The Nineteenth Byte

6:55 AM

helllloooooo

so apparently we only have 1500mb data usage left for the rest of the month, after that it's $30/gb

and yet again, my so question goes unanswered, not even a comment

and i got downvoted on an answer because I simplified away the element that the user was having a problem with...?

9:10 AM

I've got an idea for a more advanced Clacker animation challenge, by allowing 2D regions to be covered by each command. And accepting a string of arguments to apply to each pixel, ie. a bitmap in every command instead of just a bit.

9:26 AM

why not a turing complete boolean expression language to evaluate for each block? xD

I've thought about something like that, but the details elude me.

And then of course it would need to get out of the loop somehow.

I was thinking it could have "jumps" to sequences.

Then someway to modify the bitmaps in the instructions themselves. I was thinking of having a watchacallit ... register-queue-thingy ... you know the thing with delays ... dammit watsit called??

Anyway, one of those as a pipeline cache, holding the next screen's-worth of instructions, and those can somehow be the target. Maybe with negative indices.

Hmm. with 2D-bitmaps, the screen's-worth is no longer a useful measurement. ...

so you can't just loop over every pixel and set it's value to the return value of a function that takes the pixel's x and y?

and then start over?

:P

time for a pymockup

9:43 AM

I don't understand the question. :( You loop over the region and apply the ... yes, Boolean function ... with the current pixel as X and the bit from the instruction as Y.

I'm relieved that I got rid of the Boolean chains before anybody pointed-out how stupid useless and difficult it was. :)

class aThing:
	def __init__(self, w, h, rule):
		self.rule = eval("lambda x,y,v:"+rule)
		scrt = [False for i in range(w)]
		self.screen = [scrt for i in range(h)]
	def runOnce(self):
		self.screen = [[self.rule(x,y,c) for x, c in rc)] for y, rc in enumerate(self.screen)]

self.screen can be accessed from within self.rule afaik, making it turing-complete

and you write in python :P

e.g. x = aThing(10, 10, "(x+y)%2==v")

that should make a constantly inverting grid i think maybe

Yeah. That's the idea. The Bohm-Jacopini theorem for Turing-Completeness requires: Sequence, Selection, and Iteration (or Recursion). The trouble I'm running into is with "selection".

never heard of that before (noob), what exactly does selection imply? :P

I suppose the simplest is to add a callstack and jump and return instructions. ... but is that the "clackerly" way to do it?? <-- that's my conundrum. :)

ah

9:52 AM

Decisions. if/else, cond.

Conditional jumps, generally.

aaahh

well, you can access all of the screen from within the function,

you can run python code from within the function via exec,

I was trying to get something like that with the Boolean-chaining of statements, but I did a poor job of it.

meh, i'm not too good with these things :P

also i just had an idea for a hash function o: o: o:

gonna use in my terragen because why not

But, I don't want to specify the whole of Python!

i guess that's a bit of a setback ^^'

10:00 AM

Hmm. I could maybe do something like MIIX or pdp-1 and have the function call save the return address into a jump instruction card. ... Hmmmmmmmm.

@luserdroog You're thinking about an advanced version when you don't even have submissions yet? (Which I think is a shame, the challenge is pretty nice, imo. I didn't read it properly before, but might work on an answer now that I did.) Also just extending to 2D regions doesn't sound different enough to me to justify a follow-up challenge. But how about a code-challenge for producing the minimal number of instructions for a given frame?

I've got an example of that somewhere. If you've never seen it, it's really wild.

This wouldn't have to be a challenge, maybe just a development of the format into something approaching useful. :)

oh okay

A: What are the uses of self modifying code?

This wouldn't have to be a challenge, maybe just a development of the format into something approaching useful. :)

6

1960s-era assembly languages used self-modifying code to implement function calls without a stack. Knuth, v1, p.182: MAX100 STJ EXIT ;Subroutine linkage ENT3 100 ;M1. Initialize JMP 2F 1H CMPA X,3 ;M3. Compare JGE *+3 2H ENT2 0,3 ;M4. Chang...

my connection is terrible tonight.

timeout retry/cancel

makes jack dull boy

Oh. I see. If I actually type the same thing three times, it eliminates the dupes. :/

i'd really like to submit a challenge but i can't think of anything besides the mmorpg koth and it's a massive project to take on

10:06 AM

shoulda used shift-enter.

