Chat:World/2021-02-21

From CG community
Jump to navigation Jump to search

MasterCoderxD: not me bro, but i wanna try it

DaBestest: I've done like hackthissite basic stuff, but nothing other than that...

Default avatar.png Saad-py: ohhhhhhhhhhh

Default avatar.png Saad-py: Hi sup DaBastest

Default avatar.png Saad-py: Remember me

Sayach8: how long does it generally take for a clash of code contribution to get accepted or rejected?

Default avatar.png Saad-py: idk

Uljahn: less than a month, why do you care?

Sayach8: submitted my first contribution. so was wondering.

Uljahn: clash approvement process is pretty wild and low-effort most of the time, so don't keep expectations high

Sayach8: ok. Thanks.

Default avatar.png Saad-py: Well how do I contribute

Default avatar.png Saad-py: do I need to pay

Default avatar.png Saad-py: or is it free

jacek: you pay with your soul

Default avatar.png Saad-py: lol

MSmits: best way to spend it imho

jacek: soul for a clash?

MSmits: nah everything but that

MSmits: so i spent a day trying to fit parameters for a new random rollout and it was doing somewhat poorly (-5% WR). turns out I was always picking the first empty move on the miniboard =/

MSmits: I can't believe this wasn't worse than that

MSmits: just -5% ...

MSmits: ohh, maybe it was because i only did this on a free move, limited damage

Westicles: Sorry Sayach8, it looked like you put a lot of work into that

Sayach8: yes

Sayach8: but turned out that someone came up with a similar idea

Westicles: It was really identical, even the other solution worked for yours

Sayach8: btw is there any way to know if there exists a similar solution?

Sayach8: i understand

Westicles: You can look through here

Westicles: https://eulerschezahl.herokuapp.com/codingame/puzzles/

Sayach8: ok. thanks. needed this very much :relaxed:

jacek: stupid clobber, it's so unnable

jacek: MSmits might like clobber. there is many endgames to calculate and store

MSmits: cool

jacek: so yeah, a math :unamused:

Default avatar.png GiorgosGiorgakis: hi guys

Wontonimo: hello there internet person

IamQuan: hi :))

jacek: ohai

Default avatar.png UnnamedCodinGamer: when using bit boards to represent game states is there a way to get the "from" cell without iterating over each bit?

Default avatar.png UnnamedCodinGamer: I am looking at clobber at the moment

jacek: dunno. im iterating

Default avatar.png UnnamedCodinGamer: hm

Default avatar.png UnnamedCodinGamer: i tried iterating over bits for the same reason trying breakthrough

Default avatar.png UnnamedCodinGamer: at the end it was slower than the no bit board implementation

Wontonimo: what's the 'from' cell?

Default avatar.png UnnamedCodinGamer: move from cell A to cell B

Default avatar.png UnnamedCodinGamer: the position of the cell that is about to move

jacek: well bitboards allow for quick checking isOver, getWinner and other things

Default avatar.png UnnamedCodinGamer: I used it for move generation

Wontonimo: lets say 5th position to 8th position board = board - (1<<5) + (1<<8)

Wontonimo: to test 5th, you could use (board & (1<<5))>0

Default avatar.png UnnamedCodinGamer: yes

Default avatar.png UnnamedCodinGamer: but in move generation you get a board with all available moves

Default avatar.png UnnamedCodinGamer: you have to then find which of the moves corresponds to which piece

Default avatar.png UnnamedCodinGamer: you can do this easily by iterating over the bits

Wontonimo: There's a great PDF titled BITBOARD METHODS FOR GAMES Cameron Browne1 QUT, Brisbane, Australia. Hope you can google and find it. It has some great bitboard optimizations

Default avatar.png UnnamedCodinGamer: I am reading it at the moment

Wontonimo: ah. okay, then i'm out of bullets

Default avatar.png UnnamedCodinGamer: below figure 8 it says that the from cell can be recovered as the cell that must exist in the opposite direction

Default avatar.png UnnamedCodinGamer: which as I understand it needs iteration over bits

darkhorse64: hint: iterate over set bits

Default avatar.png UnnamedCodinGamer: how can I get only the collection of the set bits?

darkhorse64: there are functions to find leading or trailing ones

darkhorse64: and then you xor the current bit and iterate again

jacek: hmm

Default avatar.png UnnamedCodinGamer: is this going to be faster that going through the bits?

darkhorse64: having issues ?

darkhorse64: @jacek

Default avatar.png UnnamedCodinGamer: *than

jacek: i havent thought about that before

jacek: that would be much useful in amazons when you only have 4 bits

jacek: set

Astrobytes: http://chat.codingame.com/pastebin/d6edf8a5-561d-4142-b038-04dd634857d6

darkhorse64: All my loops on bits are built this way. Of course, if all bits are set, it's overkill

Astrobytes: Yeah, you should check for that

Astrobytes: nice clobber bot jacek

darkhorse64: ^

jacek: clopper

Astrobytes: :o

jacek: its just mcts

Default avatar.png UnnamedCodinGamer: thanks guys, I will try this out

Or64bit: Hey guys

Or64bit: Anybody up with some basic postgresql knowledge?

