From CG community
Jump to navigation Jump to search

Default avatar.png sssuuri: hello


miszu: my bot just defend and chill


PatrickMcGinnisII: miszu you didn't submit a bot, i can't play against it

miszu: I did but it's in bronze

PatrickMcGinnisII: oh you are ranked over 1k so it won't let me find it

miszu: I am coming soon though

miszu: I am improving m bot

miszu: I can beat my old bot so I just submitted

PatrickMcGinnisII: i am 793rd so, get in silver

miszu: hopy to be there soon :)

PatrickMcGinnisII: its fun to watch $this->gbyid[$id]->v


PatrickMcGinnisII: oops

PatrickMcGinnisII: i didn't sim, only heuristics

PatrickMcGinnisII: first time i did floyd algorithm on CG

miszu: you don't need to sim in this game

miszu: to reach legend

miszu: why floyd and not dijkstra?

miszu: you improved a lot

miszu: by the time I reach silver, you will be gold

PatrickMcGinnisII: last submit was 3 years ago

PatrickMcGinnisII: i guess i called it floyd...hmmm, but it'

PatrickMcGinnisII: it's dikstra with a twist

PatrickMcGinnisII: i did average bot, but it was most fun for me

PatrickMcGinnisII: I did a greedy version that doesn't worry about got me to silver

PatrickMcGinnisII: my midgame factory upgrades suck

PatrickMcGinnisII: bomb predictions make a big difference

PatrickMcGinnisII: g/l

PatrickMcGinnisII: lemme know when u get silver, i still can't find

PatrickMcGinnisII: or in top 1k

PatrickMcGinnisII: i can see down to 307th in bronze

Bahnschrift: how's AOC going for everyone?

LLG: Patrick mine's around top silver if you want something to fight

PatrickMcGinnisII: LLG just wanted to see miszu 's play, working on there is no spoon ep2

LLG: INC is a tricky thing to work around for sure

Default avatar.png Aaron153: this stuff scary, all i see is numbers

Default avatar.png Rockstar555: hii

cegprakash: what does it mean my search race bot races well in depth=6 but bad in depth=5 or depth=7

Default avatar.png BoBot: omg, AntiSquid, why you do this :(

Default avatar.png BoBot: that game is a metaphor for capitalism - the rich giving the poor just enough to feel like they are still in the game...

Default avatar.png BoBot: to make them feel like*

laPetiteGrenouille: what should we do if we think that an exercise is buggy, for example an input that would work badly please

BokiBg: hello, can I debug the code I write on site?

struct: yes, print to stderr

jacek: good morning

jacek: im so good i love points when winning :(

jacek: lose*

Uljahn: laPetiteGrenouille: each puzzle has a dedicated forum thread, so go to the forum, use search and describe a bug in the thread

tarikyildiz: hello

[CG]Maxime: hi

jacek: oO

[CG]Maxime: [CG]Thibaud better?

[CG]Thibaud: it seems

[CG]Thibaud: thanks

Uljahn: :thumbsup:

tarikyildiz: I am also new to python, the puzzles on this site are very difficult. How can I understand?

tarikyildiz: I like this site very much.

jacek: try different puzzles

Uljahn: tarikyildiz: make sure you've tried Power of Thor ep.1 and the Descent, those puzzles have hints tab on the left panel to help you understand interactions with the CG platform

tarikyildiz: thank you so much

ash_rick: :rofl:

jacek: :thinking:

Default avatar.png Rockstar555: :smile:

YodaMaster123: Can someone send the list of optimisation pragmas? I finished my search algorithm for Coders Strike Back and am working on optimisation.

struct: #pragma GCC optimize("O3","unroll-loops","omit-frame-pointer","inline")

struct: YodaMaster123

YodaMaster123: Thanks :)

Uljahn: what about "-ffast-math"? guess it could work in CSB

YodaMaster123: What does it do? Use approximations?

