« first day (6 days earlier)      last day (37 days later) » 

6:16 AM
@AidenChow Yes
@Steffan All of them, it never survived one round out of the 50 trials
 
@mousetail do i post that as a new answer or edit the already existing answer?
 
@AidenChow Edit your answer
You are free to submit a seperate bot too if you think it's sufficiently distinct and want to see it compete against the other version
But if you just want to update it edit your answer
 
@mousetail yea i dont think its gonna be too different, i just notice that my bot seem to mostly be eliminated when first semicolon is removed, gonna see if i can do anything about it
 
Ok I fixed the vyxal bot but now both jelly bots have 0 score again :/
@Steffan I found the issue I think, some environment variables where leaking from another bot's execution
 
6:32 AM
The quality of the answers here is just amazing! E.g. codegolf.stackexchange.com/a/251329/108721
 
@mousetail i update my answer, hopefully it do better lol
 
@ais523 I think the issue with your bots is other bots are still setting LC_ALL to various values and depending on which set last your jelly bots don't work. What value should I set it to to make it work consistently regardless of existing environment variables?
I tried setting it to "C" but that just made both bots fail every time
 
6:50 AM
:| i just check leaderboard and i drop to 4
wait its still using my old bot
 
Yes, I started the run before you updated yours
I'm running it again right now
It's taking ages
 
@mousetail riiip
i swear, my bot better not do worse after ive updated it
ayo i got 2nd
niiiiiice
 
Tapestry is entirely blue lol
 
wait how is the leaderboard ordered? median score?
 
Yes
average is tiebreaker
 
7:03 AM
yea that make sense, i was confused for a sec cuz my bot got 107 score even tho 3rd place got 108 score
 
the total scores fluctuate wildly between runs
median a lot less
 
ye that makes sense, median not affected by outliers
 
@NumberBasher I'm glad to see you have become smart
 
@NumberBasher ooh shit im in trouble
 
Maybe I should do it in python instead
 
7:24 AM
why
LOL
 
thats actually not good for my bot lolol
gonna have to start making a v3
XD
 
@NumberBasher I was hoping that your newfound intelligence would involve saying less LOL but it seems apparently the phrase is actually directly correlated to IQ
 
LOL IDK but I just like to use LOLOLOOLLOOLLOLOLOLOOLLOLOLOLOlo
 
@NumberBasher O_O
 
Why not submit something in LOLCODE? @NumberBasher
 
7:27 AM
Gotta admit idk what that is
 
@NumberBasher esolangs.org/wiki/LOLCODE Admiting you don't know things is a sign of inteligence
 
wtf... LOl
 
I love how "if" is "O, Rly?"
 
ok there we go i updated my bot so that i dont get destroyed by removing newline :P
 
I'm not going to do it b/c you have to start it with "HAI" so even the dumbest bot can probably do it
 
7:33 AM
ya i feel like a LOLCODE bot wouldnt do so well
unless maybe u just add a bunch of whitespace or smth
 
Doesn't mean you can't have fun writing it
Wait I think I found a bug which explains why example C bot is so good
 
what bug?
 
ya im interested to know how the example bot is placing so high
 
When the compilation fails it will not overwrite the old executable so it just runs the old version
It's probably really bad actually
 
that
's a big problem
 
7:37 AM
Agreed
 
bruuuh
ya that should be fixed
was wondering how it survived 2k+ rounds total
 
I just added rm -f code.c code.o to the start
 
@mousetail what dat do
 
remove the compiled version
every round
 
@AidenChow Yep I was wondering why the only weakness was the line break rather than any of the executable code. If everything was commented out it would compile but do nothing useful
Only way it could fail
 
7:40 AM
u running it again?
 
