The Nineteenth Byte

9:01 AM

@algorithmshark oh yes, we did have a lot of fun with that one ^^

1 hour later…

10:05 AM

@user23013 What do you mean by most new ?

10:22 AM

Ah... I meant most answers, which are newer than 25.

Ahaha, well I've posted the only one I'll post :P (my list has been exhausted)

now its done!

btw @user23013 Just saw your bounty - the regex in it seems to match most of the latest ones, but have you tried matching the earliest 4? They're the hardest, after all

I'm finding that #4 alone takes 6 chars minimum, so you kind of have to match it along with other things to make it efficient

[]^]P$ matches #2, #3

I think there was a version matched #4, but I didn't save that one and couldn't get it again

\[..].$ matches 3, 4, 13 - if that's what you mean

10:32 AM

It was something like ^.(?>...<..|.){5,15}..$

Ah, right - that's what you mean

Yeah currently I'm just trying to beat 35 - I don't want to post another rehash of the past 6 answers :/

I think the genetic algorithm should work better for similar solutions

Maybe - so far the GA only seems to output things close to what's already been posted though

(although there have been some interesting changes, like [^-$])

1 hour later…

TwiNight

11:57 AM

SE question

If I see a Java question on SO from 2011, and happens to have a better answer than all the existing answer, should I necro it or just start a new question and answer it myself.

overactor

@TwiNight Answer

grc

yeah definitely answer the existing question

overactor

Even if your answer wasn't valid in 2011

grc

if you start a new question it will get closed as a duplicate

overactor

It might not get you big rep, but it's the better option for SO as a whole

and, if it's a common question and your answer is good, you might rake in some rep over time

Peter Taylor

12:56 PM

Did anyone write a stack snippet for the 95 movie quotes question? I'm not sure I want to work through 70 answers to figure out which quotes are still available.

1:16 PM

I'm not sure how up-to-date the list is, but I assumed that it was up-to-date before I posted mine because the last answer's quote has been marked off

Peter Taylor

1:36 PM

Ah, I'd missed the prefix.

2:03 PM

@MartinBüttner Hey Martin, how long do you intend for regex golf to go on for?

@Sp3000 what do you mean? we'll see how long for people come up with answers.

I'm really tempted to do something about our current situation :/

as in...?

The past 7 or so answers essentially have []^C]\w$|\?[^w$]*\$$ in common, so all people are doing is just matching the remaining few with a few short alternations

and what do you want to do about that?

2:09 PM

I was wondering if it'd be really unfair if I tried to invalidate the second half of that regex

heh

well, it's your call

Sorry :(

do eet

Hmm this is harder than I thought, trying to get a regex that works and does what I want

I have one that invalidates the first half, but that might be a bit too mean...

well, we need to start having ~40 byte solutions

2:23 PM

We only have 38 bytes right now though :P If I kill it too much, we won't continue at all!

Sebb

Find an answer and check wether you can get another answer as follow-up

Interesting... :P might try that

@FryAmTheEggman U had me scared there, 3 score!

A: Sandbox for Proposed Challenges

Yeah, that was a typo :P

New Sandboxed Posts

Print a Negative of your Code code-golf source-layout (quine?) Consider a square of printable ASCII characters (code points 0x20 to 0x7E) for side length N, like the following (here, N = 6): =\ g \ 7 m+y "g L ~ e> PHq We also require each row and each column to contain at leas...

2 hours later…

New Sandboxed Posts

4:17 PM

A: Sandbox for Proposed Challenges

Code Yourself a DFA code-golf or code-challenge In this challenge, your task is to implement a deterministic finite automaton (DFA for short) that recognizes the following regular language L. The alphabet of L is {x,1,2,3}, and it is the set difference of (1x* + 2(xx)* + 3(xxx)*)* and (x+1+2+3)...

3 hours later…

7:24 PM

Why does Python 3 return (6.123233995736766e-17+1j) for (-1)**0.5)? Surely it's (0+1j)...

7:57 PM

Floating point precision ;p

@FryAmTheEggman Is there any way around that?

You could check if a number is less than 1e-12 or something and just set it to 0. But generally no.

Ahh okay then

8:16 PM

Has anyone done a complex plot challenge yet?

I couldn't find anything easily.

Any more info?

8:38 PM

@BetaDecay By the way cmath.sqrt(-1) seems to work.

8:49 PM

en.m.wikipedia.org/wiki/Color_wheel_graphs_of_complex_functions