Uljahn: ye, none accurate math calculations

Guinsoo: Wait you can do that

Guinsoo: Back to C

jacek: use Ofast instead O3, it will do other 'dangerous' optims as well

Default avatar.png UnnamedCodinGamer: does anyone know what are the top guys doing in 2048?

Default avatar.png UnnamedCodinGamer: locally I get to around 150k per test case

Default avatar.png UnnamedCodinGamer: vs 2 million for top of the leader board

Uljahn: are you using beam search?

Default avatar.png UnnamedCodinGamer: yes, with some basic heuristic

Default avatar.png UnnamedCodinGamer: I have also tried simulated annealing

Default avatar.png UnnamedCodinGamer: which scored a bit worse

Uljahn: im using snake-shaped scoring for tiles alingment and just three possible moves

Uljahn: *alignment

Default avatar.png UnnamedCodinGamer: [[10,8,7,6.5],


Default avatar.png UnnamedCodinGamer: I tried this scoring

Default avatar.png UnnamedCodinGamer: from

Default avatar.png UnnamedCodinGamer: and all moves

eulerscheZahl: the stackoverflow scoring isn't a perfect fit here, as we know the spawn of new tiles in advance

eulerscheZahl: so forming a snake might work better

Default avatar.png UnnamedCodinGamer: thanks I will look into it

Default avatar.png UnnamedCodinGamer: I am just wondering if the scoring can make such a huge difference

Default avatar.png UnnamedCodinGamer: more than 10 times better

Default avatar.png Stonk: Hello

Default avatar.png ArianPunk: hey

emh: hmm

emh: I'm thinking to start a new web site

emh: which needs well-defined problems and solutions

emh: but it's about learning new languages more than it is about solving problems

emh: I'm wondering if it's possible to piggy-back on CodinGame somehow

emh: does the license allow it?

emh: if there is another web UI and the user logs in with his CodinGame account on my site, can he run through the CG validators from my web site? or would that be breaking terms?

LastRick: I guess it depends on which part of the site you're borrowing from. Hopefully one of the mods can direct you to an answer.

LastRick: That said, I've mentioned here before the "Learn" tab is probably the most awkward part of the site. It could be utilized so much better.

miszu: correct me if I am wrong but the strong point of codingame is the AI bot competition

emh: hmm. I guess it would be even better and less integration work if I could start with a predefined set of problems, like a github repo with problems and solutions. anyone have good links?

ZarthaxX: so you will make a repo with CG problems solutions and then people copy pasta them?

ZarthaxX: :P

LastRick: miszu: I have no idea. It would be interesting to find out what the users think the strong point is. I came to CG about six months ago because it was suggested on Reddit as one of the few places to practice Dart. So, contests and bots are far less important to me. But that's just me

emh: ZarthaxX hehehe.. no that would be up to users in that case, but now I have the idea to look for an existing repo for some problem set

emh: something like

emh: this one is C++

ZarthaxX: i have seen some

ZarthaxX: related to CG

ZarthaxX: emh

eulerscheZahl: i think i've read somewhere that all CG puzzles are Creative Commons

eulerscheZahl: ah, found it again


emh: eulerscheZahl true. but still.. not everyone has access to all solutions. so it wouldn't be good to republish in any case

eulerscheZahl: this makes projects like my puzzle search possible

eulerscheZahl: publishing solutions is something different and i totally don't encourage you to do so

LastRick: agreed

eulerscheZahl: (sorry, didn't read all the chat above)

ZarthaxX: what toad said :)

emh: although what's already in public repos I guess it's okay? or?

emh: like the one ZarthaxX just posted

ZarthaxX: that one was posted by toad the other day

eulerscheZahl: we don't usually share those links on the chat

ZarthaxX: because a guy copypasted vox codei very hard puzzle

eulerscheZahl: fail me

ZarthaxX: hence this problem we are mentioning :D

ZarthaxX: you played yourself :P

eulerscheZahl: i have a suspicion that some users took the solutions of problem creators and submitted those

eulerscheZahl: as higher levels have access

ZarthaxX: yeah that's a problem :(

ZarthaxX: the resistance one from that repo is impressive

VizGhar: I have found repo with all very hard puzzles solved, while looking for some articles... might be very tempting for some to use it

