The Nineteenth Byte

1:07 AM

Wow! We've made it through a trial, so we have some results. We're nearing the end of trial 2, and then trial 3 should start running automatically. After that, barring exceptions thrown or timeouts in the last trial, we'll have the official results.

Doorknob

:D

Rainbolt

omg

Java wrote some jibberish to a file for only one of the pods....

ര䘊楲湥獤楨Ɒ〲㘵਍歓灥楴Ᵽ〲㤰਍潆牲獥䝴浵Ɒ〲㠶਍楐獵楴楴畯ⱳ〲㠶਍慌杵楨杮慍Ɱ㌲㔳਍敃汬汵牡畁潴慭潴Ɱ〲〷਍慔䅧潬杮㈬㠰ര䠊灩瑳牥㈬ㄶര倊潲慢汢䍥畡敳㈬㈰ശ‌䌊捹楬瑳㈬〶ല䰊瑡獥䡴瑡牥㈬ㄶര匊档穩灯牨湥捩㈬㐰സ䠊獡䅨摤捩ⱴㄲㄲ਍慐楲祴桃捥Ⱬㄲㄲ਍慌祺畔湲潣

What happened?!

Doorknob

Hmmm

Rainbolt

I can see the normal text in the preview window

But I can only see special characters when I open notepad lol

It's weird that this only happened for one pod in one trial

Ok so it looks like Notepad thinks that my file was supposed to be binary

I wrote a lot of zeros and ones, so it fooled Notepad

Now the question is... how do I get my results back lol

Doorknob

Lol

This is the hexdump

For some reason every 6th character is e

Rainbolt

1:16 AM

Oh that was a tiny portion of the actual output

Most of it is unprintable squares

You got a tiny portion of the chinese symbols

Doorknob

Oh :P

Doorknob

10:21 AM

0

Q: Slime: The Territory War

You are a glob of slime. Naturally, being slime, you want to ooze over as much area as possible. But there are 3 other slimes who want to do the exact same thing. Who will be the superior slime? Description All slimes will be gathered into one arena. The judges (i.e. the control program) will g...

king-of-the-hill

New Meta Posts

10:31 AM

0

Q: GolfScript has been removed from Wikipedia

GolfScript wiki article has been removed from Wikipedia as it doesn't meet Wikipedia General notability guideline and Stackexchange isn't a reliable source. Please read the discussion at Wikipedia. I don't agree with that decision, as it's popular and the best ranked esoteric programming langua...

discussion community-wiki golfscript

Doorknob

12:32 PM

Oh, btw @MartinBüttner, I realized that my slime thingy was probably too complicated because of your comment in the sandbox. :P I'm working on another, less complicated challenge (i.e. I can write a bot in one line of code even though it will perform terribly), but I'll probably be lazy and post it next Monday.

Martin Büttner

you did post the slime one though, right?

Also which comment was that? :D

Doorknob

Yes, see ^^^^^ a few messages ago. The one that said something like "Your 'very simple bot' is already very long :D"

Martin Büttner

oh right

well, you'll see how it goes

Doorknob

Yeah, I just now realized that if I saw my own challenge, I'm not sure whether I would take the effort to write up a fully working bot to it :P

overactor

@Doorknob I'll probably give it a shot, I've got an idea ready already.

Rainbolt

1:02 PM

Morning

Martin Büttner

afternoon

Doorknob

night

Rainbolt

I managed to botch two more trials so I have a total of three botched trials and two good trials. My antivirus scanner started running around 3 AM and it used so many resources that three more bots timed out. I don't think I can accept that :(

Doorknob

@Rainbolt Antivirus? Eeek.

Rainbolt

I call it an antivirus scanner, but I suppose it is actually a virus scanner

Doorknob

1:05 PM

On my old Windows computer, I once had one antivirus scanner detect a different antivirus scanner and they started fighting each other. (No idea why there were two antivirus scanners.)

Rainbolt

That makes a lot of sense, because antivirus scanner perform quite a few suspicious activities.

Martin Büttner

this guy codegolf.stackexchange.com/users/26997/calvins-hobbies ... member for 14 days and he asked 12 pretty decent questions without ever sandboxing one of them... what on earth...

Rainbolt

I asked four decent ones without ever sandboxing on my first run here

They were in decreasing order of quality, such that the first one got the most upvotes and the last one got 5.

