The Nineteenth Byte - 2016-03-31 (page 2 of 7)

« first day (1888 days earlier) ← previous day next day → last day (2957 days later) »

2:20 AM

@PhiNotPi ping. Looking for advice on my KoTH framework if you're there

An example.

awesome!

oh, you want an example?

That I'm here?

yes, thought you were afK

Yeah, as in implement some generic KOTH, and show what parts of the KOTH need to be written where, in order to get it to work.

2:22 AM

Thanks @Doorknob & @KennyLau :)

no problem

@PhiNotPi the one I've been thinking about is quite a bit different than yours (games are only 2 players, a fixed width triangle map). I'm assuming that's ok?

Sure

If I add bases and multiple bots to my KOTH, how should I do it?

As in, if I let them control 5 bots, should I allow them to edit each bot individually?

yes, but similar to how the rules do it

aka, only let them edit 1 bot per turn

maybe all 5

not sure

Anyways, I was thinking about your skip command, and while its really cool...I think its too powerful

When I first looked at the game, it appeared that one of the core concepts was code reuse

I am considering cutting back on the maximum number of instructions, to like 8 or something.

Code reuse is a big concept.

2:29 AM

but with skips, you're able to simply able to segment your code

reducing the total number of instructions would definitely help, but I'm still not sure what you see that is appealing about skips

I wanted to encourage the creation of more complex programs.

I can see the appeal: perhaps a fixed number skip?

skip 2 instructions?

too few to be overpowering, and too many to be useful

:P

I want to avoid the problem of people limiting themselves to super-simple (1-token) programs, since those are the most "responsive" (quickly changeable).

very true.

Maybe, here's two statistics to keep track of: average # of insertions/removals per turn, and average # commands executed per turn.

2:38 AM

that was definitely a problem I ran into with codebots

I want to have both number be high.

although, super-simple programs are at a major disadvantage

they perform less commands

waffles

That's the main advantage to having longer programs.

Which is also the reason I want to limit program length to 8 instructions: imagine a series of 24 forwards/turns/grabs by a single bot in a single turn.

I have another window open and the only things I saw on screen were waffles and That's the advantage to having longer programs.

2:42 AM

The original game had 3. I might need to make it even shorter than 8.

I think that 8 is fine

I mean, commands are built up one at a time (ignoring initial loading)

so, making all 8 useful is really hard

the more you reduce the code length, the less "code reuse" matters

I don't know if the skip command is that overpowered when there's only 8 instructions.

I don't think so either

using it uses 2 spots in your code

and the more you skip, the less commands you execute

Question: how should Z,J,H,L work? (where J is jump/skip) Should it work the same as Z,L or simply do nothing?

The robot zaps, and then skips, and then turns left.

Should the skipping "preserve" the zapping or expend it?

zapping basically says "hey bot, execute my code here"

so, zapping says to the other bot

"J,H,L"

so, it jumps

and then you Jump, then left

2:53 AM

That's actually not how I have zapping working.

I should have said "execute the first command of my code here"

but how do you have it working?

but yeah, zapping a jump (in my mind) is a no-op

Z,F,L (an arbitrary example) is executed like so...

There's the "host" bot (whose turn it is, since it's where the program is hosted) and the "mover" bot (which is the one that actually moves in response to the command).

At first, the host and the mover are the same.

The first command, Z, changes which bot is the mover, so now the mover is the one being zapped.

found an error in the userscript background!

user image

escaping FTW

The F is the next command, and causes the mover to move forward 1. The mover is reset to be the host.

that doesn't change anything

2:59 AM

The L command is executed on the host, which turns left.

So ZFL causes some other bot to more foward and the original one to turn left.

in your original example, J is the next command, which doesn't do any moving

oooh, clever idea

maybe ZJ will cause the moving bot's instruction pointer to start at the next skip

no, I don't like that

And so, the non-zap command at the end of the zap chain is the only one executed on that other bot, and then the zapping is "reset"

@Cyoce I thought I removed those... :|

@Downgoat ¯\(ツ)/¯

I was feeling that you were misunderstanding that ZL causes another bot to turn left and then the original to turn left as well.

Although my spec description of that behavior is probably not the most effective.

3:03 AM

yeah, I assumed that zapping didn't progress the pointer

Zapping doesn't progress the pointer, it changes where the next instruction is executed.

a KotH

4

A: Sandbox for Proposed Challenges

Meta Code Bots I'm just giving this a "code bots" title because that's the most similar previous KOTH. This version is based on the Twin Tin Bots board game, and will probably be given a better name before posting. On a far away planet, dozens of small mining robots compete for resources in the...

