The Nineteenth Byte - 2014-09-04 (page 1 of 2)

« first day (1314 days earlier) ← previous day next day → last day (3522 days later) »

12:29 AM

@MartinBüttner I'm listening to a podcast about the king of beaver island. I think you could supplant him

1 hour later…

1:30 AM

I'm currently working on college applications.

This is a sad time.

RIP my childhood.

@professorfish I just noticed that the sandbox is owned by you. Are you okay with getting... lots and lots of notifications from it, or would you prefer that the owner be changed to a dummy account so your inbox doesn't get flooded?

@PhiNotPi ... now I'm scared. Luckily, I still have a few more years. :P

I'm horribly indecisive when it comes to stuff like this.

2:26 AM

You could also post it on Programming Puzzles & Code Golf beta and see what people come up with. :) — Joshua Taylor 2 hours ago

^ :(

1 hour later…

3:31 AM

bah...

I'm not sure how best to code this. I'm trying to parse the function names and arguments, and I don't know regex well enough to do it

for example, I want func1(func2(arg1),arg2) to return func1 as the function name, and an array of args [func2(arg1),arg2]

The biggest problem I'm facing is splitting on the comma

@Sparr, do you know regex?

4:21 AM

yes

things that nest are not a regular language, and cannot be represented by a regular expression

you need a more general purpose parser

any regular expression you write for that is going to choke if func2 takes multiple arguments, or if arg2 is a function itself, etc

@NathanMerrill are you absolutely sure that there won't ever be any more commas, such as in func2(arg1,arg2) or even if arg2 is itself a function with two parameters?

2 hours later…

6:08 AM

@MartinBüttner I figured more or less the same in the end, thanks anyway

It's a good thing these people won't be teaching me German it seems.

6:27 AM

@MartinBüttner Thisa actually turns out nicely for you, gets people warmed up for tetris.

6:47 AM

@NathanMerrill, which regex flavour? For stuff like this it makes a difference. E.g. .Net has special bracket-matching syntax.

7:17 AM

@VisualMelon Hi. are you about?

hi @PeterTaylor

7:28 AM

@Doorknob yes please; I keep getting 'question edited' notifications

Martin Büttner

7:50 AM

@NathanMerrill I do. But as Peter said, the regex flavour is very relevant here.

Also, do you want to do that recursively, or will you just reapply the regex to all arguments?

Martin Büttner

8:28 AM

@githubphagocyte In fairness, you should edit Peter's meta post ("Keep them"), too.

hi @MartinBüttner

@VisualMelon "your example solution seems to presume that you can't wrap the distance (i.e. 99 is closer to 80 then 1) (for T = 100, that is)" This is a bug if that is what I wrote but I don't see it. Can you spell out where I give that impression?

8:57 AM

@user2179021 howdy

@VisualMelon OK let's fix the question :)

the method circle_dist has the line dist += np.abs(closest_number - t)

I just updated it a little to make high scores better

which (regardless of what the closest_number is) can yield values greater than T/2

oh let me worry about that :)

8:58 AM

unless I'm missing something, I think it ought to read something like min(np.abs(closest_number - t), T - np.abs(closest_number - t))

takeClosest is broken basically

it isn't reporting the closest number in the case you say

let me fix it

actually

let's make a test case

I think you are right but I need a test :)

timesX = [1]
timesY = [99]
T = 100

that gives the right answer

which case do you think wouldn't?

I think takeClosest isn't broken now

remember it can return a negative number

ah, I see

ok, yeah, take_closest does look right

ok . do you still think there is a bug in the code?

no, I see what it's doing now - sorry to make fuss out of nothing ;)

no problem at all!

there is probably another bug :) there is always another bug

@VisualMelon Do you think you will attempt the question?

