The Nineteenth Byte

PhiNotPi

6:00 PM

@githubphagocyte Yes, that does make sense as an ambush tactic.

Martin Büttner

@githubphagocyte pentagons are beautiful anyway ;)

PhiNotPi

The pentagon does look really nice.

Martin Büttner

φ all over the place :D

githubphagocyte

@MartinBüttner yes I love the hints at pentagrams inside too.

@MartinBüttner you can write the letter phi in here?? Unicode? It's interesting that the OP phrased it as an integer maths challenge, but it's gone far beyond simple integer functions...

although technically I suppose they are all still integer functions really

Martin Büttner

yes Unicode... I modified my keyboard layout a while ago and put a whole bunch of greek letters and mathematical symbols on Alt Gr and Alt Gr + Shift

githubphagocyte

6:06 PM

thanks @PhiNotPi :)

I need to go and stretch my legs before the next tweetable image...

Martin Büttner

haha, what's coming up? :)

githubphagocyte

@MartinBüttner I'm torn. I have a few ideas but none that I'm confident are possible, so whatever I choose will be a gamble...

I still haven't completely given up on buddabrot

I'm also wondering about a maze

maybe pacman...

Martin Büttner

you might be able to do the buddhabrot by sacrificing two channels

doesn't have to be a nebulabrot ;)

githubphagocyte

I'm also trying to work out if I can use the colour functions as functions of 4 variables by splitting the int inputs into two char inputs each

Martin Büttner

probably

githubphagocyte

6:09 PM

I'm not sure if the gain in flexibility is worth the loss of space...

Martin Büttner

I think you might be able to do something like my coordinate mapping with that

githubphagocyte

I think it would only cost a few characters to join and split them, but a few out of 140 is a lot

Martin Büttner

(the one for that circular thing I posted in chat yesterday)

githubphagocyte

The circular one looked great

If we can get projections then simple repeating patterns can be enough for something fascinating

Trimsty

sigh

but seriously, how do you view output? D:

githubphagocyte

6:12 PM

good news @Trimsty?

I used the GNU Image Manipulation Tool

Trimsty

@githubphagocyte what format is it in? :P

githubphagocyte

It is happy with ppm files, even if they don't have the ppm extension

PhiNotPi

how about a dragon curve? en.wikipedia.org/wiki/Dragon_curve

githubphagocyte

However, I couldn't get 1024 levels per colour component to work, so I used the alternative one with only 256 levels per channel

Trimsty

ahk

githubphagocyte

6:14 PM

Then I opened the file in GIMP and saved as png before uploading

Most will get converted to jpg and lose some detail on uploading, unless there is a lot of flat colour

Martin Büttner

imagemagick can do the conversion on the command line very simply

then you can just write a little script that does "compile, run, convert"

githubphagocyte

I might have to look into imagemagick then...

Yes the dragon curve is definitely achievable. If there's space left you could even look at portraying it in relief like the examples on wikipedia

I'd love to see some other L systems, maybe ferns or trees

@Sparr any luck with the ray tracing?

PhiNotPi

Do you want to continue working on the tank battle challenge?

Trimsty

6:30 PM

I need HSL to RGB for my math art thingy

i'm scared now :P

githubphagocyte

@PhiNotPi I'm about to go out for a bit. I'm guessing it's going to be refined gradually over time though so I'll definitely be interested in questions as you develop it

PhiNotPi

Okay

Trimsty

Python has it in it's standard library.

Trimsty

6:42 PM

Okay, I wrote a script for conversion from HSL to RGB,

but it leaves me exactly 9 characters to write my actual calculation script.

chopped some off, 16 chars for the calc now

wow this sucks

Martin Büttner

6:59 PM

@Sparr seeing how much fun you had with the bracket numbers, I think you might like this question: codegolf.stackexchange.com/q/35664/8478

Sparr

I'm trying to decypher that...

VisualMelon

@MartinBüttner do you still reckon the domino circuits testbed will be ready some time tomorrow?

Martin Büttner

@Sparr yeah I had to read it very carefully, but once I got it, it did make a lot of sense... and it's a really interesting problem I think

@VisualMelon I'll be trying my best, but I can't make any guarantees... sorry... I got so distracted by the tweetable maths yesterday ^^... and today I was playing tour guide for a friend of mine.

