The Nineteenth Byte

12:47 AM

@xnor I think you weren't online earlier... I'm planning to throw out a huge bounty as a test run for your meticulous-python-golfing proposal. I'm taking suggestions for the question to bounty.

xnor

@MartinBüttner oh, thanks for that

Martin Büttner

(I'm thinking Optimizer's Chinese Shrines, or my Scale up a Diamond Tiling... I'd also like Chinese Checkerboards, but I can't bounty that before tomorrow night or Monday morning, and I want to post the bounty earlier)

xnor

I had been thinking of putting a bounty on De Bruijn sequences, though not having attempted it, I don't know how hard it is

Martin Büttner

ah okay

xnor

i'm personally actually not as much of a fan of optimizing ASCII arts in Python because of how cumbersome its string manipulation is, but that's just me -- perhaps that's just the reason to put a bounty to encourage it

Martin Büttner

12:50 AM

Sp3000 suggested ascii art with symmetries might be a better testing ground because it allows for more different approaches

xnor

i see

that definitely is true

Martin Büttner

also a conceptually simpler challenge might help in getting more people to participate in the brawl :D

xnor

the magic thing would be something that's both conceptually simple and have a lot of potential approaches to optimize, and it's not clear when an optimum is reached, but that's the holy grail of code gold questions

Martin Büttner

indeed

if you can think of/find anything, let me know

xnor

I think the options you listed are good

i'm also a fan of Sp3000's suggestion of Fifteen Puzzle

i have a preference against chinese checkerboard because i've already looked at sp3000's solution, which gives me an advantage

actually, do you know how you'd deal with improvements on posts?

Martin Büttner

12:59 AM

what do you mean? if someone takes your post, golfs out two bytes and posts it as a separate answer?

xnor

yes, or just posts a comment

Martin Büttner

well if he posts a comment, you'd edit it into your post and compete with the shorter score

if someone reposts your answer with a trivial golf (and I notice) I'd probably award it to you instead of them

in general, I'd hope for people's sportsmanship

xnor

ok, that sounds reasonable

Martin Büttner

if, say, you and Sp find the shortest solution collaboratively in chat, you let me know and we'll sort something out

if you're worried about people stealing your ideas, don't post until just before the bounty ends

xnor

i wasn't worried about people being mean and stealing ideas, but being nice and giving unsolicited improvements :-)

Martin Büttner

1:04 AM

I think if they're doing that they're well aware that they're increasing your chances of getting the bounty

@xnor Out of chinese shrines and diamond tiling, which one would you prefer?

xnor

just from first impressions (having tried neither), I'd prefer diamond tiling

the scaling up thing is novel, so there's more room to search for new approaches

randomra

Morning challenge idea: A dove/hawk (rabbit/wolf) style evolution simulation where you control the animals with all having only individual memory. Your goal is to keep the ecosystem stable/alive.

This could be either a single player game or a multiplayer cooperative KotH (players all get the same score in a match). Both version would be with a controller.

Martin Büttner

hmmm... do I post a weird 450 rep bounty to get my screenshot, or do I post a 500 rep bounty and hope that I don't get more than +50 rep overnight :D

xnor

1:19 AM

wow, you have almost as much rep as the next two users combined

Martin Büttner

the next milestone is 2*Doorknob :P

randomra

@MartinBüttner Do you have a collection of nice rep-numbers?

PhiNotPi

My next milestone is 8,604.

Martin Büttner

yes

randomra

@PhiNotPi you will need some downvotes if you want a screenshot :)

Martin Büttner

1:23 AM

got the screenshot... sorry for the weird bounty... I couldn't take the risk :D

@Sp3000 bounty posted on diamond tilings

PhiNotPi

You're at 44,449

Martin Büttner

it's okay, I've got my screenshot ;)

also, I just noticed that I missed 33,333 apparently... could have sworn I got all of those.

apparently not

PhiNotPi

Time for a 11,116 rep bounty.

Martin Büttner

;)

PhiNotPi

I wouldn't mind if you put a couple grand of that on some of my questions.

Martin Büttner

1:28 AM

I've got 5,555 6,666 8,888 11,111 22,222 44,444 as well as all multiples of 10k (and also 24k for some reason)

@randomra btw, I'd love to see a Retina answer to that diamond challenge as well ;)

PhiNotPi

There you go.

randomra

@MartinBüttner will check it out, I'm looking for good Retina tasks, the Chinese Checkers wasn't too interesting, concatenating lines in reverse order was the most interesting part

@PhiNotPi Martin hasn't collected much badge since...