Yep
 
I editted mine @AidenChow now you are in big trouble
 
@NumberBasher bruh i saw
thats also not good
 
Ok now C bot is in a more natural place
 
good, update it @mousetail
you know what i think we need a new room
 
7:43 AM
I'm gonna wait 30 minutes before another run to give any other bot a chance to update.
 
dude i feel like we just gonna have an edit war or smth, with me defending against all ur targeted edits lol @NumberBasher
im already thinking of ways to counteract ur new edit of removing ='s
 
Ultimately the edit war is going to hurt you both since you will rarely be paired against eachother
 
@mousetail how does the pairing work anyways
 
@AidenChow There are 100 iterations. Each iteration robots are shuffled and placed in a circle. Each bot takes turns attacking the bot to it's right
The iteration ends when 1 bot remains
 
oh huh thats how it works
yea ok im probably just gonna leave my bot as it is then, was gonna have to remove some of my redundant chars from my other functions just to counteract number bashers bot
i assumed the pairings worked another way lol
i thought it was like all the bots get to delete a char from every other bot, random order of which bot goes first
 
7:51 AM
@AidenChow That would give a huge disadvantage to bots later in the loop since they can count on like 10 characters being removed before they even get a turn
Also it would be impossible to tell which character was the "killing blow"
 
@mousetail ye i suppose, but i was thinking like theres a random order each round so the disadvantage is spread out
i feel like how the bots are paired should really be in the challenge body, cuz there could be different strats based on the pairings
 
It is though?
Second paragraph
 
@mousetail ok fair enough lolol
somehow just skimmed right over that
ok bruh how come this time the 'n' in 'null' is causing the most fatalities for my bot, like what?????
 
@AidenChow It could be the final straw after a lot of other characters where removed first
 
@mousetail hmm idk, the leaderboard is still showing my v2 bot, removing the first newline should kill it. i wouldve expected that to be removed the most often because its near the beginning, but it doesnt seem like thats the case (of course i fixed the newline vulnerability in my v3 bot)
 
8:03 AM
@AidenChow The newline is blue meaning it's also fatal a lot
 
@mousetail yea but not the most often, surprisingly
nothing much i can do to defend against that unless i knew which bot(s) were removing that char
 
@AidenChow Can you include a base64 version of your bot btw?
 
@mousetail idk how to :|, how come u need it? i dont have unprintable chars or custom codepage
is all ascii
printable ascii
 
@AidenChow It prevents extra characters slipping in when copy pasting for bots that are very close to 256 byetes
 
ok how do i do it?
 
8:07 AM
There are online converters or the base64 console command
 
base64 encode right?
 
yes
 
atob
 
@mousetail ok edited my post
 
Wait no btoa
 
8:10 AM
@emanresuA wats dat
 
Open the console
 
@AidenChow Your base64 encoded version is 279 bytes
 
@mousetail da fuck
how
 
Probably the same newline issue that number basher had, you used \r\n instead of just \n
 
8:11 AM
ok im decode base64 and see what i get
uhhhhhh its 256 bytes
 
@emanresuA what i do in the console
@mousetail bruh i didnt know i put \r in the code
how i take that out
bro its literally carriage return, unprintable
so i have to take out some other bytes to fix the problem?
 
@AidenChow uninstall windows
No, just use regular unix newlines
 
NVM I did it already
 
@emanresuA so like how i do dat
 
8:16 AM
@AidenChow Do you have python?
 
@mousetail ye
why ask?
 
Use import base64;base64.b64encode(code.replace('\r\n', '\n').encode('utf-8'))
 
@mousetail ok
i edited it
it look correct now?
 
@AidenChow yep
 
ayo back in 2nd :D
also number bashers newline killer bot somehow got 0 points???
 