Or64bit: I have to log onto a port that I have username and password to but it says I have an auth error

Or64bit: any idea why,

Or64bit: ?

Default avatar.png userdanydan: it is probably because there is an auth error

Wontonimo: :( my bot got up to 35th in uttt temporarily and got my hopes up but then finished its arena contest at 66th... sigh... the ups and downs of watching the arena test live

Wontonimo: the scores are pretty smooth up here with about 0.1 trueskill difference between positions

jacek: :+1:

Default avatar.png UnnamedCodinGamer: 66th is pretty good

Default avatar.png UnnamedCodinGamer: best I have seen is 144th :)

Default avatar.png UnnamedCodinGamer: but I guess this is what you get with such small iteration count

Wontonimo: last night i finally made a local test bed to compare different versions locally. Makes 2 versions compete for 80 battles, 40 each with first move. Hasn't yet given me a huge insight other than that when I picked my hyperparameters I totally lucked out and picked exactly the right ones.

jacek: 80 games? i usually use 1000 games

Default avatar.png UnnamedCodinGamer: :)

Wontonimo: i was hoping that there was optimization available there

jacek: yeah, most my arse-pulled handtuned parameters are already close to optimal

Wontonimo: jacek thanks, i'll try upping. the count

jacek: most often than not ill get 47 63 initially just to get 550 450 in the end

Wontonimo: right ... stats works best with bigger numbers :)

jacek: of course it takes more time to complete

Wontonimo: well then, time to burn more cpu!

Wontonimo: I need a cluster of 1000 raspberry pi and it'll complete in 15 seconds

MasterCoderxD: @Wontomio

MasterCoderxD: I am currently doing your Challenge ("Beach Volleyball")

MasterCoderxD: https://www.codingame.com/ide/puzzle/beach-volleyball

MasterCoderxD: http://chat.codingame.com/pastebin/bf3f2173-19cb-4e5e-a47a-b053a681cba4

MasterCoderxD: I presume the shortest route is travelling in one straight line

MasterCoderxD: but the question says he breaks his initial (straight line) path and makes another

Wontonimo: right, so there are two legs to the travel, the first on land to a point where the beach meets the water, and a second leg from that point to the ball. Both are straight, but not necessarily parallel

MasterCoderxD: ohk

Wontonimo: This may help https://en.wikipedia.org/wiki/Snell%27s_law

MasterCoderxD: I am just thinking on how to approach it, using trigonometry or calculus

Wontonimo: I added that as one of the tags, but it appears to have been removed prior to final publishing (not by me)

MasterCoderxD: oh thanks for the link

MasterCoderxD: ok

MasterCoderxD: ty

Wontonimo: it is possible to do it without calc

MasterCoderxD: ok

Wontonimo: hey, thanks for doing my puzzle! It was my first contribution

MasterCoderxD: ok np

geppoz: spoiler alert: T = L* √((x-x1)²+(y-y1)²) + W*√((x2-x)²+(y2-y)²)

geppoz: find min for x

MasterCoderxD: what's L and W?

geppoz: speeds

MasterCoderxD: x and y are?

MasterCoderxD: ballx and bally?

geppoz: x is the output to find, y is beach_y

MasterCoderxD: okie thank u

MasterCoderxD: lemme try by myself too though :joy:

geppoz: so seems it is needed also derivative calculus

MasterCoderxD: yeah

MasterCoderxD: i am still in hs, so i am fresh with calc

MasterCoderxD: what's T?

MasterCoderxD: time?

geppoz: y, is the value to minimize

MasterCoderxD: k

MasterCoderxD: still getting wrong results bro

MasterCoderxD: something wrong with my code?

MasterCoderxD: http://chat.codingame.com/pastebin/ae32a935-47c5-43c9-aaa2-41bf4fabefda

Default avatar.png tridvajedan: guys

Default avatar.png tridvajedan: how do you calculate a median?

geppoz: I haven't solved it yet :D

Default avatar.png Pon4ik: guys, what do i have to do with "nextCheckpointDist" in race game?

MasterCoderxD: how do i approve a contribution?