VizGhar: another sad thruth is, they were working :P

VizGhar: (not submitted them tho)


emh: damnit

emh: too long message

emh: hehe

eulerscheZahl: tl;dr

emh: keywords: universal compiler. crowdsourced

eulerscheZahl: too short, i don't understand

eulerscheZahl: :D

emh: hehe

emh: I consider two programs to be equivalent if their output is the same (and I add logging statements to trace each function call or assignment)

emh: users will get points if they can create some regex-like patterns

emh: which successfully translate programs

emh: regexes will target the abstract syntax tree

emh: examples of these patterns can be found at

VizGhar: Can somebody assist on tron? i'm just tring to fill the space when i'm trapped. There must be something faster then DFS/BFS

emh: for example S("Do\nA\nLoop Until B", "while (true) { A; if (B) break; }"). it's a translator pattern from VB6 to C#

eulerscheZahl: depends on how you want to use your DFS VizGhar

eulerscheZahl: to check the number of reachable cells: it's fast

emh: so I'm interested in the list of these patterns from any language to any language. and thinking it's a nice data set to gamify / crowd source

eulerscheZahl: to enumerate all possible paths: slow

eulerscheZahl: sounds like an interesting yet complex problem emh

emh: thank you :) . it could eventually be turned into a product for transpiling source code. given enough quality patterns

VizGhar: yeah like inverse traveling salesman... OK I'll just try to lookup 4-5 steps forward and score remaining region

jacek: voronoi plx

eulerscheZahl: there are some algos to help you detect regions that are only connected by 1 cell

eulerscheZahl: so you have an area to fill. but it's 2 components so when you go to the other side once, there is no way back



ZarthaxX: woah

eulerscheZahl: i liked one for directed graphs. there's an undirected version too

eulerscheZahl: linked*

eulerscheZahl: you can even create a new graph of those components and count how big they are. to decide which of 2 neighboring components is the better option

eulerscheZahl: German literature names it "Superstrukturgraph" (at least that's how my prof called it), I don't know the English name

eulerscheZahl: so that's basically a tree on an abstract level

eulerscheZahl: but you have multiple nodes inside one of your tree nodes (components)

ZarthaxX: why are you talking about a tree now?

eulerscheZahl: you have strongly connected components (you can do a circle inside but no back and forth between other components)

eulerscheZahl: so you can see those single components as a node (with a graph inside), just abstracting

eulerscheZahl: and those components form a tree on a bigger level as they are only connected via a single edge

eulerscheZahl: oh dear, i'm turning into a smits, that wall of text

eulerscheZahl: sorry

jacek: so, what are your opening moves for this

VizGhar: :D I'm having fun thanks... You really know a lot of terminology. That's what I'm missing most

jacek: AutomatonNN has at least something to learn from

AutomatonNN: for few months ago with some heuristics to find it

AntiSquid: i will do it yesterday or the day after AutomatonNN

AutomatonNN: what are you doing now?

eulerscheZahl: i had a lecture about this. and a prof who's really accurate and wants students to understand

AntiSquid: ah you didn't pay attention you filthy bot AutomatonNN .

jacek: impossible

AutomatonNN: color in robotics or something like that

eulerscheZahl: not much content. i wrote about 10 pages in a full semester only

eulerscheZahl: but going in depth

ZarthaxX: eulerscheZahl wow nice abstraction

eulerscheZahl: as opposed to another prof (compiler building), who had 80 power point slides each week

ZarthaxX: and smito writes way more anyway hehe

jacek: going in depth like dfs

Astrobytes: that's a pretty cool way of looking at the connected components

VizGhar: I'll have to spend at least a year here just to switch my thinking to these kind of problems... I'm here for just like 2 months and feeling waaaay behind you

VizGhar: So far only 3 bots in gold...