PhiNotPi

I chopped off some badges for extra realism.

Martin Büttner

not enough I'm afraid ;)

PhiNotPi

Courtesy of Chrome's "inspect element"

randomra

1:36 AM

@PhiNotPi opps, I just assumed you didn't, I remembered 13 gold, but apparently Martin collects them quickly

PhiNotPi

One time I made it look like a Wanted poster.

Martin Büttner

@randomra Lab Rat Race finally hit 10k views a week ago or so, and then there was Voronoi.

anyway, I should get some sleep.

PhiNotPi

Are you sure about that last statement?

Sleep tight, don't let the Pyth bugs bite.

randomra

@PhiNotPi I like the voting koth, thought about it, and here is why I haven't felt like writing a code (just to give some feedback):
ExpectantBot is optimal if you assume bots vote randomly, any improvement would assume some kind of strategy from enemy bots. If most bot would be "strong" you could do this without checking the bots but currently the "random logic" bots outnumber those and you would have to know what they do (which is slightly against the rules) to make a considerably better bot than ExpectantBot.

@MartinBüttner I'm already up :P

PhiNotPi

@randomra Since it does appear that most bots are effectively random, it is going to be pretty hard to beat ExpectantBot.

I think Strategic Voting is starting to wind down. There were no answer updates today (except for some trivial things which didn't actually affect the code).

I'm looking forward to Catch-the-Cat.

randomra

1:50 AM

@PhiNotPi you can beat Expectant by adding a little more chance for bigger total payoff votes, but finding the proper amount/function is just trial&error as it depends on the rest of the bots

feersum

2:32 AM

can anyone explain the shape of the board in catch cat?

I don't know what a toroidal hexagon looks like

randomra

3:22 AM

@feersum haven't read the specs, but with even number of rows&columns you can fold it into a torus just like a normal squared grid

randomra

5:01 AM

I just fully golfed the diamond tiling with a fixed scale-up of 2... :'(

and now I realized scale size is input too...

xnor

5:15 AM

awww

Sparr

5:27 AM

@feersum toroidal just means the edges connect. go off the top to reach the bottom, off the left to reach the right.

feersum

5:44 AM

@Sparr but with a hex grid, there are not simply 4 sides?

it would probably clear if it said what the coordiante system is

xnor

this does, by the way, give a topological torus

randomra

6:00 AM

@isaacg you won

0

A: Can numeric input/output be in unary?

Yes, but only if that is a standard representation in your language Receiving input in a format your language is not prepared for makes input conversion major part of the challenge but some other language might take advantage of the specific representation. I think this answer could be applied ...

if you want to merge some stuff, do it, then I will delete mine

or keep/edit mine if it makes sense to you

isaacg

@randomra I've incorporated some of your ideas into my post.

Sparr

6:44 AM

@feersum I expect the map to appear rectangular, so that two of the directions extend the same distance.

a hexagon shaped map made of hexagonal cells with edges that wrap is... unusual and not an easy shape to visualize.

New Sandboxed Posts

8:09 AM

0

A: Sandbox for Proposed Challenges

Graph planarity code-golf Given an graph, determine whether it is planar, meaning that it can be drawn in the plane without edges intersecting. See: planarity testing. Input: An undirected graph without self-loops in whatever format you find convenient (adjacency matrix, list of edges, bit-stri...

randomra

8:36 AM

^ I wonder how large grid you need so that you can put the points of a planar graph to integer coordinates on it (in a way that the layout would be planar).

Unihedron

Hi PPCG! I just got the yearling badge!

Sp3000

Hi Square!

Actually.... that's not quite accurate. What would a unihedron be?

New Sandboxed Posts

0

A: Sandbox for Proposed Challenges

Hook length product code-golf A Young diagram is a collection of square boxes in left-justified rows and top-justified columns. For each box, all the spaces above it and to its left are occupied. XXXXX XXX XXX X The hook length of a box is the number of boxes in its hook, which consists of it...

Sp3000

8:51 AM

@NewSandboxedPosts Hook length looks like a nice simple golf

xnor

yes, it's really a lot more straightforward than all the background implies

Sp3000

Oh, you're accepting unary counts. Interesting... I wonder if that makes it any easier for general languages

randomra

9:06 AM

@Sp3000 for array based languages I would think so

definitely helps for J

Unihedron

in Maid Café (メイド喫茶), Apr 26 at 16:16, by Unihedron

I'm just as real, except in a dimension of my existence, you all don't exist.