I am really hoping someone will be the first :)

9:07 AM

I may do, I don't have a strategy in mind, but I'm sure I can write something mindless get do the job

sounds perfect :)

the benchmark is terrible :)

which should help

it just does a local optimization from some starting point

I don't even know which starting point :)

visualmelon.onl/spindist.png looks fun

@VisualMelon for the visualisation, it would be nice to have all the points on a circle

which one set in one color and the other in the other

I think

@

yeah, it's probably misleading that the X values are on circles of different radius

2 hours later…

New Sandboxed Posts

10:55 AM

0

A: Sandbox for Proposed Challenges

Penalty shootout Welcome to CG world cup. Your job: Bot a penatly shooter and goal keeper! Rules of the game: The goal is 700 cm wide and 200 cm tall. If I have to make a triangle out of it, the goal is ABCD where A is (0,0); B is (0,200); C is (700,200); D is (700,0). The shooter gives me th...

Martin Büttner

11:23 AM

@overactor Found this while browsing through old sandbox proposals: meta.codegolf.stackexchange.com/a/425/8478

Sounds very similar to King of the Walls though

@MartinBüttner It is essentially the same

KotH was still extremely new back then I take

Martin Büttner

:D ... in 2012 Gareth suggested an RPSLV KotH and Peter said something along the lines of "Nah, that's boring, random is the optimal strategy."

hah

oh btw

how would we do a speedchess koth?

Martin Büttner

as successfully as we do a simplifiedchess koth?

I'd say you call both processes sequentially and time their moves. and maybe each entry also get to run one background process?

possibly

though it seems some people are working on it

Martin Büttner

11:29 AM

In fact, that challenge already has a 2 second limit on moves

in speed chess you have an overall limit though, right?

Martin Büttner

yes

you get 5 minutes in total

Martin Büttner

but if that's the only difference I'd close as duplicate

mate or out of time and you're out

@MartinBüttner really?

I'd make it shorter than 5 minutes btw

maybe 1 minute

Martin Büttner

11:31 AM

yeah, the difference between 2s per move and 5min for the entire game doesn't make the challenge too different

even then... if that question gets answers, they are likely not to use all of the 2 seconds and hence could still play a game in a couple of minutes

so they could be easily reused for your challenge

@MartinBüttner but would they do well?

Martin Büttner

possibly well enough

I do admit that there's some strategic difference between the two timing systems, but I don't think it's enough

@MartinBüttner I think the biggest difference is being able to take a long time for some decisions

also, background thread.

if that is feasible

Martin Büttner

@EricTressler I have no idea what you're talking about.

I wouldn't want to post it though

probably wouldn't work out too well

11:52 AM

@MartinBüttner I had to have a look around to figure out why that was amusing.

I can't find the proposal in my meta answers. Has it been deleted?

Martin Büttner

@Gareth yes

I think it was just above the one I commented on

actually no, it wasn't

meta.codegolf.stackexchange.com/a/476/8478

@Gareth interesting, you wanted to invent cops-and-robbers two years ago :) meta.codegolf.stackexchange.com/a/528/8478

Gareth: a golfer, way ahead of his time.

12:18 PM

No. I am not @Sparr

I figured out a solution to my argument parser @MartinBüttner

I ended up iterating character by character, counting the open parentheses, and if there isn't a currently open parenthesis when I hit a comma, split on it

Martin Büttner

12:38 PM

yes, of course the non-regex approach always works ;)

hi

hi

Martin Büttner

12:55 PM

yay, first time I discovered something worth putting in a tips question that wasn't already there

0

A: Tips for golfing in Mathematica

You can stick an expression in Break which can save one or two characters. Example (other details not golfed for clarity): result = False; Break[] can be turned into Break[result = False] to save one character. If the expression in question does not have lower precedence than function appli...

congratz

Thank you, kind stranger, who just upvoted all of my questions.

Martin Büttner

all of them? o.O

@PhiNotPi Is that not chain voting?

Martin Büttner

I was just gonna say

12:58 PM

@MartinBüttner are golf-tip questions/answers always community wiki?

Martin Büttner

those upvotes are likely going to be removed again

@overactor yes

@PhiNotPi I've been getting upvotes for my really old questions lately, a couple here, a couple there. Odd.

Martin Büttner

for some inexplicable reason...