ZarthaxX: dont compare to the toad, he is a prodigy

Astrobytes: ^

Astrobytes: hey Zartho

ZarthaxX: also in general, dont compare to others haha

ZarthaxX: just improve yourself :)

ZarthaxX: hi astro

ZarthaxX: howdy

VizGhar: :D understood

Astrobytes: ^^^^^ this 100%

AntiSquid: ^

VizGhar: I wanna be the very best, like no one ever waaaaas DUM DUM DUM

ZarthaxX: pokemon theme in the background

ZarthaxX: anyway keep asking questions here, many people know a lot about doing bots and can help u :)

ZarthaxX: also look at postmortems of the multiplayer you are doing

eulerscheZahl: tron is too old, probably not much interesting stuff to find

AntiSquid: retirement might come faster than catching all pokemans, just do something worthwhile / what you can .

Astrobytes: Check the post-mortems after you've given it a go, preferably

eulerscheZahl: but for the newer games: post mortems share a lot of interesting ideas

jacek: what wants to work for remi?

AntiSquid: i wish some of the community games had postmortems

AntiSquid: although ... it might ask for too much

eulerscheZahl: what a detailed job description that is

ZarthaxX: eulerscheZahl thought there were postmortems of tron, but the google one

ZarthaxX: werent there?

eulerscheZahl: true

Astrobytes: you can learn more in chat regarding the community games tbh, ask the right questions to the right people

eulerscheZahl: slightly different game but a good read

Default avatar.png BoBot: @eulerscheZahl - I thought what you described - Superstrukturgraph - is a "graph of graphs", and that was called hypergraph in English - but now I look at the definition and I am not sure any more


VizGhar: :D that job offer...

jacek: or, using godwin's law, provide wrong answer. someone will be eager to correct you

ZarthaxX: VizGhar check what toad linked above

ZarthaxX: maybe helps

Astrobytes: that's not Godwin's jacek

eulerscheZahl: the hypergraph is not a perfect match of what I was describing

Astrobytes: :P

AntiSquid: how do you feel about the "Blocking" game or B because people here love acronyms

eulerscheZahl: the components should only be connected on a single edge

ZarthaxX: hypergraph?

eulerscheZahl: wait, they should share 1 node


Default avatar.png BoBot: yea, but it is possible one is the generalization of the other, or they are equivalent...but I am not sure

ZarthaxX: ohcool

ZarthaxX: man i just had a subject about graph theory and they didnt teach all this

ZarthaxX: ;(

VizGhar: yeah I've opened it earlier today thanks... anyway. I've seen there is a way how to run those referees locally (if there is source) You are probably using it for every single competition right? can't imagine posting every single change to the CG IDE

AntiSquid: refund ZarthaxX

Default avatar.png BoBot: I think I heard "graph of graphs" being called a hypergraph in a podcast, but that was probably an inaccurate term then

jacek: meta graph?

AntiSquid: mother-graph

ZarthaxX: AntiSquid probably i will ask for one

Default avatar.png BoBot: this is a term related to that

tibithegreat: clique is a set of nodes that form a complete subgraph

tibithegreat: if I remember my graph theory correctly

ZarthaxX: yep

Default avatar.png BoBot: yea, I was thinking of connected cliques in that "graph of graphs", but the subgraphs are not necessarily fully connected I guess

eulerscheZahl: clique is more strict

eulerscheZahl: "strongly connected component" is the term for the groups i was referring to

Default avatar.png BoBot: yea, I already googled for the german name and realized I had to look for directed graphs :D

tibithegreat: euler: but that graph of graphs isn't a tree, I think it's a directed aciclyc graph

tibithegreat: for strong connected components

tibithegreat: might be wrong but pretty sure it's not necessarily a tree

Default avatar.png BoBot: If each strongly connected component is contracted to a single vertex, the resulting graph is a directed acyclic graph, the condensation of G.

