« first day (1139 days earlier)      last day (4001 days later) » 

00:09
@JonathanVanMatre I'd like to point out that my question contains potential cycles, whereas the other does not. The algorithm to solve such a problem falls into a completely different class of hardness
Digging in to this APL/Lukasiewicz thing.
0
Q: How to understand the F function in Burks/Warren/Wright's Lukasiewicz Logic Machine

luser droogFrom the bibliography of chapter 1 of the 1962 A Programming Language, I found this intriguingly concise description of a forward-Polish (Lukasiewicz) Logic Machine. And I think I'm with it up to this part on the Logic Function F: What does (2a) mean? How is this a function? Here's my implemen...

00:22
What is our policy for answers that are near-identical to others?
0
Q: What qualifies as "enough difference" in posting similar answers?

QuincunxThis user likes to copy people's ideas and translate literally into other languages. For example, this post: http://codegolf.stackexchange.com/a/23951/9498 from my post: http://codegolf.stackexchange.com/a/23822/9498 (note that I golfed mine more after he copied it). His code was in TI-83/84 Bas...

00:46
I have seen that they are supposed to be either golfed more or given a different twist, and credit must be given to the original poster.
Oh yeah, I got a "Nice answer"
@TimWolla Nice answer, then.
I have both Nice and Popular question, but no nice answer badges yet.
01:05
I've got "Good answer" on a bad answer. :(
You mean the one with a smiley?
It isn't that bad.
I just know C++ and a little Python, so I can't do much (or anything) cool with my programs like that.
I wonder if APL had to wait until the invention of the shift-reduce parser. The 62 book only shows how to parse fully-parenthesized expressions.
Knuth's LR(k) paper came out in 65.
0
Q: Write the most terrible and annoying website possible

DoorknobYour challenge today is to write the worst website you possibly can! Rules: You must use HTML combined with any other language(s) of your choice. (For example, CSS and JavaScript might be common choices.) Your goal is to create the worst website possible! This includes but is not limited to: ...

enjoy ;)
If I can get this Lukasiewicz thing working, I think I can implement the Turtle Graphics via implementing LOGO.
@hosch250 Which one with a smiley? I'm talking about the Hello World as a big echo with @ signs.
Yes.
01:14
Oh. right that smiley.
I do put them everywhere.
01:27
Hmm. Is there a language called "Brainfsck" yet? (with an 's').
0
A: Proposed questions sandbox - Mark X

TimWollaMinesweeper Solver code-golf algorithm We already generated Minesweeper fields, but someone really has to sweep those generated mines before they explode. Your task is to write a Minesweeper Solver interfacing the accepted solution of “Working Minesweeper”. Rules: No Standard "loopholes" wh...

It's already used as a euphemism.
@TimWolla That sounds neat.
@luserdroog What's your opinion on my question in the last paragraph?
I'm still loading the page, but the excerpt looks interesting.
01:45
Woah. I missed this one.
4
Q: Create a random 3D asteroid

Sebastian NegraszusImagine you are working on a 3D adaptation of the classical video game Asteroids. Your task is to write an algorithm that can create a random 3D asteroid. This is a popularity contest, not code golf. Write nice code and create interesting results. Add other cool features if you want. Rules: Y...

I wonder how you'd bust them up. Tetrahedral partitioning?
02:44
@Rusher @PeterTaylor OK, OK. I hadn't given a full read to Rusher's question; I was just using it as an example based on Rusher's description of it.
And I am aware of the differences when the graph is acyclical. Negation of the graph enables solving for the longest path by solving for the shortest.
Lukasiewicz Logic Interpreter. gist.github.com/luser-dr00g/9519896
Might make a good challenge.
Anybody know how to get the epsilon that looks like "within" as an html entity? & epsilon ; looks like ε
or not.
ε
03:03
@luserdroog Any opinion yet?
:)
Oh I think it's good. It's not quite my area, so I'm not sure how I'd go about solving it. Barring that, I can't really evaluate it. Except to say that I do like linking-up new challenges with the previous ones, making a network of related implementation goodies.
Maybe I'm distracted. It needs some kind of backtracking search, right?
Any opinion on my Lukasiewicz madness?
@luserdroog Backtracking is not possible. Do you back track in Minesweeper once you blew up yourself? ;)
What's your opinion on the question at the end? a) Communicate directly with the other program or b) Only use the format as input / output of a function?
Not exactly. But planting flags and then removing them once the bombs are determined, is, I think.
Oops. I didn't see the question. I'll give it some thought.
You'd need 2 pipes.
@luserdroog Well, I never remove a flag once I placed it. I did not take the challenge myself yet, but I would probably go for an interative solution:
And there may be issues getting the handshake right, I think.
03:13
until solved
  for cell in field
    if cell is obvious
      fill in
      break
I'm having trouble thinking of an example where I used flags this way.
@luserdroog So rather a function: string nextStep(string situation)?
It's when you've got an even row and an odd number, and one end connects to another partially-known piece.
I guess the tricky situations turn up more often in Expert mode on the classic MsWindows version.
But I've definitely used flags speculatively.
Just fired up my fancy new Windows 8 Minesweeper App
Let's see if I can solve expert :D
The "adventure mode" is loads of fun, but steals too much time.
And it won't share the screen with Netflix. :)
03:18
@luserdroog Yeah
Aaaand I blew myself
To bad :(
I can't remember the details but I have encountered difficulty getting a program to both send output to another program and receive its output via pipes.
But I also remember getting it to work eventually, somehow.
So I don't see any issues with the challenge, except, the question text is rather short. It could use some jazzing-up. :)
Usually if you blow after the first few clicks, you call "mulligan" and play again.
It's only after you open up a patch that the "game" starts.
@luserdroog I suspect using pipes / popen / whatever would kick out to many programming languages
Though it would be nice for GolfScript
GS, and js
Maybe not too many.
JS can communicate with other processes when run inside node.js
(I do test my CoffeeScript solutions in node.js)
04:15
@luserdroog I hope you didn't use named pipes? Those are terrible. Who even invented them
Maybe not there, but I have used them.
Btw, I managed to solve the 'Write a function that returns something different every time it is called' in haskell
I made .signature a named pipe, with a demon in perl to produce random quotes.
My login script would check if it was running and launch it if not.
or .profile or whatever
@luserdroog Oh my. You're evil.
I even ran an http server the same way.
on the university unix server no less!
Never got in trouble until the server crashed when I was downloading porn into a private subdirectory of /tmp
04:25
Bye o/
 
