Chat:World/2020-10-30
Laminator: @Articuno share your code please
Laminator: We need to knowwwwww
father3301: https://www.codingame.com/clashofcode/clash/14412270e36f1f164976927fb39726f20f915a0
soutenu: so glad they added D lang to this finally
soutenu: now it's just leetcode that's lacking
CopperFr: i prefer Amiga E
Chazfer: im new;-;?
Chazfer: bruh i was bored and my mom just told me whatever i want here so idk wut to do next lmao
DgH: im mew
Chazfer: im still 12 tho
Chazfer: hewwo {MEW} Chazfer ismyreal name
Chazfer: ehh the chat is dead
Chazfer: uhmmm
Bachus: i am looking for some job
Bachus: if you have something pm me <3
phantom1299: Hello Ak47aman, I'm World
Ak47aman: thank phantom1299 for your response
Chitchen: It's funny seeing what conversations go on here.
FredericRANDRIANARIVONY: hello everyone
LukeCloudwalker: Has anyone done the ASCII Art problem? What is the expected output on the last test? Meaning what should I do with '@'? It doesn't seem to specify in the rules...
LukeCloudwalker: Nevermind, I am supposed to use the questionmark
RromanSmoll: funny you shoulf ask this question)) I am staring in Ascii art exercise right now with no idea where to start)
The_Auditor: Good Morning everyone ! including Automaton2000 !
Automaton2000: i submitted a bot that can be done in python
SPDene: Automaton2000 : what's a "bot"?
Automaton2000: its not really a fan of it
eulerscheZahl: good afternoon to you as well
Zenoscave: moin
Astrobytes: Och aye the noo
eulerscheZahl: ohai
Zenoscave: おはようございます。
Zenoscave: eulerscheZahl Do you sim c4l?
eulerscheZahl: yes
eulerscheZahl: https://www.codingame.com/forum/t/code4life-feedback-strategy/2835/21
Zenoscave: Would you care to critique a design for c# sim in a bit? My code layout is weak
eulerscheZahl: did you read my post mortem yet?
Zenoscave: several times
eulerscheZahl: oh :D
eulerscheZahl: let me pull up my code and try to remember what I did
Zenoscave: It doesn't need to be c4l I'm just trying to get a general template or pattern down for sim in C#
eulerscheZahl: depends on the kind of search. my code layout is different for beam search vs that monte carlo thing i was doing in C4L
Zenoscave: Monte Carlo is what I'm aiming for
Zenoscave: Do you pass a game state object around?
eulerscheZahl: I have a function Simulate that I call as long as I have time remaining
eulerscheZahl: and I pass the initial state to it
eulerscheZahl: as a copy so I can modify it
eulerscheZahl: for C4L I do completely random simulations. In other games (especially multi-agent) I try to modify an existing solution instead
Zenoscave: Ok that should be enough to work with
Zenoscave: have you worked with ref much in C#?
eulerscheZahl: rarely
eulerscheZahl: if you pass objects, you pass by ref anyways
eulerscheZahl: and such things as ref int myNumber
eulerscheZahl: meh, I rarely have use for that
eulerscheZahl: afk, cat time
Patela: Hey, anyone can help me understand why Am i failing test 6 of ascii art? (java)
Zenoscave: Meow meow euler. thank you
Zenoscave: What is test 6? I only see 5 Patela
Patela: when you press "submit" 2 more show up i think
Patela: and i fail on the number 6
Zenoscave: Ah those are validators. they are not available publicly. (To stop hard-coding solutions)
Uljahn: validator 6 is called "Test full alphabet in CAPS..."
Uljahn: kinda hint
Zenoscave: That's as much as we can give though. Uljahn is correct
Patela: how do i test the code tho?
Zenoscave: You can't only on submit
Astrobytes: Make a custom test case with all caps
Zenoscave: click custom next to test cases
Astrobytes: Settings on left, Expert mode (if not already) then what Zeno said
Zenoscave: if it doesn't show that option you need to click settings on the left and enable expert mod
Astrobytes: Hi Zenscave & Uljahn btw
Zenoscave: Beat me to it
Zenoscave: Meow
Uljahn: to enable custom test tab switch the mode to expert in the settings on the left
Zenoscave: hi Astro and Uljahn
Uljahn: pewpew
Patela: should i write the full alphabet?
