The Nineteenth Byte - 2014-09-02 (page 3 of 4)

« first day (1312 days earlier) ← previous day next day → last day (3529 days later) »

3:00 PM

...footprint.

Is there any concise, defensible, easy-to-describe way of handicapping languages so that even the clunky ones have a decent chance of winning (if a "normalized" competition is what's desired)?

That's been discussed, but the general idea seems to be "who cares?" from what I can tell. There's no need for it except the shiny green checkmark, but that doesn't matter so much (IMO).

@COTO Not really, you can of course score by byte count rather than character count

If a C# answer is impressively golfed (for C#), I'll upvote it even if it's much larger than the accepted answer.

but even then, CJam, GS or maybe Bash will usually take the checkmark

but as @Geobit says, that's not very important

Martin Büttner

hence, by-language scoreboards to give more people the feeling of accomplishing something by being good at golfing in their (verbose) languages

3:04 PM

Re per-language summaries: they bloat the question and go out of date. Would a better solution be to write a userscript which adds a heuristic "sort by score" (using the last non-struck-out number in the first header of each post)?

The shortest answer is often boring, also. If it's the obvious way of doing it, and doesn't take much golfing to make it the shortest, I'm much less impressed.

So basically per-language scoring, or "it doesn't matter anyway". I suppose that's what I expected. :P

Thanks for the feedback.

Completely off topic: What's a good way to teach chess to a 7-year old? My son wants to learn, and I'm trying, but I find it hard to play without crushing him.

@PeterTaylor But the all the concise languages will be at the top

Maybe I should cross post this on both Parenting and B&CG :D

3:07 PM

@Geobits Would be interesting to see what the differences in answers are

Martin Büttner

@Geobits here's what my dad did: every time he loses he can take out one of your pieces for the next game (counting rooks as 5 pieces and so on). when he wins, the handicap is decremented again

alternatively, teach him Go :D

(that comes with built-in handicap rules)

and the rules are much simpler anyway

@MartinBüttner I did something similar when he learned Go, giving myself a huge handicap to start with, but he seem to play his strategy to the handicap, if that makes sense.

If I try the missing-piece handicap in chess, I'm afraid it will reinforce him to ignore those pieces.

Martin Büttner

@Geobits well ideally the handicap should diminish over time, which would mean he has to adapt to less and less handicap

@overactor So a userscript to generate a contents list. Either way, I see an advantage to not denormalising. (And I see making the bloat opt-in as a good thing too).

Yea, we normally play Go with a 2-3 handicap on a 9x9 board now. He still hasn't gotten the big-picture enough for a 19x19 match IMO.

Martin Büttner

3:10 PM

@Geobits in the meantime there's 13x13

@Geobits Use a clock and force yourself to play blitz?

how many moves qualifies as crushing ?

Martin Büttner

also if you use 3 handicap in a 9x9, there's probably no way to put enough handicap on a 19x19 anyway

but certainly for a 13x13

@PeterTaylor That might work eventually, but he's only played a few games so far, so "blitz" is basically what I'm doing without trying.

@PeterTaylor: What is a userscript? Is there a way of writing scripts to act on SE pages?

3:12 PM

@MartinBüttner 13x13 is weird. Also, I have both other boards :)

Would allowing him to make an extra move every now and then make sense at all?

Martin Büttner

@Geobits :D no it's not

how well do you play yourself?

@overactor that would affect strategies way more than having Geobits start with fewer pieces

@MartinBüttner I suppose it would

@MartinBüttner Go? Not too badly, but I'm not a pro or anything. Chess? Decent, used to play a lot more, but it's been a few years.

@overactor An extra move turns a check into a checkmate :o

@Geobits, when you play with a handicap, you would take random pieces, no?

Martin Büttner

3:14 PM

@Geobits Well I didn't assume to be talking to one of the few western "pro"s :P

@overactor no, I'd let him choose them

I haven't played rated games, just with a local group of friends.

Some online, but nothing serious.

Martin Büttner

k

I like the handicap idea of choosing pieces based on points, though. Will give it a shot.