RalseiOwO: you get the median from a sorted array, then get the value in the middle of the array tridvajedan

Westicles: You can only use approved tags, but the autofill list of tag includes every tag anyone has ever tried

MasterCoderxD: i don't understand

MasterCoderxD: i wanna approve a contribution, what should i do?

Westicles: A clash?

MasterCoderxD: no a contribution

MasterCoderxD: this one https://www.codingame.com/contribute/view/507070b7925f661e2d2835c14f950b9d157e

Westicles: It is a WIP, so you can't approve. Plus you have to be level 20

Default avatar.png GiorgosGiorgakis: why am I stuck in this https://www.codingame.com/ide/puzzle/coders-strike-back

MasterCoderxD: ohhhk

Default avatar.png GiorgosGiorgakis: it doesnt progress

Default avatar.png GiorgosGiorgakis: blah

AntiSquid: MasterCoderxD why are you desperate to approve it?

MasterCoderxD: it's for the xp thing :joy:

Default avatar.png GiorgosGiorgakis: where is the next button?

AntiSquid: doesn't mean much and besides you can get xp from other sources

MasterCoderxD: i am a hackerrank guy, i am just trying this out

MasterCoderxD: plus i thought his contribution was cool

AntiSquid: forget hackerrank culture, this is CG

MasterCoderxD: hackerrank is more popular

MasterCoderxD: bruh

MasterCoderxD: :rolling_eyes::joy:

AntiSquid: facebook and twitter are more popular, popularity isn't a good thing

AntiSquid: it just means it caters to more people, doesn't reflect quality

MasterCoderxD: k not into arguments

AntiSquid: sure, but take CG as it is, forget other sites

Astrobytes: Yeah, it's not really comparable in that respect.

MasterCoderxD: you r ranked 11?? in the world?? gg Antisquid

MasterCoderxD: :scream:

MasterCoderxD: gg to u too Astrobytes

AntiSquid: when you do a contest you get the overall idea of this site

MasterCoderxD: i am only 16, u guys r my inspiration :joy:

AntiSquid: at the moment it's quiet

Astrobytes: It's just a number. There are far more skilled players here with lower levels, they onyl play what interests them

Astrobytes: *only

MasterCoderxD: my clash of code rank is 2000, that good?

MasterCoderxD: u r right

AntiSquid: play some multis, see what you like, when you get stuck as questions, try to engage in meaningful conversations

MSmits: hey Astrobytes I just tested hashing of late game states in uttt. I get about 30% hits with 5 boards finished or more. Not worth it

MasterCoderxD: k but antisquid u didn't answer my question lol

AntiSquid: i think clash is just to get going with coding something

Astrobytes: Ah. Not much has really worked with your new UTTT bots MSmits

MSmits: that's uttt for you

Astrobytes: Yep. Defies all reason.

AntiSquid: the end goal imo is to play a multi, probably even make a book for UTTT, linearly optimize the crap out of it and submit spam

MSmits: nice early solve here though: ply 37 out of 67, p1 hashbot, p2 bot without hash

MSmits: https://www.codingame.com/share-replay/530048846

Astrobytes: Nice early solve yeah but not exactly an improvement over the other D:

Astrobytes: * :D

MSmits: my thought exactly

MSmits: i'll submit the hashbot to test it, just in case I'm missing something

jacek: so you mastered endgames. now try to get to them

MSmits: thats the idea :)

MSmits: I also tried biased randomfor selecting a random miniboard in sim, had a good system for it, slowed down my bot, same winrate

MSmits: I think what hurts my hashbot is that I had to let go of my "ab pruning" in my solver. it's hard enough to combine TT with ab in a minimax, trying to both with a mct solver seems too hard

AntiSquid: how come there's no referee for hanabi (fireworks)

Wontonimo: jacek - about the 1000 tests, thanks again. turns out that indeed the early results were wrong

jacek: you can choose if you want to show the referee to download

jacek: Wontonimo you need more confidence

Astrobytes: tric trac didn't link it AntiSquid, I don't think it's really needed though

MSmits: yes Wontonimo, stand up straight, be assertive!

Wontonimo: jacek , like personally or are you talking stats :O

Wontonimo: haha

jacek: yes

MSmits: it's annoying that hashing states is so often just not worth it or just barely worth it. On which boardgames on CG did it turn out to be worthwhile?

Default avatar.png Omer_ctr: why is this giving me nothing ?

Default avatar.png Omer_ctr: http://chat.codingame.com/pastebin/65277fd6-f9b9-4de6-ac4a-69752248db87

Astrobytes: afk for a bit

Default avatar.png Omer_ctr: ???