Astrobytes: pewpewpew
Patela: i forgot one letter of the alphabet
Astrobytes: :)
Patela: is there a better way to do an arrawy of the alphabet ? ( I'm kinda rusty)
Astrobytes: Well, technically you don't even need an array with the alphabet
Astrobytes: I gotta go for now, but think about ASCII and how you might go about using it to your advantage here
BiMathAx: Who know Franz Ferdinand ?
Clagus: Franz! One of my favourites in my 20ties.
BiMathAx: Hi, if someone want to test my contribution ? the link : https://www.codingame.com/contribute/view/5581506ad04331f4020bef765773d3759587 Thank's
tekki: hi world
tekki: something is going wrong?
tekki: if so don't call me ask AutomatonNN :)
AutomatonNN: and it's basically the same thing
Zenoscave: euler you still here?
eulerscheZahl: no
eulerscheZahl: watched a southpark episode. back now
Lachrymosa: Hey CG, quick hypothetical. If I wanted to make a wiring diagram that let me click on a wire and it would highlight where that wire flowed to through the system. What types of solutions exist, short of coding it from scratch?
AndrewWellington: i'm new to coding
AndrewWellington: but I have a little experience with javascript, if anyone is down to help me with some basics I would greatly appreciate it
AndrewWellington: or can guide me to a place that could help me
eulerscheZahl: for drawing these plans I got to know eplan. but was disappointed about it. e.g. the lack of simulating that kind of stuff
eulerscheZahl: i can say that now, as Siemens and Siemens Energy are different companies and eplan is from the first while I work for the latter :P
Zenoscave: I have a template MC Searcher finally. It builds and is ready for game specific dev
Zenoscave: The search also is capable of other variations like beam or MCTS just did MC first
eulerscheZahl: c4l isn't the easiest one to get it working
eulerscheZahl: i was close to giving up
Zenoscave: Due to the time move diffs?
eulerscheZahl: yes. and late rewards
Zenoscave: Right
eulerscheZahl: you score at the laboratory but have to take a few steps before
Zenoscave: What search depth did you end up doing? r if full depth howmany rollouts?
eulerscheZahl: if you score wrongly, you will analyze the first sample, then see that you can complete it, go to molecules and then lab
eulerscheZahl: random, about 5k plans iirc
Zenoscave: ok. good to know
eulerscheZahl: a plan means visiting 3 devices
eulerscheZahl: and random choices at the molecules
eulerscheZahl: weighted random, depending on my own samples
eulerscheZahl: and a bonus if blocking the opponent
Zenoscave: interesting
Zenoscave: I was goindg to just due full random current state possible actions
Zenoscave: could probably adjust weights in sample too
eulerscheZahl: full random is bad for picking molecules. takes too long to find a good one
eulerscheZahl: i also have weighted paths. e.g. from the molecules i prefer going to the lab
Zenoscave: and each device you do multiple actions if available?
eulerscheZahl: yes
eulerscheZahl: and stop randomly too
Zenoscave: why stop randomly?
eulerscheZahl: because i don't know how many molecules i want to take
eulerscheZahl: i only see that when i simulate the lab
Zenoscave: Ah
eulerscheZahl: told you: C4L is a pain for Monte Carlo :D
Zenoscave: Then why did you chose MC?
eulerscheZahl: i also ran a big fat MC to tune the scoring parameters
eulerscheZahl: because MC seemed viable to me. not easy but still doable
eulerscheZahl: no way i'm willing to write a huge if-else of that
Zenoscave: didn't you end up with a massive number of ifs anyways?