Martin Büttner

yeah so he can basically decide if he wants to take out three pawns or a knight

@COTO It's a way of writing scripts to act on pages in a given domain. Some SE ones are collected under stackapps. See e.g. stackapps.com/questions/4342/…

3:18 PM

Oh crap, I'm gonna end up playing him with no pawns at all ><

@Geobits Someone probably already said this, but you should handicap yourself with time AND force yourself to look away when he is taking his turn. That way, you can't think during his turn.

@MartinBüttner Random would allow more variety though

Martin Büttner

re Go... I stopped playing actively when I moved to England... I might pick it up when I move back to Germany. I really wonder how much my skills have degraded in 4 years :D

@overactor and less insight into the game

@PeterTaylor: Neat! I'll definitely use that when compiling leaderboards in future.

@Geobits What you said below this comment

3:20 PM

@Rainbolt Maybe, but I have a fair memory, and normally have a few turns in mind, especially now while he's still learning.

@MartinBüttner Why is that?

That should change over time, but I generally already do other things if I'm letting him think for a minute.

Martin Büttner

@overactor well if he can deliberately experiment with removing different pieces, he might figure out how valuable they really are

Other times if he seems stuck I may get him to talk it out.

@MartinBüttner I doubt he'll sense a lot of that

seems more likely he'll pick a few favorites to remove

3:22 PM

@overactor At first, sure. The first time I beat him with a rook+queen missing he might change his mind.

I have no doubts he won't think to remove my entire row of pawns until after he's tried a couple other "better piece" games.

When I first started to play Chess, my only goal was to build the largest pawn triangles.

It worked well in middle school

my tactic at school was to talk a lot, had varying success

Yep, at school/friend games smack talk works better than anything.

I wonder if there are any good behavioral studies on the circles that people form in school. I used to watch the quiet people get shouldered out while the loud people just walked up and were instantly absorbed into the circle.

Martin Büttner

@PeterTaylor I really wanted to argue against reverting to a userscript and for including the by-language board in general, but my argument always would have been based on "it's more about being on the list to feel some accomplishment rather than being interested in the list oneself", but that's the worst possible argument for bloating up the question...

3:27 PM

@Rainbolt What do you love most in this world?

@Geobits I love being able to change what I said last minute to better reflect what I was thinking.

Martin Büttner

userscript sounds good then. if it gets popular enough we might even have enough people to pay attention to all answers being formatted decently

@overactor where's your short and simple code golf? :P

@MartinBüttner How is Tetris coming?

Martin Büttner

@Rainbolt I spent an hour yesterday rerunning vector racing, and then I left my laptop for my girlfriend to use because her laptop's power supply decided to break (and tending to her blog-turned-online-business seemed a bit more important than tetris)

not sure, I'll get around to it tonight (having a guest over) or tomorrow (going to a gig) either... I suppose I'll try to spend some time on it over the weekend.

Alright. The longer you wait, the more likely it is that I will have a library worth using.

Martin Büttner

3:38 PM

speaking of tetris though... can I get some feedback regarding the random generator?

You get 7 pieces, they are shuffled, and the first iteration skips a random number of pieces right?

Martin Büttner

yes

xnor pointed out that that makes perfect algorithms even easier

Is that how real Tetris works?

Martin Büttner

@Rainbolt except for the skipping, that's one of the two common implementations

I could instead go for this one: tetris.wikia.com/wiki/TGM_randomizer

or just use uniform randomness which would make the game harder than actual tetris

what I liked about the Random Generator is that it adds a possible optimisation level by trying to figure out the bounds of each batch over time

Personally, I like the random bags one after another algorithm. It's just simple.

Uniform randomness would be even simpler, but at the expense of using any good research that's been done on Tetris

Martin Büttner

3:42 PM

which may or may not be a good thing

Right. I'd say that it's a design decision and up to you

Martin Büttner

well I still think that optimal algorithms being feasible is not a problem

a) it would still be a fastest-code challenge. b) a heuristic might perform better by being much faster and hence going for longer c) you could be even more sophisticated and play optimal as long as time permits and then switch to a heuristic

anyone volunteering to generate nice GIFs from a protocol of the games? (@VisualMelon ? :))

You could just convert the output to HTML and display it in a panel

It would be mindbogglingly slow, but you could easily capture it in a gif and speed up (skip some frames) the framerate.