@luserdroog So can I assume that the output will always be 20x20, or do I have to infer the dimensions from list of instructions?

Your implementation may choose whatever size, minimum 20x20. The examples don't even push that boundary.

okay, I'll probably give it a try tonight... I have an idea for how to do the boolean function in Mathematica, but the string handling is probably going to kill my score

maybe ruby would be better... I'll see when I get down to it

awesome. I've been pondering how to do it in C.

There's an ioccc program that draws in X11 using just send().

my C obfuscation skills are so far from paying the ioccc any attention, to be honest

10:20 AM

Understood. But if you find one that does something you're interested in, ....

There is a certain joy in the teasing-apart process.

Particularly if you've written something similar.

Now I loot it for ideas.

There are also some fascinating answers on SO: stackoverflow.com/questions/tagged/ioccc

For golfing, though, the bottleneck is the GUI interface. All those calls to build every little thing that makes up a window.

ah nice, didn't know about the tag

Here's the shortest window program I've written:

1

A: Hello World window

C+xcb+Cairo 2540 (82 lines) Not exactly short. :) This is literally the PNG hello world from the Cairo docs, with just enough xcb to open a window and find the parameters needed to register the window with Cairo. //xcr.c //cc -o xcr $(pkg-config --cflags --libs cairo xcb xcb-icccm) xcr.c -lcair...

... damned connexion.

haha

11:08 AM

theoretical reverse-stack-based lisp-like language might become ena

pretty okay for golfing

Graham's number:

=|G_.64{k.3^.3 1.3 .4

only because it has k and _ built in, but whatever :P

2:01 PM

When did there start being a notification for the Suggested Edits review queue? Is that new? I've never seen it before.

2:13 PM

Should we vote to close bad posts? Or just ones that don't fit on the site?

For reasons other than being bad, I mean

As in? Examples?

sorry to interrupt, but I was wondering whether to close codegolf.stackexchange.com/q/12188/12205 as being not having an objective winning criterion

This doesn't seem like a great challenge (I think it could be modified to be good, maybe, but it isn't now) but now that the OP has specified a win condition I don't think it's rule-breaking. Unless I've forgotten something.

@ace I'm assuming the [puzzle] tag implies "first correct answer," but meh

@Doorknob If that's true it should be in the tag wiki.

2:23 PM

Yes. And I thought about this question because the one that undergroundmonorail linked to is very similar

The question I'm talking about now has two win conditions, if that's the case, because it now has code-golf and puzzle.

Yes, I tagged it as puzzle. If puzzle implies first answer perhaps we should let the OP decide?

Related but not very helpful: meta.codegolf.stackexchange.com/q/1588/12205

Yay, finally managed to solve a code golf challenge purely in regex!

1

A: Recursive acronyms

Regex, .NET flavour, 65 bytes (?i)(?<=^\w(?<c>\w)*) (\k<c>(?<-c>)\w+ ?|(?<w>\w+) ?)+$(?(c)(?!)) You can test it here. If the input is a recursive acronym, this will yield a match, and capturing group k will contain all function words. If it isn't, then there will be no match. This does preser...

@Doorknob the .NET-bug that prevented me from solving your building shadows challenge in regex is still not fixed :-/

2:46 PM

I've used this site for 99 consecutive days. hype hype hype hype hype hype hype hype

@undergroundmonorail the question you linked to is now closed as "general programming question"... but i really don't see how that's a general programming question...

Yeah, if it's going to be closed that really seems like the wrong reason.

I've voted for reopening

3:04 PM

@ace I did, too, now that it's specified as code golf.

<><

subliminal christian messages on ppcg?

I thought he was referring to the Fish language

ha, didn't even know that existed, but apparently it's spelled ><>.^^

3:23 PM

foosh

@TheDoctor what do you mean by "Uhhhhhh, Whaaat???" in the comment here?

