« first day (5007 days earlier)      last day (132 days later) » 

00:45
@Ginger Here's something that is probably unrelated to the pronouns.page issue: If I click "load to my last message" it doesn't even add the slash to the old messages it loads
Also, for a workaround to the pronouns.page issue, you could presumably spread in he/she/they at least, like { he: 'he/him', she: 'she/her', they: 'they/them', ...pronounsPageData }
 
2 hours later…
02:29
There's something so beautiful about a true quine
Eval quines and their like are fine and all, but lack beauty
It's that aesthetic relation they have with having to actually duplicate/encode their own data
Whether that data's transparent in the source or not
And whether it's encoded simply or across something long and stilted, it always means there's pervasive structure through the entire program
03:17
@ATaco what's the difference between a true quine and an eval quine?
Eval quines have a debated validity as they execute over data, which is a convenient inverse of generating data from the currently executing code
ah ok
 
1 hour later…
04:31
based lecture notes
05:19
context please
I need a lecturer like this
It's just cardinality stuff
cool
 
3 hours later…
08:46
I have never seen such passion about a subject brought so well. Definitely give this video a watch
09:16
12
Q: What programming languages implement memory safety?

user5984One of the most known ones is Rust, but what others implement this?

You can count the wrong answers on one hand yet people managed to answer two of them
True lol
Regarding "modern" C++, I'll quote user on this one:
> Some people get a fancy new car every few years. Others keep fixing up and modifying their decades-old car until it's simultaneously a race car, monster truck, tank, and spaceship. In their hands, that is: every time they hand over the keys to a friend, the friend either can't start it or ends up igniting the rocket fuel and burning themselves alive.
That souds a lot like my bycycle
@mousetail wow that is shockingly incorrect, crazy what HNQ will do to a question
Mousetail, you have access to (not) 10k tools on PLDI right? Is the first wrong answer at the top of the "most commented" list?
10:04
what the hell
This is normally an amount exclusive to SE doing something really unpopular and featuring it
10:33
Contraversial option: Steve is right. You do normally care about aspects of a program beyond the inferface, like security, performance, etc.

