« first day (1368 days earlier)      last day (3493 days later) » 

6:47 AM
Awwww, thanks so much for the downvotes guys :P
 
 
1 hour later…
7:51 AM
Anyway, we can graduate now. :)
 
Noooooo! I've just realised that Quincunx's polynomial factoring question doesn't specify that the polynomials are monic.
 
@MartinBüttner Congratulations on getting 20k!
3
 
thanks :)
@PeterTaylor The implication being? Wasted work based on a false assumption?
 
The implication being that the factors can be non-monic, so the division has to be able to do reciprocals over the base field. Relatively straightforward in finite fields of prime order, but over Z that means I need to implement rationals. It's a whole new layer of complexity.
Actually, over fields of prime order I can forget about it, because I'll just get a constant multiplier at the end.
 
I got prime fields, Z's a pain though. Had a heuristic but apparently it's complete false.
If I use brute force, is there any way of knowing when to stop?
(If there is, then 2 hours is a very generous time limit)
 
8:14 AM
@Sp3000 Yes. I found an excellent review paper on the subject: Bounds on factors in Z[x]
 
Wow they're.... pretty big bounds
 
Yes, so in GolfScript I may have a tradeoff between speed and length to meet the 2 hour requirement. In Python you should be fine with an ever looser form of Mignotte's bound
 
It wouldn't be hard to implement, but I wouldn't be very satisfied with my program :(
I'm already thinking of ditching the code golf aspect just to get a "good" program in
Since it's a hard question
 
8:29 AM
0
A: Sandbox for Proposed Challenges

Beta DecayEvolution of Tetration domino-coding N.B. This is another version of Calvin's Hobbies new challenge idea. I won't go ahead and post it until I get a full thumbs up. If not, I'll convert this to code-golf. Tetration Tetration, represented by na, is used to represent large numbers. Basically, ...

 
@MartinBüttner What're you thinking of as a restriction?
 
Nothing in particular. I just think that Calvin's Hobbies chose Levenshtein distance + simple task to test the waters, and not to make that the thing of the challenge. If we keep doing that, that's exactly what Doorknob warned of in the chat yesterday: it's gonna derail into the next code-trolling very quickly.
 
Ahh okay.. I'll have a think about it then
 
I was thinking of something like "Print the previous program" at one point (with char limit restrictions), but I think that'd turn into a horrible kolmogorov :/ But something different in that sense?
 
That would just be people getting the previous answer's code from online...
 
8:39 AM
Ahaha true - would be patchable, but that was just an example :P
 
Maybe have a rule that your answer must contain 50% of the previous answer's characters with comments banned.
 
never mind...
 
Nah, that regex might go down well
 
@BetaDecay what for? fetching results from the internet is a standard loophole?
@BetaDecay no it's too trivial... unless you do something like "a regex that matches all previous submissions, but nothing else"
 
Oh yeah. That idea might work then.
 
8:47 AM
score would be most submissions, ties broken by sum of regex sizes (less is better)
gotta head off
 
See you
 
I just realised on the way to the train that the above idea is still broken. What could work is "match the previous submission but none of the ones before it".
 
Yeah. Matching all of them would be quite a feat...
 
9:12 AM
No, it would require matching all of them literally with | in between
 
Ohh. That'd be rubbish
What do you do for the first answer though?
 
 
1 hour later…
10:24 AM
@BetaDecay [] in ECMAScript flavour (matches nothing at all)
 
10:39 AM
0
A: Sandbox for Proposed Challenges

Martin BüttnerRegex Domino Golf domino-coding code-golf regular-expression I just wanted to record this idea here before I forget it. However, I will hold off posting it until I'm convinced that domino coding is actually a good idea for PPCG (which I'm currently not). Feedback under the premise that domino c...

 
@MartinBüttner Ahh
 
Do you think it's possible for someone to ruin the fun with a stupidly large regex that, say, contains the superstring of every possible n-gram or something in order to always get matched?
 
no, just use anchors
 
Ah true :P
 
also, not every possible n-gram is valid regex syntax
I'm off again
see you
 
11:13 AM
0
A: Sandbox for Proposed Challenges

Calvin's HobbiesSo I had written this up as another domino-coding contest but the more I wrote the more I felt it was too complex rule-wise and might not work anyway. I might scrap it but I might as well see what others think. Help Grow the Communal Code Blob [code-challenge] [printable-ascii] Intro Defi...

