feersum

That's actually pretty weird that they have a database of which IP addresses visited every link.

I see

I wonder how I got the 'link visited by 300 IP addresses' and the 'link visited by 1000 IP addresses' badges in the same day from some old link.

In the large table limit you can just do rejection sampling (completely start over if you get an overlap) and get a totally uniform distribution.

Apparently it's called the "rectangle fit problem".

In the cramped table limit, the problem of just determining whether the cards can fit at all is NP-hard.

Can they be rotated 90 degrees?

So what is your desired distribution?

@cairdcoinheringaahing I predict arbitrarily large scores will be found.

Hmm, I suppose they don't. Maybe they do now though...

Why watch it? It doesn't seem like a very entertaining form of entertainment.

[ 2>/dev/null; echo 'you can put a bash command here' # ]: # (generates nothing in markdown)

@El'endiaStarman ^

[ 3 < 2 ]: # (as a start, this line generates no output in either language)

[ // ]: # (why is this is a line comment in markdown)

You don't have to have a hashbang, as long as bash is your default shell.

I run on a populist platform of banning golfing languages, and win on the support of the 200-rep demographic.

That challenge certainly lacks a proper spec, but it's not impossible to write one.

@A_ I suppose that's the kind of anarchy you get when moderators resign en masse.

@ngn Yeah I was focused on the SSE2.

@Anush You're discussing multiversioning on x86? Why? Sholudn't you be using a 64-bit built if you care so much about performance?

Perhaps, we always posted this in chat but no one remembered to actually write it in a challenge.

@HyperNeutrino Don't we have a standard solution for this: require the answerer to run the program on all test cases before posting

Well, we know what to do next time we have a feature request.

Anyway you could have some kind of conditional stack opreation or soemthing if it's really needed.

@JoKing Eh? I know you're not new here...

No jump. It's like, barely even esoteric if you have jump.

I wonder if that would be enough for all programs.

You could get golden ratio by having it be root of a polynomial.

How do you specify the number?

Or perhaps there is a more fun approach, to hard-code the gcd(1,2,3,...12) root of pi.

@Anush Maybe not, in which case you'd spend a few more bytes to hard-code pi.

@Adám Nope.

It has a big decimal library. Probably like 60 bytes.

@Fatalize Yep.

At least in Python 3.8, you can assign variables inside expressions, so newlines will probably be entirely obsolote.

Well, a professional programmer needs to be able to handle writing under multiple style guides. Here we just have a slightly different one which recommends one letter variable names and 0-space indentation :)

No, how so?

It sets a bad example and requires the participants to manually examine teh numbers.

It's still bad.

It requires potentially arbitrarily large precision.

"First 20 digits should be correct" is a bad way to specify it.

@Bubbler This.

Is it actually interesting for you to get an answer like bigdecimal.power() or something?

You're likely to get only builtins answers.

We should set something on fire.

@cairdcoinheringaahing I understand what golfing languages are like and still find them uninteresting.

Blame caching?

Works fine for me.