Martin Büttner

@Rainbolt what panel? what output?

oh

I see what you mean

I think even if I did it myself I'd be faster with actual 2D graphics

but I know VisualMelon has looked into GIFs in C# quite a bit ^^

for me it would be easiest in Mathematica, but then people couldn't use it

I just thoughtTetris was really suited to being displayed in HTML, and applying styles to it would be easy that way.

3:47 PM

I'd gladly spend some more time making gifs ;)

Then we could all come up with our own Tetris.css

nothing I want to golf at the moment

Martin Büttner

the best thing would be JS, so you can just paste on a web page and generate the GIF there.

Would make a decent (i.e., terrible) pop contest. "Come up with the best CSS for this other challenge."

Martin Büttner

@VisualMelon I'll let you know when the protocol is ready, but most likely it'll just be a stream of characters for all the events. which would mean you might have to re-implement part of the tetris game.

@Rainbolt you meant art contest

3:49 PM

What's the difference?

Martin Büttner

well...

Ok fine I'll go read the tag

Doesn't exist :(

Martin Büttner

:D

At least art-contest doesn't exist.

I know I know. "Made you look."

Martin Büttner

I think if someone created it Peter would lose all faith in PPCG :D

4:05 PM

random-winner

6

Or perhaps there could be an anti-popularity-contest. Winner is the submission with the most downvotes.

Martin Büttner

@PhiNotPi Makes me think of a challenge where you need to devise a PRNG which satisfies some randomness benchmark, and then all the PRNG's vote on winning submission by using n different seeds.

(so you need to make your submission random but still more likely to spit out your user ID when given one of the relevant seeds)

it's probably impossible to make this work and unhackable though

@PhiNotPi: Obscene Tetris.

Although, to think of it, certain parties might not downvote that.

Martin Büttner

I don't feel like googling that right now

It's not a real thing.

"Hell Tetris" is, however.

kongregate.com/games/banthar/hell-tetris

Martin Büttner

:D

I don't even get points for pressing "down"

4:13 PM

We could have a voting-style koth, where each submission votes for another, and the submission with the most votes wins. It would be played in an iterative style.

Martin Büttner

O.O I never picked up on the fact that there are corresponding Hell and Heaven Tetris xkcd's

@MartinBüttner I came up with nothing

Martin Büttner

there hasn't been a challenge for 16 hours

I don't know if that's better or worse than the days where we have 8 closed ones

I think it's worse (provided those 8 closed ones aren't "ANYONE SEND THE CODING FOR TSC CODEVITA")

4:29 PM

Bah. Quit complaining and write a challenge then :D

Martin Büttner

my challenges tend to take days of preparation :D

Mine's coming soon

within 3 days, possibly today

Martin Büttner

@NathanMerrill is that Code Bots 2?

Yes

Martin Büttner

k

I'd already be happy about something simple to golf

4:30 PM

@MartinBüttner Mine are either days of prep or straight off the cuff, never in between it seems.

Martin Büttner

@PhiNotPi could get his averages ready :P

@Geobits I'm not good with coming up with simple golfs which are fun but not duplicates

There are a lot of very bright people in this chat room. We've gotta come up with something, right?

Ok, brainstorm time. Step one: name any algorithm or game.

Martin Büttner

(or general topic)

okay I've got an idea

Glad to be of service.

Martin Büttner

determine if there's another move on a hexagonal match-3 game board

4:38 PM

@MartinBüttner codegolf.stackexchange.com/unanswered

Martin Büttner

I actually had solve this one when prototyping such a game, but I could never be bothered to actually implement it, because so far I haven't encountered a situation where there wasn't another move ;D

@PeterTaylor yeah, I've looked through that list several times

(and I've removed two or three from the list myself)

but most of them are there for a reason

what about frogger?

turn based

Martin Büttner

what would the challenge look like?

what about anti-frogger?

ascii art?