MSmits: I'm not even sure what language that is Omer_ctr

Default avatar.png Omer_ctr: thats javascript

MSmits: is that output or error printing?

MSmits: because if you're trying to debug, you need to error print

Default avatar.png Omer_ctr: it gives me nothing in output

MSmits: but do you want normal output or debug output?

MSmits: it matters

Default avatar.png Omer_ctr: normal output

MSmits: is it asking for several lines of print?

MSmits: output i mean

Default avatar.png Omer_ctr: it says Found: Nothing

MSmits: probably you were supposed to give one line

MSmits: and your first line was nothing

MSmits: kinda hard to know when we dont even know if this is a clash or puzzle or somethign else, better link what you're doing

Default avatar.png Omer_ctr: this is clash

Default avatar.png Omer_ctr: bro im tired with javascript

MSmits: i was tired of javascript before even starting

MSmits: must be those glowing testimonials

RSBat: MSmits do you treat all equal states as equal? ie finished miniboards regardless of moves in it?

MSmits: yes

MSmits: i even go further than you i think

Default avatar.png Omer_ctr: i learned python in like 3 weeks

MSmits: there are only 8629 unique miniboard states in my system

Default avatar.png Omer_ctr: 2 months of learning js but still begginer

karliso: TT in uttt? I don't know about that.... 30% seems quite high.

MSmits: i have a special miniboard system that turns the 19683 variations you get with a ternary representation into a 8629 variant representation, where a board is just a number between 0 and 8629

MSmits: these are as unique as they can be

karliso: There are only 8629 variations?

RSBat: what about last move position? sometimes it can be stored differently too

MSmits: yes

MSmits: let me show you

MSmits: http://chat.codingame.com/pastebin/05f37f17-df3f-4b6d-b471-a77f801780eb

MSmits: these two are identical

MSmits: in both cases p1 wins in 1 move

MSmits: and p2 draws

MSmits: RSBat i do this too yes

MSmits: a free move is a free move, no matter where it's targeted

MSmits: i give it index 9

MSmits: as opposed to 0-8 being regular last moves

RSBat: ok, cool

MSmits: there are ways to make the 8629 even smaller, but it's extremely expensive

karliso: Would a board with empty cell 4 would be identical as well?

MSmits: no

MSmits: different possible futures

MSmits: all possible futures and the way the board affects the big board, needs to be identical

karliso: If they are both free moves then they are same.

MSmits: you can have identical boards with 2 empty squares

MSmits: yes,

MSmits: thats what i meant by making the 8629 even smaller

MSmits: you cant use that in a live bot

MSmits: there will be 512 different coordinate systems then, depending on the makeup of the main board

MSmits: i calculated all of those

MSmits: the 8612 goes down to 2k/3k realistically in some cases

MSmits: with 1 board left it's only like 110 or so, but whats the point

MSmits: you could use this system for your lookups karliso

karliso: no, thanks

MSmits: from what you shared before you use a biased lookup for picking moves

karliso: :D

MSmits: or at least you lookup your random moves

MSmits: so i guessed it migth be biased

MSmits: for example, giving winning moves more probability

karliso: I always take a winning move in rollouts

karliso: Never even tried otherwise

MSmits: you mean a miniboard win?

karliso: Yes

MSmits: weird

MSmits: when i tried it, it was worse, not sure why

MSmits: I have a win/block random

MSmits: to get a bit more variation in the rollouts

karliso: More variation is not necessarly better.. and winnig boards fast makes for faster rollouts

MSmits: yeah, that does make sense

MSmits: and you also avoid giving free moves when you win, if there is a non-free move win available

MSmits: that i know

karliso: yeah, and if multiplesafe wins I pick lowest... I think I tried to randomize way back, but it didn't give anything

Or64bit: Hey guys

Or64bit: Anyone up that knows something about DNS stuff? I'd have a pretty simple question that I do not know the answer for

Andriamanitra: https://dontasktoask.com/

Or64bit: I'll keep that in mind :D

Or64bit: so why could it happen, that I cannot connect to webpages ending with .hu (My country's ending thing) but I can connect to .com sites?

Andriamanitra: what do you mean by cannot connect? your browser doesn't load those web pages? have you tried getting the ip manually from command line?

Or64bit: To be honest this is just a question I'd have to answer, the only extra info I know is that the page is not responding when I try to visit it

Or64bit: but .com sites do load

Or64bit: and also most of the pages that are on my university's network

Or64bit: only .hu endings don'T

Andriamanitra: are you sure it's ALL .hu sites? and is there any way you could verify it's not a problem with your browser (maybe try `curl google.hu`)?