The answer doesn't answer the question at all though and is excessively condescending, especially since it's just a matter of a different perspective
10:56
"Unsolved problem in physics: What is dark matter?"
Well, dark matter is dark matter. So I’ve solved the problem! Yahoo!
11:24
milk inside a bag of milk inside a bag of milk moment
11:50
so I've been thinking more about what I like about golf and I think the things I enjoy the most are trying to improve others' answers (and seeing mine get improved) and come up with novel approaches to problems; I really don't get the competitive part. SE does this fine, but I think the concept could be brought to an even more community oriented system
It's not so much competitive as a goal to optimise for
would anyone be interested in a site where the "goal" is not only to golf your own stuff, but also to improve others' solutions, and there is like one answer per approach per language that everyone works on?
att
att
weekgolf kinda had something like that
as far as I could find all sites either have the QA system, where the original author has to approve the changes, or competitive
att
att
in that after solutions were revealed you could submit based on what you saw from others
11:53
week.golf was mostly competitive though right? solutions were secret for the week and the goal was to make your own solution the shortest?
I want a focus on the community aspects
att
att
yes
so only kinda
yeah(:
@RubenVerg you could get points afterwards for improving though
after a challenge closes on YQ (name TBD), it will be a bit like that. All solutions are revealed and now you can try to combine the best aspects of all solutions to colaborativly try to get an even better one.
@lyxal based on how much you improved on the actual amount of bytes?
11:55
I think it was just "points for an improvement"
I don't think there was any consideration for improvement size
@mousetail hmm okay, I guess I just want that from the beginning, also already expressed my dislike for full program golfing which takes away the interesting (to me) part
att
att
idk how it was scored but i think the github is still around
I'm considering points for each byte/char/whatever taken out, plus some bonus if your solution hasn't been improved for some time
(except by yourself, of course)
A issue is collaborative solutions tend to be very low quality. Just look at this site. To get good golf answers you need to know a better solution exists and spend a lot of time alone trying to find it.
att
att
yeah i do think i learned a good amount from spending more time on code.golf recently
11:58
personally knowing that somebody will improve it makes me want to search for novel approaches
I do get the quality concern though and don't really have a solution
I'll start using code.golf when they get rid of the global leaderboards and add, like, golflangs
@emanresuA so never, basically :p
not sure if it's enough to mean the whole concept won't work though
att
att
@RubenVerg and similarly here I'm in the same boat but can't deny it's taught me things
I mean it won't work because nobody but me will use it, like anything else I've ever done, but that's beside the point(:
12:01
I still find it so, so stupid how their only argument against adding newer, golfier languages is a ranking system that we got rid of ten years ago
and it's not even a good argument
@emanresuA very glad someone else finds code.golf arbitrarily restrictive
how does the ranking work?
att
att
I'm hoping they add mathematica some time
probably will be miserable to golf with all string input though
Isn't mathematica non-free?
Licensable (TIO has a license)
att
att
12:02
they'd need a special license like tio
yeah
@emanresuA plus, now they've added languages like BQN and Uiua as experimental, making it even more stupid to not add golflangs
like you're 90% of the way there
It's kinda funny how much the code.golf and cgse community hate eachother despite consisting of 90% the same people
that said automated testing and permissive I/O do not mix well
I suppose my problem is that I golf for learning smart and concise approaches to doing things, not directly for shortest in bytes, which is more of a consequence
which is why I don't like full program golfing and also why I don't like competitive sites, which focus on the wrong (to me) aspects
I'd also want a discussion space where people can together come up with new approaches or discuss cross-language potential saves
@emanresuA imo if you just let people write how the interpreter is called from, say, bash in the header / footer, this solves itself.
12:06
only issue is people cheating by preprocessing data in the header
The whole point of automated testing is to ensure solutions are valid. If you let people control that, why have automated testing in the first place
In fact, here and code golf codidact are pretty much the only code golf sites without automated testing, and from my experience that lenience just leads to so much more creativity and clever golfing, especially since you can actually see what other people have done
It also leeds to an excessive amount of loophole abuse to the point it's no longer fun
define "excessive"
We have a whole list of things that are no longer considered fun
the main problem of loophole abuse anywhere is that it's very fun to learn about loophole abuse for the first time, which means that as long as there are new users, there are users for who loophole abuse is still fun
@emanresuA fair enough when you put it like that
Competing on this site is mostly a matter of combining a bunch of ancient meta posts in such a way to trivialize the problem
12:09
disagree
I have no idea where you're getting that idea from
modern answers really don't abuse anything (except mine, i'm a new user on a technicality), if anything i find they're quite dry and to the point
Based on just reading answers
att
att
I definitely make good use of permissive i/o
Find me one undeleted answer from this decade that "abuses loopholes"
12:12
I don't mean it like that. I think almost all answers play too fast and loose with IO standards
i disagree again
so you'd rather have a mandatory a,b=map(int,input().split()) at the start of every python answer
^ why i'm not on anarchy golf, the bash boilerplate is horrendous
@mousetail So what specific issues do you have with permissive I/O standards? What, if anything, is "unreasonable"?
take https://codegolf.stackexchange.com/questions/252082/reconstruct-matrix-from-its-diagonals as example, since it's at the top of hot and happens to have somewhat heavy use of flexible IO:

i see no issue with any IO requirements imposed by answers. they're all very reasonable ways of requiring IO to a matrix-related function
12:16
I think flexible IO is good because, again, your bytes focus on the actual algorithm and not on adapting to whatever weird system the author wants
It's weried being in this chat again, in the code.golf I'd be the one arguing for slighly more flexible IO.
I don't think it's fair to save bytes by changing your IO method
didn't SE use to be like this and then everyone went "this is awful we need more flexible IO"?
No, it was just unenforcable here
You are correct the standards used to be tighter, but it didn't work for a veriety of reasions
It was just an absolute mess because of how question writers had to specify things
afaict the main issue people have with old challenges here is the bad IO rules
12:19
@mousetail I think it is fair. i see no reason why one must be locked into an IO standard once they submit one answer with it. would you rather they submit another answer if they change IO formats?
att
att
fwiw I occasionally intentionally make answers longer than they could be with more lenient io
One aspect of this site is it allows literally every language to participate, no matter how esoteric. I think that's great. But that requires flexible IO that makes it less fair when using a normal-ish language
att
att
and definitely have gone too far (subjectively) with the flexibility before
all challenges here are secretly half popcons anyways, you don't win anything by writing the shortest answer, you do win rep by writing the shortest interesting answer
Yea, another aspect I really dislike about this site
There is no incentive to do a better job golfing since an abitrary answer will get the points
12:21
good to know people don't always think like me(:
that's fair
att
att
oh definitely
Yes, I do consider golfs by changing I/O methods to be somewhat "cheap", but having those options of flexible I/O available lets you actually focus on the problem at hand instead of spending seventy bytes parsing it from stdin
att
att
just look at my top answers
;)
@RubenVerg Yes, but that doesn't mean you shouldn't try anyway
and that's just an aspect of the SE model
@att (or anyone's really)
12:22
I don't enjoy wasting time parsing stdin either
yeah so that's why I want my points to be awarded based on improving stuff while keeping the same basic idea, you both incentive "outgolfing" but also cool solutions
@mousetail on the other hand consider that competition is per-language anyway (other why bother with anything but golflangs), and that the winner in any language is very likely to be the only participant
att
att
@emanresuA i think my top answer is particularly egregious
I feel here the sol that gets the most votes is rarely the most "cool" solution. Neither is it the best explained solution, or the shortest. Often it's a solution that doesn't even work. It's just arbitrary
@Themoonisacheese yeah sometimes in like python there's multiple solutions but mostly there's one approach and when improvements come from someone else it's encouraged to comment them
which does sometimes not give you enough recognition, I think, and it's part of the reason why I think scoring on improvements could be useful
12:24
@Themoonisacheese I think having only one solutions per language is an inenvitable downside of public answers
@mousetail it's very often the solution that got posted first tbh.
@mousetail that's obviously an issue and I agree
@Themoonisacheese Really no, unless it's the only one posted within the first dayish
att
att
like, RandomPoint@Sphere[] is really uninteresting but it's straightforward and has a pithy comment
reddit had (has? idk i havent been there in years) a settinng where the default sort for comments was random, that'd be great
12:26
I swear we had a feature request somewhere to make the default sort by active
On code.golf, it's incredibly satisfying to find a byte save and go from #11th place to #10th place. That's what makes golfing fun for me. On this site people mostly don't even bother posting if it's not better than the existing answer
@att tbh i'd upvote that. it's interesting specifically because (yet again) mathematica has a builtin
It feels ike joining an exclusive club
30
Q: Could PPCG benefit from a new sorting mode for answers?

Martin EnderThe default sorting for answers (by votes) seems somewhat inappropriate here on PPCG. In any case it tends to put early answers over late answers, because once a few are there, few people scroll to the bottom to see newer answers and even fewer people sort by activity. If you consider that bette...

