« first day (1605 days earlier)      last day (2904 days later) » 

12:04 AM
@Lembik does the program need to exit successfully, if it's printed the correct output ?
i have something that is matching with brute force for the tests i'm throwing at it, although i'm not entirely confident it's correct
atm being able to exit with some junk on stderr in some cases allows a 5 byte savings
6 hours later…
6:06 AM
And Beta Decay is back with a bang ;). Haha nah, joking, but does anyone know why 5318008 is so popular?
6:51 AM
@MitchSchwartz great answer!
although junk on stderr isn't great :)
@Lembik thanks :) do you want me to change the answer?
@MitchSchwartz It's up to you really...look inside yourself for the answer to that :)
on anarchy golf exiting with a nonzero status is not penalized
atm exiting cleanly costs me 3 bytes
now only 1 byte
@Lembik was that the approach you found ?
@trichoplax I guess CJam is not big enough to flee in the first wave
Hopefully for CJam the junkware is only being bundled with the most popular mass downloaded files, but I thought it would be worth mentioning here just in case.
(and even smaller projects may see this as an ethical reason to leave)
7:31 AM
Hi everyone
Hi @Katenkyo
@trichoplax In the open-source world, ethic is everything. It's more than a big reason to leave.
Part of the reason I posted it is so that people who didn't realise it was happening could find out and decide whether to move their project
@MitchSchwartz Yes exactly. How did you work it out out of interest?
And it's a good reason, I wouldn't know that without you !
(I have no projects on SourceForge, but it's a reason to always track the official site when downloading :))
7:38 AM
I only found out because I happened to search for "free software" while writing an answer on Area 51 for the upcoming Open Source SE site.
I wonder if it's worth pinning to the starboard?
It's old-ish news. I guess, though, if nobody objects
@trichoplax The problem with the internet : if you don't look for information, it will not come to you... So you were lucky this time ^^'
I starred it, it can't hurt
@Lembik well i'd already used that indexing before, and then Sp3000's solution used KMP so it didn't really take any ingenuity to think of it, then just a matter of checking whether it works out
i mean, ingenuity for me to think of it, given that i'd read Sp3000's answer
This article is from the day before yesterday, but yes it does seem to be a recent example of an older problem
8:26 AM
I think I might change the spec of summoner wars a little...
I can't find a way to implement the clone/merge properly, there's always some bug with them, or insercurity...
and can't find a way to balance the clone either :/
Do you mean cloning is too much of an advantage at present?
Yeah, even if I cut life by two for the cloner and the clonee, it's still to much of an advantage (you just produce glass cannons in fact)
I would recommend basing decisions on what makes for the best game - bugs can be raised here and attacked collectively by the community
I think I'll come with a more "basic" version, without the clone/merge system at first. Everything else will be kept.
There's still plenty strategic option as it is
Could cloning depend on some analogue of "food", so it can't be done indefinitely? Maybe cloning requires gaining something that either takes a lot of time, or requires killing other demons, or some combination of the two
8:31 AM
I can't do it by killing deamons, because that mean it would be easier to clone your strongest than your squishiest
same for "food" on the board, a stronger demon could more easily go take it
it end up with only the strong can clone, where, for balance, it should be harder for them to do
(or at least, as hard)
I think I'll do a first version without cloning/merging in off, I'll see how it is, and then start a reflexion about cloning if it is needed
(If I can't in this one, I can always come with a v2 later, and some more options/constraints etc :))
Maybe cloning (and possibly exploding?) could be capacities rather than basic moves?
@trichoplax So it would be even more limiting regarding the capacities, because cloning would be the must-have :/
Ah yes
Maybe there should be some incentive for not cloning
Perhaps some stat that increases each turn but does not increase on turns you choose to clone?
are there nontrivial python programs that remain valid under the deletion of one single char from the code? let's say nontrivial means producing output
yeah, but I can't think of a good fair one