Andriamanitra: if it's not a problem with your browser it could be a dns problem or a firewall

Uljahn: or hosts file

Or64bit: Sadly I cannot verify, though you're right, it only states I cannot connect to to like five other specific we services

Or64bit: but it also does not say that it could connect to other ones

Or64bit: Here's the quetion translated, I hope you guys will understand it

Or64bit: Between two lectures you decide on visiting something.hu, but the page does not respond. You'd like to figure out what's going on. Doing so you notice, that you also cannot reach a.hu b.hu c.hu and d.hu, altough you can open facebook, gmail, youtube, and also most of the university's sites load in. What do you think the problem could be?

Uljahn: i guess, there is no valid dns record for them or those sites are offline or blocked by ISP

Saphine: https://www.codingame.com/clashofcode/clash/16041572369f7fe2dd542488e28af80ed82e4c9

AntiSquid: use #clash for links to clash Saphine

Or64bit: ThanksUljahn, the ISP is a very good idea

reCurse: sigh

jacek: oO

jacek: reCurse did you use convnet for uttt?

reCurse: Some aspects of it

jacek: training a0 style?

reCurse: See above

jacek: hmph

reCurse: Sorry not leaking sensitive info for free :)

jacek: then i will make my own nn with blackjack and hookers

reCurse: Don't forget the cocaine

Laserbeak43: In Power of Thor - Episode 1, you have to change Thor's position, but the only variables for Thor are intialTx and initialTY. Do I use those?

struct: you can use those

struct: but you need to update them

jacek: you need to calculate thor's current position by yourself

Laserbeak43: Thanks

struct: Boss even has comments, depth 3 minimax

jacek: what boss

struct: C4

jacek: and you can download it

struct: oh

struct: Did not even notice it

darkhorse64: 17 moves win against the boss

darkhorse64: C4 is ready

darkhorse64: for approval

jacek: MSmits should approve it. he used most cpu power for it

reCurse: Ready for detonation?

darkhorse64: :smiley:

MSmits: i did do 1.5 million games in meta mcts :0

MSmits: (full games)

MSmits: i'll go approve

reCurse: Not touching that one with a ten foot pole

Andriamanitra: i'm getting captchas between each clash again :(

Andriamanitra: pls fix

MSmits: you dont need a ten foot pole reCurse, apparently even dogs can play it. jacek found proof

Astrobytes: afaik they will fix the clash captcha issue next week Andriamanitra

jacek: im two foot pole

Astrobytes: hahaha

Scarfield: xD

MSmits: ah damn, you went so long jacek. Now I have to reset my time-since-jacek-talked-about-his-pole timer

Scarfield: he was reffering to being pole-ish

MSmits: ohhh

reCurse: :door:

MSmits: good thing I didnt reset it yet

MSmits: keep going jacek

jacek: do you have :soccer: timer as well?

MSmits: yea it's at 13 seconds because you just talked about it

jacek: about what

MSmits: you're just trying to get me to talk about it too

MSmits: lol i got random time out on first test of connect 4 code

MSmits: spent a few minutes trying to find the bug

MSmits: it was just one of those random timeouts cg has

darkhorse64: The boss is having a hard time

MSmits: yeah

jacek: its on?

jacek: i didnt get notification

MSmits: it's on!

darkhorse64: join the party

MSmits: happened 10 mins ago

MSmits: thanks for leaving me the approve btw

jacek: nah, ill just give random bot for now

jacek: so the boss wont be lonely

MSmits: didnt you already have a better one?

jacek: yeah

jacek: ill wait for more people. then ill mess with leaderboard :imp:

MSmits: btw, as promised, wont counterbook anyone. So far have general book of 486 moves that doesnt really go beyond ply 5. That's after 1.5 million games

jacek: going for column 1 after all?

MSmits: yeah based on results

MSmits: http://chat.codingame.com/pastebin/e90b8754-f885-47f6-abf2-108b9ceff1c2

MSmits: closest to 50%

MSmits: i doubt you will notice a difference if you start at 0

MSmits: because it didnt go below 50% until after hundreds of thousands of games

MSmits: and you're only playing one game at a time so..

MSmits: just dont start with 3 or 4... that would be bad for your winrate

MSmits: ohh

MSmits: there's another league

MSmits: I was thinking, where did the tric go?

Scarfield: you need to trac him down

MSmits: :grin:

Astrobytes: lmao

MSmits: well at least the boss is better than random

MSmits: it has jacek's random bot beat easily

MSmits: nice to have a reasonable bar of entry to wood 1

Scarfield: elStructo mentioned a depth 3 minmax boss

MSmits: ah yeah thats a good strength to aim for I think

Laserbeak43: "Timeout: your program did not provide an input in due time. Earth was destroyed!" Provide an input? I thought I was supposed to write to console


Laserbeak43: could someone please clarify for me?