@PhiNotPi but it does, because the zapped command progressed the pointer

anyways, now were just getting into semantics

Well, maybe I should say that all commands progress the pointer, after execution. Clear? Moving on.

3:05 AM

yes

The original question was "how should Z,J,H,L work?" and, as of right now, the zapping doesn't do anything.

:)

The J command still jumps the instruction pointer, as if the zap wasn't even there.

I could make it so that Z,J is a no-op, so that Z,J,H,L just does nothing overall.

Maybe a better way to phrase the question: Should ZJHL cause the bot to halt, turn left, or cause other bot to turn left?

Causing it to halt is probably the most consistent.

progression of the instruction pointer always happens on the host bot

not the mover

therefore, I think its reasonable that J would still occur on the host bot

and therefore, turn left

and the other bot to not move

Okay, in that vein, what should ZHL do?

3:12 AM

Halt

Right now (and in the spec) the zapping does nothing, but causes the original bot to turn left.

Okay

wait, no

hmmm, I'm indecisive on that one

perhaps we do what is more interesting?

halting is probably more consistent

the most interesting is that zapping a jump makes the other bot do the command after the jump

and ZH halts the other bot (no-op), but not your own

The 1-letter abbreviations are both convenient and slightly confusing.

3:16 AM

In CJAM, is W=-1 documented?

Chris Jester-Young

@KennyLau Yes.

Where?

Chris Jester-Young

sourceforge.net/p/cjam/wiki/Variables

Thanks

Chris Jester-Young

Always my pleasure. :-)

3:18 AM

dl.dropboxusercontent.com/u/15495351/cjam.pdf

ftw

@Downgoat stuff is broken

@MarsUltor isn't it always D:

@PhiNotPi What language is this?

a KoTH

5

A: Sandbox for Proposed Challenges

Meta Code Bots I'm just giving this a "code bots" title because that's the most similar previous KOTH. This version is based on the Twin Tin Bots board game, and will probably be given a better name before posting. On a far away planet, dozens of small mining robots compete for resources in the...

lol, I got unpinged

waffles

3:24 AM

-.-

would flag for edit abuse, except that simply goes to the perpetrators

@AlexA. Dutch babies >= that

El'endia Starman

@HelkaHomba .... the Dutch eat their babies?!?

._.

@El'endiaStarman When they look like this:

user image

Wow, Dutch children are very flat and appear to be covered in a fine layer of powdered sugar

El'endia Starman

3:27 AM

Reminds me of crepes. Flat and covered in powdered sugar.

@AlexA. There go your Dutch voters

@El'endiaStarman Aren't those French children?

@HelkaHomba ono orlp y

@AlexA. tres tres bon croissant

Should I allow tabs for indenting in Cheddar?

@Downgoat no

3:29 AM

@Downgoat No

Tabs should be a syntax error

that's what I thought

@El'endiaStarman No, dutch babies are eaten. There is a crucial difference: anyone can enjoy a dutch baby, not just a dutch you racist.

Except in strings

@AlexA. even if it appears in a string

3:29 AM

@AlexA. they will return a CheddarError.UNEXPECTED_TOKEN

Chris Jester-Young

@AlexA. Except in GolfScript.

Then it's just another variable.

Oh wait

@Downgoat yes

El'endia Starman

@Downgoat ABSOLUTELY.

\t in a string != tab

3:30 AM

@NathanMerrill Yep. No tabs anywhere. Tab-free 4 life.

okay, tabs do not count as whitespace in Cheddar anymore

@ChrisJester-Young Oh right. I forgot that you can do that kind of evil stuff in GolfScript. :P

El'endia Starman

._______.

although, life would be better if we simply got tabs right the first time

Chris Jester-Young

And because you can define the value of tab to be different from the value of space, you can do Whitespace-like behaviour in GolfScript. :-P

3:31 AM

@AlexA. forces Alex to watch patronizing documentary of a World Without Tabs

Chris Jester-Young

@AlexA. Remember, before I embarked on the idea of making an independent implementation of CJam (because it's saner than GolfScript), I actually wanted to make an independent implementation of GolfScript.

Anyone want to take the pain of explaining how the second level of this diagram is being constructed?

Chris Jester-Young

@AlexA. In fact, I created some code for that project too: gist.github.com/cky/3392271

@AlexA. Tabs are silently stripped in Jelly.

@Dennis \o/

> package nz.kiwi.chris

Beautiful :D

3:33 AM

@Dennis +1000

Chris Jester-Young

@AlexA. These days I'd use nz.cky, but I wrote that before you were able to register second-level domain names in .nz.