have you considered that maybe public, flexible golfing like here and codidact are just not your cup of tea?
12:27
^
Haven't I said literally that
att
att
i enjoy both
it's a tradeoff
I'm not against this site existing or trying to change the rules here
@mousetail that's what I really don't understand about competitive golf I guess
att
att
also please join my python number spiral club it's lonely
12:28
Just trying to explain the value of competetive golf
I've been active on this site for a long time and had a very good time
i know, but then it really feels like you're just ranting for ???
Fair, I guess I have been ranting a bit. Sorry
we all have our days
remember to drink water :)
true!
true...!
att
att
12:37
actually that's an interesting thing it feels bad when I don't feel like my solution is particularly inventive
anyone who thinks my idea could work and would actually enjoy contributing to community solutions?
att
att
i would
yeah that sounds good
13:35
Have had a factoring challenge with restricted byte length?
14:13
so I'm trying to fix sechat and, uh. requests to /ws-auth are 500ing
@RubenVerg I like it but I don't like this one-solution-per-language thing
Part of the fun of this site is multiple people will approach the same problem with different ideas in the same language, and some will be shorter but that doesn't mean the longer idea shouldn't be there
no no not one solution per language
one solution per approach per language
It's hard to say if one solution uses the same approach as another
is it really?
Things can be ordered differently, or have certain sections of code that do the same thing written differently, etc
14:20
approach is a very high level thing
Yeah
Also two solutions may use the same "approach" but be written totally differently and use different golfing tactics
i think in that case i'd prefer to only have the shortest one but this is still debatable
@Ginger works fine for my port
satisfying message id
I’m just thinking, what if I make a challenge that merges cops and robbers with KOTH?
@RubenVerg yeah I'm pretty sure it's my fault :p
I may spend a bit making sechat slightly less fucky
14:45
why on earth does chat need the sechatusr cookie and the fkey to do anything
0
A: Sandbox for Proposed Challenges

aeh5040Name that logic gate (Is this a duplicate? Seems very basic...) Given the truth table of a non-trivial symmetric 2-input logic gate, output its name. The 6 possible input/output pairs are: 0001 -> AND 1110 -> NAND 0111 -> OR 1000 -> NOR 0110 -> XOR 1001 -> XNOR (where the 4 symbols represent t...

15:11
@Ginger because chat is a giant amalgation of 10 years of tech debt?
15:27
@Themoonisacheese yep
15:48
CMQ I want to minimise (c*r)^r over positive integers r and 0 < c < 1. Can I always just minimise it over the reals and round to the nearest integer?
 
2 hours later…
att
att
17:37
no, take c=0.249
min over real r is at r~=1.477 but c=0.249>(2c)^2=0.248004
17:59
@att that's great! Thank you
How did you find it?
 
1 hour later…
att
att
19:18
it's fairly straightforward to get (cr)^r minimal at r if c*r=1/e
at r=1.5, c=1/(1.5 e)~=0.245 and (1c)^1>(2c)^2 so I rounded c up to 0.25 (making r<1.5)
but 0.25=(0.5)^2 so anything between 1/(1.5e) and 0.25 should do
 
1 hour later…
20:22
( ͡° ͜ʖ ͡°)
@att Thanks! That is cool
Are there any other ranges that work?
att
att
21:12
I'd guess possibly all ranges (1/((n+1/2)e), n^n/(n+1)^(n+1)) for integer n>0
mathematica says yes all such ranges
21:59
Congrats to our very own @RGS for setting a programming-related Guinness World Record!
11
nice!
22:37
Next up: world's smallest audience for a programming lesson
ChatGPT talks to itself for an hour
22:51
0
A: "Hello, World!"

noodle personTinyAPL 0.9.0, 21 bytes (17 chars) ⎕←"Hello, World!"

23:05
@TheEmptyStringPhotographer you wouldn't be the first to propose such a challenge :p
23:20
@RubenVerg I think it's an interesting idea--vaguely like a code-golf version of Rosetta Code. I might participate.
@Themoonisacheese loud correct buzzer
23:38
@UnrelatedString I don't think we discussed clicks in my Intro to Linguistics class and now I'm feeling vaguely cheated :P
is there an alternative to sys.argv[1:] in python?
Not exactly, that I know of, but what's your use case?
23:54
code golfing :P

« first day (5007 days earlier)      last day (132 days later) »