MSmits: please don't destroy Earth before I get nr 1 at uttt

jacek: yes, you write to console

jacek: but did you print more than neccessary? or less?

Laserbeak43: just made a big if/else

Laserbeak43: that always sets a string variable to one of the directions which is output by Console.WriteLine() on line 60

jacek: really? can you show the code

Laserbeak43: and why do I not see this channel on the discord server?

struct: maybe you have an infinite loop?

reCurse: I might just keep an overfit bot to spam submit in spite

jacek: i bet you have Console.WriteLine() in one if

jacek: and it doesnt print

MSmits: overfit to whom?

MSmits: reCurse?

reCurse: To perform better against you than everyone else when you get rank 1

MSmits: mmmh

jacek: hmph

MSmits: do it :)

reCurse: You could say I'm in service of protecting the planet

MSmits: oh yeha, thats right

Laserbeak43: https://pastebin.com/U4pEB8ms

jacek: Laserbeak43 CS1002: ; expected

jacek: compile error

Laserbeak43: just noticed that...

Laserbeak43: like literally when u pinged me lol...

jacek: :duck:

MSmits: is this thor 1?

struct: Also you need to update coords

MSmits: the basic thor?

MSmits: I thought my first C# solution was big, but this is even bigger

MSmits: I like it

Laserbeak43: update my coords?

MSmits: when you output

MSmits: you move to new coords

Laserbeak43: initialx and initialy?

MSmits: you need to remember them (= update)

MSmits: yes change those, or use a different variable

MSmits: so when you move to the right you do initialx++

MSmits: for example

Laserbeak43: was wondering if the code I don't see handles that or not

MSmits: initialx is never changed

MSmits: so no

Laserbeak43: yeah i did that "initially" :) but removed it

Laserbeak43: what what?

MSmits: a single iteration of the while loop corresponds to a single step

Laserbeak43: i get that

Laserbeak43: but

MSmits: your initialx and y are set before the while loop

MSmits: and never change

struct: Only stuff inside the while loop get updated every turn

MSmits: unless you do it

Laserbeak43: exactly

Laserbeak43: but you said no?

MSmits: i said you current code doesnt

MSmits: but it should

Laserbeak43: ah

Laserbeak43: ok

MSmits: sorry confusion

Laserbeak43: np np thanks for the help

MSmits: np

Scarfield: i just looked up my python solution for thor, good laugh xD

MSmits: it's good for your confidence when you look at those old solutions

MSmits: wow, i learned a lot...

Laserbeak43: lol.... I really want to be doing scala

Laserbeak43: but I'll try them both

Laserbeak43: yeah I know there's gotta be a more elegant way to do this, just being lazy I guess

MSmits: connect4 seems to be really rare on draws

Scarfield: yea it was probably the first puzzle i made after onboarding

PatrickMcGinnisII: Whoo hOooo!

darkhorse64: fierce competitors do not accept draws

MSmits: happy Patrick incoming

PatrickMcGinnisII: lvl 35! ;) :P

Scarfield: Patricks day?

MSmits: grats :)

PatrickMcGinnisII: I should be ashamed, no contributions

PatrickMcGinnisII: heh

MSmits: same

MSmits: but all in all i think CG has quite a lot of them already

MSmits: testers are more needed than contributors

MSmits: 10 : 1 win ratio :)

PatrickMcGinnisII: xp is getting harder to come by

PatrickMcGinnisII: fewer competitions

MSmits: still plenty of multis you havent tried i'm sure

MSmits: I found a draw! https://www.codingame.com/share-replay/530096589

PatrickMcGinnisII: only connect 4

PatrickMcGinnisII: hmm, of course i haven't gone "all out" on some of them

MSmits: wait, you tried every other multi?

MSmits: the xp is in the legend promotion

MSmits: isnt it like 500 ?

PatrickMcGinnisII: yea, promotions are a b*tch

darkhorse64: it is

PatrickMcGinnisII: achievments give a little xp, but i have to do more langs

darkhorse64: languages gives many achievements but few points

PatrickMcGinnisII: i should make a puzzle, prolly easier xp

MSmits: only make a puzzle if you have a good idea fo rit

darkhorse64: no fibonacci in it

MSmits: you dont like fibonacci?