 
 
2 hours later…
1:07 PM
0
A: Sandbox for Proposed Challenges

Vi.Traffic lights Easy challenge Write a program that that graphically shows traffic lights on a crossroads. It should output animation. There should be red, red+yellow, green, blinking green, yellow. Bonus - pedestrian traffic light. Complex challenge Write a program that manages traffic lights...

 
1:19 PM
Woah, I've never seen programs so short...
0
Q: Print the ASCII table

Forlan07The task is to display n characters of the ASCII table. You may write a function (or a program that takes the argument as a parameter, STDIN is allowed as well) that takes a parameter n, which will be the index of the last character to print. The task is quite simple, so as an example here's a ...

 
1:48 PM
@Rainbolt Can I ask a serious question? Why are pushing so hard to defend that question? It was a pure code dump with a title that didn't match the actual code. You argued that it was perfectly clear before you realized/read what the code did, and refuse to change your position in light of that. What am I missing here? I know you're argumentative (hell, so am I), but I can't help but feel it's just not worth it in this case.
 
2:18 PM
@Geobits Can you please explain how pushing to get the close vote reason changed from "Unclear" to "We don't golf code for other people" is "defending the question"?
@Geobits Like I said at least five times already, the fact that it was code at all meant that it is inherently unambiguous. I didn't have to understand what it did, and I'm still waiting for you to explain why it's important. My stance all along has been that what the code does is not important.
 
@Geobits +1 for your MSE question!
 
Even if I agree that the code itself was unambiguous, the fact that it doesn't match the title or the OP's obvious intent makes the post unclear.
@ProgramFOX Thanks :)
 
No, it makes the OP's intent unclear, not the post.
 
If I title a question "Sort my array" and dump code that prints pi, that's unclear.
 
The task was clearly to shorten the code that was dumped. I'm sorry that you had trouble figuring that much out.
I guess we'll agree that one of us was able to see a clear spec and one of us was not.
 
2:29 PM
You can be patronizing all you want ("I'm sorry you couldn't see that..." Really?), but that doesn't really help anything. If you think it's clear, vote to reopen it and let the community do its job. I'm done with it.
 
@Geobits Speaking of being patronizing, who was is that was using my clearly nonexistent confusion as the basis of their argument?
"It took multiple comments to figure out that it was Kolmogorov." No, not really. That's just you projecting your own confusion.
@Geobits And why would I vote to reopen the question? At what point did I advocate that? You are just full of "put words in your mouth" today.
 
2:49 PM
hola
0
A: Sandbox for Proposed Challenges

durron597Keep the unique characters down The challenge It's very simple: Your program should print the following text: Elizabeth obnoxiously quoted (just too rowdy for my peace): "THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG," giving me a look. Scoring The number of bytes in your code multiplied by t...

 
0
A: Sandbox for Proposed Challenges

durron597Keep the unique characters down The challenge It's very simple: Your program should print the following text: Elizabeth obnoxiously quoted (just too rowdy for my peace): "THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG," giving me a look. Scoring The number of bytes in your code multiplied by t...

 
Excuse me bot, I just said that.
 
@durron597 For us lazy folks, is every letter of the alphabet in the sentence?
Both cap and lowercase?
 
I'm a bit late, but... ^ hooray! Congrats @MartinBüttner ;)
 
3:06 PM
@Doorknob If that were all in one comment I'd star it
Ah, there we go. I found one to star :)
 
@Rainbolt Yes, both cases.
the quoted part has all the caps, and the sentence has all the lowers.
Anyone have any comments before I post it?
Originally I wasn't squaring the unique bytes part, but then I decided that wasn't giving enough weight to the uniqueness part of the challenge, which is the whole idea.
I actively want a language like brainfuck to have a chance to win.
But I don't want it to be a shoo-in, either.
My completely stupid Java program that just has one print statement has a score of 773955 under this scoring system
This means a brainfuck program would beat it if it were under 12093 bytes.
 
@durron597 And what happens if you don't square it?
 
@Rainbolt Java score: 12285
 
Which means the brainfuck solution would have to be how long?
 
3:19 PM
Brainfuck would have: 1535 characters to beat it.
This is the java program:
class S{public static void main(String...a){System.out.println("Elizabeth obnoxiously quoted (just too rowdy for my peace): \"THE QUICK BROWN FOX JUMPS OVER THE LAZY DOG,\" giving me a look.");}}
 
@Doorknob Thanks :)
 
Using just a basic Ascii to brainfuck translator, i get 1160bytes long
 
Sounds to me like not squaring is better, then.
 
