The Nineteenth Byte

... nvm

What the heck

Anonymous

@ETHproductions Looks like it to me

I'm missing some revs all of a sudden

dammit

Anonymous

I'm thinking of being lazy for the : delimiter

3:21 AM

@Mego It seems to me like "output the Nth number that satisfies rules A, B, and C" vs. "output N run through a formula given by rules A, b and G"

Which could be a dupe

Anonymous

Find the longest string of characters that will successfully eval to a complex, float, or int, or 0 if a string of length > 0 cannot be found

Anonymous

Seems like a reasonable strategy for this

OK, I'm not sure how this happened

SuperJedi224

Goodbye guys

But I guess I'll just revert revert and then implement the one missing feature

Anonymous

3:22 AM

@ETHproductions It's close enough that it'd probably get closed as a dupe

Yeah, darn :(

I did a thorough search for dupes before writing it, but only a few minutes ago found this other one by searching test-cases

Anonymous

@phase I'm looking at it, but not comprehending

Anonymous

I'll figure out a web ide once I get the interpreter working :P

Oh well, back to the drawing board...

@ETHproductions Probably a bit iffy imo - I wouldn't close it as dupe because the challenge itself (order) is different and there's wraparound, but I dunno

3:24 AM

@Mego That's exactly what I did for Minkolang.

I should really call these Tuesday Mini-Golfs :P

@phase I believe I have fixed everything.

Can you do me a favor and run FizzBuzz in it just to make sure I didn't fuck everything up

Anonymous

@El'endiaStarman It's probably the least elegant way of doing it, but it'll produce results

Alrighty then, goodnight everyone

Anonymous

Night ETH

3:26 AM

@Calvin'sHobbies The sad thing is that my Changeling interpreter has a bug that prevents me from using it for this challenge.

Good night, ETH!

@Mego: I have (let's say) a and b. They start off as a=0 and b=len(<input>). Then on every failure to convert to complex, float, or int (in that order), I decrement b. If b gets down to a, then I increment a and reset b to len(<input>) and repeat. If nothing in the input is a valid number, I push -1.

@Mego El'endia forgot to mention the dropping invalid chars part btw, so ab3cd evals to 3

Oh yeah, that.

(with cd still waiting on input buffer, I'm assuming)

yup

So nn on ab3cd will give you [3,-1] on the stack.

3:27 AM

Ah, that's better.

Also, I wonder who Hipe99 is on PPCG...

Ah good the FizzBuzz program works.

So everything is ok

btw @El'endiaStarman have you tried Fibonacci word with Minko? I'm having a lot of trouble with that one

not yet

I'll take a look.

I wish there was a Skype that wasn't owned by Microsoft

@phase basically

@phase ooVoo?

3:34 AM

I pretty much have always used Steam voice chat for talking with my friends, but I had to use Skype a lot for school junk and I always hated it

I've used that a couple times for free more-than-two-people video-calling.

Anonymous

@El'endiaStarman I'm doing something slightly different. Numeric literals are prefixed with :. When I encounter a :, I start with a,b=code.index(':')+1,len(code), and attempt to do eval(code[a:b]). If the eval returns a ComplexType, LongType, FloatType, or IntType, we're good. Else, we decrement b. If b==a, push 0.

@Mego Oh, in code. Gotcha.

(Oh, in code)

Anonymous

@El'endiaStarman What's the distinction?

3:36 AM

I thought you were talking about parsing numbers from input.

Anonymous

Lol no, that's easy

Anonymous

I'm using Python 2's input()

Not if you want it to behave the way mine does.

Anonymous

If it throws, then it's treated as a NOP

Anonymous

I meant in the context of Seriously :P

Anonymous

3:37 AM

I should do numeric literals a lot smarter

How do you mean?

Anonymous

And use : as a delimiter, rather than a prefix

By adding an ending delimiter?

Anonymous

So that it matches the functionality of " and ` and []

Yeah, that's what I did.

Anonymous

3:38 AM

Self, why did you design numerics like this?

Not too late to change it...

@Dennis Re: TF2 - I even pinged Conor with CJam permalink with the exact " "_~ construct :/ apparently it was "okay because it's still interesting" or something -_-

(although I never got a reply after the permalink, paraphrased quote happened before)

Anonymous

I just did change it, it's so much simpler now

@Doorknob I need a bit of Ruby help.

Anonymous

I'm just wondering why I thought the original design was good when I wrote it

3:42 AM

@Sp3000 I would rather have had it fixed, but if he won't... ¯\_(ツ)_/¯

¯\_(ツ)_/¯

@quartata sure, with what?

With Ruby.

...

Just trying to help.

3:43 AM

Basically, I want to make a hash with the keys being every character in ASCII and the values being some random characters I generate.