9 hours later…
13:16
wow, the number of rooms increased lately.
 
7 hours later…
20:34
Some mistakes in my tests?
21:03
Checking now
What would a computer use the question marks for?
@TimWolla In case you needed to be tagged to see this
Just saw it
@Rusher Just in case a computer would need them to store something
The Python program claims to support ? as well
I was just thinking.. I use them as a memory device when determining what move to make
But a computer would always make the best move right?
And wouldnt need a memory device
Like... I'll find a move where my odds are 50% and plant question marks
Then i find another move where my odds are 30%
I'll mark those instead
etc etc until I am satisfied that I have a good move
As I said: Just in case a solution wants them it can use them
If not: Fine as well
@Rusher Do you miss some good moves in the tests?
In what tests
Oh... I was getting there. Sorry
I assume a good move is the one with the best odds right?
A good move is one that is 100% correct
Say a number 3 with no attached flags and 3 unknown neighbours
-> all 3 of them have to be a mine
21:11
So all of your inputs are guaranteed to have a good move?
Because not all minesweeper puzzles are solvable with 100% confidence
@Rusher It may take any field if there is no clear move
"otherwise pick any untouched field and open it."
The tests serve the purpose of testing some basic logic
The program should be able to solve any solvable situation. Is that unclear from my wording in the question?
"Working towards a solution" and "sane move" are both subjective, so I had to clarify
Any suggestions how to make it clear in the question?
Sure. Remove the ? option and then say "Your choice must follow the best odds for survival."
That way, if multiple 100% moves exist, any can be chosen. If not, a 75% move will be chosen over a 50% move, and so on
And nobody can simply place ? everywhere and claim they survived
Your last test case appears to be impossible, but it's hard to explain why in few words
Possibly, but it should only test, whether the both given choices are taken
Both of them are clear choices
21:21
Err ignore that. I followed all of your test cases.
They all work
So, I edited the rules based on your suggestions. I think it's ready to go?
Can you check me on this?
On what exactly?
0 0 and 0 1 obviously cannot both be mines, correct?
Which test case?
21:23
Last one
@Rusher Yeah, correct
but one of them is a mine
So that satisfies the 2 at 1 1
There are now two mines touching the 2
Hu? I cannot follow, sorry.
21:24
The 2 at 1-1
Oh, nvm
I was being blind
It is now satisfied, right?
Yes
Looks like that test case is broken again
So the output 1 0 2 is probably bad
Because that would unsatisfy the 2
Yep, I'll try to construct something
21:26
lol just make a board in the actual minesweeper program (probably be easier than doing it in your head)
Thanks
@Rusher Oh, I did, but failed at minifying it
NP... I might answer if time allows this weekend. Good luck
Thanks for your help
If this is not the place for discussions like this, someone tag me and let me know. I'm not really sure what topics are encouraged here. Thanks.
@Rusher I think it would suffice to change the two 2's in the second row to a 3, wouldn't it?
21:54
How can we undo a bad post review?
I just reviewed this:
0
A: 4, 8, 15, 16, 23, 42

JaimeSomeone please make this in ArnoldC !!!

I switched tabs for a second, then came back and accidentally clicked Looks Good.
I couldn't find a way to undo it, so I flagged it again.
22:08
@Rusher It gone out, Thanks again
22:20
@hosch250 I've just downvoted and voted to delete.
22:54
@Gareth thanks.

« first day (1139 days earlier)      last day (4001 days later) »