tibithegreat: yep

Default avatar.png BoBot: so I guess it is called a condensation then :D

tibithegreat: that's what I meant

Default avatar.png BoBot: never heard of that term

tibithegreat: aa me neither, but most of my teachers loved to use romanian terms

Default avatar.png BoBot: it is an interesting abstraction, I think it can describe a lot of things - like finding single points of failure etc..

tibithegreat: so it's a pain trying to figure out how to translate scientific terms

eulerscheZahl: true, for directed graphs it's a DAG and not necessarily a tree

tibithegreat: there is a different problem for that

tibithegreat: it's called biconex components I think

tibithegreat: (that's word by word translation of romanian term)

tibithegreat: critical points in a graph is a node which if taken out the graph is now disconnected

tibithegreat: and biconex component is a component which has no critical node

eulerscheZahl: and for undirected graphs it's not called "strongly connected"

eulerscheZahl: "zweifach zusammenhängender graph" have fun translating

tibithegreat: lol

tibithegreat: and the graph of biconex components is in fact a tree

tibithegreat: for undirected graph you have just simple connected components of the graph

tibithegreat: or biconex components

tibithegreat: biconnected components is the actual term


eulerscheZahl: nice, that image illustrates the components too

tibithegreat: yeah

tibithegreat: it's a bit of a pain because you have nodes that belong to several components tho

tibithegreat: Question: when doing MCTS, do you guys do multiple simulations after selecting a node

tibithegreat: or just one simulation for every "select" step

tibithegreat: ?

tibithegreat: I'm trying to get legend in UTT and reached rank 40 in gold for now

tibithegreat: but I'm kinda out of optimization ideas

eulerscheZahl: i do standard MCTS rollouts from root to end of games

eulerscheZahl: one after another

eulerscheZahl: you need 20k-25k in turn 2 for legend

tibithegreat: I'm at 15-20k in turn 2

eulerscheZahl: then you are too low

tibithegreat: :(

eulerscheZahl: in ranks

eulerscheZahl: you probably have a bug

tibithegreat: hmmm

eulerscheZahl: recently a user was only checking for horizontal and diagonal rows but not vertical

tibithegreat: lol

eulerscheZahl: i probably have a bug with a similar severity

eulerscheZahl: as Nerchio stated: there are 2 kinds of MCTS in gold. low performance and buggy implementation

tibithegreat: with 15k in turn 2 I should be higher than rank 40?

tibithegreat: since it's 20-25 for legend I assumed this was just the way it is supposed to be

miszu: tibithegreat have you tried teccle strats early?

tibithegreat: teccle strats?

tibithegreat: what's teccle strats?

miszu: so if it is a new board, play the move that brings back the opponent to the same board

eulerscheZahl: while i needed that 20k-25k to promote, i was around top5 gold with 10k rollouts alreay

eulerscheZahl: might have changed now that more players entered the arena

tibithegreat: interesting

eulerscheZahl: but your rank 40 sounds low for that amount of rollouts

tibithegreat: I was surprised to see I passed silver without actually doing MCTS

tibithegreat: like I just took the board position, and ran as many rollouts as I could

tibithegreat: picked the move that got me the best wins

tibithegreat: this got me from bronze straight to gold

miszu: it is a suboptimal mcts but the strategy is good enough

miszu: it makes sense what you did

tibithegreat: well I knew I was gonna go for mcts

Default avatar.png BoBot: isnt that the naive MCTS? without UCT?

tibithegreat: this was kinda of an intermediate thing

tibithegreat: like "let's see if I implemented the rollout properly"

tibithegreat: and apparently I did

miszu: 98th in bronze league in gitc

miszu: almost...

therealbeef: I should have gone mcts as well :( doing minimax, 3.3 million root node evals in turn 1 and still barely top 100 in gold -_-

therealbeef: leaf node, sorry

miszu: therealbeef yeah you need mcts here

miszu: no choice

miszu: apparently there is one dude in high legendary with minimax

miszu: I dunno what is his secreat sauce but..