@Gareth no he really just got 15 or so

@MartinBüttner So I should propose some way-out-there stuff and then I can expect to see it in a couple of years. :-)

Martin Büttner

Yup, seems to work quite well. :D

1:01 PM

@Gareth AFAIK, that's exactly how it works.

Feel free to just make up words as you go along without explaining them too, we'll figure out your genius given enough time.

It never came to me.

Martin Büttner

@Rainbolt Tilda Swinton didn't come to you?

I'm pretty good at Madgab, but I failed this one

OMFG

She's a famous actress

I didn't think to Google her. I thought it was a madgab

Martin Büttner

:D

You tricked me!

Martin Büttner

1:05 PM

(I didn't)

You're in cahoots with whoever tricked me!

+1 for using 'cahoots'.

1:22 PM

Hi.. I would like to advertise codegolf.stackexchange.com/questions/37269/… . I am hoping for a first answer!

1:48 PM

First rule: never use the word advertise in an advertisement.

Second: Try to slip it organically into the conversation. "Hey look at this!" Doesn't work as well as you'd think.

Try cahoots. Nobody will ever click on that

Disguise it as an xkcd link :D

I don't understand "You should run your code for T = 1000,10000, 100000, 1000000, 10000000". What are these numbers?

Martin Büttner

I think I need to start doing my triangle-centres program in trilinear coordinates. There will be more boilerplate code to set this up and do the conversion, but that should allow me to implement many centres trivially, as well as adding isogonal and isotomic conjugation.

2:03 PM

ugh nevermind. T was defined in the first paragraph. Sorry!

githubphagocyte

@Martin oh yeah - I can edit other people's meta posts now. Good point. I'll make a trivial edit to unlock the votes for balance.

Martin Büttner

@githubphagocyte make the same as in your post

2:16 PM

I almost want to ask a "Tips for golfing in Manufactoria"

hi again

Hi (first!)

:)

2:31 PM

Hi

@Geobits Funnily, this is relevant: http://xkcd.com/870/

Martin Büttner

@overactor how meta

@MartinBüttner There's at least 3 layers there I think

Why does SQL 2012 Management Studio have a "Close" and a "Close All But This" but not a "Close All"?

Spite?

I think a lot of applications have experienced spite-driven design

Martin Büttner

2:44 PM

@Rainbolt That's always been annoying me so much in Visual Studio

@MartinBüttner Visual Studio 2012 has Close, Close All, Close All But This, and Close All But Pinned

Chrome has Close, Close All, Close All But This, and Close To The Right

(The close all is the x button btw)

@Rainbolt But not close only pinned?

Right. They swapped close all but pinned for close all to the right

But guess what

Pinned tabs fall to the left

that is pretty neat

Martin Büttner

@Rainbolt I think you have some plugin installed that adds Close All. I remember one of the popular ones has it.

2:48 PM

Ah that's possible

I've got close all too in VS12

:^)

I don't have any weird extensions that would do that

Unless it came bundled with another unrelated tool. I was looking for some sort fo productivity extension

Martin Büttner

hm maybe they added it in 2012, but I thought I was missing it there as well... don't have VS here to check

that was strange

i had unused variables in eclipse

so there was a warning

I had eclipse supress those warning automatically

one by one

then it gave a warning that I had an unecessary token in the supress warning line

so I had eclipse solve that too

now I ended up with exactly the same code I started with, but no warnings

Sometimes Eclipse will show suggestions for lines you have highlighted, regardless of whether you have suppressed the warning or not.

If you click off of them, and the warning isn't important, it will retreat.

It may or may not be relevant to your situation

2:55 PM

eclipse does seem to be funny that way

it wasn't relevant

the warning sign were just in the way of my breakpoints

This is why IntelliJ is awesome

3:08 PM

what's that

A trendy IDE mainly for Java

this is part 5 of my caveman question series, after "How food"

@Rainbolt How quickly should I jump on this bandwagon?

not so fast that you bounce off

I use notepad++ for my code, but I don't work on major projects

but for <10k lines of code, it works great

I used vim for a long time, before I switched to notepad++

vim is intimidating

Martin Büttner

can anyone else hear this faint blissful voice chanting "Subliiiiiiiime" in the distance?

