« first day (5061 days earlier)      last day (77 days later) » 

00:09
mfw something suggests using poetry but my local installation of poetry is so fricked I can't use it :p
would that something perchance be Vyxal Bot
fuck. I have become the very thing I mock
@Ginger no
I should probably write a README for VB3 if you're intending to actually contribute to it
I do intend
it's just not what I'm working on rn
01:06
Hi I've found a new pair of tags to complain about: and
both are maybe used for "execute a specific sequence of steps", but often not even then, and that could apply to half of the challenges here honestly
yep :3
A lot of other tags would fit that prescription here
Honestly, unironic take, maybe it'd be better to have only tags for OWC and potentially language tags too
And things like source-layout, restricted-source too
tags like math, string, integers could go too
Have only tags that objectively and helpfully define attributes of the challenge
Rather than high level categories
To be fair, we have lower standards for tags here, considering that unlike most SE sites it's difficult bordering on impossible to objectively define attributes of a question
decision problem and Kolmogorov complexity are examples of objective attributes
Or rather challenge types
Yeah they're challenge types not attributes
OWC and challenge type tags are fine
01:20
Those high-level categories still serve a fair amount of purpose - [string] tells you that a challenge involves string manipulation, [primes] tells you that a challenge requires computing primes in some way - and when searching through challenges, which is the main purpose of tags, they do do their job - they help massively with looking for duplicates or related challenges. I do agree that there's a distinction between "objective attribute of the challenge" and "high-level category", but (cont)
those high-level categories are still useful
@lyxal The definition of meta-tag used here is "doesn't clearly describe the content of the question", and I'd argue most of the tags we have do.
It also contains:
> If the tag can't work as the only tag on a question, it's probably a meta-tag. Every tag you use should be able to work, more or less, as the only tag on a question.
OWC requirement aside, a challenge type tag could stand on its own
... you know that's not how cgcc works
Something like math I'd say couldn't
like with half of the things for SE in general
@emanresuA that's why I said OWC requirement aside
01:24
@lyxal sometimes there's nothing more specific to use
Of course non OWC tags can't be the only tag
@emanresuA which is handled by having an OWC
it's a nice fall back
wdym? I'm saying there are quite a few questions tagged with just , because there's no more specific subset of applicable to the challenge
So under the assumption that there doesn't need to be an OWC, but that posts can effectively have 0 tags, I think one could say math is a sort of meta tag
@emanresuA I'm saying drop the math altogether
Have code golf as the only tag
From the title and challenge description it should be obvious it's about math
That's not the purpose of tags lol
I believe it should be :p
01:27
tags are a convenience, not extra information
@thejonymyster Retina 0.8.2, 10 bytes
@UnrelatedString then there should be no problem with the algorithm tag or the simulation tag
math, in a similar way, could apply to over half the challenges on the site
20 mins ago, by emanresu A
both are maybe used for "execute a specific sequence of steps", but often not even then, and that could apply to half of the challenges here honestly
math is also used as a general subject
01:30
There's a spectrum of how much information a tag provides about a challenge, and at least those two are at the "almost nothing" end
@lyxal math is a bad example and imo it's kinda borderline but still way better than algorithm
Like I said there's room for challenge type tags
But I would be interested in exploring a purging of a larger set of tags than just simulation and algorithm
I think that's my ultimate point
@lyxal Aside from the occasional case of people putting on challenges without explicitly specifying a winning criterion in the body, tags don't add any extra information to the challenge. What they do do is a) aid searching b) on pages where only the title+tags are shown, e.g. the main page, provide some extra context, especially since our titles tend to lean more "interesting description" than "technical description"
@lyxal Agreed, but I'm not in favour of purging non-objective tags entirely
@emanresuA What's to say I wouldn't want to find a post that's about implementing a specified algorithm or simulating something?
Algorithm and simulation would help me search for those kinds of posts
There's definitely a niche for at least the first one, but it should probably be e.g. a separate tag [specific-algorithm] to avoid the confusion that's resulted in 440 questions tagged algorithm. I don't know about the second but there might be something there as well. But currently, those searches are next to useless because of the vagueness of the existing tags
01:36
Well that's what I'm saying about tags like math. Even though it's sometimes the only applicable tag, there's still a level of vagueness about it
Fair enough, and I agree that's one of the vaguer tags, but I still think it's nowhere near as bad as algorithm/simulation
 
