century])

.

Optimizer

--

Dennis

ideone.com/GC0MXa

quartata

@Dennis Floating point amirite

Dennis

4:34 PM

Argh, I'm trying to explain my BF answer, but I don't remember how it works...

quartata

@Dennis Which answer?

Dennis

58 mins ago, by Dennis

0

A: Sign that word!

Brainfuck, 40 bytes ,[>>+>>,]<<[[<<]>>[-[<]>>[.<<->]>+>>]<<] You can try the code on brainfuck.tk.

quartata

@Dennis whoa

Fixed a typo by the way.

Dennis

Thanks.

quartata

Showering in upvotes

That's shorter than PHP

wow

Optimizer

4:38 PM

21 hours ago, by Dennis

okay seriously stop abusing upvotes

TimmyD

@quartata Why would you do that? Most normal people use soap and water ...

Dennis

Hahaha.

Cᴏɴᴏʀ O'Bʀɪᴇɴ

I spent way to long on this:

0

A: Find the Interweaved Source Codes (Cops)

Vitsy, 4 programs, 228 bytes; Score 17/144 (if safe/cracked) Heeeyyyy @VoteToClose. I don't expect this one to be up long, but you never know… I love your language, now that I'm using it a lot! Interweaved program 1E2P-0P1E^l*312359121E4/13+1*-205+-5Pv*92t\9\[931[2-\D+\*12/]1N5*[t22*^50\+[\72C...

quartata

@TimmyD Just like how people like showering in money.

It's a status symbol thing I suppose.

Cᴏɴᴏʀ O'Bʀɪᴇɴ

I'm still waiting for the day when Code Golfing etc. becomes profitable ^_^

quartata

4:43 PM

1 rep = 1 dollar

TimmyD

Dennis

@quartata Yes, please.

quartata

SERIOUSLY?! _____ — Alec Teal Aug 20 at 9:36

44

A: Sign that word!

GolfScript, 1 byte $ Yes, only 1 byte. Try it here.

@Mego Someone from the past wants you to beat this

GamrCorps

@quartata I propose a different system: 1 rep = 1 dollar, 2 minutes coding = 1 dollar, 1 minute wondering how Dennis got his code so short = 5 dollars.

4

Anonymous

@anOKsquirrel Online interpreter or local interpreter?

Anonymous

4:54 PM

@quartata Unfortunately that's 2 bytes in Seriously

Anonymous

It seriously needs implicit input

Dennis

@GamrCorps Hey, today I'd get money for that.

Anonymous

@GamrCorps I'M RICH! I spend hours baffled by Dennis!

GamrCorps

@Mego lol same

Anonymous

Wait a second, we're getting $30/hr for coding?

GamrCorps

4:56 PM

we SHOULD but we dont

Lets start a Coder's Union

Equality for all coders

Anonymous

If we spend that hour simultaneously coding and being baffled, do we get $150/hr?

SuperJedi224

Hi guys

GamrCorps

Yes

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Scoring an Asynchronous Prisoner's Dilemma Game In a round of the prisoner's dilemma exercise, two players each decide whether to cooperate or defect for that round. The scoring for a round is: Player A and Player B both cooperate: 1 point for both Player A and Player B both defect: 2 points f...

Cᴏɴᴏʀ O'Bʀɪᴇɴ

@SuperJedi224 hiya

Anonymous

4:59 PM

I need to shave off 4 bytes from the Seriously math grader so I can out-golf Dennis

GamrCorps

strawpoll.me/6098815

Anonymous

Lol you voted for everything

GamrCorps

yep

flawr

Any mathematica guys here?

Anonymous

Woohoo dropped a byte

Anonymous

5:05 PM

Just 3 more needed to beat Dennis

TimmyD

@Mego Dennis golfs 10 bytes

Anonymous

noooooooo

Anonymous

2**5-2 is a much shorter way of spelling 30 :P

Anonymous

(by much shorter I mean 1 byte)

Cᴏɴᴏʀ O'Bʀɪᴇɴ

peace out

Anonymous

5:10 PM

girl scout

Anonymous

TFW you think you discovered a clever way to golf your code more but it comes out to be the same length

TimmyD

Worse, when it comes out one byte longer

Anonymous

My code for error detection is 43 out of the 62 bytes :(

Anonymous

Err, invalid input detection

VoteToClose

Vote to reopen?:

0

Q: Make the PPCG Logo

Come November 1, Programming Puzzles and Code Golf will graduate, so in the next 11 months, we'll want to save some memories from when we were ungraduated. Write a program that produces the PPCG logo, seen here. You can use whatever methods you want, from an image to ASCII or anything else. ...

Vote to reopen?:

0

Q: Make the PPCG Logo

Come November 1, Programming Puzzles and Code Golf will graduate, so in the next 11 months, we'll want to save some memories from when we were ungraduated. Write a program that produces the PPCG logo, seen here. You can use whatever methods you want, from an image to ASCII or anything else. ...

popularity-contest graphical-output

Stefnotch

5:26 PM

@VoteToClose But the rules are rather vague...

Anonymous

No

2

Anonymous

Nothing has been changed since it was closed

Transcript for