8:26 AM
@AidenChow Oops I guess I broke it
 
also bubbler posted a new bot which u didnt include
 
Yes I know he posted it just after the run ended
I'm interested who would win, bubbler vs tapestry
similar strategy
 
damn tapestry 4574 total rounds survived
 
Tapestry has the worst K/L ratio in the game
 
55 rounds median wow
@mousetail trueeeee
if it got more kills it would dominate, it literally survives for so long
 
8:33 AM
@mousetail At least for score, I don't think median makes sense, just because there's such a limited range of possible scores
 
@emanresuA Mean is the tiebreaker
 
@mousetail oh no
 
@NumberBasher ato.pxeger.com/…
You need to pass a argument to exit()
 
I ran it in tio no problem
just add that
rhgu
 
GCC or Clang?
actually it works in neither
 
8:40 AM
oh, i see, then i will add that
 
it looks like C (tcc) from his edit history
 
tcc?
 
except its for his old code
 
Ok hmm I guess that specific C compiler doesn't require a argument for exit. Should have been specified in the post if it required a very specific compiler
 
nvm it doesn't matter
You should run a script to start on boot and automatically grab the data and run every 10 minutes lol
> If a program reaches 0 bytes in length it will always be eliminated, regardless of if it produces valid output or not. Otherwise it would be unbeatable.
What if a program produces invalid output given an empty program?
 
8:48 AM
@NumberBasher ummmmmmmmm.... it just gets eliminated
all empty programs are eliminated no matter what
 
oh i see
 
tho i would be very shocked if a 0 byte program could actually produce valid output
 
via some fancy tags?
or just to eliminate others
shrug
 
@NumberBasher Rules state a invalid program is illegal
even if it would otherwise produce valid output
@NumberBasher In this case I'd consider the "tags" to be the program
 
@mousetail I see
@mousetail Hmm...
 
8:55 AM
yea that would be like a loophole i feel like, just put a 256 byte "program" that doesnt actually do anything, but the actual program is through flags, and boom, easy win
 
@AidenChow Yep, I hope people would downvote such a solution
 
Since flags are considered part of the definition of the language, that would probably count as using a language specifically made up for the challenge, which is a standard loophole.
 
There's probably some esolang implementation with a debugging print statement that prints an index and a char
 
@pxeger I could argue that it might just be a delicate combination, not that the flag was "made for such a purpose"
Or even better, use MegaGolfLang...
 
@emanresuA perhaps, but to also be lucky enough to get the correct char from the input...
 
8:59 AM
Not necessarily, idk
 
shrug
 
9:18 AM
Um, can I request a re-run, @mousetail
 
@NumberBasher This afternoon
Want to wait till Kevin submit's his answer
 
OK
Wait, when is afternoon
LOL
Which time zone
 
CEST
 
ok
That does not seem long
 
I'll run when Kevin submits
not sure how long that will be
 
9:24 AM
ok, let's hope that's not 10 thousand years later
 
I'll run it end of the day if nothing happens
So far every time I run today someone submits or edits immediately so I need to start again
 
lol
I would say run once every hour
and dont care if they submit.
BTW, SOMEONE IS PROBABLY GOING TO MOVE ALL THIS STUFF
 
It takes a long time to run and while it's running I can barely do anything else
You are free to run it yourself if you like
 
Honestly yeah we should make a room for this KOTH
3
 
me too thinks so
@mousetail oh, i thought you could just put it in the background
@mousetail too much configuration + stuff to do
 
@NumberBasher It's not that hard, it runs in docker
 
9:41 AM
403 messages moved from The Nineteenth Byte
 
@emanresuA thanks, probably took a while to find days worth of messages
 
Why am I here?
 
I probably moved a message of yours by accident
 
kthxbai
 
9:43 AM
403 messages, god
 
@AdĂ¡m APL submission incoming?
 
10:28 AM
Hello?
 
10:43 AM
New version of C bot:
aabbccddeeffgghhiijjkkllmmnnooppqqrrssttuuvvwwxxyyzz = 0x1234567890;
k,j;main(i){for(j=getchar();i<255&&k>=0;i++){k=getchar();if(!((k&168)^'(')){printf("%d %c",i,k);exit(0);}}printf("0 %c",j);}
 
Burrowing wheel still one-shots it
 
true
maybe I should change the alphabet with a more random array of characters
 
Won't touch it then
Most likely it'll attack the printfs
Never mind, it's still targeting exit (and/or main)
It's a really clever algorithm...
 