@AlexA. Can I turn your treehouse hill into a forest?

@HelkaHomba Uh, why?

Chris Jester-Young

(Yes, I own both cky.nz and chris.kiwi.nz.)

anyone?

3:34 AM

Also it was your hill, I just invaded nearby.

Very close by >_>

@AlexA. It looks so bare since I moved

C: clone. Creates a bot on the square in front of you that will keep executing your string of commands until the next H (halt) — Cyoce 1 min ago

^ worst idea ever

@PhiNotPi C:

@DemCodeLines I'd try to help if I could but that's well outside of my area of expertise.

@PhiNotPi No. best idea. You can create bot-ception. Who doesn't love bot-ception?

3:38 AM

:(

.=...
.^^..
.....
.....
.....

GRAB
.....
.^^..
.....
.....
.....

RIGHT
.....
.>^..
.....
.....
.....

ZAP
.....
.>^..
.....
.....
.....

JUMP
.....
.>^..
.....
.....
.....

FOWARD
..^..
.>...
.....
.....
.....

DROP
..^..
.>=..
.....
.....
.....

HALT
..^..
.>=..
.....
.....
.....

^ example of execution of GRZJHFD by the left bot.

`P` (`POSSESS`): the bot in front of your current bot works for you now, bitch! (the owner of that bot now controls yours, but ppphhhhhh, who cares about that? The new one is so much cooler!) (unknown
side-effects)

user image

4:04 AM

Would anyone care to create esolangs.org/w/index.php?title=J&action=edit&redlink=1 ?

@KennyLau It's an esolang?

@MarsUltor Why is it not?

It's a normal language

El'endia Starman

J is an esolang in the same way APL is. Which is to say, it's not, but it almost is. :P

alright

4:15 AM

Which should I use for OR: or, |, or ||?

| = bitwise, or = logical w/ short-circuit, || = logical w/o short-circuit

or something

El'endia Starman

Is there any language that distinguishes between short-circuit and not?

R and Julia

& and &&

El'endia Starman

Is it ever useful to have the distinction?

@El'endiaStarman C#

Yes

for unreadable chaining stuff

foo & bar always performs both operations

useful sometimes, probably not very often though

4:19 AM

@El'endiaStarman In R, & applies element-wise over arrays, so if you have two arrays of booleans, you can efficiently do logical calculations on each element.

On the other hand, foo() && bar() only performs bar() if foo() returns true

waffles

Yes (whoops)

I smell the lua's ternary idiom

El'endia Starman

@AlexA. Huh, that's neat.

4:22 AM

@AlexA. that was random, why not pancakes?

@El'endiaStarman Yeah! Like x <- c(TRUE, FALSE, TRUE); y <- c(TRUE, TRUE, FALSE); x & y returns the vector c(TRUE, FALSE, FALSE).

El'endia Starman

Nice. That would be x \pair(&) y in Pytek.

Ah, so infix operators are specified like \pair(<op>)?

El'endia Starman

What \pair(<op>) does is that it applies <op> to the left and right arguments in a pair-wise way.

You can define additional operators yourself.

I had an example in the Pytek room yesterday...

Isn't applying an operator to left and right arguments in a pairwise way exactly what an infix operator is...?

El'endia Starman

4:26 AM

Oh wait, I posted it in here.

@AlexA. Not necessarily.

(Also why was that starred?)

El'endia Starman

>>> [1,2,3] + [4,5,6]
[1, 2, 3, 4, 5, 6]

\pair() modifies the operator you put in to become pairwise.

@El'endiaStarman Oh right, stuff like that. In R land, + and similar operators are always element-wise over arrays.

El'endia Starman

Gotcha.

@El'endiaStarman Oh that makes sense. That's neat.

El'endia Starman

4:29 AM

I will also have \each() which basically splats the right argument (probably). So [1,2,3] \each(*) [1,2] would be [[1,2,3], [1,2,3,1,2,3]].

I'm very confused why the result has only two elements

(sublists)

El'endia Starman

>>> [1,2,3] * 2
[1, 2, 3, 1, 2, 3]

Oh

Python is weird

El'endia Starman

lol

It's a natural extension of using + to mean concatenation when applied to lists.

In R and Julia, scalar * vector = [scalar * v for v in vector]

El'endia Starman

4:31 AM

I figured.

Overloading arithmetic operators for stuff like lists turns me off a bit when it comes to Python, but it's not too bad.