VisualMelon

ok, I've just go to decide if I'll write tomorrow off or not ;)

given that, I'll probably not be able much tomorrow, so don't hurry for my sake

Trimsty

hey guys, what do you think of my mathematical art thingy? should I enter it?

not very complex xD

Martin Büttner

7:06 PM

looks very nice :)

Trimsty

i used up almost all the characters xD

Sparr

@Trimsty yes

githubphagocyte

If it's under 140 characters, post it!

You did that in just 9 remaining characters??

Sparr

@MartinBüttner I need to learn numpy if I'm going to keep doing math challenges, and I need to get a lot better with normal python first.

Trimsty

I scrapped the hsl thing actually,

Sparr

7:07 PM

my goal in doing these challenges at all is to get up to speed on algorithmic problems in python

Trimsty

that's why it doesn't look like I want it to

githubphagocyte

How did you get the swirls?

Trimsty

@githubphagocyte It's from an algorithm I wrote ages ago -- I really don't remember how it works xD

I chose some reference points for it, and what it does is it calculates the distance and then does some strange trig division stuff

posting the code now :P

githubphagocyte

Well post it and we can use it for inspiration for the next entry...

Martin Büttner

@Sparr numpy is fun!

Trimsty

7:10 PM

What to call it, what to call it...

Martin Büttner

I only started using it 2 weeks ago

Trimsty

0

A: Tweetable Mathematical Art

Some swirly pointy things Yes, I knew exactly what to name it. unsigned short RD(int i,int j){ return(pow(pow(abs(73.-i),2)+pow(abs(609-j),2),.5)+1)/(pow(abs(sin((pow(pow(abs(860.-i),2)+pow(abs(162-j),2),.5))/115.0)),.5)+1)/200; } unsigned short GR(int i,int j){ return(pow(pow(abs(160...

that was the fastest 3 upvotes that I have ever gotten, ever

Martin Büttner

the C framework has macros for square and square root

Trimsty

xD

Martin Büttner

I think it's only fair if you do the same

Trimsty

7:15 PM

orly?

Martin Büttner

well sqrt is built-in and there's a _sq macro

githubphagocyte

You'll actually save characters if you use them

Martin Büttner

that's what I meant

Trimsty

@MartinBüttner wait, i actually used pow(x,2) instead of _sq(x).

how stupid

githubphagocyte

And avoid anyone who might downvote for needing to modify the outer program

Martin Büttner

7:16 PM

yup

githubphagocyte

You might even be able to squeeze in something extra with the characters you save...

Trimsty

;o

Wow, that made it a lot shorter

githubphagocyte

:)

Trimsty

and it runs like 100 times faster, wow

githubphagocyte

you can also add your own macros inside one of your functions and use it in all the following ones. If Martin hadn't pointed that out I wouldn't have been able to post my second entry

_sq is implemented as multiplication rather than pow - I don't know if that affects the running speed

Trimsty

7:20 PM

@githubphagocyte i think it does, a lot o:

Sparr

the macro trick is only especially useful if you can do less work in red than in green and blue

that would work really well for something like the flag entry, not so much for most of the entries with equivalent channels

githubphagocyte

@Sparr the only way round that I can think of is to make the macros in the red function and then call the blue or green (or both) functions to give the red ouput

Trimsty

Also, uh, the fact that there are only ~4 different shades for each color in my entry is unintentional

it looks good, but i have no idea how to fix it xD

Sparr

I wish we had recursive golfing challenges

githubphagocyte

Of course calling green and blue from red wasn't even possible until yesterday when the forward declarations were added - it's got even more interesting...

Sparr

7:23 PM

I want to take the tweetable art challenge and then make a new golf challenge based on each entry for the best tweak

githubphagocyte

@Sparr how would you define how much is permitted before it's no longer a tweak?

Sparr

popularity contest, let the voters decide

I mean, maybe change half the bytes?

Trimsty

guys

i just hit 1000 rep from those 3 upvotes

:DD

<< see look yay

githubphagocyte

Nice one

Trimsty

now i can edit other people's posts

muahaha

githubphagocyte

7:28 PM

lol

Trimsty

this power that i yield
i must use it for good

Sparr

@githubphagocyte I don't think the ray tracer is going to work. I can't work out a good way to get the recursive sampling/tracing to work within the three-function framework in the challenge

githubphagocyte

with great power comes thousands of watching eyes

Trimsty

wait, you can enter using something other than c++

and python has a builtin for hsl->rgb...