4:41 PM

play the truck drivers with multiple frogs, try to kill them all efficiently.

>>>___>>_
__<<F_<<<
>__>>>>>_

etc

Martin Büttner

would my above idea be a duplicate, given that this exists? codegolf.stackexchange.com/q/26505/8478

the problem is a little more advanced on a hexagonal board

It's edgy

I was wondering about that. It seems more advanced, but really the only thing different is neighbor-matching, right?

(as in what pieces are neighbors)

Martin Büttner

well and dealing with 3 principal directions

btw, if anyone wants to play said prototype: mbuettner.github.io/hexagonal-zero/public

the controls are quite horrible (click on first piece, then click on adjacent piece)

4:43 PM

True, but making a hex roguelike I didn't see much difference once you add a small bit to deal with the hex.

roguelike or roguelite?

roguelike

would Centipede be convertible?

@Geobits hardcoregamer.com/2013/07/06/…

I think I'm gonna get my averages ready. Probably turn it into a rosetta stone.

Martin Büttner

4:51 PM

thumbs up for that!

I actually wanted to solve that in multiple languages anyway :D

(get rosetta-stone right though: one post per user with all languages, the overall winner is the person who wins in most languages, tie breaker being the combined size of their winning submissions)

there is a challenge!

0

Q: Write a run-length encoded Brainfuck interpreter

Write a program that takes a run-length encoded Brainfuck program as input, and interprets it. A run-length encoded Brainfuck program uses the same operators as regular Brainfuck, however a number can be appended to an operator to repeat the operator that many times. For example, this program: ...

code-golf interpreter brainfuck

@overactor Yes. Rogue-like. No persistence, procedural, etc.

Martin Büttner

sounds like a close duplicate of the brainflow one though

wait no

I'm mixing that up with extended brainfuck

the RLE is still a trivial addition to a normal BF interpreter

@Geobits what's the pitch?

It was done on a hex grid, and you had two characters. Six classes to choose from.

It was brutally hard to balance.

are there skill points?

Rpg elements in general?

4:56 PM

Very basic ones

linear progression?

If by "linear progression" you mean climbing a near-endless tower floor by floor with no real/engaging story, then yes.

@Geobits I meant as opposed to a branching skilltree

linear story progressio is fine for a roguelike imo

Oh. Mostly linear. There are a couple decisions that branch it, but it's more like a class "specialization" or subclass.

what sort of gameplay does it have?

or maybe better question

what are the classes?

5:00 PM

Standard RPG fare. Different names, but you had your basic warrior, tank, dagger, mage, healer, and bow-guy.

It's been a couple years, I'd have to get some code/docs out later to be more specific.

(couple = have no idea really, but more than one)

ha

did you get far with it?

I got it to playtesting with 40-50 people, but I couldn't find a good enough balance to release it. Dropped it when life got in the way.

Martin Büttner

afk... I expect to see an averaging rosetta-stone challenge up when I get back :P

@Geobits What's this "life" everyone is talking about? It sounds awful.

@Geobits Life being your kid?

Life meaning divorce, changing careers, moving back to the US. Little things, really :D

5:04 PM

@MartinBüttner The graphics are great, but it's riddled with bugs and the npc's are annoying /terrible and old joke

@Geobits That does sound terrible

@overactor and griefers broke the economy ;)

I have never gotten anywhere near even being close to almost halfway finishing a game.

I have no perseverance :'[

@overactor Yea, the last 10% is a bitch.

@MartinBüttner: One challenge I've been toying around with is based on jQuery. Your only tools are jQuery traversal and manipulation methods (e.g. .parents(), .each(), .next(), eq()), as well as .attr(), and .length.

Functions are allowed, but no operators except for string concatenation, user function calls, and jQuery calls.

No literals or variables allowed except for 'this'.

@Geobits I usually get stuck after the first 10%

5:12 PM

At runtime, your code is dumped into an HTML document with some tags of the form

<output id="out" value=""></output>
<input id="in1" value="4" />
<input id="in2" value="4" />
<input id="in3" value="2" />
<input id="in4" value="9" />
...

where the value attribute is always between 1 and 10.