I have the random character generator.

in an array?

(the random characters)

Now I need to figure out how to generate the hash.

undergroundmonorail

@Dennis i don't understand why you were downvoted for the op's mistake. you play to win and if people don't like how you won they should take that out on the person who made that possible

@Doorknob No, it is a function that outputs them.

ok, so make it into an array

3:44 AM

So run it 128 times and put the results in an array?

sure

OK, but then how do I make the hash out of the values?

then you can probably do something like Hash[[*' '..'~'].zip arr]

Ahh I see.

(didn't actually try that, but something along those lines)

3:45 AM

Thanks.

Oh wait a second.

Would converting the char to its char code and then using it to access the array of outputs work?

huh?

Since they are in sequence.

Yeah it would.

sure, that would work too

How do you convert a char to its char code in Ruby?

.ord

3:46 AM

gotcha

@undergroundmonorail I agree wholeheartedly. As I said before, ¯\_(ツ)_/¯.

hifives Dennis for making the same mistake

Anonymous

I fixed complex values \o/

o/\\o

Anonymous

More importantly, I fixed 99 bottles of beer! \o/

3:49 AM

I made a very nice-looking fractal via supersampling

Anonymous

@Doorknob now animate it

and I realized I'm an idiot and was compiling with -O0. Turning on all optimizations made my code about twice as fast >.<

@Mego that would take all night...

Anonymous

@Doorknob Run it while you're asleep

not a bad idea :P

Anonymous

Now that complex literals work, I get to implement all of the complex functions

3:53 AM

Does ~2+ add 2 to the inputted number in GolfScript?

As a full program? Yes.

Cool.

What are you doing?

I'm making a modification to GolfScript called TwistScript.

I'm testing the input.

(It piggybacks onto GolfScript in kind of a strange way)

Are you making it better or worse?

3:55 AM

I'm making it far, far worse.

Think Seed but worse.

Unusable language?

No it isn't for that challenge

It uses RNG so it couldn't, I don't think.

OOK

(eval):1:in `block in initialize': undefined method `+' for nil:NilClass (NoMethodError)

Strange error..

That means the stack is empty.

3:57 AM

How does GolfScript distinguish between program code and input?

@Dennis Hmm

PhiNotPi

CSS is hard

well, not really, just time-consuming

@quartata If you supply a file as command-line argument, STDIN is input. Otherwise, STDIN is code.

So it can't take input if the code is provided on STDIN?

Not in a straightforward manner.

Shucks. That means I have to write my code to a file.

3:59 AM

Why does the code have to come from STDIN?

$ golfscript <(echo '~2+') <<< 42
44

(In Bash, anyway.)

@Dennis Because the way I have it currently is that it redirects STDIN to a pipe, pushes the code followed by input and then runs require_relative "golfscript" which essentially runs the interpreter if it is in the same directory

@Dennis Interesting.

So just pipe "input"code.

GolfScript has a builtin for input?

No.

Huh I didn't know that I thought the implicit input was the only way.

@Dennis Ohhh I see what you mean.

You mean push the input as a string first.

4:01 AM

^

Gotcha.

As long as you escape backslashes and double quotes, you should be fine.

Yay it worked.

\\o/

Thanks.

4:04 AM

If it is important that your program behaves exactly like GolfScript, you should put a ; before the input string.

It shouldn't be.

Even if it reads the code from STDIN, the interpreter pushes an empty input string on the stack.

Oh huh.

I'll do that then.

Well jeez I think I've finished my TwistScript interpreter.

Only problem is that testing it could take a long... long... time

If it's worse than Seed, that's probably an understatement.

Cat is easy at least though.

I'll use that to test the input stuff.

4:16 AM

It takes 10 minutes and a paper clip to turn off my Nokia Lumia 720 because it is the worst phone in the world

@Hipe99 oh wow that guy hi

Anonymous

I am not proud of this code

@Mego If your code is for golfing, you should never be proud of it.

Anonymous

@phase This is for interpreting a golfing lang

Anonymous

gist.github.com/Mego/dc3759387be59e58f2d4

Anonymous

This is the hack I had to pull to be able to easily support complex values in my already-written commands

Anonymous

4:18 AM

(for the math commands, at least)

Looks groos, but at least it has support for it

Anonymous

Also it doesn't work for math.sqrt(-1)

Anonymous

I should fix that

Oh dude I'm adding this in O

Anonymous

@phase Complex number support?

Anonymous

4:21 AM

Overrated

I made so many bad design decisions that there is no hope for it any more...

Everybody does, but then you fix things and the language evolves, right?

but backwards compatibility

Anonymous

ideone.com/y970cX

Anonymous

By some miracle it works

Anonymous

4:24 AM

I'm shedding a manly tear for Python, for I have abused it in terrible ways

Anonymous

Please forgive me, Guido

Why le J?