The question is not understandable

@TheDoctor So basically, add code in front of, in the try block, and/or at the end, such that it will print "You won!"

@TheDoctor what's not to understand? You're supposed to reach the outer catch by writing code inside the inner catch or outside of the outer block.

ohh

i dont like it

3:30 PM

:D

3:48 PM

hey

yo

@Trimsty how's your progress on the terrain?

@m.buettner I got lazy and decided to wait for an answer to the SO question, gonna start working on it now :P

if i keep waiting i'll never make it

haha fair enough

now i just have to figure out how to vertex map a boolean 3D noise function and force it onto a given heightmap (true 3d, allows for overhangs/really steep cliffs/etc)

then i have to simulate erosion for a bit

3:57 PM

Why do you want 3D noise if you're going with height-map terrain anyway?

btw, I found this yesterday code.google.com/p/fractalterraingeneration/w/list

i'm not, actually, the heightmap forcing is so that it's not like aerated chocolate

I'd do midpoint displacement but i have no idea how to render that consistently and with good block tiling

right... I'm not entirely sure what you're doing but it sounds like a fun hybrid

i'm not sure either ;333

the trick with voxel-based terrain is usually to have a noise function where positive corresponds to solid and negative to air... then you can just start with the function f(x,y,z) = -z to get a solid boundary at z = 0 (rock below)... then you distort that with whatever height map function you want, plus noise plus warping plus whatever... to render it, use marching cubes

interesting

4:02 PM

rendering stuff like that is horribly complicated though, because you need some form of paging algorithm

http.developer.nvidia.com/GPUGems3/gpugems3_ch01.html

maybe i can midpoint-displace a previously rendered heightmap?

takes away my tiling worries

sure, I don't see why not

i could displace the midpoints with 3 noise maps similar to the base for all 3 axis

and also use a blurring function

so it's all smooth-like

ooohhhh

thinks furiously

i've always loved multiplication matrices and image imaginary coordinate mapping and stuff like that to make it look more realistic

seeing the direct application to a 3D position map is exciting

@Doorknob ux.stackexchange.com/questions/57929/…

:(

Oh, nevermind, you saw it :P

okay wait

i can generate random chunks of noise and grow/pinch/displace them in a way that they shove each other around

and do some light erosion computation along the edges

canyons, mountain ranges and other things ahoy

realizes he just invented simulated tectonic plates

ohai @TheDoctor

4:17 PM

hi

welcome to my sky!

??

xkcd reference

ohhh

see

i'm in the shibe storage business ;333

4:18 PM

lol

i like how a stackoverflow meta post asking about offensive content and including an example of it got locked because of the quoted content

and the post they quoted didn't get deleted

lol

:D

But... we have feelings too :'( — Doorknob 2 hours ago

:P

Hey @ProgramFOX :D

ProgramFOX

4:24 PM

Hi all!

TARDIS

kk

5:12 PM

so far i've written some lines of code

such progress

class terrainManager {
	public:
		terrainManager(int blockSize) {
			this.blockSize = blockSize;
			std::vector<int[blockSize][blockSize]> rmap;
			this.pointMap = rmap;
		}
		void gen(int x, int y) {
		}
		int blockSize const;
	private:
		auto pointMap;
};

probably gonna throw errors

woo, errors

i don't see any in there

but c++ always finds a way

2

C++... I like to call it... "The German among programming language."

:P

is lisp lojban?

or clojure?

that's probably rather prolog?

5:16 PM

ahk

but seriously

clojure is best programming language

yay, itunes is playing my fav song

how nice of itunes

5:33 PM

I have an idea for a 3d KoTH

ProgramFOX

Nice!

I'm going to try to get a discussion going on meta about puzzle. I have two answers written up in different tabs for this question, one containing a proposal for a "real" definition of a "puzzle" for this site and a seperate one detailing some options on how we could make the tag fit on this site. Is it kosher to post them separately or should I try to work them into a single answer?

hard to tell without reading your posts, but I'd say post them separately

Alright, going for it

@TheDoctor sweet! you're not the only one though. ;)

@undergroundmonorail if someone complains you can always merge them

5:36 PM

Very true.