Sparr

I could probably do it with a matte instead of reflective sphere, but then it's not nearly as interesting

githubphagocyte

7:33 PM

@Trimsty I think it might be frowned upon to use any libraries not also provided in the C++ framework

Trimsty

@githubphagocyte aw :cc

githubphagocyte

There are plenty of c++ features not available in this challenge too

Sparr

if C++ had nested functions I could try

there's another entry that needs that functionality, too

githubphagocyte

I'm not sure exactly where the lines lie, but I'd guess an image manipulation library would be off limits...

@Sparr is there any way around it with recursive macros?

I was disappointed to find that

#define D define

Sparr

recursive macros won't help :(

githubphagocyte

7:35 PM

doesn't work

Sparr

heh, I wonder if he would consider it cheating to define new functions below red()

that is, put "} newfunc(){ blah" somewhere "inside" red_fn

githubphagocyte

@Sparr you can define what you like as long as you stay inside the function body

Sparr

C++ injection :)

Trimsty

you could hide it by opening the function on the same line as red's body and ending it on a new line

githubphagocyte

lol @Sparr that's devious

it's not the OP you have to fool - it's every voter who passes...

that many eyes will probably show you up

Trimsty

7:36 PM

unsigned short red() {
malicious;}unsigned short injection(){code;
}

githubphagocyte

@Sparr if there really isn't a way round it with self recursion, my only suggestions are:

1. mutual recursion so that the image is monochrome but shiny

2. An iterative approach and forget recursion altogether

Sparr

@githubphagocyte mutual recursion is easy, the problem is the data types. I need to be able to define operators, and pass non-int return types.

Trimsty

wait

i wrote my code in python by accident and chopped off half of the expression?!

githubphagocyte

@Sparr I think you can simulate a 4 argument function with one of the colour functions by decomposing the 2 int arguments into 2 char arguments each. Costs characters to convert back and forth though

@Trimsty do you mean the one you posted?

Trimsty

@githubphagocyte yeah...wait, does return even need parens? :P

githubphagocyte

7:41 PM

not in c++

Trimsty

@githubphagocyte ahk, ty :P

so i didn't kill it

Sparr

@githubphagocyte the data type being used in the raytracer I'm trying to shrink is a struct of three floats

githubphagocyte

parentheses won't hurt anything except your character count

Trimsty

@githubphagocyte :D

githubphagocyte

@Sparr I'm guessing it'll need to be approached from scratch - trying to shrink an existing approach is probably too much to ask

Sparr

7:44 PM

yeah. I'm not up to the task of designing one from scratch right now

especially with these constraints

give me 420 bytes to use however I want, inside and outside of the function bodies, and I could make it work, just barely

Martin Büttner

yes... two days ago I was still hoping to finally catch up with my SO account in badges and today I've surpassed it in silver and bronze :D

githubphagocyte

The only thing I really want is reflection (to arbitrary depth). Everything else is optional. I keep trying to think of ways to make it simpler. I considered shiny cubes but I think spheres is so much simpler than anything else. Unless you have mirrored planes and matt objects between them

Sparr

an int can hold a cast pointer on most systems... :)

githubphagocyte

@MartinBüttner if your SO self finds out won't it try to pull away again?

Martin Büttner

@githubphagocyte at the moment it doesn't seem like it... badges are more fun on PPCG than on SO for some reason

maybe when I catch up in rep... but that's gonna take a while longer

(although my rep/answers ratio here is much higher than on SO)

githubphagocyte

7:48 PM

@MartinBüttner the ratio is certainly favourable here. My highest rep is here.

Martin Büttner

I went crazy on SO for the first three months and another two or three months later on... that's gonna take a while to catch up with ^^

but it earned me an invitation to their new-office party in london although I was only a user for 3 months or so :D

githubphagocyte

Wow

Martin Büttner

well it wasn't that exclusive any more after someone tweeted it :D

but I can feel good about actually getting invited ^^

Trimsty

Well, TIL that the bitwise and operator renders the Sierpinski triangle.

unsigned short RD(int i,int j){
	return (i&j)==0?100:0;
}
unsigned short GR(int i,int j){
	return RD(i,j);
}
unsigned short BL(int i,int j){
	return RD(i,j);
}

githubphagocyte

It's amazing how many different things lead to the Sierpinski triangle

Trimsty

7:55 PM

@githubphagocyte yeah, especially given how unnatural it seems

githubphagocyte

Great to have a pattern in so few characters - means you can add it as a backdrop to anything that has that many characters free...