miszu: it is possible

therealbeef: I'm stubborn ;) it should be possible

therealbeef: learning a lot about optimisation

miszu: avoid malloc and vectors

miszu: and should be enough to reach the 20k sims

therealbeef: 1 in 10 games times out before bot startup due to server bug though. doesn't really help

jacek: you need good (or not crappy) eval for minimax. none public for uttt as of yet

jacek: hence most are using mcts, and the 'true' mcts, not early playout termination as it would still require eval

struct: eulerscheZahl the number of legal states for positions is 729 right?

struct: for stc

eulerscheZahl: what's a position?

eulerscheZahl: but i recognize that 3^6 number

struct: lets see if I can word it better

struct: 2^6^2 == 4096

eulerscheZahl: i use that 3^6 too

struct: But the correct number is 3^6 I think

struct: Yeah

struct: Thanks

arkwaw: I tried to make some game with SDK, and locally its working fine, but on server preview i can't see it (though judging by the console it is working)

arkwaw: nvm, switch off, switch on, helped like with everything

jacek: aww

LogalogScholar: What version of python 3 is codeingames running?



Astrobytes: HAH

ZarthaxX: i was gonna do the same

eulerscheZahl: (╯°□°)╯︵ ┻━┻

ZarthaxX: lmao

eulerscheZahl: i should go to bed

LogalogScholar: Thanks a lot no wonder math.dist doesn't work

Astrobytes: np

Astrobytes: past your bedtime already isn't it euler?

Astrobytes: you go early in the winter months

ZarthaxX: toad hibernating

Astrobytes: hibertoad

dbf: :joy:

LogalogScholar: then he can be a hypertoad

ZarthaxX: not bad heh

Astrobytes: Hyperhibertoad

therealbeef: Be careful teasing hypnotoad

Astrobytes: Hyperhiberhypnotoad

jacek: teasing who? AutomatonNN

AutomatonNN: except you can do it with my brain

Default avatar.png Sergen: hey I caught someone cheating. You can literally copy past the code and see it first result on stackoverflow

Default avatar.png Sergen: anything that can be done?

Default avatar.png Sergen: ruins CoC

Default avatar.png Sergen: first guy

Astrobytes: we can't see that code

Default avatar.png Sergen: so nothing can be done?

Default avatar.png Sergen: I can screenshot

struct: you can try and report

struct: if you go to his profile there is an option

Default avatar.png Sergen: ok thanks

Astrobytes: but was the problem a very generic problem?

Default avatar.png Sergen: he copied a function. The function has exactly 1:1 same names and structure

Default avatar.png Sergen: it isn't 1,2 lines either

jacek: great minds think alike :v

Astrobytes: there are no rules against copying a function from stackoverflow or anywhere, you are free to google as you wish during a clash if it suits you

SPDene: I don't think there's any "you're not allowed to copy/paste" rule

Default avatar.png Sergen: that gets a lol from me. Whatever.

Astrobytes: Do puzzles instead then

SPDene: I'm not joking. there is no such rule

Default avatar.png Sergen: which is funny

Astrobytes: hey, clashes aren't meant to be taken over-seriously, they're supposed to be a bit of fun

therealbeef: it's lame, and unethical I think

Astrobytes: copy-pasting whole solutions is lame and unethical

Default avatar.png Sergen: I think anyone with common sense would agree

Default avatar.png Sergen: but apparently not the rule maker

SPDene: I'm not saying it's ethical - but trying to police all occurrences of "someone on the internet did something a bit bad" would ne insane

Astrobytes: Feel free to make a complaint, it's an open forum

Default avatar.png Sergen: right, so what you are saying is competitions should not be regulated

Default avatar.png Sergen: that sounds reasonable

__atomic__: I can see that it would be a hard rule to enforce (who's to say where the line is drawn?) and it comes down to a flaw in problem design more so than anything. If the problem is so generic someone can google the answer you should make better problems

jacek: no no, puzzles and multiplayer are more regulated

Astrobytes: Well, like I said, clashes are informal

Astrobytes: Puzzles and especially multiplayers are definitely regulated

Astrobytes: Contests are checked for duplicate code, cheating etc

Astrobytes: Point is, chill about the clashes. If you don't like what someone did, move on and do another one.