@trichoplax Games would be endless, plus, the winning strategy would be clone-spamming
You clone-spamm untill surrounded, then wait the increase
You end up with your rear over-buffed
8:44 AM
While you are cloning your enemy has been building up a huge reserve of that stat, allowing them to decimate your clones
Maybe the stat that increases could be attack?
So the increase should stop for at least 5 turns
@trichoplax No, because it would make defense useless
same goes for an overall buff : attack/defense increase cancel each other
so you just end up doing the same damage, but against an opponent with more life
Range would be the only thing that matter
But it would make explode too strong
you wait the up of your "bomb" and wipe the ennemy team with raw damage at once
too basic to be a good "optimal" strategy, it would show a hole in the specs
And without raw damage, explode could be removed, rendered useless
So range increase is not a good idea x)
What if you don't have any stat points to start, and get a small number each turn that you choose how to distribute differently each turn?
@trichoplax Could be an interesting idead, a great one that excite me !
but it doesn't solve the cloning problem ^^
(put everything on the same guys -> he clone each turn, a golden army :))
8:50 AM
I was thinking that maybe you need to spend all your turn's points to make a clone if the clone will have that many points (so you are cloning the demon, not it's points)
So if you have 8 new points per turn, you can use them to give 8 demons one point each, or to clone one demon that currently has 8 points, or to clone 2 demons that currently have 4 points
@trichoplax So it wouldn't be a demon doing an action, but you deciding to "cast" a clone spell... Have to think about it
If you have a demon that has 9 points, then 8 points is not enough to clone it. You would have to save up points or wait until the demon has lost points
If you want it to be the demon's actions, you could say that all of the stats are shared between the demon and its clone
So they have half each
Then turn based points would gradually increase them again
@trichoplax I love this idea, would be very interesting !
(screening the conversation to keep it in mind)
This should mean that it is an advantage to make a small number of clones (to give strategic advantage from position) but making a large number of clones is a disadvantage as they are all weak
@trichoplax But as you can't do less than 1 damages, it could also do it, very nice !
Thanks a lot !
8:57 AM
If your attack drops below 1, I take it you can't attack anything until you get your attack back up to 1 again?
@trichoplax it depends of which attack
if you're using the move CROSS for instance, you get to damage with attack+1 (attack won't drop below 0), so you'll do 1 damage, no matter the defense of the opponent
Ah yes. But not all demons have Cross
but if you use EAT, it's purely (attack-def<0)?0:attack-def;
@trichoplax actually, ATTACK also has this +1 bonus as it is basic and has a 1range no matter your stats
It seems there is lots of balancing to think about
Haha, yeah, it's all about balancing when you have to play with stats/characters/spells ^^'
9:03 AM
It's interesting
@trichoplax I love Game Designing for this reason ! It's kinda interesting, and pretty every idea can become true !
If you don't reduce stats when cloning, I guess you could produce a clone army and then merge them all back into one with huge stats
@trichoplax That's the point why a merged unit was disallowed to clone ^^'
@Katenkyo yes I was just looking at that rule - it doesn't prevent doing all the cloning first and then merging with them all afterwards
@trichoplax It was to disallow this one guy with huge stats from cloning again into an army :)
I never was against using clones to buff yourself :p
9:10 AM
Ah OK :)
Making me think of it, if you have ways to buff yourself other than merging, then merging become pretty useless...
9:39 AM
so, for points each turn, something like 6+nbTurn/50 should be ok, right?
9:58 AM
So you get 6 points to distribute on the first turn, then the same for each of the first 50 turns, then you get 7 points per turn for turns 51 to 100?
10:13 AM
@trichoplax that's it. I think you'll also get points at start, and each demon you summon at start would cost you 2 (you start with 12)
10:37 AM
This way, you still can chose to start with between 1-6 demons with an advantage if you start with less
11:27 AM
Well, did some big edits on the sandbox :3
A: Sandbox for Proposed Challenges