3:12 PM

shazam

Martin Büttner

wat?

@overactor I haven't yet, only because I am comfortable with NetBeans' interface

But I wore carpenter jeans all through middle school and freshman year because I was comfortable with those too.

@Rainbolt that's okay with me.

scientists give 0 shits how you look

0 shits is the right amount of shits for most occasions.

What's the SI abbreviation for shits?

3:18 PM

SH, I assume

For fecal matter mass in general it's Courics.

if you ever read about Newton, he was a very peculiar dude too

though he may have had heavy metal poisoning

Hmm. I must not have listened to enough heavy metal to be poisoned by it.

Newton was very much into alchemy. Lots of mercury and lead

Martin Büttner

@Geobits relevant

3:28 PM

\

^

@

Martin Büttner

I don't think I understand your ascii art

\/

Martin Büttner

I also broke it

Me either

that's okay, it was a weak attempt

@MartinBüttner pardon me if I've asked you this before, but where are you?

Martin Büttner

in London

3:32 PM

Look at his profile.

that's exotic for us america-types

Your top voted answers list is insane btw

what

Martin Büttner

@Rainbolt it's not very diverse, is it? :D

@EricTressler it's moderately exotic for a German from a small-ish town as well :P

I think my ancestors were German too

but I don't really know

I'd like to think that in the best of all possible worlds, it doesn't matter

Martin Büttner

3:39 PM

countries... such an outdated concept

go tell the EU that

and provide popcorn

I don't disagree with you, by the way

Martin Büttner

I'd say the EU is the right step in that direction though, isn't it?

probably; but it all breaks across religious boundaries, unfortunately

Religions support countries how?

Martin Büttner

I think you don't even need religion here... just cultural differences are probably too big to reconcile under one roof at some point

3:47 PM

Wouldn't it be neat to have a democratric system where you not only vote on the law, but the areas which it affects?

It could even be granular so that moving from town to town changes laws

Wait this can't work

Because then I could impose strict laws on you.

well, okay. I suppose my own predilections are anti-religious

There needs to be some fair way of making laws affect only the areas it needs to affect

Then we can all live under the same roof

Martin Büttner

I think that's how having separate countries/states works

It's different though. With countries, my laws are ALL mine, and yours are ALL yours.

This way, they all overlap

I guess you need a common enemy to make something like that happen though. Or maybe I've been reading too much Shadow of the Hegemon

4:11 PM

@Rainbolt Why should people impose laws on people in other areas? Aren't the people best suited for making laws in a given area from that area?

Maybe the ideal system is one where there is some kind of inheritance system. Something like City extends County extends State extends Country extends Planet?

oO

somebody needs to take down Lockheed

It got 8659 points. Second place, AttackOrElse got 1774

I have no clue what that means. I just looked up Lockheed Martin news, but apparently that's not it ><

CodeBots

27

Q: What's wrong with public variables?

Code Bots I hate private and protected variables. I just want to access anything and everything! If you are like me, then this challenge is for you! Write a bot that works together in harmony with other bots makes other bots do what you want. You are a programmer, and you know how things are...

king-of-the-hill

4:28 PM

the bot that defeats Lockheed had better be called F35 (hopefully everyone will be bankrupt by the time it ships)

So.... Lockheed's own F-35 will be what eventually takes it down? Nice.

@Geobits But then you can't have a law that affects only two states.

@Rainbolt Why is that a necessary thing? The people in both states can decide on their own.

If they both vote one it and it passes, then both states have it.

I think we're saying the same thing...

New Sandboxed Posts

4:44 PM

0

A: Sandbox for Proposed Challenges

A game of chance Hello! Thought I would try out the sandbox, a rather useful tool that I haven't utilized yet! This is a game I thought up with some colleagues, basic in theory but has some deeper algorithms to win. I'm thinking of tagging this as code-challenge, any opinions welcome, especia...

0

A: Sandbox for Proposed Challenges

BS king-of-the-hill The goal of this challenge is to implement an AI for the game of BS, also known as Bull Shit, Cheat, Bluff, and numerous other names. The game is outlined in this wikipedia article. The Rules of the Game For the purposes of this challenge, the game will work like this: ...