Sp3000

D:

Unihedron

// backstory: chat.stackoverflow.com/transcript/message/…

Suig

9:31 AM

I have an announcement to make.

My name is now appropriate.

Optimizer

10:24 AM

in*

Optimizer

11:04 AM

@randomra how much in J ?

randomra

11:15 AM

@Optimizer 26

Optimizer

fully golfed ?

randomra

mostly golfed, but I haven't checked other methods

how long in CJam?

Optimizer

I have 2 23 . 1 takes unary, 1 takes row lengths

randomra

then I'm fine with 26, will check the row lengths method later

but the unary seems to suit well J

Optimizer

hmm

Sp3000

11:29 AM

I've been stuck on 216 for diamond for the past 3 hours :(

Martin Büttner

wwxd

Optimizer

hgjd

feersum

Ascii art isn't my favorite, but python-only challenges are

Optimizer

is it the "speak a random thing" day ?

Martin Büttner

@xnor ha, yet another diamond tiling related challenge ;)

Optimizer

11:38 AM

hook length is diamond tiling related ?

xnor

@MartinBüttner where?

Martin Büttner

well, diamond tilings of hexagons are essentially isometric renderings of 3D young diagrams

xnor

ah, right

New Sandboxed Posts

11:56 AM

0

A: Sandbox for Proposed Challenges

Flash Cards Viewer code-golf In this challenge, your goal is to produce a flash card viewer. Since this is a Code Golf challenge, it won't be a particularly featureful flash card viewer - in fact, it'll be the bare minimum a flash card viewer can be. We'll implement the following features: Re...

randomra

has someone done a Java-wrapper for Python KotH before? or was it just the other way around?

Martin Büttner

I only know about arbitrary-language wrappers for Java KotHs

randomra

:(

Martin Büttner

@Katya eeeehm, coincidence?

randomra

I would like to write my KotH in Python, but I don't think there would be many entry in only Python

Optimizer

12:06 PM

@MartinBüttner its not. My jenga KOTH idea on chat here was also converted to sandbox by someone else..

@randomra more than cjam

Katya

@MartinBüttner Maybe............ ;-)

randomra

@Optimizer that would be more PvP than KotH

Optimizer

@randomra who knows. maybe some format can be made .. (or CJam gets stream support)

randomra

Pyjnius seems to be doing what I need

Martin Büttner

Python + Java seems odd though... at that point you might as well just do the KotH via stdin/stdout and allow all languages

randomra

12:11 PM

@MartinBüttner programs shouldn't keep memory between round

Optimizer

can be done

randomra

and launching the program at every turn takes a lot of time (like 0.1 sec/launch)

@Optimizer what do you mean?

Optimizer

@randomra you answered it.

randomra

:/

or trust in programs/submitters a lot...

^ maybe I will go in that direction

and I can do a test run where I check if they answer only based on the stdin

Optimizer

wow. 50mbps connection feels so luxurious

Sp3000

12:25 PM

Just golfed isaac's down 3 bytes, so now I need to take off another 2 :P

Martin Büttner

12:52 PM

more Fission golfitude

@BrainSteel do you wanna have a look and see if you can beat this? ^

if ? would continue to work after returning EOF, I'd have 13 I think :/

anyone know a tool to create a gif where each frame is just some ASCII text?

like the ><> animations @Sp3000 includes occasionally

Optimizer

he would know, right ?

Martin Büttner

I think his method was specific to ><>

(like included with the interpreter or something)

@Sp3000 I think you'll need a bit more than 2 bytes now ;)

Sp3000

Hah, I'm working on it :P

Also, my method for ><> gifs was frame-by-frame Adobe Flash, actually

Optimizer

450 bounty is too much.

Martin Büttner

@Sp3000 D:

Optimizer

1:05 PM

specially when we have already had python-only bounties in past

but no other language-only bounties yet

Sp3000

I think Martin's disappointed :P

Martin Büttner

by what?

Optimizer

that xnor hasn't posted yet ?