2 hours later…
I went to the messages surrounding the one you replied to and:
Jul 7, 2023 at 14:20, by RubenVerg
til your name is Lyxal and not iyxal
I did not need to remember that was said
pffffffffffffffffffffffffffffffffffffffffffff
@DLosc but yeah, you can do this with empty groups if you leave one as a "flag" to "set" anywhere that would generate an equals sign then end the program with a disjunction of backreferences to each of those empty groups
 
3 hours later…
06:24
The whole reason I got stuck in this rabbit hole in the first place was because I thought Regenerate would be a fun language to do the Nice challenge in, until I realized there's no recursion or .NET-style balancing groups, which then just got me so stuck on wondering if it's possible at all to generate Dyck words... I'm pretty sure it's not possible to do with clean output, but
okay I've been TRYING to do it with extra junk for uhh
3 hours
and this is the closest I've gotten
gn y'all 🙃
nice, g'night :3
:3
for some reason it just. refuses to let the depth go above 1
and I have no idea if that's a flaw of the program or something inherent to Regenerate's execution strategy
OHHH maybe it's uhhh
yeah good night
06:40
...and then I tried constraining the length, and it just timed out...
oh that part's because of how stupid a hack I'm trying to use to do that lmao
yeah i'm actually eeping now :3
i have gazed into the abyss too long
third time's the charm :3
was nearly asleep until I woke up remembering that. I can just make it loop that many times instead of making this stupid counter to make extra loops fail LMAO
gnnnnnnnnnnnnnnnnnnn
nini
07:00
... fourth time's the charm? g'night :3
OHHHHHHHHHHHHHHHHHHHH I FINALLY GOT IT
Junk: ((${$2+#7-#6}!0)(X{1-$2}!)(($4!)(\)?{1-#3})(\({1-#6})!)){$~1*2+1}$9{$2}
Actual Dyck word: "$5"
The problem all along was literally just that I had assumed that negative repetitions were empty instead of failing 😭😭😭
ONE exclamation mark was all it took
nice :3 now actually go sleep once you've posted it
I don't even have anywhere but TNB to pist it :D
I am crashing RIGHT now
after getting out of bed literally three times already to try shit
:3
 
3 hours later…
09:53
Got to play around with liquid nitrogen
KGB has an event called LN2 where we make liquid nitrogen ice cream (using equipment from CMU's experimental physics lab class), but we also have hours and little supervision
We did have one prof there, who's also our organization's advisor
I did not take the same approach she recommended when it came to safety
To put it mildly
I did some extremely cool stuff, she was really not happy at first but quickly realized I was too much of a force of nature to do anything but watch with awe and horror
At one point I accidentally made a bomb
Or at least, a very very pressurized soda bottle
Apperently a soda can can hold 20 atmospheres of pressure
 
3 hours later…
12:49
@rydwolf ahahahahaha
I feel like that's just what everyone goes through knowing you :P
gotta love how the google youtube api requires an api key just to get the video title from a video id
noembed.com the real mvp
💀
yeah this is why everyone loves yt-dlp so much :P
scraping for the win
well see I can't exactly use yt-dlp
given that I need this on a frontend
and uh yt-dlp, from experience, is a backend thing :p
I just need the title
http://noembed.com/embed?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3Fv%3D4jXEuIHY9ic
beautiful
exactly what I need
12:55
sweet
13:23
@UnrelatedString ...somehow I already had a whole 96 tabs to close mostly just from this after I COMPLETELY nuked my tabs in this window (where by completely nuked I mean procrastinated on restoring because I have been planning to set some bookmarks in folders and then actually remove most if not all of my pinned tabs and try to keep myself to only 2 or 3 browser windows at a time so I'm less distractible)
huh
Anyways while it's fresh in my mind, I guess I may as well explain how the program works :P
0
A: Sandbox for Proposed Challenges

caird coinheringaahin gRemiND me WHat MY PassWord Is Unfortunately, I've forgotten what my password is. I wrote it down somewhere, but, because I know it is bad security to just have passwords written down, I instead wrote it down in hints. In these hints, I wrote down two things about the casings of each letter in my ...