Wait how, it's just run length encoding right?
 
Idek how the compression works but it's more than just RLE
 
10:57 AM
@emanresuA Ok this defeats it
tLdvKw,sWaTTLMdnoMiApkYQjhYsiIbZCDurvwZVcZNcELjVqnAyaTPvvYoXl,*qtMPsnpwK,xw,FwmgtmKIljxSgxyxqAjdOAAoY=0x17312B2;
k,j;main(i){for(j=getchar();i<255&&k>=0;i++){k=getchar();if(!((k&168)^'(')){printf("%d %c",i,k);exit(0);}}printf("0 %c",j);}
 
It lasts three rounds
(in my testing)
 
Honestly good enough for me
It won't face wheel that often
 
Sometimes only 1
 
Is the leaderboard updating?
@mousetail the Newline Killer was updated 2 hours ago, not reflected on the leaderboard
 
Yes I know, I'll update it near the end of the afternoon or when Kevin posts, whatever happens earlier
 
11:16 AM
ok
I wonder if we can use -1 to indicate the last character?
 
No
 
Must match \d+ regex
 
I'm writing a new bot
 
Cool
What language?
 
11:18 AM
A mix of sample C, targetting, and distributed damage
written in old fassion python
 
Cool, interested to see it
 
is it possible to just use input()
i dont think so
 
You can
but it will read just one line
and might fail for non-unicode but that depends on factors
 
oh
sys.stdin.read()
should work
right?
 
Yes, if you use binary mode
 
11:20 AM
wdym
 
I use open(0,'rb').read()
sys.stdin.read() should work depending on encoding env
 
oh ok
161 bytes, time to add some comments lol
You should add the b64 encoding link into the question
 
The site isn't actually a very reliable way to do it though
Using python is better since you can check the length
 
explanation check
0
A: Radiation Hardening KOTH

Number BasherPython3, Key Characters Attack, 255 bytes ### repetitionrepetitionrepetition ### =lxn_1";=lxn_1"; ### $(>+%cstststststststststststst from sys import stdin as s; i=s.read(); for c in'\n )}.01+\n )}.01+'[len(i)%8:][:8]: if c in i:print(f'{i.index(c)} {c}');exit(); print(f'{len(i)-1} {i[-1]...

good
Can you help me look for bugs or feedback for github.com/TvoozMagnificent/SunSip ? Kinda need it soon
 
@NumberBasher I'm not gonna run your code
 
11:29 AM
why just asking
 
@NumberBasher this looks ok
 
@mousetail why
hmm
 
Why did you put the code twice?
 
to clarify the tabs
 
Ok
make sure when you base64 it that the tabs are encoded properly, they sometimes get lost when copy pasting
 
11:32 AM
I don't want to base64 it i dont see the reason
 
Because you have tabs is a important reason
It avoids mistakes by me
 
oh, ok
 
It makes you responsible for describing exactly how it should be encoded, so I don't have to spend a long time fixing other people's mistakes
 
I can't paste the tab into a b64 though, is there a python program or something...
 
11:35 AM
thnx
 
11:46 AM
That might be a reason for you to run it again LOL
 
11:59 AM
@NumberBasher Ok It's updated. How tf did sorry get so many more kills than tapestry despite living for shorter. They have the same basic strategy
 
 
1 hour later…
1:13 PM
quick fix on the go...
 
 
1 hour later…
2:34 PM
DONE!!! @mousetail
Gotta go, just ping me
 
 
8 hours later…
10:30 PM
Yoooo my bot is still in 2nd, idek how it’s doing so well lolol
 

« first day (6 days earlier)      last day (37 days later) »