Sp3000

That I didn't have some awesome tool for GIF making :P

Martin Büttner

@Sp3000 oh yeah, indeed :D

Optimizer

1:09 PM

you can screen record then convert to stop motion

Sp3000

Wait did isaac forget to update? I don't see how the new one is 195, even after spaces -> tab

Martin Büttner

@Optimizer hm yeah looking into that right now

xnor

@Optimizer i'm working on it!

isaacg

@Sp3000 Forgot to update'

Martin Büttner

1:32 PM

oh, I've got a challenge idea: Fission programs without portals are generally invariant under (toroidal) translation and rotation (since all components that depend on a direction have versions for all four directions). so the last step when golfing in Fission is to find the shortest version by minimising the number of lines and rotating as many spaces as possible to the end.

so the challenge would be, given a Fission program (just a string really) find the shortest equivalent version under correct rotations and shifts.

I might wait with posting that until Fission 2.0 though... I'm really hoping C0deh4cker publishes his changes one of these days

New Main Posts

0

Q: Supersonic domino tilings

Task Write a program that reads two integers m and n either from STDIN or as command-line arguments and prints all possible tilings of a rectangle of dimensions m × n by 2 × 1 and 1 × 2 dominos. Dominos of an individual tiling have to be represented by two dashes (-) for 2 × 1 and two vertical ...

Dennis

Good morning everybody!

isaacg

Ni hao!

PhiNotPi

'Allo!

BrainSteel

@MartinBüttner That sounds like a neat challenge! Though, it seems like the kind of thing that would attract 10 bruteforced solutions and 1 or 2 clever ones. That's not necessarily bad :) And I'll try to shorten that fission code, but I can't promise anything :)

Howdy!

Optimizer

1:49 PM

@Dennis so the product of m and n is always less than 65 ?

Dennis

Yes, I wanted compact representations to fit in 64-bit registers.

Optimizer

ok

wow, is 70 bytes too long ?

Dennis

For what?

Optimizer

yours

Dennis

The fastest-code challenge? Go nuts.

Optimizer

1:51 PM

oh wait. Its fastest code !!

2

damn, 10 minutes of my life wasted

BrainSteel

"... free (as in beer) ..." Ah, the best kind of free.

Optimizer

@Dennis the 1 hour limit is crazy big then. my cjam 70 byte solution can do 1 hr for the 64 test case

Dennis

I didn't want to leave solutions unscored if I could avoid it.

And processing time is cheap. :P

Optimizer

So even the native bruteforce approach is not an invalid entry then :P

Dennis

Wanna post the very first fastest-code CJam submission?

Optimizer

1:55 PM

it won't win

Dennis

So? You used to answer code-golf question using JavaScript.

:P

Optimizer

and I switched to CJam.

will not make a similar mistake :P

Dennis

Good point.

Optimizer

though, let me run it for some 50+ numbers on the java version

isaacg

2:08 PM

1 less character on diamond.

There's so many more opportunities for golfs in a non-golfing language. It's a lot of fun.

New Main Posts

2:56 PM

2

Q: Cryptographic hash golf

This challenge is a search for the shortest implementation of a hash function that is collision resistant, i.e., it should be infeasible to find two different messages with the same hash. As a cop, you try to invent and implement a hash function finding the best compromise between code size and ...

code-golf cryptography cops-and-robbers hashing

1

Q: Cryptographic hash golf (robbers)

Companion thread of Cryptographic hash golf As a reminder, here are the rules for robbers from the main challenge: Task Crack any of the cops' submissions by posting the following in the robbers' thread]: two messages M and N in I such that H(M) = H(N) and M ≠ N. Scoring C...

code-golf cryptography cops-and-robbers hashing

New Sandboxed Posts

3:09 PM

0

A: Sandbox for Proposed Challenges

Output every halting program The goal of this challenge is to (eventually) output every possible halting program in a language of your choice. At first this may sound impossible, but you can accomplish this with a very careful choice of execution order. Below is an ASCII diagram to illustrate ...

Sp3000

3:24 PM

Man, CPython takes about 10 seconds just to calculate the number 2^(2^30)

Martin Büttner

@BrainSteel got it down to 12

(same idea, different layout)

:21934506 input is in range 2^2^30, but the time limit is on 2^2^20

Sp3000

Ah right, that's why I'm getting confused :P

BrainSteel

@MartinBüttner I tried something totally different, and it's currently 20 bytes. I thought I had a good idea but it got out of hand. I'm submitting it separately because I can't shake this feeling that it's golfable and we need more Fission in PPCG :)

Martin Büttner

:)

BrainSteel

I think the first line of it is pretty neat, but the rest of it is... ehh...

I finally found use for J and teleporters though!

Martin Büttner

3:29 PM

nice :)

I had a neat idea for using 4 teleporters the other day, but it wasn't worth it

Sp3000

Does CarpetPython's submission invalidate the no builtins rule?

