The Nineteenth Byte

12:18 AM

Never mind, beat it

12:37 AM

@primo There is this

Only matches global though

2 hours later…

PhiNotPi

2:20 AM

music I've been working on for human performance: soundcloud.com/phinotpi/quintemplate01f

this is "Section 1" of something that will be 3 or 4 sections.

2 hours later…

4:08 AM

@H.PWiz i think you got me

how long is the decoder, btw?

7 hours later…

10:57 AM

@primo My shortest ascii solution is 180 chars

11:24 AM

@flawr HTTrack is my go-to software.

3 hours later…

2:32 PM

@H.PWiz mine is 185, although i have an idea that might reduce it considerably

2:48 PM

Interestingly, I do almost exactly the same as my Haskell answer.

@flawr does it not save a copy every 60 seconds anymore?

that's what i'm going to try next ;)

@ETHproductions it does, but that just saves it on their server. I also want to be able to save my stuff locally in case they decide to discontinue their site.

@flawr HTTrack?

Or a Python script with requests?

@wizzwizz4 thanks for the suggestion, didn't see the notification for some reason:/

2:58 PM

If you're good at regular expressions, should be pretty easy to set it up with a depth of 2 and a filter just to get the full-screen versions but crawl the IDE versions and let it loose on your fiddle list.

That'll get stand-alone rips of all your fiddles, iirc.

@wizzwizz4 unfortunately I'm terrible at regex :D

@flawr It's ok, you can get help pretty easily.

I'll show you. Observe:

Can anyone help me with this RegEx? h*ppy I want it to match "happy", "hoppy", "hippy" etc. but it doesn't work.

@flawr There. Now all the RegEx people will flood in to fix the obvious error, and you can jump on them with a harder problem.

hehe:)

crowdsourcing

caird coinheringaahing

@wizzwizz4 replace the * with [...], where ... is every single unicode character /s

@H.PWiz i actually can't a bigint solution to work

3:06 PM

can we find a regex that matches all valid regexes?

@primo Interesting, too slow? I currently can't get a bigfloat solution to work, I wast to be able to do @evalpoly(a,<Array>...), but it doesn't seem to work like other varargs

for whatever reason, bigfloats keep popping up

i'm not using evalpoly o.O

I'm not either

maybe i should be

182 now, btw

Leaky Nun

21 hours ago, by Leaky Nun

CMC: We define a : nat -> nat by writing n = 2^v1 + 2^v2 + ... + 2^vk and defining a(n) = v1 | v2 | ... | vk where | is bitwise-or. Given n : nat, return n | a(n) | a(a(n)) | a(a(a(n))) | ...

3:15 PM

@cairdcoinheringaahing You mean [\x00-\U10FFFF]?

@wizzwizz4 I think you just want h.ppy

@ETHproductions I do, thanks.

@flawr .*?

Matches all valid regexes, and some things that aren't.

Oh thanks, I have become the fool once again :P

@primo I really can't get a bigfloat solution to work. I suspect I'm loosing accuracy at some step.

I have almost no idea what the larger regex question means though

3:22 PM

@H.PWiz you setprecision beforehand?

Yep

then i don't know

do bigints requre div, or can you just use / ?

I've been using div yes

maybe that's my problem >_<;;

BigInt(1)/BigInt(2) gives me 5.0e-01

El'endia Starman

3:29 PM

@ETHproductions Almost got me too. I had to read the messages above for the actual context.

@wizzwizz4 you could probably answer 90% of all regex questions on SO with this :D

3:50 PM

@primo My BigFloat one takes 192, the problem is that I take too many to compute the derivative

ngn

4:16 PM

@LeakyNun ngn/k, 17 bytes: {2/|/'2\x,&|2\x}/

if n represents a set X, the positions of the 1-bits of X are its elements, and a(n) is its union

or their union

n|a(n)|a(a(n))|... is X united with its elements, united with its elements' elements etc, which is the transitive closure

Leaky Nun

4:33 PM

right

@H.PWiz one thing that saved me quite a few bytes is the realization that x/x==1

evaluating the polynomial and its derivative "properly" isn't strictly necessary

Hmm.. I'll think about it. Currently, my integer one is at 178, only one more byte and I'll save a byte. (what I don't like about compressed source)

lirtosiast

@H.PWiz what challenge is this for?

@lirtosiast code-golf.io/scores/%CE%BB/julia

2 hours later…

6:26 PM

@H.PWiz i can deduce from that that your decompressor is 2 bytes shorter than mine ;)

but i do know what you mean. i recently reduced by ruby solution by 5, but was unable to get it aligned properly to save even one