Katenkyo Sandbox note : Big change in specs, mainly for balancing issues (thanks to @trichoplax for your sugeestions :)) Major changes : Starting, Turn deroulement (2 phases from now), tie-break, Cloning. Merging may be removed due to the new specs (rendered pretty useless). With the current ...

12:01 PM
TIL: you can refer to meta tags with [meta-tag:] in exactly the same way as [tag:] for tags on main
@trichoplax hum?
[tag:king-of-the-hill] gives a linked tag icon
When writing a meta post the tag icons still link to tags on main, not tags on meta. If you want to create a tag icon that links to a tag on meta, you can by using meta-tag instead of tag. For example
Ho, yeah, never noticed Oo
12:19 PM
Well, actually, the model seems ok, only the controller is remaining ! \o/
Things are always highlighted better by the presence of a controller...
Yeah, but as I was redoing a lot of things, and had a nearly working controller, I knew what to change on the model
hm I wonder if this could make a challenge:
reproducing it?
@MartinB├╝ttner 3d stack overflow ?
12:32 PM
yeah sort of... although it's hard to get exact angles and coordinates... I created that by tweaking the parameters by hand until it looked believable.
@Optimizer yes.
I thought that's obvious :D
That would be interesting as an animation...
Like a slinky spring :)
is there a PBM like format for gif ?
There's a poorly supported one for PNG, but I don't know of anything as simple as PBM
12:51 PM
I wonder if there's a simple "draw a cube from this angle" challenge
Probably a bit simple though
@Sp3000 I just asked a question on Mathematica.SE about that. :D
Throw a series of error messages that viewed from a distance most resemble this image
1:06 PM
That sounds almost as insta-close as "Write a Java program which looks like a heart and outputs a program which looks like a duck, which in turn outputs the original program, without using any file-writing methods and without any of the letters in "HEART" or "DUCK""
Oh, and no semicolons.
@trichoplax Closed as unclear. The viewing distance is unspecified.
@Geobits or commas
@Sp3000 lol
Write code that makes toast using no ASCII or Unicode characters and no bread
user image
1:11 PM
Wow, I left 2 minutes, and it's already speaking of non sense !
@Katenkyo You particularly don't help either :P
Who said I want to help? :)
I love non sense !
I eat non sense
I drink non sense
I AM non sense !
Change your name or didn't happen
See? I am non sense :)
Hum, chat isn't updating my name :'(
1:18 PM
So... how long does it take until you can change your username back?
within some time or 30 days
It took Bigtoes 30 days for sure.
Not a big deal, I don't care of my nick ^^
1:46 PM
is there a pyth built-in to stringify something?
I can't seem to find it
Why do people star the stupidest things I say? I'm going to have to start saying sensible things.
Saying sensible things is a definitely a slower path to stars.
A: Sandbox for Proposed Challenges

trichoplaxMultiples - a wrap battle king-of-the-hill Overview Change cells in multiples to wipe out your opponent, while avoiding being wiped out yourself. This is a 2 player game, played on a linear string of cells of length L that wrap in a loop. Counting along the loop eventually brings you back...

@trichoplax Because people star funny things :)
A new KotH with cool 1d graphics
1:59 PM
@trichoplax Yeah, I like the Idea of this koth, I can't find anything to say actually :)
Thanks :)
I need to think about what numbers to use, but apart from that it's pretty simple so hopefully it should work out
I'm trying hard to think of an optimal solution but I think it's safe from that
@trichoplax btw, thanks for the forgotten 4 moves, fixed :3
@trichoplax I don't think there's an optimal solution, because it could always be destroyed by the opponent.
It seems that extreme approaches are less effective, so that gives me hope that it will be about finding an effective approach in the middle, rather than being an obvious optimal solution. For example, neither using all of the stockpile every move nor saving all of the stockpile every move are optimal, since the more you save the more you can use next turn, but the more you use the more you will gain next turn
For the range, is that the diameter rather than the radius? Otherwise radius 9 would be enough to wipe out a 17 height board
2:18 PM
Everything is expressed with radius
I'm counting 17, because you could hide in a corner, you need 40% life to explode, so it's dangerous to be in the middle
maybe log(range) would be better?
Ah I see
@trichoplax From what I'm seeing, I agree, extreme approaches are, at best, leading to a draw
Can a player see their opponent's information? Will they know that a demon is sitting in the corner saving up for devastation?
@Runer112 nooooooo u moved to pyth too :(
(that would mean a radius of 4+1 instead of 9+1 for 81 points)
@trichoplax Yes, you can see the position, the stats and the life of each of your opponent demons
but you can't see their name or their capacity (still thinking about that, I don't like it, I want bot that can predict everything è_é!)
2:23 PM
Name? You mean you can't tell which demon is which after they move around?
So if you spot that one has a capacity from what they did, you can lose track of which demon it was if they shuffle around next to the other demons?
@trichoplax If I hide the names, I'll allow the opponent to label them for sure, it would be stupid other way ^^
hiding the name is to prevent the meta-gaming
So they could be numbered?
Will a given player choose their demon types before knowing anything about their opponent?
like "hoooo, when I see a demon with this name, I know it's this player and this one will be used like that"
@trichoplax Yes
Oh you mean names of the players - I was thinking names of the demons. I get it now
@trichoplax They could, I'll go for that( not hard to do :))
@trichoplax Nah, name of the demons too
2:27 PM
Or the players could choose one demon at a time, and be able to see what the other player chose before deciding which demon to choose next
@trichoplax I mean, if you name your demons after their roles (all your healer are called "healer", your surprise mofo bomb is called "explo" etc, it could be spotted)
Most KotH answers include an explanation of the strategy anyway - it isn't usual to hide that
But if you just give them numbers you could assign different numbers each time so an opponent can't just hard code "the enemy demon number 3 is an exploder"
It's for preventing the use of those explanation, to make sure a bot will have no way of telling who's he's dealing with
@trichoplax If I do number, it will be with a static counter, and each demon generated will increment the counter
yes that makes sense. So I could make my player generate the same demons every time, but in a random order so that the counter numbers don't come out the same
that's it :)
2:34 PM
Next question V is so cliched.
Q: Stretching Words