eulerscheZahl: 125 including comments :D
eulerscheZahl: my bot still has some funny actions that i don't understand
Zenoscave: lol
eulerscheZahl: https://www.codingame.com/share-replay/432144883
eulerscheZahl: check the last few turns
eulerscheZahl: that useless intermediate stop
eulerscheZahl: and i do that all the time when I can't get anything done before the end. No harm but I don't understand it :D
Zenoscave: yeah that is weird
Lachrymosa: Euler, I am currently working on your minesweeper puzzle lol
eulerscheZahl: hm, frame 86 in that replay above was stupid :thinking:
eulerscheZahl: hi Lachrymosa. Glad you give it a try :)
Lachrymosa: Yea hopefully I'll make it work today
NitMpez: welp euler just made me feel bad about my 941 line c4l with comments
eulerscheZahl: i'm around 1.3k lines
NitMpez: oh, i thought you said 125 lines of code...
Shahmuizz: can anyone tell me how to sort without using array in c++?
NitMpez: apparently someone bumped my decision tree into legend... was nice of them
eulerscheZahl: https://www.youtube.com/watch?v=5hvRrhhzpZc&ab_channel=EricRosen this is getting too far
NitMpez: that just hurts my head
eulerscheZahl: we should port it to CG. just for reCurse's reaction
NitMpez: is there any good tutorials on basically running your sim, then running it for enemy, then using there position, and reevaluating with sim with their new positions.
Astrobytes: dear god
yassineH: that sounds like fun
NitMpez: cause i swear any time i try to predict any movement... my sims get worse
arthurvergacas: gosh
arthurvergacas: this ui
arthurvergacas: is so
arthurvergacas: weird
arthurvergacas: i spend like 15 minutes just figuring out what I should do
wlesavo: fog onitama :smiley:
Astrobytes: Fogitama
eulerscheZahl: i had a game in mind that's heavily fog oriented
eulerscheZahl: but discarded the idea, i don't think you would like it
eulerscheZahl: you have to find a path in a maze and reach the opponent start location. but you are blindfolded
eulerscheZahl: you can produce sound which travels at a given speed
eulerscheZahl: and wait for the echo to come a few turns later
eulerscheZahl: so you can calculate the distance to each wall
eulerscheZahl: I even had a title in mind: NaNNaNNaNNaN Batman
LastRick: i like it, more map puzzles please
Astrobytes: that title :D
Lachrymosa: That does sound fun :3
Astrobytes: It's very OOC-ish
eulerscheZahl: yes
eulerscheZahl: had the idea before OOC
eulerscheZahl: part of the reason why i discarded it
Astrobytes: It would work fine, if the fog is central to the game as it is in this case (and OOC) it's cool. Just not when it's added to something like PacMan :P
LastRick: Holey moly I beat Euler to a solution! I can retire from CG.
eulerscheZahl: what are you talking about?
Astrobytes: lol
LastRick: Thank you everyone for distracting him long enough for me to solve the cardinal number puzzle.
eulerscheZahl: i wasn't even aware of that existence :D
LastRick: then my plan worked perfectly
eulerscheZahl: link?
LastRick: https://www.codingame.com/training/easy/nicholas-breakspeare-and-hugh-of-evesham
eulerscheZahl: looks tedious
eulerscheZahl: -2^64 ≤ Xi < 2^64
eulerscheZahl: what a poorly chosen limit
eulerscheZahl: why not 2^62?
eulerscheZahl: 63, typo
LastRick: it is definitely tedious
eulerscheZahl: and a faulty stub generator
eulerscheZahl: reads int instead of long
LastRick: my favorite part is when I got 2/3rds of the way through and had to switch from int to BigInt, I love going through and mindlessly switching dozens of those
Skynse: me: wants to join a clash in time captcha: Im about to end this man's whole career
eulerscheZahl: Unhandled exception. System.OverflowException: Value was either too large or too small for a UInt64.
eulerscheZahl: i kinda saw that coming
eulerscheZahl: 100% on submit. it's just that one number (-2^64) that makes me crash
Astrobytes: bit of a fail that one
eulerscheZahl: still i'm not happy about how the moderation process went there. 2^63 changes nothing about the puzzle but removes the need of BigIntegers
eulerscheZahl: but am I pissed enough to complain on the forum?