Trimsty

@githubphagocyte indeed ;D

Martin Büttner

so I'm currently running the quadratic residues challenge on 24 bits... and I'm monitoring the memory footprint so I can kill it before bad things happen (Mathematica is horrible with system resource, it can easily hijack your entire machine if you're not careful)

githubphagocyte

@MartinBüttner can you specify an upper limit on its memory or do you just have to catch it in time?

@Trimsty do you have enough characters left to put that sierpinski triangle into perspective? So you can see it stretching off into the distance?

Martin Büttner

I've been looking at this for a while, and now I've figured out that each of the small peaks is probably determining the "score" for a base (when I'm sorting the bases by efficiency)... and then the slow peak is the one where I'm actually removing the numbers caught by the best base. so now that allows me to figure out how far the program is just from looking at the memory usage pattern :D

I've got two bases now... if the cover size doesn't blow up for larger numbers of bits, this may actually complete in a reasonable amount of time

githubphagocyte

8:01 PM

beats a progress bar...

Trimsty

@MartinBüttner you would be great at cryptanalysis

Martin Büttner

@githubphagocyte you'd have to think about a progress bar before starting the computation... that really makes it less useful :D

when I'm running these huge mathematica computations I often only notice that I'd like to know how far I am by the time it's done the computation for half an hour or so :D

Trimsty

@githubphagocyte well i only used 45. xD

Martin Büttner

in mathematica it shouldn't even be a problem to abort and pick up where you left of with minimal changes to the code... but I'd rather not risk it

@githubphagocyte that might be possible, but again, I'd have to remember that before starting the computation :D

githubphagocyte

"Write a program in 140 bytes that solves the halting problem for arbitrary input. Include a progress bar."

Martin Büttner

8:03 PM

:D

there was once this link on hacker news about a job offer where someone basically wanted to outsource the halting problem

I don't remember if it was a joke or someone being stupid

but the responses were hilarious

githubphagocyte

lol

Martin Büttner

half of them were obviously auto-generated replies

then there was a ridiculous number of replies which seemed to be handwritten, where people were like "we're the right people for the job. we'll do this and deliver. blah blah blah"

and then there were the funny ones (references to the amber room etc :D)

Trimsty

"My name is Fermi. It's almost 'fear me'."

i just looked on his Enrico Fermi's wikipedia page and saw this

"Alma mater Scuola Normale Superiore"

really? Superior Normal School?

githubphagocyte

supernormal?

I'm lost...

extraordinary?

Trimsty

@githubphagocyte xD

We're not just extraordinary.

We're insanely ordinary.

githubphagocyte

8:08 PM

lol

@MartinBüttner I wonder if anyone actually ended up with the job.

Sparr

@Trimsty the number of ways to render the sierpinski triangle is staggering

Trimsty

@githubphagocyte Unfortunately, they'll never finish their work.

HA

@Sparr indeed

githubphagocyte

Unfortunately for who...? They'll keep getting paid forever

Sparr

@Trimsty there are half a dozen elementary cellular automata that render variations of the sierpinski triangle. lots of them come out of my entry to that challenge :)

Trimsty

@githubphagocyte that's one way to look at it ;3

githubphagocyte

8:11 PM

You could have a sub-competition within the question to see who can post the answer with the highest number of apparently unrelated sierpinski triangle approaches incorporated into one image

Trimsty

@Sparr it's so weird how other elements like the Von Neumann neighborhood works with it

githubphagocyte

I'm trying to work out how to get reflecting spheres in 140 characters per channel. Have you ever seen the reflecting spheres sierpinski triangle? Three spheres touching and one behind them.

Sparr

so the gap and the reflections make a nested triangle pattern?

githubphagocyte

indeed.

it looks wonderful

I'd only ever seen it in a ray traced example, with the reflectivity set unnaturally high to emphasise the effect. However, I've just searched for "reflecting spheres sierpinski" and there are now plenty of real photographs showing the effect too

If someone can get reflecting spheres to work in 140 bytes then the sierpinski triangle effect with be a free bonus...

Trimsty

I was playing some more with the tweet art challenge,

and i accidentally rendered sheet music...

githubphagocyte

8:20 PM

????

Is it under 140 bytes?

Trimsty

@githubphagocyte quite far

githubphagocyte

will you post it?