I still have to think really hard before I do anything like that though

Should probably also be noted that Julia uses . to indicate element-wise operations as well, e.g. .* is the element-wise version of *. Just if you wanted something to compare to \pair

El'endia Starman

@AlexA. How do you mean?

@Sp3000 Ah, cool.

@El'endiaStarman If I write [1,2,3] * 2 I sure as hell don't expect [1,2,3,1,2,3] offhand; I have to remember that what I'm using isn't what I'm used to. :P

El'endia Starman

And, y'know, I haven't really implemented much of anything beyond working on the parser. I'm probably gonna make concatenation ++ (it won't be an increment operator), so I could probably do +* for repetition and make + and * vectorize by default.

4:36 AM

@Sp3000 Yes. Julia is great and does all things.

@El'endiaStarman +1

@AlexA. I'm starting to appreciate it a bit more, don't worry :)

@Sp3000 Worried? Hell, I'm thrilled.

El'endia Starman

Pytek is not going to have parseInt though. That's just silly. :P

So you won't be able to parse integers out of strings? :/

Btw, does anyone have a good resource on ASTs and stuff like that that are useful for making (pars|lex)ers?

@AlexA. dinosaur.compilertools.net/yacc/index.html

lex & yacc is useful

El'endia Starman

4:41 AM

@AlexA. 'course you will! int

@El'endiaStarman Oh, then what was your comment about parseInt about?

@Downgoat Those are tools I'd have to integrate into something though, right? I'm looking for a way to build my own.

El'endia Starman

@AlexA. Trips me up in JavaScript. I was thinking that Julia had that too, but it's something a bit more complex, isn't it?

@AlexA. I could tell you how I wrote mine.

Not sure if you want to do it all by hand though.

@AlexA. no, that is a compiler compiler, so you write your compiler and it compiles your compiler so you have your own compiler

@El'endiaStarman parse returns an Expr unless it's super obvious, like parse("1") makes 1. Otherwise you give a DataType as an argument, like parse(Int, "1").

@El'endiaStarman Sure!

@Downgoat What

El'endia Starman

@AlexA. Oh, I thought they changed it to something ugly in 0.4.

4:44 AM

Yeah, they deprecated int in favor of parse(Int, )

@AlexA. It is a compiler compiler 1. you write your compiler 2. Compile your compiler to a compiler in your language of choice 3. Profit

I hope that made sense

@Downgoat That's not clearer than your other sentence. :P

El'endia Starman

@AlexA. Ah, not as bad as I thought. Still, less convenient.

Eh, I don't like int. Too Pythony. :P

Essentially you write your compiler in an intermediate language and it generates a compiler.

It's so you don't write an entire compiler by hand, unless you want to be stupid like me and write your entire compiler by hand

4:46 AM

I want to write an entire compiler by hand

._.

why

Wat

To learn how to do it

no need to reinvent the wheel

Did you know: VS is written using VS

4:47 AM

I'm not trying to reinvent anything, just trying to learn.

I have no idea whether text editors are written with themselves

@AlexA. Hand = asm?

I highly recommend against it, if you heed my advice, you will save yourself from weeks of head bashing on your desk and crying

Wait

Compiled compiler = any compiler

I think "compiler" isn't what I'm going for; I'm not trying to generate machine code or bytecode. I'm just making an interpreter.

@AlexA. well you still need to compile it to something so your language can interpreter it whether it be an object or array.

4:49 AM

Were you thinking of something like how Element converts the program to Perl code and evals it?

So are you going to use Python as your interpreter interpreter?

@PhiNotPi Not really, though I had considered that

@MarsUltor Nope

@Downgoat I know

> interpreter it

El'endia Starman

@AlexA. Aww, but Python interpreters are all the rage! :P

Mar 22 at 4:33, by Downgoat

ugh, stupid spelling. we should all just communicate in code...

4:51 AM

@Downgoat What is you in code?

@El'endiaStarman I thought JavaScript interpreters were what the cool kids were using :P

@El'endiaStarman pffft, Python? Node.js is where it's at! :P

@AlexA. JS is good as well

@MarsUltor u

@MarsUltor "you"

4:51 AM

@MarsUltor That's false

@AlexA. ?

@MarsUltor +1

@AlexA. ಠ_ಠ

@MarsUltor JavaScript is not good

pls stop spam replying

Nobody is spam pinging. We're using replies as they're intended to be used.

4:52 AM

0

A: Merge two sorted lists

JavaScript (ES6), 69 79 bytes f=(a,b,c=[])=>(x=a[0]<b[0]?a:b).length?f(a,b,c.concat(x.shift())):c.concat(a,b) How it works f = (a, b, c = []) => // `f' is a function that takes arguments `a', `b' and `c' - // `c' defaults to `[]' - which returns the fol...