Astrobytes: Yeah, I just read the comments and checked it out
Astrobytes: Up to you. It seems like something that should've been picked up as not helpful during moderation as opposed to just being part of the puzzle
eulerscheZahl: puzzle discussions share my concerns. including the crashing default code
eulerscheZahl: screw it. it's anarchy now
Astrobytes: Yeah, saw the forum post
Astrobytes: Honestly.
eulerscheZahl: dwarfie has a valid point about string manipulation being viable. i'll change the stub at least
Astrobytes: Doesn't change the fact that a lot of contributions are slipping through without due care and attention in the approval process (again)
eulerscheZahl: yes. thumbs up for the few users how try to keep it clean and prevent the worst
eulerscheZahl: I gave up and just spam my own brainfarts instead :D
Astrobytes: Yeah, absolutely. lol, keep brainfarting!
eulerscheZahl: not before the next contest. my next one will get complicated if i really decide to go for it
Astrobytes: Ah yeah, not long til the contest now
eulerscheZahl: hype
eulerscheZahl: also i'm trying the topcoder marathon
eulerscheZahl: did a bit of quick coding today, will really start tomorrow
Astrobytes: I saw someone post about that, might have been you
eulerscheZahl: it was
eulerscheZahl: you have a final table from a soccer tournament (goals scored+received, points gathered) and you have to guess the score for each match
Astrobytes: lol, you got this one for sure :)
ParticleBeam: Dots and Boxes proving tricky
Zenoscave: I completely forgot about enemy movement in my Template. It always assumes a game is single player...
Zenoscave: need a better multiplayer version...
Astrobytes: Pass in who's turn it is to your playTurn method, iterate through players... multiple approaches
Astrobytes: *whose
AntiSquid: dots and boxes is annoying to structure imo . i find my implementation not "clean"
ParticleBeam: I have a rather clumsy structure and its very expensive when trying to run simulations.
AntiSquid: i didn't add in yet all the tricks. :P
DesmondW: .
Zenoscave: Anti when did you hit lvl50
Zenoscave: gz
vtenneke: Would anyone have any pointers on Unfolding paper(very hard) I can pas 5/7 of the tests but I'm failing on the ones where you need to fold a lot. Currently using folding and floodfill so probably very inefficient but can't seem to wrap my mind around a formula.
LastRick: I know I'm late to the discussion (had to go carve a pumpkin) but maybe there could be a setting that doesn't allow you to approve the post without solving it yourself? I know that's a question when you hit approve but maybe it should be mandatory
LastRick: vtenneke: I haven't done this one yet but certainly there must be a math way to calculate the number of pieces (based on where the # are) for any given N. That's my guess.
RohanasaurusRex: Wow level 50 is amazing lol
MaliciouslyCrypticUsername: The "Unlocked certifications" part of the home screen lists certifications that you've already unlocked, correct?
MaliciouslyCrypticUsername: Well duh
MaliciouslyCrypticUsername: But it says I have "Artificial intelligence" gold level
MaliciouslyCrypticUsername: "The developer knows how to develop an advanced AI: the AI optimizes its movements and adopts an appropriate strategy according to context..."
MaliciouslyCrypticUsername: I don't remember doing anything like that
Scarfield: hover over the dots on the path: "Reach top 2% of the Bot Programming Leaderboard" to get gold
MaliciouslyCrypticUsername: Yea it says completed
Scarfield: just above the yellow line it says how you got it
MaliciouslyCrypticUsername: Oh I didn't see the hat symbol next to it
Scarfield: yea its a little weird
MaliciouslyCrypticUsername: I wonder why this whole quest map was introduced...
jthemphill: I'm mid-gold in UTTT. The hot-spot in my code is in calculating available moves. I'm working on bit-boards and planning on returning a u128 representing the set of available moves I can take. Do you think bit-boards are an important optimization to do in UTTT or do you think I'm wasting my time?
inoryy: it's very important, yes
jthemphill: sweet, I shall carry on then
RohanasaurusRex: So how long have you guys been programming for
twitlydoof: yes