Trimsty

i think so

i mean, it looks exactly like sheet music o:

githubphagocyte

It reminds me of the punched paper you feed into a music box

PhiNotPi

As someone who knows how to read sheet music, that is definitely not sheet music.

githubphagocyte

8:23 PM

lol

@PhiNotPi you may have to squint somewhat

PhiNotPi

It does kinda look like the punched paper for music boxes.

githubphagocyte

It does have what looks like crotchets dotted around too

Trimsty

and it has connecting bars o:

0

A: Tweetable Mathematical Art

Sheet Music Sierpinski music. :D The guys on chat say it looks more like the punched paper for music boxes. unsigned short RD(int i,int j){ return ((int)(100*sin((i+400)*(j+100)/11115)))&i; } unsigned short GR(int i,int j){ return RD(i,j); } unsigned short BL(int i,int j){ return ...

and measure dividers

PhiNotPi

This is music box paper: encrypted-tbn0.gstatic.com/…

Trimsty

@PhiNotPi yeah, i dunno why it's so low-definition :/

PhiNotPi

8:29 PM

hi-res

Trimsty

@PhiNotPi ty :P i meant my render :D

it also does that on the other one, like, is it from short's rounding errors or...?

PhiNotPi

I think I heard something about SO automatically compressing large images.

githubphagocyte

yes, the upper limit on image size is 500KB

anything above that gets put through the jpg scrambler

Trimsty

yeah, it's also bad offline

i found it though,

the 11115 in my code, i turned it into a float, and this is the result

githubphagocyte

handy to be know you can truncate to integer without needing to use an explicit cast

Trimsty

8:33 PM

i think I like the low-quality one better xD

on the other hand, I can now verify my "Sierpinski music" claim

githubphagocyte

You can put both in the post so that it has some context

PhiNotPi

I guess the "sheet music"-look was a result of rounding.

Trimsty

@githubphagocyte good idea ;o

edited!

PhiNotPi

By the way, I liked your "swirly pointy things" a lot.

Trimsty

@PhiNotPi thanks ;D

that would explain the constant stream of upvotes

PhiNotPi

8:46 PM

in your code for that answer, why are you using _sq(abs(844.-i)) instead of _sq(844.-i)? (or am I completely wrong?)

githubphagocyte

sounds like 5 characters saved to me

Trimsty

@PhiNotPi i have no idea

githubphagocyte

half way to being able to have room to add a sierpinski backdrop...

PhiNotPi

more than halfway, since he used that 4 times in each function.

githubphagocyte

room for backdrops in 2 different shades per stream then...

Trimsty

8:50 PM

edited 'em out, thanks ;3

Trimsty

9:21 PM

hey guys, how's this one? it's based off of the swirly stuff

woah hey, I got +100 today

Martin Büttner

congrats! :)

you've got 2.5 hours to crack the rep cap ;)

Trimsty

@MartinBüttner ty :D it's amazing considering I've been here for forever and I've made 1000 total, now 100 today

@MartinBüttner challenge accepted

githubphagocyte

Image questions and popularity contests seem to do well with votes. Put both into one question...

A few more answers could do it...

VisualMelon

I'd better get on with one then

Trimsty

I've already made 3

lolol

githubphagocyte

9:31 PM

@VisualMelon it's only 420 characters - how long could it take...? :)

It took me hours to get my second one down to 420 characters...

Trimsty

i managed to stay up to 11:30 with those 420 chars

VisualMelon

I've been having fun with something like Martin's beautiful patterns in C#, certainly too tired to get C++ to work

Martin Büttner

speaking of which

I just added a random walk in both directions to that, and it looks amazing :D

if people have algorithmic suggestions let me know

so far I've tried: stripes in different directions, isotropic colour spots... diffusion of different degrees... using only one or two of the channels... and coordinate mappings to create the circles and spirals

githubphagocyte

I tried to get a diffusion limited aggregate effect but it came out too similar to your painting ones. I couldn't get it to branch

Martin Büttner

9:47 PM

what does that even mean? :D

githubphagocyte

I might try again now that I'm more familiar with the memoisation approach

Martin Büttner

I wonder if I can get stripes in different directions that actually cover each other instead of looking like a weird overlay

githubphagocyte

DLA crops up everywhere - I've always found it fascinating

Yes interwoven paint stripes would be great

Martin Büttner

oh that looks sweet

githubphagocyte

They are easy to grow in a program - I used to make them in BASIC :)

but I couldn't get it to work with these three colour functions - I think it can be done though

Martin Büttner

9:54 PM

hm that didn't go according to plan but it looks quite nice anyway

that looks a bit like coloured fabric

Trimsty

@MartinBüttner nice :D

That would work nice for a text overlay

imagines Hendrix t-shirt

gtg now, cya :c

Martin Büttner

see you

githubphagocyte

see you

@MartinBüttner that one reminds me of tree bark

Martin Büttner

btw, in that picture you can see the effect of the proper random colour walk

@githubphagocyte ha that was my first thought too, but then I remembered what colours I was looking at :D

githubphagocyte

What do you mean by proper? Is this a different approach?