You could update your past answers (slowly) to use version numbers. I don't like breaking backwards compatibility much either, but sometimes you just look at your code and go "well, that's just plain broken"

Anonymous

@phase ?

@Mego 1j instead of 1i or just i

4:25 AM

I don't mind breaking backwards compatibility at the moment. I'm the only one using Minkolang. :P

@phase Engineering origin.

Anonymous

@phase In my language design or the ideone output?

Anonymous

Python uses j as the imaginary unit

@Mego holy god no gross ew get it away from me

.......Google it.

There are reasons for it.

Anonymous

@phase That's actually pretty normal outside of the field of pure math

4:27 AM

@Mego If I had a choice, I'd probably do a try/except - if it doesn't work with math, then do cmath or something

> Python follows the electrical engineering convention.

Anonymous

@Sp3000 I might do that, that would be smarter

Anonymous

I wish cmath would automatically convert a+0j to a, though, for its functions

Anonymous

Then I could just replace every call to math.* with cmath.* and it would work perfectly

Anonymous

Though I can see good reasons not to do that

4:31 AM

There's no reason it would, because the 0j could have arose from floating point precision, for example, and you might still want it as complex

Anonymous

Or have a builtin module that does basically what I've done: call the math function if possible, else the cmath function

Anonymous

@Sp3000 All functions except for sqrt will return a real value for a real input

Anonymous

Only complex inputs (and negative inputs to sqrt) will yield complex results

I was saying why I don't think cmath should convert a+0j to a, not sure what you're getting at...

PhiNotPi

Apparently companies are starting to buy TLDs: home.barclays

Anonymous

4:38 AM

@Sp3000 Every function in cmath will return a+0j for all real inputs, with the exception of sqrt. It would be nice if they recognized that and returned a float for real input, rather than a+0j.

Yeah that's what I mean, there's no telling whether the 0j was caused by floating point precision (and is actually 0.000000000001j)

Also being able to return one of two types means that people will do a lot of checking to see which type was actually returned

@Mego fix your git email

Update index.html

Yay, my avatar finally updated \o/

Anonymous

@Sp3000 But if type(arg) in [IntType,LongType,FloatType], then the result is guaranteed by mathematics to be a real value. All I'm saying is, it would be nice if cmath's returns could match math's for inputs where the functions are identical.

Anonymous

That would allow cmath to work as sort of an upgrade to math

Anonymous

4:49 AM

(this problem is non-existant in Python 3 afaik, so this discussion is a bit pointless)

Anonymous

@phase My git email is borked?

Oh you mean a+0j is the input, not the output?

@Mego This should be you, not a blank account.

Anonymous

@Sp3000 No, I mean a+0j is the output, for input r.

... shrugs

Anonymous

4:51 AM

>>> math.sqrt(4)
2.0
>>> cmath.sqrt(4)
(2+0j)

I swear every time I run my IDE the functionality of it changes. Sometimes strings don't work, sometimes multiple inputs don't work, sometimes spaces don't work, and sometimes I just want to go cry in the corner.

\o/ ShapeScript just beat C#.

Anonymous

@phase fixed it, thanks

Anonymous

@Dennis That's like celebrating because Mike Tyson beat up a toddler

Thomas Kwa

@Dennis noice

A: Biplex: an important useless operator

4:55 AM

@Mego ShapeScript is not a golfing language. See for yourself:

0

ShapeScript, 186 bytes '"1 "$0?_1-"+"*":2"@~@"0 "~"0 "$" "~+'1?_*!"2"$"0"~" "$""~":"1?'@1?"+"$_1>"+"*+@1?"0"+$""~'1?_*!#"0"+@":+"'1?1?"0"+$_2<"+"*+'1?_*!"0"+$"1"~@$"1"~":"$""~"+"$""~'" "@+" 0"$""~'1?_*!" "$""~ An important, useless language for an important, useless operator. I/O is in binary...

@Dennis pics or it didn't happen

Anonymous

@phase That's like me, except I always want to go cry in the corner

Thomas Kwa

@Mego More like a Mike Tyson that, in a week-long inspection, appeared to lack arms.

Anonymous

@Dennis Yeah but C# is the exact opposite of a golfing language. It's a Microsoft language.

Anonymous

@ThomasKwa A kick to the face is still effective

Anonymous

4:56 AM

Now I get to make functions work in Seriously, woohoo

Anonymous

But I think I'll put that off for a little while

@Mego ShapeScript has no integer literals other than 0 to 9, no loops, no maps, no conditionals, no variables... Heck, there's not even a straightforward way to reverse a string. (Would have been handy here.)