PatrickMcGinnisII: last puzzle i did was fib...centrifuge tuning, rofl

darkhorse64: There are too many of them in Coc

MSmits: ah

PatrickMcGinnisII: coc isn't my jam

Laserbeak43: So I've updated the coordinates. and still the same error though.

Laserbeak43: Expected a movement (one of N, NE, E, SE, S, SW, W, NW) but found

PatrickMcGinnisII: 4 people in connect4, ...o k a y...must be super new

MSmits: share the code again, sounds like an output problem

MSmits: yes PatrickMcGinnisII < 1 hr

struct: maybe it doesnt enter any if

darkhorse64: do your ifs cover all cases ? I would say probably no

Laserbeak43: https://pastebin.com/f5Ps9QWm

MSmits: you're supposed to compare against light x

MSmits: and forget everything else

MSmits: so check if your x > light x or x < light x to determine if E or W is necessary

MSmits: those if statements contain a lot of unnecessary comparisons

Laserbeak43: so lighty never changes?

Laserbeak43: trick question?

MSmits: light is stationary

Laserbeak43: for some reason I thought i had to chase it

MSmits: nope

MSmits: x and y are both stationary

MSmits: your thor moves

MSmits: lightx and y i mean

Laserbeak43: either way, what I've done in regard to comparison should break the code?

MSmits: it's a bit hard to tell, if your bot doesnt output anything it means none of the if statements are true

MSmits: if you simplify them it should be easier to fix

MSmits: thats a better path than trying to figure out why this one doesnt work

PatrickMcGinnisII: Laserbeak43 the move is null because you haven't met the criteria in all the if statements to change it

Laserbeak43: hmm...

PatrickMcGinnisII: if you setup a deltax and deltay before the loop

PatrickMcGinnisII: you will have better luck

PatrickMcGinnisII: and since the Y direction always comes first in the move output

PatrickMcGinnisII: you can just move=N or S

PatrickMcGinnisII: then test the x, and move=move+ W or E

PatrickMcGinnisII: instead of 8 ifs you can reduce it to 4

Laserbeak43: ok thanks. I'll try it

PatrickMcGinnisII: so, if deltay!=0 move+= N or S

PatrickMcGinnisII: if deltax!=0 move+= W or E

PatrickMcGinnisII: you'll get it

PatrickMcGinnisII: I gotta clean my house, cookin' steak for the gf in a few hours

PatrickMcGinnisII: you had two of these: (xy[0]>lightX) && (xy[1]==lightY)

PatrickMcGinnisII: should be a <

Laserbeak43: changed the whole code block to this: http://chat.codingame.com/pastebin/eebe59b1-b8b7-470e-802b-8057e07a94b9

Laserbeak43: oops

Laserbeak43: sorry

Laserbeak43: http://chat.codingame.com/pastebin/ac4a7a64-3cae-49d3-bb73-ed8b204bd15c

Laserbeak43: ok i can't copy and paste code from the editor into chat

Laserbeak43: https://pastebin.com/VZ9Gd1YE

Laserbeak43: oh no i see what was happening.. lol

PatrickMcGinnisII: remember you are appending a character to the move string

Laserbeak43: hmm

PatrickMcGinnisII: ok, u found it

Laserbeak43: I was wondering. I don't see why I would append a direction to move?

PatrickMcGinnisII: do y first as its the first direction to be output

PatrickMcGinnisII: so

PatrickMcGinnisII: move="";

Laserbeak43: 0 +"E"?

Laserbeak43: no...

PatrickMcGinnisII: if (deltay!=0) { if (deltay<0) { move+="N'; deltay++;} else {} kinda thing

PatrickMcGinnisII: I dunno what lang, what is it? c++?

Default avatar.png tridvajedan: how to make a puzzle

Default avatar.png tridvajedan: like a code puxxle

Default avatar.png tridvajedan: and upload it

jacek: contributions

Laserbeak43: C#

PatrickMcGinnisII: so strcat(move,"N"); ?

PatrickMcGinnisII: move.push_back("N"); ?

Laserbeak43: I believe move "x" + "y";

Laserbeak43: or move += "Y";

PatrickMcGinnisII: so move+="N"; should be fine

Laserbeak43: should work

Laserbeak43: i feel like im being stubborn about something...

PatrickMcGinnisII: ok, i get mixed up between JS and java therre for a sec

PatrickMcGinnisII: hmm, oh works the same

PatrickMcGinnisII: unless you are using stringbuilder

PatrickMcGinnisII: ok, i'm done giving advice, second guessing myself

PatrickMcGinnisII: prolly should have said "S"

Laserbeak43: hehe thanks...