BrainSteel

I'm starting to get the feeling that they make the code cool (see: harder to read) but they rarely really improve byte count.

PhiNotPi

Does seeding the random number generator involve a hash? I think it would.

Martin Büttner

@BrainSteel I was able to save a lot with teleporters in Fissile Numbers

at some point it gets hard to route your atom paths around each other

also, when you're not golfing, they're useful for modularising your code

Raphael

hi.. anyone mathematically inclined about?

Martin Büttner

3:31 PM

just have separate blocks that perform a task each and join their inputs and outputs with teleporters

BrainSteel

Oooohhh, yeah, I forgot about that. Yeah, if you construct a bunch of independent parts that work, that is super useful. Probably not so much in the little golfs.

@Raphael What's your question?

Lembik

hi

trichoplax

Hi

Martin Büttner

@BrainSteel posted the 12 byte version

Lembik

I want to calculate the probability that the inner product of two vectors of length 10 is 0 by enumeration (that is not using math :) ). In the first vector the prob that each element is -1 is 1/2 and the prob that it is 1 is 1/2

BrainSteel

3:38 PM

Oh, that's cool! Very compact.

Lembik

In the second vector the elements are -1,0,1 with prob 1/4, 1/2, 1/4

I can do this easily if the probs had been 1/3,1/3,1/3 but how do I do it when the probs are not equal?

I am sorry if this is really obvious :(

BrainSteel

That is a good question.

Hmmm...

Lembik

maybe I should set it as a code-golf challenge :)

PhiNotPi

I just cracked CarpetPython's answer, I think.

Lembik

but then I wouldn't understand the answer :)

BrainSteel

3:42 PM

It's a good thing we have a somewhat established precedent for writing explanations for the things we do :D

trichoplax

@Lembik Do you mean by a method that doesn't involve evaluating all the possibilities and their respective probabilities?

Lembik

@BrainSteel :)

@trichoplax evaluating them all is fine.. How exactly do you do that?

when the probs are not equal

BrainSteel

There's a lot of possibilities.

Lembik

@trichoplax I don't think this is higher math :) I am just confused

there are 3^10*2^10

so not too many

BrainSteel

Well, more than is doable by hand. And doing that feels like cheating.

Lembik

3:43 PM

60466176

sorry I meant write code to do it

trichoplax

@Lembik If you just want to brute force it, you just go through each possibility and add its probability if it meets your criterion

Lembik

this is programming question

BrainSteel

There's got to be a more elegant solution though, right?

trichoplax

I would hope so :)

PhiNotPi

The following number should be a collision with CarpetPython's example of 5.

Lembik

3:45 PM

@trichoplax so iterate over all 60466176 possibilities and for each where the inner product is 0, calculate the probability of getting that number of 0's in the second vector ?

and add some tiny probability to some sum?

PhiNotPi

pastebin.com/h6Nzs0Xr

Lembik

@PhiNotPi what does that relate to?

trichoplax

@Lembik I don't understand why you say "calculate the probability of getting that number of 0's in the second vector"

PhiNotPi

The hash-golfing Cops and Robbers.

Lembik

@trichoplax what do I add each time I get 0 inner product?

@PhiNotPi ok :)

trichoplax

3:46 PM

@Lembik The probability of that combination of vectors being chosen

Lembik

@trichoplax right but how do I calculate that?

if someone could show me some code that might make my life easier :)

trichoplax

If A.B = 0, calculate p(AB) = p(A)p(B) since they are independent.

Lembik

what is p(B) though?

where B is the one with -1,0,1 values with uneven probs

BrainSteel

Well, if <x,y> = A1 + A2 + A3 + ... + A10 you can find the probability of each AN being a certain number, and then find the probability of the sum being 0.

trichoplax

the probability of choosing B from all the possible -1,0,1 vectors

Lembik

3:48 PM

@trichoplax right but explicitly.. say I have got to 1,1,1,1,1,0,0,0,0,0 for example in my exhaustive enumeration

what is that prob?

trichoplax

p(1,1,1,1,1,0,0,0,0,0)=p(1)*p(1)*p(1)*p(1)*p(1)*p(0)*p(0)*p(0)*p(0)*p(0)

Lembik

is it (1/4)^5 * (1/2)*5 ?

trichoplax

I believe so

It's a long time since I used probability though

Lembik

ok so that is the part where I said you had to calculate the prob of having that many 0's

trichoplax

Provided they are chosen independently

Lembik

3:50 PM

they are :)

trichoplax

Ah OK - but it's not just whether they are zero - you could have just 1s and -1s and still get a zero result, right?