@Dennis +1 for JavaScript

Chris Jester-Young

@MarsUltor Click the speaker button next to "all rooms".

5 replies in one page :(

-1, not enough jquery.

Chris Jester-Young

@MarsUltor People really are using replies correctly.

It's stars that we're not using correctly

Chris Jester-Young

4:53 AM

but you should silence your speaker if you don't want that nasty sound with every reply.

El'endia Starman

@MarsUltor That's not bad. And, the replies are not without purpose, as others have said.

...

@Downgoat Oh the irony...

Chris Jester-Young

/me elevens the stars off that post. :-P

@ChrisJester-Young I do

Just I have desktop notifications on

4:54 AM

@ChrisJester-Young :(

El'endia Starman

@ChrisJester-Young Wat. [restars]

Fuck, I just realized literals can start property chains ;-;

@Downgoat "you"?

El'endia Starman

...I can't?! ಠ_ಠ

$.message("How is this meant to reduce spelling errors?")

4:55 AM

@Downgoat my stars :(

$.message($.like($.this))?

Chris Jester-Young

@El'endiaStarman Nope, you can't. In SE chat, stars are treated as votes of score 1 in the backend. When stars are cancelled, their corresponding vote scores are set to 0, but the votes aren't removed.

@MarsUltor jQuery spell checker

@Downgoat My bad, grammar errors

@ChrisJester-Young What if you clear the stars, and then someone else stars it?

Does that count?

Chris Jester-Young

4:57 AM

@MyHamDJ Then the new stars have score 1, and the original stars still have score 0.

The number showing on the starboard is simply the total score of all votes on that message.

Ok, so it will be on the starboard again?

Chris Jester-Young

Yes.

What if the stars are cancelled within the timeframe that I'm allowed to undo my star?

Could I undo the cancelled star to star it again?

Chris Jester-Young

@PhiNotPi Good question. Since I don't have the code any more, I can't tell you the definitive answer. You'll just have to try it and see.

@Downgoat Dark times...

5:02 AM

ಠ_ಠ

Any plan for April Fools' Day's Special Question?

Yes!

HAH GOTCHA! April Fool's! ( ͡° ͜ʖ ͡°)

No

Hahaha! You totally got me!

@AlexA. ಠ_ಠ

I should change my avatar to upgoat for april fools

Or I could become Chatgoat...

What do I change my avatar to then?

5:14 AM

It'll be another Maria Tidal Tug

Uh oh, election begins on April 1st...

._.

That seems incredibly fitting for PPCG somehow.

^

> nor am I fond of avocados

>:U

@Dennis your nomination says "almost 70,000 reputation", you should update it now that you've hit 70K

I wonder if Martin will get to 400 in the primary before the election begins

5:20 AM

$ rm -r julia-*

@Dennis WHY WOULD YOU DO THIS

Because I managed to install it from the repos.

@Dennis From source?

No, a packaged binary. 0.4.5 is the latest version, right?

Yes

5:21 AM

Sweet. openSUSE science repo has it.

Oh

Nvm

@Dennis +100 to openSUSE

+100 to science

@Downgoat status-completed

Dr. openSUESS

lel

5:42 AM

Anyone know a famous tune that can be played in exactly 27 or 28 beats? Like the first two lines of Ode To Joy, but that's 32 beats.

8

A: Element names above 118

JavaScript (ES6) 164 (unhexquadium) 171 The capitalization is the tricky part. (Test in FireFox) Edit 3 bytes saved thx user2786485 f=n=>[([...n+(l='')].map(c=>(c='nil,un,bi,tri,quad,pent,hex,sept,oct,enn'.split`,`[c],n=l?n+c[0]:c[0].toUpperCase(),l+=c)),l+'ium').replace(/i(i)|n(nn)/g,'$1$...

According to the leaderboard, this is 3 bytes. ¯\(ツ)/¯

user image

The stack snippet is messing it up I bet

@Cyoce Are you using the userscript?

@AlexA. yeah.

@Downgoat [bug] ^

5:46 AM

I like it a lot, ~~but I'm finding bugs.~~ finding bugs makes it more fun

« first day (1888 days earlier) ← previous day next day → last day (2957 days later) »

Transcript for

Mar '1631

The Nineteenth Byte

The Nineteenth Byte: General discussion for codegolf.stackexc...

join this room
about this room

00:00

06:00

12:00

18:00

all times are UTC

site design / logo © 2024 Stack Exchange Inc; legal

mobile