Stretch ManiacWrite a program or function that duplicates characters in a word, so that all the duplicated letters arranged in order would form the input array. For example: input: chameleon, [c,a,l,n] output: cchaamelleonn Input The starting word (e.g. chameleon) An array of characters ([c,a,l,n]) or a ...

2:50 PM
Q: All possible subsets of array, with specific size

CQMThere exists an array with 8 objects in it, create all possible combinations subarrays containing 4 objects in any order.

1 hour later…
4:09 PM
we were 6.3 once, now we are 3.6
It was 3.5 a couple days ago :(
@MitchSchwartz hi
4:33 PM
@Lembik hi, i'm going to be away from my computer for maybe a few hours
4:51 PM
@MitchSchwartz just wondered if you would so kind as to put up a (slow) j =3 solution
so I can check any future efforts I make
By default, can I submit a C #define macro instead of a function?
@MartinBüttner Yeah I don't think Retina is possible without a market
5:31 PM
let's try this again, it looks like more viable responders are around
You have a question?
in Pyth, is there an operator to convert a list to a string?
Ah, I'm not viable...
as in, [1, 2, 3, 4, 5] -> 12345
I know solutions like joining with empty string work, but that's two bytes :p
5:54 PM
A: Sandbox for Proposed Challenges

Stretch ManiacMerging Words Input Two strings, and two numbers: hlwl, elo_ord!, 1 2 These may be function parameters, STDIN or language equivalents. Output The second string merged into the first string. The numbers specify how many characters to add when merging. This is what happens: h l w l ...

@Dennis oh yeah, I did edit the -s thing in when I changed it to the (broken) 28 byte version
@Sp3000 it's possible, but longer than 33 bytes
Any ideas how I could fix my More is Less? I don't really want to make it a pop contest, but every scoring metric I can think of would probably favor the least interesting solutions...
6:28 PM
A: Sandbox for Proposed Challenges

trichoplaxHigh Precision Metallic Means code-golf other tags? Background The metallic means, starting with the famous golden mean, are defined for every natural number (positive integer), and each one is an irrational constant (it has an infinite non-recurring decimal expansion). For a natural number...

6:46 PM
@Dennis I think it's probably better if it is a pop contest. I don't see a way to score this that won't end up in a trivial/boring solution.
Q: Rocket Bots KOTH

CainRocket Bots The year is 3024. People have become too scarce a resource to risk in war, so combat has shifted to the robots. Your task is to construct a bot like no other, who's rockets will rain destruction down on your foes and shoot down all incoming threats. Gameplay Bots Gameplay takes pl...

7:07 PM
@Vioz- Me neither. I've made it a pop contest for now. If nobody comes up with a better idea, I'll post it as such.
Why are there so many old challenges being dredged back up to the active page? Is there some new policy that I've missed?
Community bumps questions without accepted answers from time to time.
Doesn't really make sense for PPCG, but elsewhere it means the OP still needs help.
@Dennis There seems to be an abnormal number of these questions at the moment though
Mostly with 1/101 rep users answering them
Not a bad thing. Everybody has to start somewhere, and the older challenges are usually easier.
Q: Implement a FuzzyFinder

kirbyfan64sosInspired by this link I found on Reddit. A FuzzyFinder is a feature of many text editors. As you start to type a file path S, the FuzzyFinder kicks in and shows you all files in the current directory containing the string you entered, sorted by the position of S in the file. Your task is to imp...