((${$2+#7-#6}!0) [...] )*$9{$2} is basically a while loop, where $2 is a decimal number tracking the cumulative number of open parens minus close parens and $9 always fails because there is no ninth capture group, so $9($2) succeeds if and only if $2 is zero and the program just keeps trying different numbers of repetitions of the Kleene star but only outputting balanced results
(X{1-$2}!) is capture group 3, which is just a goofy little hack for testing if the depth is nonzero--X{1-$2} is X if the depth is 0, empty if the depth is 1, and failing if the depth is greater than 1, so the length of (X{1-$2}!) is 1 if the depth is 0 and 0 otherwise
(($4!) [...] ) is... I'm actually not entirely sure why I have to do it that way or how I even thought to, but it, uhh...
Oh yeah, it's an artifact from when I was making the depth start at 1 or trying to force it to terminate at -1
So un-nesting it just adds an extra junk opening paren to every output but doesn't change anything else fundamentally, and if the depth calculation was reordered to the end of the loop it would be fine
so (($4!) [...] ) is itself capture group 4, and ($4!) is capture group 5, which I treat as the actual final output after the loop is done so that there's a lag in building it--every character that gets added is in limbo in $4 before the next evaluation of groups 4 and 5 saves the previous $4 into $5
(\)?{1-#3}) is group 6, which optionally adds a closing paren if group 3 is empty (i.e. the depth is nonzero)
and (\({1-#6}) is group 7, which adds an opening paren if and only if group 6 is empty, using length/repetition instead of a normal | so group 2 has an updated count
and that about does it
 
2 hours later…
16:08
@UnrelatedString what this about?
16:27
@rydwolf nice! always wanted to do that
 
2 hours later…
18:26
@Simd Generating Dyck words in Regenerate, an esolang DLosc made that’s basically a regex flavor that outputs its potential full matches instead of matching an input
It’s more powerful than true regular expressions because it does have backreferences, but its one and only ability to actually store state without outputting that state is by using empty groups as Boolean flags that can’t be unset once set except by backtracking
I suspect it’s TC, but it’s still completely unusable for a lot of challenges without lenient I/O
For like 5 minutes I thought I had a concept for translating FRACTRAN into Regenerate last night but then it just completely left my mind
Like a lot of it should be very simple considering Regenerate’s arithmetic expressions do support modulo, but the thing that clicked was how to actually implement the main loop and I can’t even guess how I thought I was going to do that
…I guess just with a chain of match length conditionals the same way as I did with the Dyck words lol
Okay now to eat lunch before I start actually implementing this on my phone because I would SO do that if I don’t get up now
19:01
where are vyxal's flags documented? I can't seem to find anything
v2 or v3?
@UnrelatedString TC just means Top Cat to me (not really :) )
CMQ given a square matrix, is there a non brute force at to see if you can permute the rows and columns to make it symmetric?
19:39
v3? sorta self-documenting in the online interpreter
19:54
thanks
 
2 hours later…
21:38
0
A: Sandbox for Proposed Challenges

EonemaHow many 2x4s can you make from a log? code-golf geometry You have a pile of logs that you want to mill into two-by-four inch boards to build a new stand for your grilled-cheese sandwich business. You've counted and measured all the logs, but can you get all the lumber you need out of them? Chall...

 
2 hours later…
23:16
I'd like to eat butter like a butteronaire
23:57
> build a new stand for your grilled-cheese sandwich business
topical given what I'm listening to

« first day (5061 days earlier)      last day (77 days later) »