Once the DOM is built, your jQuery script is run. One or more simple computational objectives is provided, such as:
- set the "value" attribute of 'out' to the sum of the value attributes of the 'in..' elements
- set the "value" attribute of 'out' to the maximum of the value attributes of the 'in..' elements
- etc.

The shortest query script that accomplishes the objective for any set of inputs wins.

I know that at least some of the challenges are possible. For example, a simple script to implement the sum would be:

$('body').append('<b id="1child"><a /></div>').
append('<b id="2child"><a /><a /></div>').
...

$('input').each( function() {
$("#" + $(this).attr('value') + "child").children().appendTo('body');
} );
$('output').attr( "value", $('b').length );

I know all integer-based computational tasks would be possible, since jQuery plus HTML tags implement the full set of operations covering a sigma algebra, and a sigma algebra can be put into 1:1 correspondence with the standard real algebra over the integers, but what I haven't had due time to contemplate is whether there are any obvious shortcuts or hacks that would completely undermine the challenge. One obvious one that comes to mind is that no operation...

...could write out a <script> tag at any point. Also, I'm not sure that certain challenges (such as the "max" challenge) would be implementable in a realistic amount of code.

If anyone here wants to take the time to iron out these details, make sure the exact set of traversal and manipulation methods allowed is soundly codified, think up a set of solvable computational challenges, and write the whole thing into a programming challenge, I think it would make for an interesting one. By all means go ahead.

If nobody is interested, I'll write it up myself in a week or two, or whenever I get time.

ETA: The comment length restrictions on this site suck.

Got to run. ;)

and then it was quiet

2

Martin Büttner

5:41 PM

@overactor Same problem. The closest I ever got was that match-3 prototype (and another one of similar quality)

@COTO multiline messages can be huge

Recent changes to my averages problem: meta.codegolf.stackexchange.com/a/2132/2867

I would want more test cases, but I am on mobile and using Wolfram Alpha to find the answers.

@MartinBüttner ^

I like the problem

Martin Büttner

@PhiNotPi You need a tie breaker for if the two people win the same number of sub-competitions.

(as suggested before: total size of winning submissions)

I can add more test cases for you if you like, Mathematica has all 5 of them built in

To break ties beyond that, how about the number of languages that person has submitted programs for.

Huh. My browser shows all five of those Wikipedia "mean" links as 'visited'. I webs too much.

Martin Büttner

5:55 PM

@PhiNotPi that's still more likely to tie than total code size, but it would be more in the spirit of the challenge.

you could do both

btw, if you don't want me to use those Mathematica functions in my actual submission, rule out built-in functions for the mean. the loophole is no longer valid. although I'd be sad if you did... the function names are way too long to be competitive anyway

It's okay to use built-in functions, because each language is competing with itself.

Martin Büttner

ah good point

maybe I'm blind but you're not stating the order in which to output the results

Haha, I love Mathematica... just tried the {1,2,3,4,5} example and got:

{3, 2^(3/5) 15^(1/5), 300/137, 11/3, Sqrt[11]}

(it looks nicer with Mathematica's formatting)

wait, "shortest total byte count" over all submissions or the winning ones?

Which do you prefer?

Martin Büttner

hm, I don't know

my first intuition was the latter, but the former but the latter incentivises golfing down losing submissions as much as possible

It would be smarter to delete them, but that doesn't seem "fair".

Martin Büttner

6:05 PM

@Geobits no because, then you lose the first tie breaker

Smarter to delete what?

The losing ones that add to your score without contributing to wins.

Martin Büttner

@Geobits the first tie-breaker is total number of submissions

Ah, didn't see that.

Well then the smart thing to do is enter a few bad golfs to win that one.

Then character count doesn't matter.

I don't think that's a good incentive.

Martin Büttner

well the main idea is to show as many languages as possible anyway

6:07 PM

I think just smallest byte count of winning entries should be the first tiebreaker, to prevent these strange incentives.

Martin Büttner

@PhiNotPi hm, I think "shortest total byte count over all submissions" as second tie breaker doesn't have those incentives either

@MartinBüttner If the first tiebreaker is total number of submissions, then if I'm tied, the obvious thing to do is add ten more entries in other languages. Then it doesn't matter if they're golfed (or even good) since the second tiebreaker never happens.