PatrickMcGinnisII: your original code would have worked with 2 minor changes.... the < and tx++ .... but there is alot of unnecessary conditions in the statements

Laserbeak43: http://chat.codingame.com/pastebin/d16474af-4d49-4f6e-b76c-dcdfd0ffece7

Laserbeak43: that worked... for straight line

PatrickMcGinnisII: you can always ctrl+z, to undo things

Laserbeak43: think i'm just going to do it this way, since I don't know how the blackbox works

PatrickMcGinnisII: remember do Y first

jacek: y

PatrickMcGinnisII: y !

PatrickMcGinnisII: ;)

PatrickMcGinnisII: or is it ~y

Laserbeak43: :cold_sweat:

PatrickMcGinnisII: stupid tilde doesn't show up in chat

PatrickMcGinnisII: '/u223c

PatrickMcGinnisII: grrr

jacek: maybe the site is made in php

struct: its due to your font

AbundantPuddle: ỹ

Laserbeak43: `

PatrickMcGinnisII: the tilde in html might think it's a css operator

PatrickMcGinnisII: U+007E

PatrickMcGinnisII: grrr

PatrickMcGinnisII: ~

PatrickMcGinnisII: alt+0126\

PatrickMcGinnisII: grr

PatrickMcGinnisII: i give up

PatrickMcGinnisII: glhf

AbundantPuddle: I don't know if you'rer referring to your text that says "or is it ~y," but I can see the tilde there.

PatrickMcGinnisII: ahh ok, i can't

PatrickMcGinnisII: all i see is -

AbundantPuddle: Then as struct stated, it's probably due to your font.

PatrickMcGinnisII: or a longer -

struct: -~

PatrickMcGinnisII: prolly

PatrickMcGinnisII: ok struct, you win

AbundantPuddle: struct just did a dash and a tilde

PatrickMcGinnisII: winner winner :poultry_leg: dinner

AbundantPuddle: Lol

PatrickMcGinnisII: all i see is a long dash

jacek: rainbow dash?

PatrickMcGinnisII: &#126

PatrickMcGinnisII: grr

PatrickMcGinnisII: laterz

Laserbeak43: Thanks. Later

Laserbeak43: https://pastebin.com/wJreDU6E

Laserbeak43: i'm not understanding something about the flow of the program... this always stops 1 coordinate short of 1...

Laserbeak43: of Y*

struct: Laserbeak43

struct: on the last if

struct: line 62

struct: will just append text

struct: with your current code it can match one previous if

struct: and then match that if

Laserbeak43: @struct Thanks. Not exactly sure what you mean, but I'll focus on that block

struct: https://i.imgur.com/GgDjja4.png

jacek: oO

arhariss: Hallow

Andriamanitra: fuck these shit clashes with only 4 test cases :rage:

jacek: reverse?

Andriamanitra: shortest

Andriamanitra: i saved one character too many

MSmits: tihs clashes

Andriamanitra: i had 43 (44 without the bug) and the next person had 95

MSmits: sucks =/

MSmits: so it worked on all the testcases but not on the validators/

MSmits: ?

Andriamanitra: yup

MSmits: that's annoying indeed

Laserbeak43: @struct I've fixed that before (and now), and in my current code I am trying to figure out how to get ThorY to equal lightY when deltaY == lightY. he always goes past the bounds

struct: Whats your current code?

Laserbeak43: https://pastebin.com/Uw8TkeYG

Wontonimo: Laserbeak43 I was just told last week you don't need to put the @ infront of the name. just passing along this wealth of knowledge

struct: laser

Laserbeak43: Wontonimo thanks

struct: you should use else if

Laserbeak43: hmm

Laserbeak43: ok i'll try that

struct: if(deltaY < lightY){...} elseif(deltaY > lightY){...} elseif((deltaY == lightY){....}

struct: Something like that

struct: the last elseif can be just an else

Laserbeak43: struct omg...that's all i needed to fix it

Laserbeak43: i really need to learn program flow...

struct: check other players code

Laserbeak43: yeah I'm feeling a bit silly

Laserbeak43: thanks for the help

Laserbeak43: Going to try doing it in Scala now

struct: The logic is the same

Stormalix: is there any way to stop all of the captchas im getting.. its really annoying and ive already done like 20 of them

Default avatar.png jtsimmons: Is there an easy way to see old clashes after you close out?

brandex123: Is there a way to check out the test outputs of old clahes after you close out?

Wontonimo: @Stormalix i think that's been raised as a bug in discord

Default avatar.png RyanGilbert: yeah if you do too many clashes in a row it gets sussy of you

Default avatar.png jtsimmons: Not a bot, just have nothing better to do.

Wontonimo: do some puzzles