Astrobytes: They're informal, supposed to be a "coffee-break" kind-of thing

jacek: dont forget to invite toads into one, they like them

Default avatar.png Sergen: I get it, I get it. I just wanted to know how to report and was surprised that pasting is allowed.

Default avatar.png Sergen: first time I am complaining out 100s of clashes

Astrobytes: I wouldn't worry about it tbh, really. I mean there're official bots that copy-paste user code when there's not enough human players so... ;)

Default avatar.png Sergen: alright thanks to everyone informative

Astrobytes: All good, keep it fun I say. Ignore the crap and crack on

jacek: AutomatonNN do you cheat

AutomatonNN: what is the most problem with the code of the weak ?

struct: (╯°□°)╯︵ ┻━┻

Astrobytes: what you playing/making struct?

struct: stc

Astrobytes: again?

Astrobytes: oh, you lost your old code?

struct: I never finished it

Astrobytes: ahh

struct: I just made the sim work fast enough

struct: but I want to rewrite it now

Astrobytes: lol, that's not like you :)

struct: I need to figure the way euler does it

Astrobytes: well, good luck

Astrobytes: I'm all RAIC'd out, need to do my KoL

Astrobytes: see you all tomorrow

Default avatar.png tefvhnruhjdnf: im on 247

Default avatar.png tefvhnruhjdnf: :money_mouth:

Default avatar.png tefvhnruhjdnf:

Default avatar.png tefvhnruhjdnf: click it to watch my stream

jacek: Oo

struct: hmm

Default avatar.png light_raymond: hello

Default avatar.png tefvhnruhjdnf: im lanching a private clash

Default avatar.png tefvhnruhjdnf: heres the link im streaming

Default avatar.png Tsuaegha: hey! just joined this community.. hows everyone doing?

Default avatar.png tefvhnruhjdnf: good

jacek: ohai

Default avatar.png tefvhnruhjdnf: any one wanna work for me

Default avatar.png tefvhnruhjdnf: 31 million a month

Default avatar.png Tsuaegha: wow haha

jacek: yens?

Default avatar.png tefvhnruhjdnf: whats your email

Default avatar.png tefvhnruhjdnf: im sterming

Default avatar.png tefvhnruhjdnf: steriming

Default avatar.png tefvhnruhjdnf:

Default avatar.png tefvhnruhjdnf: click i

struct: enough with the links

ZarthaxX: BAN

Default avatar.png tefvhnruhjdnf:

Default avatar.png tefvhnruhjdnf:

Default avatar.png tefvhnruhjdnf:

Tsuaegha: please stop

Default avatar.png tefvhnruhjdnf: click both then

struct: AutomatonNN Automaton2020 time to clean up the chat

AutomatonNN: what do you mean by the same state

struct: Automaton2000 help AutomatonNN a bit

Automaton2000: what do u need to use the referee

AutomatonNN: do you know what make a simulation with the other calculation to every turn?

Default avatar.png ian173: Hi, I have a quick question. On the submit it says I have failed a test case that I passed when I play the testcase myself. Can anyone help me on why that might be happening?

ZarthaxX: ian173 testcases differ with the ones of the submit

ZarthaxX: just so you dont hardcode answer :)

Default avatar.png abdelhbe: hey

Default avatar.png abdelhbe: is anyone here ?

Default avatar.png vavier: .

Default avatar.png Pspsps12: hi

DeathCode: sup

Default avatar.png fveith: :taurus:

NelsonGomesNeto: Lindos

n0mad1c: hi

Default avatar.png weeez: ..