But Calvin's definitely got me beat lol

Doorknob

1:30 PM

Heh, he even has one of the highest voted on the site (the arranging pixels one)

I don't plan to sandbox the new KotH that I'm making, but you can see a draft here

Martin Büttner

@Doorknob while that's funny, it's horribly tedious to read after the first 3 sentences :D

Doorknob

Heh I was kinda worried about that :P

Also the WiFi is terrible here (especially at night) so chat responses might be slightly delayed

Rainbolt

Oh wow... Google Docs workbooks can't have more than 200,000 cells

I am hitting so many speedbumps withthis freaking challenge

Geobits

It's almost 2015. Once everyone has a flying car, speedbumps won't matter so much.

Rainbolt

But even in the clouds, you are still limited to 200 thousand cells!

Doorknob

1:40 PM

@MartinBüttner I changed it so that only the intro and description are written in cave-speak. :P Will be updated in next commit

Geobits

That's just what they tell you to limit your control. Once you can see beyond the veil, the 200k cell limit doesn't exist.

Martin Büttner

1:51 PM

@Rainbolt there is no spoon

Geobits

@Doorknob Do slimes only jump in a straight line, or is a knight's move allowed? "exactly two squares away" could mean either, I guess.

Doorknob

@Geobits "Two squares away" counts diagonals, so yes.

Geobits

Thought so, just wanted to make sure :)

Rainbolt

Someone scored 499500 in 1000 rounds...

Martin Büttner

over all trials?

Rainbolt

2:01 PM

No just one

Martin Büttner

phew :D

that's pretty insane though

Rainbolt

Average score is around 2000

Martin Büttner

o.O

Rainbolt

But these two submissions just went crazy

Martin Büttner

oh

you mean over all participants lol

Rainbolt

2:02 PM

The sad part - the same submission scored 1912 in trial 2

Martin Büttner

are you still gonna do the pod with the winners from the three other pods?

Rainbolt

I don't think so

Should I?

Martin Büttner

I'm not entirely sure... I think being in the winner-pod you should probably even use a different strategy than in a general pod

so maybe not

Rainbolt

So all of the highest scorers were in a pod where people just pretty much voted one way all the time

So two submissions just kicked ass in that particular pod

Leading to a massive score

Martin Büttner

right

Geobits

2:04 PM

499500 is a perfect score, not just a massive one.

Rainbolt

I wonder if I should run more trials and take the median instead of the average

Say 5 trials, median score

Martin Büttner

actually, I almost forgot, I wanted to post the domino challenge today

any final comments? meta.codegolf.stackexchange.com/a/1861/8478

(particular whether adding the forwarders seems like a good idea or an unnecessary complication)

overactor

@MartinBüttner I'd opt for leaving them out, I think it would detract from the fun

The challenge is nice and concise without them, if difficult to explain

Martin Büttner

Thank you, finally someone with an opinion on that :D

overactor

Other than that it looks really cool, too difficult for me, but cool

Martin Büttner

2:15 PM

Okay, I'll wait another 15 minutes for feedback, then it goes on main.

Doorknob

2:28 PM

Alright, one last push and now I'm going to sleep

Hmm, the only reason I now know basic C++ is that I decided to learn it to write control programs :P

Martin Büttner

That's similar to how I know Ruby ^^

(I think it might even have been a challenge of yours, where I learned it)

ha, yes it was :D

Rainbolt

I started learning Ruby over the weekend a little. Enough to fix some people's submissions

Martin Büttner

:D not bad

@Rainbolt you know how much I value your opinion... would you have quick look at the domino thing before I post it? (I must sound really desperate by now)

Rainbolt

3:41 PM

@MartinBüttner Sure

New Sandboxed Posts

0

A: Proposed Question Sandbox - Mark XIV

I'm posting here first so I don't clutter up the site in case people are tired of pi questions. Thoughts? Yet Another Pi Question code-golf Today (actually tomorrow as I post this) is Pi Approximation Day (22/7), so here is a variant of the "compute pi" golf that I don't think has been see...

ProgramFOX

Just wondering, does @Rusher still ping you?

Geobits

No, he doesn't ping me any more. Sometimes Rainbolt does.

ProgramFOX

:D

Rainbolt

Nope

ProgramFOX

3:45 PM

Ok, thanks

Rainbolt

Do you miss my evil half?

Martin Büttner

I know it's still there somewhere....

Geobits

Wait, this is the good half?

Rainbolt

@MartinBüttner Am I correct that a diagonally falling domino affects three other cells when it falls?

The two that are adjacent as well as the one diagonal from it?