7:13 PM
Yeah I guess so. I wonder at which point, all simple challenge ideas have been used up... Only horrible complex ones like that MtDNA challenge
Don't bother trying new ones as @Dennis would have already answered them before you even come to know of the question
2 to 3 times
7:25 PM
@Runer112 No, jk is the shortest.
If the first element in the list is a string you could simply sum with s.
7:41 PM
I found this question on math.se, and I find it quite interesting. For an n-sided die and a number from 1 - k, is there a general method for finding a solution that takes the least number of rolls, on average?
7:51 PM
What would you define as a solution?
As in the number of times the dice needs to be thrown to approach a uniform distribution?
8:12 PM
@BetaDecay Come in here to discuss your code, there's too many comments as-is
8:38 PM
@Vioz- I'm here
I called you before you fixed everything :)
Your code works now, you might want to think about changing return to print to save a byte easily, and adjusting indentation to save bytes
Oh yeah, thanks haha
I've forgotton how to perform simple mapping in Pyth again
how to apply a built-in unary function to every element of a list?
8:55 PM
@Runer112 hMY, where h is the unary function and Y is the list.
I'm an idiot
that's what I thought it was, but I was using the wrong symbol for the unary function
@Changming If you need any help in Pyth, I'm happy to help.
A: Sandbox for Proposed Challenges

DJ McMayhemFind the optimal nim move This is a code-golf challenge. The game Nim is a mathematical strategy game, where 2 players take turns taking items from distinct heaps. On your turn, you must take at least one item, and you may take as many as you want, provided that you only take from one heap. T...

@xnor Did you try any approaches with x.split(c,1) for Stretching Words, out of curiosity?
hmm, I didn't
I tried with find, but I like your idea
9:07 PM
def f(x,y,o=""):
 for c in y:a,x=x.split(c,1);o+=a+c+c
 print o+x
^^ That's 65, unfortunately
can you recurse instead?
Yeah, that's what I'm trying to figure out
@Lembik it feels low priority to me, and not so interesting since it is a straightforward extension of j=1 and j=2, but i'll probably get to it eventually
Q: Best place for learning GolfScript?

PhaseIt looks like a fun language, but the official site is down and the EsoLangs page doesn't explain it quite well and looks like it's missing stuff. Is there anywhere that the language is documented?

9:54 PM
Nothing like getting a question to HNQ for getting random votes on answers from months ago...
10:19 PM
Why does [1 {7}*]{+}* put 7 instead of 6 on the stack?
Is this GolfScript?
@Phase Why would it put 6 on the stack?
1 {7}* runs the block {7} once (just as 7{1}* would run the block {1} seven times). So [1 {7}*] creates the array [7]. Then {+}* just gives the 7, because fold needs at least two elements to apply the block.
oh, makes sense
I thought {7}* would be run before the 1
@Phase {7}* wouldn't make sense by itself, since * needs two operands.
10:45 PM
I think golfscript.apphb.com just went down :\
What is it with all the GolfScript stuff breaking? O_o
Would golfers prefer arbitrary precision language features to be permitted or excluded for a golf question that requires up to 65,535 decimal places?
I prefer placing restrictions on end results to intermediate steps, but this will be a big advantage so I'm torn
65,535 decimal places? Is there a language that can support that?
Peter Taylor tells me that GolfScript has arbitrary precision, which is on topic right now :)
Some languages have arbitrary-precision ints, and rationals, so yes.
10:52 PM
I'm inclined to put no restriction as there is always some advantage for some languages whatever I decide. I just wanted to double check there wasn't a strong objection to that first
@Doorknob do you have a preference? It's for the metallic means sandbox question.
I would probably just not have a restriction.
Thanks - I'll probably go with that then. Saves me worrying about the wording of the restriction too...
I love how each operator can have so many different functions. Like ~ can be NOT or evaluate a string.
11:33 PM
rAnyone know where I can download the GS Ruby parser?
nvm, got it
A: Sandbox for Proposed Challenges

trichoplaxOne transparent colour code-golf Overview This challenge will only involve a hex string as input and a hex string as output, but will be explained in terms of computer graphics. Given a fixed size background image and a smaller fixed size sprite, place the sprite on the background image all...


« first day (1605 days earlier)      last day (2904 days later) »