Anonymous

@Dennis Variables are overrated anyway :P

Haha.

Anonymous

I definitely need a break from beating up Python to make Seriously before I go back and try to make functions work

5:07 AM

almost 800 rep ;o

800 rep ;o

Congrats! :D

@El'endiaStarman fenks <4

5

5:26 AM

@phase It just got cut off because it was too long :P (I can't tell if you don't know)

@AlienG I figured that part out put it's surprising that just the # was cut off.

Also, you did search C not C#...

@Mego Now gimme teh codez! Seriously.

@AlienG Yeah, that's why I clicked on it in the first place. I saw the C because the # was cut off.

@AlienG That is literally the entire point of the gif ;-;

Anonymous

@Dennis I almost regret the language name. Almost.

5:29 AM

I'm expanding on it :C

@Mego better than the letter O

Anonymous

@phase If you say so

@phase Diabolical

Anonymous

I guess mine has the benefit of searchability

You didn't.

5:31 AM

@AlexA. SO is truly an evil mastermind

feersum

Are you surprised because of how high or how low it's ranked?

Anonymous

@phase That reminds me, I should write some docs for Seriously at some point

@phase You may have personalization of results. ;)

@Mego #1 Use ReadTheDocs, #2 Don't use MkDocs, they suck and have less features, #3 Do everything Pyth did

Anonymous

@phase I'm gonna use ReadTheDocs :P

5:36 AM

@AlexA. CHECKING INCOGNITO PRONTO

@AlexA. still the same :D

@AlexA. Looks like ~ needs fixing as well. This doesn't work at all.

Anonymous

It's #4 for me in incognito

A: Biplex: an important useless operator

1

ShapeScript, 186 bytes '"1 "$0?_1-"+"*":2"@~@"0 "~"0 "$" "~+'1?_*!"2"$"0"~" "$""~":"1?'@1?"+"$_1>"+"*+@1?"0"+$""~'1?_*!#"0"+@":+"'1?1?"0"+$_2<"+"*+'1?_*!"0"+$"1"~@$"1"~":"$""~"+"$""~'" "@+" 0"$""~'1?_*!" "$""~ An important, useless language for an important, useless operator. I/O is in binary...

Wrong link.

0

A: Fibonacci + Fizz Buzz = Fibo Nacci!

ShapeScript, 83 bytes 11'1?1?+'77*2**!""'"%r "@+@0?2%1<"Fibo"*1?3%1<"Nacci"*+0?_0>"@"*!#%'52*0?**!"'"$""~

@phase #4 for me!

O is 9th for me, but still front page

5:40 AM

Literally looks like my cat jumped on top of my keyboard while I tried to turn on sticky keys. Nice job. — phase 1 min ago

Hahaha.

> Pushing broken code for someone else to fix

Commit message: "No longer my problem"

Why I put collaborators on my projects in the first place, cause I don't want to push working code.

Words of wisdom

Here's my pseudocode, make it real (totally not a challenge I made on here or anything)

"JVM Programing Langage" Not buying that app

kthxbai

isaacg

5:51 AM

@AlexA. No, it's on the free Heroku plan.

Anonymous

My Seriously RTD is up, though it doesn't say much

6:04 AM

@isaacg Cool, thanks for the info.

@Doorknob According to Isaac, Pyth uses the free Heroku plan.

Or was it @quartata or @Mego who was wondering about Heroku...

Can't remember.

I'll just ping everyone.

Too many people name their projects "TeaScript"

absolutely slaughters google search rankings ;-;

> Because I hate coffee and I have a lisp.

Anonymous

@AlexA. Both

Anonymous

@Vɪʜᴀɴ 44

6:14 AM

@Vɪʜᴀɴ :O You're danking?

Golfish: "Did you mean: Goldfish?"

WHOOHOO "Minkolang" is unique!

Hexagony is 4th for me, after a board game of the same name

2nd for me.

~~I 'll just give TeaScript a separate domain, that better increase search ranking~~ OS X Server doesn't let you to do that :(

6:19 AM

What is OS X Server?

Apple's attempt at a server app

@El'endiaStarman But yeah going back to Fibonacci word and stuff, how to be not-terrible at string challenges without an array type...

works great on old macs laying around

@Sp3000 I'm pretty sure I've figured out how to do it with only the stack. Also, Dennis allows outputting with a constant delimiter, which is great because I can stick with N. .... Actually, come to think of it, it would be shorter to use O so I don't have to do any char -> int conversion...

I'll fix that once I get it verifiably working. 0s and 1s are good for debugging. :)

Hmm k...

I was almost wondering if this would be easier with recursion, or something

6:39 AM

Got it! Ran it for just a couple seconds aaand...44386 characters output. Well, dang.

:P