That seems easier than golfing down my losing ones.

Martin Büttner

@Geobits if that's so simple the person in a tie with you can do the same. if not, that's an achievement because you know a lot of languages, which is what rosetta-stone languages aim at.

why not tie break on rankings within the language?

If two people have the same number of winning entries, the tie break on how many second place entries they have

I don't really need to "know" a language to write some program to get averages. I could probably find examples all over the place.

Martin Büttner

6:11 PM

@PhiNotPi You might want to add a note what constitutes two different languages. JavaScript vs ActionScript? vs ES6? C vs C++? etc

@NathanMerrill that sounds quite good

@PhiNotPi I'll add three more test cases to it now

Is any consideration taken to BF-equivalents? Does every one count as a separate language?

C# vs C-Omega? (C# with some additions, pretty much)

Okay, I just edited it.

In the past, there's never been an explanation of what constitutes a different language.

I don't know if it is needed.

Martin Büttner

Add a simple note of "No shenanigans with basically equivalent submissions in different dialects/versions of a language (e.g. two or three examples here). I will be the final judge of what qualifies as different enough."

@PhiNotPi let me know if you want any other particularly interesting test cases

I guess one containing repeated numbers can't hurt

@PhiNotPi Are exact results allowed, or do you want floating-point output?

I sorta want floating point, since it's easier to understand.

Martin Büttner

6:26 PM

k... in that case, can I assume floating point (as opposed to integer) input?

I guess.

Martin Büttner

k

@PhiNotPi no external libraries I assume?

(e.g. NumPy)

No

Does the spec, as it is now, look ready for publication?

Martin Büttner

yes, unless you want scoring by characters instead of bytes (and even if not, you might want to clarify)

@MartinBüttner Closest I've come to finishing a game:

giphy.com/gifs/ylyUQl8yhoK2k1HC2Q

Martin Büttner

6:36 PM

not bad!

what the spaceship does is actually a very good representation of my engagement with game developing

Martin Büttner

the other one I did was computationally quite interesting, because it involved circular k-d trees and monte carlo algorithms as well as a lot of geometry

:D

it's a Qix/Jezzball clone with circular geometry: mbuettner.github.io/circular-zero/public click, drag, release to build walls

@MartinBüttner Did the users have to concern themselves with that?

Martin Büttner

@overactor no only I did...

ooh, it's cool

6:39 PM

I made 3D game of Khet with no AI

is that a pointcarré disc?

that's my greatest achievement

Martin Büttner

@overactor if so, that's an accident

let me look it up

whoa, Poincare

don't go adding letters

Posted the challenge.

Martin Büttner

6:41 PM

the idea was just a circle with circular walls, which I decided to be perpendicular to the arena at the intersection (because there was an additional degree of freedom)

I have to leave now. I'll be back in an hour.

Martin Büttner

@PhiNotPi working on Ruby now, so I can post it with at least two languages... will have to do more later

1

Q: Rosetta Stone Challenge: What's Average Anyways?

The goal of a Rosetta Stone Challenge is to write solutions in as many languages as possible. Show off your programming multilingualism! The Challenge When people use the term "average," they generally mean the arithmetic mean, which is the sum of the numbers divided by the number of numbers. ...

code-golf code-challenge rosetta-stone

@EricTressler You can't tell me what to do

that's true

6:46 PM

I'm enjoying your game @MartinBüttner

Martin Büttner

I'm glad you do! :)

fun-fact: the collision detection is exact (to floating-point precision)

that is, if any time step would move a ball into or through a wall, the step is intersected, direction reversed and the remaining step continued in the new direction... repeat until the step is completed

are the balls represented as points?