Martin Büttner

10:04 PM

before I was only randomly increasing colours which led to regular overflows, producing what looked like individual brush strokes and more or less continuous gradients

now am I'm random-walking in either direction

less overflows, but interestingly varying shades

that makes it look all the more real I think

githubphagocyte

Yes, despite the bright colours this has more of an organic feel to it

I'm trying to work through my list of potential tweetable image ideas, but instead I just keep adding more to the bottom. I wonder if I'll actually get to the end of any of them...

PhiNotPi

I used Mathematica to create a contour plot of one on your swirly pointy functions.

githubphagocyte

I think Trimsty is gone for the day...

PhiNotPi

oh well

githubphagocyte

Do you have an image you could post here?

if you include an @ tag it'll be there to see tomorrow

PhiNotPi

10:18 PM

wait a sec...

githubphagocyte

sorry if I'm stating the obvious. I'm fairly new to chat

Martin Büttner

@githubphagocyte I noticed the same problem... the people close voting as too broad do have a bit of a point... but the results are just so amazing.

Okay guys I added a link to an imgur gallery of some more plots and descriptions to my post.

githubphagocyte

@MartinBüttner hang on. What? Who is close voting what? You lost me.

Martin Büttner

imgur.com/a/HIHLj#0

@githubphagocyte there are two close votes on tweetable maths

PhiNotPi

@Trimsty Here is a contour plot of one of your pointy-swirly functions (specifically the green one) i.imgur.com/yTiMO3u.png

VisualMelon

10:21 PM

these are all fabulous

githubphagocyte

It's hugely restrictive. Surely it's not more broad than the allrgb question?

Martin Büttner

thank you... sorry that this prevented me from finishing the domino controller :D

@githubphagocyte yeah, I think you're right

still, people submitting 4 different answers is a bit of an exception that made me scratch my head :D

VisualMelon

no no, this is more important

githubphagocyte

Well you might be apologising to some people, but it depends on perspective. From my point of view, thank you for not finishing the domino controller - or I wouldn't have seen these.

Martin Büttner

I'm currently thinking whether I could turn this into a framework and put it on GitHub

VisualMelon

10:24 PM

I'm only going to spend days trying to improve a doomed brute force solver when you finish it ;)

Martin Büttner

hehe, yes I admit I'm very satisfied with my priorities this weekend :D

githubphagocyte

I've added a comment to the question but it's so far down I doubt the close voters will see it.

Martin Büttner

if someone else gives that another upvote it might end up on the ones displayed by default

PhiNotPi

I just tried to create a super-high-resolution 3D plot of the pointy-swirly function, and Mathematica ate up all my RAM.

Martin Büttner

I'm off to bed... see you guys

githubphagocyte

10:27 PM

see you again

PhiNotPi

bye!

3D swirl plot: i.imgur.com/LAzvsb1.png

githubphagocyte

nice to see the shape behind the image

PhiNotPi

Also, another idea for a KOTH challenge is a game like Lifeboats

The main feature of the game is that every step is entirely democratic

it is a pure negotiation game, in which you have nothing to offer.

Players basically vote on which lifeboat will progress to shore and which ones will sink.

githubphagocyte

So it would be about trying to guess who everyone else will vote for and position yourself accordingly?

PhiNotPi

Yeah, basically.

Each player has several sailors under his control, and they try to make sure that they can position their sailors into the correct boats.

If you are outnumbered on a boat, then the others will target you and throw you off when needed. If you are all by yourself, others will not vote for your boat to progress towards shore.

githubphagocyte

10:42 PM

Brutal

PhiNotPi

There is a board game review group called "The Dice Tower" which rated this game as one of the most anger-inducing board games of all time.

githubphagocyte

ha ha ha I can imagine

At least with bots it will be less personal...

PhiNotPi

When played in person, the votes are revealed simultaneously, so people could lie to your face and then vote against you five seconds later.

githubphagocyte

How much information would bots have about how others voted previously?