I don't think it really matters. When I golf (which is rarely), I just compete with the JavaScript and Python answers and ignore the others.
 
CJam has an easy score of about 3700, I think.
(And it's not if the score isn't just "How many bytes?")
 
3:28 PM
@PeterTaylor Just code challenge then?
 
@PeterTaylor "Kolmogorov complexity of an object, such as a piece of text, is a measure of the computational resources needed to specify the object." If you consider unique bytes a resource, then the challenge is indeed a Kolmogorov complexity challenge.
I took that from the tag btw
 
I think so.
@BetaDecay The tag is slightly looser than a formal definition, but "Kolmogorov complexity" doesn't mean "fixed output".
 
@durron597 This may have already been asked but is the passage counted as part of the unique character count?
 
@BetaDecay The string itself, you mean? If that's what you mean, then of course, that's the whole idea - the string itself has a unique byte count of I think 59
 
3:34 PM
@PeterTaylor Many fixed output kolmogorov-complexity challenges exist with no quarrel about the tag though
 
Yeah, 59
What is the boring perl -p score?
 
The boring GolfScript literal string scores 7560
 
I guess perl -pe is the same as the golfscript literal string?
 
Maybe a shade longer, because I believe it would need an assignment to $_
 
Won't GolfScript score the same as CJam and others? What's the tiebreaker?
 
3:39 PM
Maybe it would be better to just do code golf and cap the number of unique bytes
 
I'm 99% sure that would be a dupe
 
hmm
 
Wait... the string is being provided as input? Or am I misunderstanding what -pe does in Perl?
 
Yeah, you should stick with the unique characters.
 
3:41 PM
The argument to -e is going to count in the score right?
 
so the program would be $_=<the string>
@Rainbolt yes
regardless, the point is there are languages that can print a string as-is with almost no additional characters
 
@durron597 If they do that, then they lose right? So what's the problem?
 
@Rainbolt Do they? If we're not squaring the unique bytes bit of this
ascii to brainfuck scores about 9k
boring golfscript string literal scores 7560
 
I imagine that it would be much easier for Brainfuck to reduce the unique character count than Golfscript
 
@Rainbolt you mean reduce the total score?
 
3:47 PM
Well, of the two factors, I think unique character count has a much higher order of magnitude already. So yes.
 
The string itself has 124 characters in it.
Brainfuck, according to @Ethiraric, is 1160 * 8 = 9280
Golfscript, according to @PeterTaylor, is 126 * 60 = 7560
 
Ok, so removing a single unique character from the boring Brainfuck answer would give you 1325 extra characters of space.
On the other hand, Golfscript gets only 128 extra characters by dropping to 59 unique characters from 60.
So Brainfuck already has far more to gain from removing unique characters. The question is, how difficult is it to do that?
 
You can hardly remove one of them. How long is the string to display ?
 
@Ethiraric The string has 124 characters.
 
Then it might be possible just using + - and . to score less than 7560
 
3:54 PM
The goal is to come up with a scoring system to make it interesting with both a regular language and language with brainfuck or whitespace - where it's not obvious which will win.
 
Using 3 unique characters gives brainfuck 20.3 bytes in average to display each character of the sentance
 
How do you easily get the counter up into the lowercase letter range without loops?
then you have to go from d to space to ( to j
 
There is no easy way, it's just tons of + and -, but may be you will save enough bytes when displaying the rest of the text
Or using > and <, it can use 12.2 bytes in average
 
maybe i should restrict the source in some other way.
 
I've got to go now. I'll be back in an hour. I'll test it later. Bye :)
 
4:10 PM
@durron597 That was for the string literal. Using a simple base conversion I can score 4624 in GolfScript, and (as I mentioned earlier) the translation to CJam probably scores about 3700. (I can't be more precise without actually writing and testing the CJam program, because I don't know the language).
 
@PeterTaylor Do you think the question is a decent one as written (without squaring the uniqueness)
 
4:27 PM
If you're after a quick question for people who like Turing tarpits. I don't think you achieve the goal of being interesting for "regular" languages.
 
5:07 PM
I have just tested and a brainfuck program with only +-. is 3571 bytes long (10713 score). My apologizes i was terribly wrong.
 