Martin Büttner

@overactor nope

but they don't collide with each other

and their radius added to the hitregion of the walls

that's one way of achieving that

Martin Büttner

oh yeah, I think that's actually how the code works

6:52 PM

I did the same with my spaceship game

sadly, it sometimes passes through walls

in corners

I'm blaming floating points precision

Martin Büttner

:D

I was also once working on a 3D tower defense game when a friend who was going to do art didn't feel like it anymore

so I stopped too

Martin Büttner

7:07 PM

oh actually, I worked on a sick non-linear 2.5D platformer with some friends for a Game Dev seminar in uni, and we got to one playable level.

Seems relevant :D gamedev.stackexchange.com/a/6330/32807

Martin Büttner

@Geobits yeah :(

see, with those two prototypes I showed today, my intention was to do a third one and then try to make the best of them into a releasable game for some platform... problem was, I never had the time or idea for that third prototype.

as for that 2.5D platformer, I believe the idea is actually really good, so I don't want to touch it again until (if it ever happens) I have some actual decent experience and a few people to work on it and do it right.

I have a game in that state except it's been released. It just needs a major overhaul for modern phones.

I mean major. I believe on most of today's Androids it doesn't work at all. I should probably take it down.

7:33 PM

@MartinBüttner I also had a cool idea for a 2.5D non linear platformer once

I'd love to hear your ideas

but now i gotta do dishes

hi

I am writing meta.codegolf.stackexchange.com/questions/1847/…

but how can I make the data creation portable?

it is done in this python code

#Set the seed to make data reproducible
np.random.seed(seed = 10)

timesY = np.sort(np.random.uniform(low = 0.0, high = T, size = N))
indices = np.sort(np.random.randint(0,N, N/10))
timesX = timesY[indices]

some people might not want to run python code

Martin Büttner

7:50 PM

@overactor ah come on, you only wanna do it before I can! :P

hi @MartinBüttner

Martin Büttner

hi

specify pseudocode?

@MartinBüttner hmm.. I am not sure I know how to do that in this case

actually, I want to make the data creation even more complicated so that the optimal score isn't zero

by adding some noise to the position of the points in timesX

Martin Büttner

8:06 PM

in that case, just providing a python snippet which writes the data in some predefined format to STDOUT seems fine to me. then people can run that python code and do python generate-data.py 1000 | myprogram

8:22 PM

@COTO You said "no literals". I count 11 in that snippet.

@MartinBüttner, what colour are your enemies supposed to be? I just see a white circle.

Martin Büttner

@PeterTaylor Red... I've heard some problems with Firefox on Unix before. I've only ever tested this for Chrome on Windows. But I think at least Firefox on Window should work too

I remember that it was some oddity with the OS and not just the browser

Curious. Ah well.

Martin Büttner

do you have another browser to try?

It's the same on Chrome and Firefox. It obviously won't work on Lynx. I don't have any other browsers installed.

8:42 PM

@MartinBüttner As if I have the necessary knowledge/skill to pull it off

Martin Büttner

9:04 PM

@PeterTaylor damn, I really wonder what that could be... probably some OpenGL issues or something

the game runs on WebGL

Could be. I'm not going to try debugging it now.

@MartinBüttner Good call. I will do that!

@MartinBüttner Thanks!

Martin Büttner

no problem

@PeterTaylor I don't expect you to ;)

Martin Büttner

9:26 PM

@overactor The basic idea was to have 3 characters with different skills (similar to Trine), but you choose one for the entire game (unlike Trine), which basically picks a campaign. some levels are unique to each campaign, but most levels appear in all of them... however, in those cases, each character has a different path through the level, depending on their skills

Martin Büttner

9:48 PM

@overactor Are you working on a Marbelous submission for PhiNotPi's rosetta stone challenge? :P

githubphagocyte

@MartinBüttner I love your circular game - it's quite addictive!

Martin Büttner

:) thanks!

I'm starting to run out of languages... I know Lua, but I don't want to try to steal that other user's sub-win, because I currently have enough :D ... I know C#, but I can't be bothered to open Visual Studio... and I feel like I should know a lot more languages, but I can't remember which ones right now :D

ah, PHP!

If you know any Java the current one seems quite bloated at 457.

And I'm pretty the Groovy answerer isn't sure what golf means at all.

« first day (1312 days earlier) ← previous day next day → last day (3529 days later) »

Transcript for

Sep '142

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