Martin Büttner

@Rainbolt yes exactly

Rainbolt

3:47 PM

Ok good

Martin Büttner

isn't there even an ASCII art showing that?

ah no, we removed that again

because the rules for how to affect the adjacent ones are different from the diagonal one

Rainbolt

Ok this is complicated so bear with me

You have this setup D - A

So the domino is being pushed on both sides but can't fall

If you push it from a third side, does it fall?

Martin Büttner

haha, good question

it should, yes... I might need to revise the wording there

Rainbolt

From the current wording, it appears that it is "locked in" and "can't be pushed by future dominos"

Martin Büttner

I'll add a word and another example

one second

Rainbolt

3:51 PM

Ok

Martin Büttner

how about now?

Rainbolt

I upvoted because looks perfect

Martin Büttner

awesome

thanks for having a read and giving it some serious thought ;)

Rainbolt

NP lol

Here's an interesting one:
C -
| \

It should transform to
# #
# \

Martin Büttner

nothing happens

Rainbolt

3:57 PM

Is that right?

Martin Büttner

no C can't push over those two adjacent ones

you'd need

C /
/ \

no that doesn't work either

Rainbolt

Why not? C is pointed down and right. - can fall down.

Martin Büttner

@Rainbolt no that's what I meant by "adjacent directions" have different rules

only C| and C/ will fall

Rainbolt

Oh ok I see it now

I missed that rule

Martin Büttner

Have a look at the | above Q example.

kk

Rainbolt

3:59 PM

So for your example
C /
/ \
you get
# #
# \
right?

Martin Büttner

yes, but not because the third one is locked, but because it's not hit by the two /

Rainbolt

I thought it was hit but not pushed

Martin Büttner

no, because again, it's adjacent but doesn't point in the right direction

(it's hit by the initial C though)

(but not pushed)

okay, simpler example: C\ results in #\ because the left domino "falls past" the right one

the only possible dominos that fall to the right of a C are C| or C/.

Rainbolt

I don't understand how it falls past. The rules clearly say that diagonal affects two adjacent and one diagonal, but in your example, it doesn't

And then says "If it affects it, but it isn't within 45 degrees, it doesn't fall"

So I interpreted as "Affects but doesn't fall"

Martin Büttner

"If these [adjacent] cells contain a domino which can fall in the same direction as the original domino, or in the axis-aligned direction away from it, that domino does so in the next time step."

Rainbolt

4:04 PM

How do I calculate the "axis-aligned direction away from it"

Maybe that's the part I'm missing

Martin Büttner

basically it's the vector that connects the original domino and the affected domino

I'll just add all four cases as examples

Rainbolt

Ok that makes sense, but I still have a problem

If it isn't falling in an axis aligned direction, that doesn't mean it wasn't hit to begin with

Martin Büttner

does it make a difference?

Rainbolt

For example, "D - " would not fall. But it was still hit

I was under the impression that being struck by a domino made a difference for cases when you are locked in

Martin Büttner

not if it's a direction in which you can't be pushed anyway

Rainbolt

4:08 PM

So I would need a consistent definition of what it means to be struck. I was going by your use of the word "affected"

Martin Büttner

(as in your first example)

Rainbolt

Ok. So you only affect nearby dominos if you are able to push them. That works

I'm gonna reread

Martin Büttner

the comment thread on this challenge is going to be fun :D

Rainbolt

Ok I think my comment was invalid

"if a domino is pushed in both valid directions at once"

I didn't see that before

Martin Büttner

yes, that was my addition after your first example

(which is why I said "I'll add a word and an example" ^^)

Rainbolt

4:15 PM

ooooooh

Ok

Martin Büttner

bloody hell, why do I always come up with these concise challenge titles...

Rainbolt

4:27 PM

I have that problem all the time. I had to change "Wild" to "Survival Game - Build Your Own Wolf"

I had to change "Pandemonium" to "Good Versus Evil"

Although I believe having a more descriptive title attracts more people from HNQ, so I'm not complaining

Martin Büttner

it depends... I'm more likely to not read a three-liner properly than a catchy title where I wonder what it's about

0

Q: Golfing for Domino Day

Given a setup of dominoes your task is to figure out which dominoes fall and which don't. Input Take a rectangular ASCII representation of a domino setup. The following characters are used for the ASCII grid: (space): empty cell |, -, /, \: dominoes Dominoes can fall in 8 directions, which...

code-golf

suggestions for catchier titles are always welcome

Rainbolt

Domino Doomsday

Background - the dominos are actually skyscrapers in a city.