5:39 PM
Today, one of my lecturers wrote the most terrifyingly mismatched expression on the whiteboard. I can't let you guys get away without suffering through the same pain:
Psi(x) = C(x) [ A exp(i S(x)/h + B exp(-i S(x)/h)}
3
 
5:58 PM
0
A: Sandbox for Proposed Challenges

YpnypnLeave your mark! domino-coding In this challenge, each answer builds on the previous one ... (Lots of text paraphrased from "Evolution of Hello World") Your challenge is determined by when you answer. Specifically, for the nth answer, divide n by 5, and if the remainder is... Print pi, ...

 
6:25 PM
@MartinBüttner For the sake of consistency, is there a third option? (wrt to your recent meta post)
 
@Rainbolt I don't know. As I said... some people like to exclude languages which are subsets of others. But that's hard to nail down I think.
ah yeah, the list
 
Yea I just realized I hadn't yet posted a way to make C and C++ the same
The only way to do that is to say they belong to the same class
I wish there were a more self-sustainable method.
 
well the subset one
 
What is the subset one?
 
Saying that languages are considered the same if one is a subset of the other
which catches C and C++
 
6:37 PM
It only catches it in one direction
Which is probably ok
 
doesn't matter, right?
 
Nope, you're right
That post indicates that C is not actually a subset of C++
 
I was afraid so
it's not a distinction I'm actually comfortable with though
and in fact, if you have non-standard C which is usually used for golfing, it's not valid C++ anyway
gcc does a lot more stuff implicitly which leads to compiler/linker errors in C++
 
Ah, there's a fourth method probably not worth posting. "The challenge author is judge. Period."
 
I think that is worth posting.
(I'd probably downvote it though. :D)
 
6:41 PM
Lol heaven forbid they change the way rep works on meta.
And I suddenly bounce around 2k rep because I do this all the time
 
You know what's more fun the expressing my own opinion?
Expressing all the opinions!
 
@MartinBüttner Arguably it allows as many "C"s as there are C compilers. They all have their little quirks.
 
7:00 PM
hm, true
 
7:36 PM
sick, 29 languages on the source-layout rosetta-stone challenge
 
7:55 PM
*30
 
Wow. I just got sent a code review from a developer who recently discovered ReSharper. I'm tempted to deny the review because all I can see is red. 3000 lines of red.
 
 
1 hour later…
9:10 PM
How was codegolf.stackexchange.com/a/40532/31414 able o answer when the question is protected and required 10 rep
@MartinBüttner Mine ? ;)
too bad I hit the rep cap again :(
 
@Optimizer He answered 2 mins before protection. His answer (or another shortly after) is probably what triggered it.
 
@Geobits Ah, I see. I saw that question protected since yest. maybe not 10 rep limitation
 
@Optimizer yes ;)
 
Code golf was almost invented to trigger Protection.
 
Better safe than sorry
 
9:13 PM
@Geobits Oh good, you came back. Now we can discuss Tips...
 
I was just going off the "Protected by Community 1 hour ago" banner. Looking at the revision history, Dennis unprotected it 3 hours ago.
@Rainbolt Better tips than the other :P
 
Who has better tips than who?
 
I have better tips, obviously. They're frosted.
(no. not really. god no)
 
How do I protect/unprotect stuff ?
 
@Optimizer Well, first you get a sword and shield.
 
9:15 PM
Or get more rep. I forget what the priv level is on beta, maybe 4k?
 
I see
 
there are other kinds of protections too
@MartinBüttner I can add more too. will do tomorrow
 
Sure. For some you need to create a platform, gather local support, and run in a legislative race. Once elected, you can work to protect those who need it.
 
9:17 PM
I'm looking forward to it ;)
 
@MartinBüttner The question is super interesting .
 
I'm glad you like it :)
 
10:15 PM
@PeterTaylor ;)
 
10:38 PM
woah a rocket just blew up
in The Pod Bay, 15 mins ago, by Undo
"Be advised, something just blew up"
 
that's not so good
 
11:00 PM
I have an idea for a challenge, but it isn't going to work, I don't think.
Something about you are an international spy for country X and you are using Stack Overflow to communicate with your boss.
You need to create questions and answers where every N word is part of your message, but Stack Overflow cannot suspect what you are doing.
Your questions need to be about the language you wrote the bot in (Golfscript bots need to ask Golfscript questions, etc.).
BTW, is this somebody's bot? codegolf.stackexchange.com/users/1834/user
 
11:41 PM
@hosch250 If so, it's one dedicated bot. Possible sock puppet, but it's a useful one either way :P
 

« first day (1368 days earlier)      last day (3493 days later) »