ProgramFOX

@undergroundmonorail When posting them as two answers, people can vote separately on both answers, which is not possible if you put them in one answer.

Okay, posted them

so I just found this language called Agda

its interesting

OH oh oh

my terrain generator

right

^^'

the fun thing is, once you can actually view your terrain and get to half-decent state you'll spend a ridiculous amount of time exploring terrain instead of coding more :D

lol

Does this look like it'll work?

class point {
	public:
		int x, y, z;
		point(int x, int y, int z) {
			this.x = x;
			this.y = y;
			this.z = z;
		}
	}
}
class terrainManager {
	public:
		terrainManager(int blockSize) {
			this.blockSize = blockSize;
			std::unordered_map<int[2], point[blockSize][blockSize]> rmap;
			this.pointMap = rmap;
		}
		void gen(int x, int y) {
			int[2] id = [x, y];
			point[this.blockSize][this.blockSize] cblock;
			for(int y = 0; y < this.blockSize; y++) {
				for(int x = 0; x < this.blockSize; x++) {

just generates flat terrain for now

5:59 PM

ugh, lower-case class names

imma change that :P

isn't it int id[2]?

same for declaration of cblock

you've got some serious naming conflicts as well

parameters x and y, as well as local (iteration) variables

ah

why did i do that...? xD

if it's int id[2], how do make an unordered_map with an array of 2 ints as key?

also, if your your point map is two dimensional, it's effectively a height map... so storing x and y doesn't really make sense... you might just as well have a 2D array of z-values. if you want voxel based terrain, then your map needs to be a 3D grid, with each voxel specifying if it's air or rock (and maybe some meta information). in any case, storing points in the grid is a bit pointless, because the coordinates are implied by the position of the current cell

@Trimsty no I think that part remains int[2]

@m.buettner: the 2D mappiness is just so i know which points to connect. There's nothing stopping pointMap[origin][0][0] from having a higher X than pointMap[origin][1][0]; that's why Point also specifies x and y.

strange, but yeah, i'm new to this 3D stuff :P

6:06 PM

isn't that.... weird? :D

I don't know, you probably have something in mind ^^

i'm just creating a 3D model of the surface is all

probably should've (why is autocorrect highlighting that) said verts instead of points :P

oh okay, well if your surface can be parameterised by a square grid (but is actually arbitrarily curved) then that kinda makes sense

yay! ;D

i'm also a memory efficiency geek

B)

I'm a bit scared of doing hydraulic erosion on this map

perhaps i shouldn't even be bothering with the 64x64x64 block format?

wb @TheDoctor :3

hydraulic erosion sounds horrible for 3D terrain :D ... if you get to work I'd be interested in how, though ^^

:|

6:14 PM

:|?

i could generalize the points and simulate rigid body physics on edges and force volume preserves, but then it'd be just a bunch of big blobs of blue jelly

:D

anyone ever use Processing?

i've heard a lot of hype about it but don't have much experience

it's Java based...

but its cooler than java

6:16 PM

like clojure? xD

^_-

i dont know

i was about to switch to another song

but then shuffle brought it up for me

hehe

y u psychic itunes

Spotify ruolezz

6:18 PM

spotify is racist

they don't let me use it because of where i live

(same with 000webhost)

legal issues with record distribution under 60s media law i think

Grooveshark?

i didn't get an immediate region block message, so that's good

it has monstercat! YEEEAAAH

too bad i already downloaded the entire monstercat label

hi @JoeZ. :D

hey

as for 000webhost, I can't imagine using a free web host with that name

ikr

(although I have known people who have, and I'm a bit of a hypocrite, I used one called "0000free" in the past)

6:26 PM

so instead, i smuggled a vps from a shady man named Gertrude ;.;

lol, I just bought one from OVH

specs?