Martin Büttner

@Rainbolt That PRNG challenge might be quite interesting if the range is unknown

hi

@user2179021 Hi!

@ProgramFOX I hope you don't mind but I would like to advertise codegolf.stackexchange.com/questions/37269/… . I am hoping someone will be the first to try it

@ProgramFOX have you had a chance to look at it?

it is improved from how it used to be :)

@user2179021 Not yet, will look at it.

4:53 PM

@ProgramFOX great! If you have any questions or anything is unclear, please ask

I will hang out here

@user2179021 Nice challenge, have a +1!

@ProgramFOX thanks! Now I jsut want someone to try it :)

Martin Büttner

@user2179021 What might help is a verifier. Given input data and a rotation, it spits out the total of distances.

5:08 PM

@MartinBüttner Whatever the range is, as soon as you know it, an optimal guess exists. From there, it's pure luck. If the range is unknown, then it can still be solved optimally, because you you can make an optimal guess on the range, and then make an optimal guess based on that optimal guess.

If your optimal guess was wrong, you just got unlucky.

He would be better off requiring an optimal solution, except that someone who wrote a non-optimal solution could actually do better on specific test cases.

Haha he called me RainBolt

Maybe my ancestors did just concatenate some words together to form my last name.

It would certainly be better than the explanation I got from my grandfather. "When my great great grandparents entered the country, the United States misspelled Randolph or Randulph." They were German. @MartinBüttner, does Randolph sounds even remotely like Rainbolt when pronounced by a German?

5:27 PM

the function dist does that

@MartinBüttner sorry I meant... the function dist does that

CodeBots 2 is ready and open!

0

Q: CodeBots 2: Event Driven Design

Welcome to Code Bots 2! You've learned your lesson since the last Code Bots. You've tried to figure out more ways to fit more actions in less lines, and now you finally have it. You're going to make an event-driven Code Bot. Your bot must consist of 24 lines. Each line follows one of the two...

king-of-the-hill

githubphagocyte

@Martin I wasn't sure if it would be OK to put my edit wording into someone else's meta answer, but since you suggested it I now have. Hopefully that's all balanced now.

Martin Büttner

5:44 PM

@Rainbolt no, but I think R[ae]inhol[dt] might be a name.

I know a couple Reinholdts.

Martin Büttner

@NathanMerrill what's it with the 24? Why not 32?

Why would I make it 32?

because 32!

Martin Büttner

Why wouldn't you?

5:49 PM

Because CodeBots is also about code golfing

@MartinBüttner in CB1 it was bumped from 16 to 24 because a bunch of us made good cases that there would not be a lot of variety in 16, but 24 gave plenty of space

not only that, but with these bots, you can do more in less space

@MartinBüttner I added a comment about dist. Thanks

also, @Sparr the most convincing argument you guys gave about 24 is that it was divisible by 3. I love highly divisible numbers, and 24 is one of those

:)

Martin Büttner

5:52 PM

good point

Also, you'll like This

I don't think I was arguing specifically in favor of 3n, just pointing out that 3 is the minimum number of duplicates for an effective self repairing bot

it refers to the line the command is currently one

Martin Büttner

Any is a redundant in CB2 btw :)

not C

because of Not?

I know

Not(None())

Martin Büttner

5:54 PM

yup

Block was also removed, because there is now a much better way to do it

@NathanMerrill your Attacker bot... it moves once, then attacks when an enemy is ahead of it, moves when an enemy is beside/behind it, and randomly changes direction when it's alone. right?

Martin Büttner

no randomness this time?

Attacker uses E as a random variable

spec doesn't define E, but I assume it still exists and is random

yes, thank you

Martin Büttner

5:59 PM

regarding the If statement: does C refer to the variable C or some arbitrary condition C? in the former case, does true meant >0?

@NathanMerrill in CB2 you want people to keep backup lines and restore them rather than blocking?

« first day (1314 days earlier) ← previous day next day → last day (3522 days later) »

Transcript for

Sep '144

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