The letters represent the direction Godzilla is facing

Martin Büttner

4:43 PM

and he's really good at cell division?

Rainbolt

6:21 PM

Congratulations to @MartinBüttner for posting challenge number 2222.

Martin Büttner

HA! thanks :)

I thought Ypnypn's was later though

Rainbolt

6:49 PM

Oh man this is painful for me to write. Our customer wants settings to be saved between uninstall and reinstall. But they don't want to give the customer the option of checking a box that says "uninstall settings"

In other words, once you install our software, part of it will be there forever until you manually find and delete the files I am about to place in a discreet location.

I wonder if this breaks my Oath of Silence

Geobits

Whether it is or not, that should be illegal.

Rainbolt

Would you be upset if a web config was left hanging around? It's just the web config file, and nothing else.

It's not that major.

Maybe a class B misdemeanor?

Geobits

I dunno. If I tell something to remove itself from my computer, I tend to expect it to not leave pieces.

Not much software seems to live up to that expectation, but a man can dream.

Rainbolt

I like the option Steam gives.
Me: "Steam, uninstall MyGame."
Steam: "Alright. Check this box if you also want to uninstall all the little pieces."

Geobits

Yea, that's fine. If it does it. Some programs just lie to your face[citation needed].

New Sandboxed Posts

7:30 PM

0

A: Proposed Question Sandbox - Mark XIV

El Farol Jail : The Game Of Happyness A King-of-the-hill challenge You and your gang are a famous mafia, gaining reputation for stealing stopwatch. One day, your mafia are caught on the action when you're stealing the most expensive stopwatch on the world. You're tried and found guilty, and...

Martin Büttner

7:50 PM

@Rainbolt Isn't this very close to being a duplicate of Good vs Evil? meta.codegolf.stackexchange.com/a/1871/8478

Rainbolt

8:01 PM

Seems like it

I mean, the answers from GvE could be copied

The scoring just has a different formula

And you occasionally get red carded and can't participate for a few rounds

Geobits

8:12 PM

And it might not ever end.

If people play conservatively at the end and only visit the gym every x days, the odds of at least 60% of the players choosing the same day are pretty low.

Martin Büttner

9:54 PM

@Rainbolt how is your testing progress?

on PPCG I'm really missing out on tag-related badges because I don't have enough answers o.O

Rainbolt

@MartinBüttner Someone ended up getting paired with 10 retards who vote the same way every round. He and one other got nearly perfect scores of 495000.

The average score is 2000. So I decided to go with a median instead of a mean

But that means I need more trials to get a median worth considering

I decided on 5. So you could have two bad outliers, two good outliers, and one middle score that hopefully represents your true performance

Does that seem reasonable?

Martin Büttner

Yep (and I was around when all of that happened... so I was wondering how you were getting on with the 5 trials ^^)

Rainbolt

I have two trials completed and in a spreadsheet. I hope to arrive at home and find three more trials waiting.

Martin Büttner

ah okay

Rainbolt

However, one of the pods within a trial seems to get botched 30% of the time and output weird results to notepad

It converts my zeroes and ones to Chinese characters and other unprintable symbols

So I've been dumping those trials in favor of the ones that don't get messed up

Martin Büttner

10:09 PM

H̤̥̲͉̩̠ͅË̦̩͈̲̳̤́͞'̪̻̗̬͚͕̠̚̕S̜͒̃̽̇ͯ̒̍ ͉̩͔̲̾̋͐̑ͫ̉̓C̖̰̪̣̻͖ͅŐ̯̻̻̠̙͙̄̆ͬ̍ͤ͟M͕͖͎̟ͣ̋ͥͅI̹̦͔̩̬̠͓ͯͬN̘̯̍͑̈ͧ͂ͨ̚͝G̫̺̱͖͎̾́̋͝ͅ ̲̹̓ͨF͓̻̪͚̼͈̜ͨ̽̀ͪ͑̓̾O͌̽͗̈̂̃́R̜̘̉ͪͥ̽̾ͣ ̠̯̟̟͈̟̀̓Y̺͗͌̿́͐̓͡O͙̹͐̂̈ͫͥU̪̫͋͠!̸͖̇̒ͮͣ̑

overactor

10:21 PM

@Rainbolt you could also consider a geometric mean

It rewards people who score rather consitently

though even there the perfect score probably messes everything up

Transcript for