$2.99/month for 10 GB disk space

and, uh

that's pretty decent

ovh.com/ca/en/vps/vps-classic.xml <- here, I'll just link you

6:27 PM

i've got an openvz one with supposedly 4gb ram, 250gb space and a gbps uplink

ssd-cached

for me it's $3.38 since they charge sales tax here in Canada, but that's alright

250 GB space?

how much are you paying for it?

quite a bit more expensive than those though

$20/mo iirc

you could get a dedi with that sort of mon--oh

yeah, can't do dedi for that

there is a place actually,

Hey, JOe

6:29 PM

that sells 4gb dedis for $25

although OVH does have dedicated servers for $59.99/month

hey, doctor

but they're ALWAYS out of stock

Still stuck on 17 of Quetzalcoatl

haha, so is everyone it seems

i'm stuck on 4

._.

6:30 PM

thank god too, since I'm not done 19 yet

(making it, that is)

also, why are none of you on IRC anymore :<

wait

u haz irc channel?

o:

irc.badnik.net #toqtz

oooh

I was on other thingses

I'm stuck on 5. I think I know what the answer is supposed to be, I just don't know how to find it...

6:31 PM

okay, done with work... now I'm gonna give that steampunk challenge a go and then I'll get back to working on the terrain

what is this 'work' of which you speak

o:

let's not talk about such pesky things in this beautiful chat room

oh... hey... this isnt IRC!

Trimsty, you haven't joined the channel yet...?

you can STAR!

6:35 PM

@JoeZ. badnik scares me ;_; gimme a min :P

I've always wanted to try my hand at making an web-based chat program

they're not very hard, my friend made one for a web game in a few hours

it would take me at least a day ;_;

I'm trying to do something in Flask at the moment

just use irc and make a frontend for it

Nah, there's no fun in that. Also, you can't implement stuff like starring messages.

6:36 PM

or just copy SE's chat

I don't think they're open-source :P

or get something from github

could, but probably doesn't exist

OMG java is so slow

or maybe its Safari..

@TheDoctor Java has a pretty good code speed to development time ratio

6:38 PM

not really.

but i'm trying to run a browser applet

not develop

hehe

@JoeZ. i'm on the irc now ;3

saw you

7:50 PM

ls -l

okay, steampunk challenge sorted... that took longer than expected.

it's fun though, I can really recommend giving it a try :)

:?

wut?

nothing

8:31 PM

/me thinks it's quiet in here

A: Proposed Question Sandbox - Mark XIII

I've rewritten a fair bit of my KotH draft and dropped the entire multiple-bots-communicating component in order to simplify it. There are still things that need to be decided (and polished), and I'd be happy about all kinds of feedback, so if you can be bothered to fight your way through that wall-of-text...

0

Conquer the Solar System king-of-the-hill Here is an outline for a strategical King-of-the-Hill challenge which is loosely based on Risk (loosely enough so as not to preclude a future Classic Risk KotH). It is far from complete, but I'm posting it here to gather feedback from the community to ...

oh, not IRC

8:54 PM

hey guys

hiho

if there's a real offensive comment on a deleted post (deleted as in high-rep users can still see) should a flag it for mod attention?

*really offensive; *should I flag it

the "as***le" comment on the 1337-challenge?

I already flagged it

yeah, okay thanks

that's still a valid question though, what's @Doorknob's opinion here?

also, you know you can edit your chat messages, by pressing up (for 2 minutes or so)

8:59 PM

wow, didn't know that, thanks a lot (testing)

hi ace. you changed your avatar

Yes @TheDoctor. In fact I've changed it for half a month now

oh

It's the same rabbit though, just different perspective

which 1337 challenge?

9:06 PM

some guy posted a challenge to transliterate leet-speak

but the challenge had a few problems (mainly being underspecified and lacking a winning criterion)... people pointed that out and he got a downvote or two, and couldn't deal with it and started to flame people... doesn't matter...

yay, my koth proposal got an upvote, that's new! :D

probably me

sweet, thanks!

any comments?

Yeah, I'll look at those comments on the 1337 post when I get home (on mobile now).

pls link to this post... i cant find it