Lembik

yes!

that is 100% true

so basically it seems I need to do some calculation for every pair of vectors I iterate over and then add some potentially tiny probability to a sum

it seems a fragile method :) Say, I increase the size of the vectors to 15 for example

trichoplax

Yes. That's brute force. If you can think of any other way it will almost certainly be much faster

Lembik

thanks

trichoplax

Is it worth stepping back a level in case there's an easier way of approaching this? What is the context?

Lembik

3:56 PM

@trichoplax I would really like to extend codegolf.stackexchange.com/questions/42299/… to cases where the probs are not equal

but I feel the PPCG community may be bored of too many similar questions :)

so I thought I would do it myself

is @Jakube still around these days?

Sp3000

Jakube was just in yesterday

Jakube

4:14 PM

hi

Agawa001

4:45 PM

back from the dead

Martin Büttner

@Optimizer Dennis used fff o.O

@BrainSteel nice work with the 15 byte version. the end of the explanation isn't entirely correct though: the empty K negates the energy, leaving it at +1, such that the ~ gives 0, not -2.

BrainSteel

5:05 PM

It does?! I swear, this language has some arbitrary rules. I'll fix that :D Thanks!

Agawa001

they gave me more reputation for bein foul mouthed , thats generous

Martin Büttner

5:26 PM

@BrainSteel yeah I have no clue how he came up with some of the rules, but they definitely make for an interesting language ;)

BrainSteel

Oh, it's tons of fun to try to do things with it. There are some things I feel are sorely needed, though. A general energy negation operator would be awesome.

Sp3000

Ahaha Rule 0, nice

Martin Büttner

@BrainSteel lol yeah

I made a list of component ideas this morning

that was number one :D

@Sp3000 oh he actually uses Mod[#^2,256] so it's even every multiple of 16 that collides

BrainSteel

5:45 PM

Haha, I'd like to see that list!

Sp3000

I'm sure that even if the rule was fixed, you could probably have a long initial board which results in the same as a short one when cut down

Martin Büttner

@BrainSteel:

- Swap sign of energy
- Copy energy to mass
- Copy mass to energy
- Read and write integers to and from energy
- Repurpose lower case letters. Two options:
1. Keep functionality, but change values to `0` to `25`.
2. Ditch them entirely to make room for 26 new components. In this case, one of them could be `'` with an offset of `-32`.
- Delay component.
- Debug component.

@Sp3000 yeah there are a lot of other collisions as well. if he had used a more elaborate hash to obtain rule, initial condition and the number of steps, it would have been much harder.

another solution is to take two multiples of 99 that don't differ in the first 128 bits

then the rule doesn't matter

@BrainSteel I've patched a debug component into my interpreter already. It was absolutely necessary for Fissile Numbers.

Sp3000

Hmm in hindsight mine's pretty easy too. Will rethink in the morning :/

Martin Büttner

whenever an atom passes a `, I get some output like Atom at (1,1): [116,0]. Direction: ^

BrainSteel

Those would all be so useful. I approve this list :D

Yeah, there's only so much -t can do.

I'm a fan of ditching the a-z and making room for 26 components.

Martin Büttner

5:54 PM

yeah, I rarely use that

@BrainSteel I'll suggest all of those to him, but I want to see his update first :/

BrainSteel

For the two golfs I've done, -t did an okay job pointing out where I screwed up (which is frequent)

Is he still actively developing it?

Martin Büttner

> I actually didn't think anyone was using Fission so I never bothered completing my changes. Since I now know that at least one person is using it, I'll go back and finish making changes (and fix this bug you reported, thanks!).

> I have had a massive set of local changes for a long time that I never pushed as they weren't finished. They somewhat change the way the language works so I was hesitant to do so. I think that I'll push those in a new branch while I finish working on them and let people see what they think of the new Fission. It modifies some components and adds others.

BrainSteel

Oh, great! Hopefully a few from the list make it :)

Optimizer

6:19 PM

@MartinBüttner I stopped golfing that after looking at Pyth :P

Is it good time to post another AAoD ? Or should I wait for monday prime time ?

Martin Büttner

no clue

Optimizer

6:49 PM

you know what ?

This can probably be golfed to around 35 bytes. I am trying to figure out how .. — Optimizer May 27 at 12:37