« first day (3788 days earlier)      last day (137 days later) » 

7:28 AM
@lyxal * is outputted in my common sense
 
@Wasif why would "*" be outputted?
 
@lyxal should be nothing right
 
@Underslash because i and x both increment at each iteration
 
question like these get asked a ton in programming exams
 
or do they?
 
7:37 AM
@Wasif read the line with the for loop again
 
i don't know java may be i am missing something
err no it is missing braces
it errors badly
i am so used to python that i didn't notice missing braces
 
single line for loops exist in Java
you can omit braces
 
like with the if statement
 
then what's the error?
 
the for loop has a ; following it
which stops it
so none of what follows is in the for loop, hence why x would only be = to 1 and * would never be printed
 
7:40 AM
sighs
 
single line brnches/loops take a single statement followed by a semicolon
so basically the for loop runs an empty statement 4 times
 
for loops in java and c are cool
you can do things like: for(;;); and its valid syntax
 
need to learn java
 
do you know c/c++?
 
@Underslash not that much
 
7:44 AM
well, in any case, its really not hard to learn
its just really OOR
 
BTW is someone intersted in golfing with Notepad++?
when you golf with Vim
 
@Wasif @DLosc has many answers in Notepad++
 
wdym golfing in Notepad++?
 
i will really appreciate if someone can prove Notepad++ turing complete (/// was proven turing complete with just only a sed like replacement syntax)
 
is that a language?
 
7:46 AM
@Underslash yeah
 
oh
cool
 
it has regex too
 
/// is turing complete because it can do multiple replacements
regex all by itself afaikisn't turing complete
 
@Razetime can't Notepad++ do it too?
 
maybe with notepad's movement it's possible
 
7:48 AM
am I getting baited
 
@Wasif try making a simple interpreter
 
do you mean the actual Notepad++
 
yes the editor notepad++
 
oh ok
how would it be turing complete?
 
you can use any editor as a language
 
7:49 AM
but they dont run code, no?
 
using keystrokes and motions to get a result can be considered programming
 
@Underslash no but
err how to explain it
 
it has storage across multiple buffers, and ability to access aribtrarily deep data using a find function
+ regex
 
and many other cool tools
 
true
how would you output something though?
ex, Hello, World!
 
7:51 AM
even plain notepad can do a lot of amazing things like this codegolf.stackexchange.com/questions/97305/print-some-json/…
 
in vim, the initial buffer is considered as stdin and stdout
you simply modify the buffer
 
and after all modification the remnant buffer is output
 
fair enough
 
but Vim is the most powerful
a brainfuck interpreter was written using it
 
most good modal editors can do it
 
8:01 AM
Raze and Underslash were both right
@Razetime and it's funny you say that... I took that from a past exam from my uni
 
8:18 AM
0
A: Sandbox for Proposed Challenges

pxegerGeneralised multi-dimensional chess knight's moves code-golf combinatorics chess array-manipulation Multi-dimensional chess is an extension of normal chess that is played on an 8x8x8x8... "board". In normal 2D chess, a knight's move is a movement by a vector of \$ \begin{bmatrix} \pm 2 \\ \pm 1 \...

 
that appeared in chat faster than it did for me on the sandbox, and I posted it!
 
@pxeger Quick thing, specify that n >= 2 otherwise you could get problems
 
I have
 
wow
i am blind
 
> You should assume the input vector will always be at least 2-dimensional
 
8:23 AM
yeah
i looked over that twice
 
edited to make more obvious anyway
 
appreciated lol :P
other than that, I like the challenge
 
Hi @ngn
 
ngn
hi @Anush
 
8:44 AM
0
Q: Count all my sheep

RazetimeSandbox Given a boolean matrix representing my grass field, sheep length \$n\$ and wool thickness \$k\$, you will have to count my sheep. A sheep is a single independent unbroken line of sheep length \$n\$ and thickness \$k\$. #|##|# | # #| | #|# are all valid sheep for \$n=2,k=1.\$ ##|###|## ...

 
 
1 hour later…
10:12 AM
@ngn was there a * missing from the second line in your double to long long code?
 
ngn
@Anush oops. yes, sorry
 
@ngn no problem. Thanks for the help!
 
0
A: Sandbox for Proposed Challenges

LeopardL GDI am not a dad! Task Given a non-empty string. Check if it starts with "I am"/"I'm" (case-insensitive) and if it is, output the dad joke. Otherwise, terminate. There are some exceptions which you can see in the Examples section. P.S. There must be nothing in the STDERR under normal circumstances....

 
10:35 AM
@RedwolfPrograms you say that you pad all "odd"-length programs by suffixing a _ but wouldn't that produce ambiguous code? whereas if you prefix the _ then that would be unambiguously parsed as the identity monad applied to the rest of the program
also the docs are very sparse I mean you have four different flavours of "base conversion" with no explanation
 
 
1 hour later…
11:40 AM
if you have a million integers, what is the fastest way to remove the duplicates?
 
What's the range?
 
I would use an O(n log n) sort algorithm to sort the intgers and then check each adjacent element for duplicates
 
you can do this in linear time
just use a boolean array, or if your range of inputs is too large, a set
 
@hyper-neutrino 32 bit ints
 
too large for an array then. just use a set
 
11:49 AM
@hyper-neutrino What is "a set"?
 
but it would not preserve the order, then you might want to have a hashtable, put all elements of the array in the hash table, then iterate over the array, if element is in hash table then add the element to the result and delete the entry from the hash table
 
@Wasif I think you can do a little better than sorting and then copying to a new array avoiding duplicates
I can't understand this at all stackoverflow.com/a/1533667/2287805
 
@Wasif the set itself doesn't preserve order but you can use a vector and a set and add the elements to both
 
can anyone understand it?
 
@Adám uh, is set not a standard definition in CS? in that case i mean a hashset
 
11:51 AM
@hyper-neutrino What do I know; I'm not educated in CS, but in lower-level languages (e.g. C) you'd have to implement such yourself.
 
for maximum speed I am not sure a hash table would be fastest.. ?
 
hash tables are O(1)
 
@Adám the problem is that hash tables have poor memory locality
 
@Adám ah true. i was thinking C++ where you can just include <set> lol.
@Anush maybe not but it's amortized O(1)
so better than a sort at least
 
@hyper-neutrino Sure, and in APL, I'd just use and rely on my clever colleagues to choose the best method for the given argument type.
 
11:52 AM
@hyper-neutrino only maybe.
 
ngn
@Anush are you asking in theory (like big O) or in practice?
 
@ngn 100% in practice
 
i find it hard to believe you'll manage to get adversarial enough data to make a hash table significantly worse than constant
@Adám fair enough lol
 
ngn
sorting is unreasonably fast in practice
 
@ngn true.. But there is this nice idea where you do merge sort but don't include any duplicates in the recursive steps. That might be a lot faster
and also stackoverflow.com/a/1533667/2287805 which I just don't understand at all
 
ngn
11:55 AM
it may depend on how many duplicates you expect
 
let's say 90% are duplicates
 
i can understand this easily: stackoverflow.com/questions/3350641/…
another SO post
 
@ngn hashsets can also be pretty damn fast
in Dyalog APL the builtin for unique is ~1.5x faster than just sorting
 
@dzaima is that just because the builtin is written in C?
 
the sorting is written in C too
 
12:05 PM
then what is it doing differently I wonder?
 
@Anush not using sorting, presumably
 
and instead?
 
a hashset or something
 
ngn
@dzaima are you sure?
 
@ngn does k have hash tables/sets?
 
12:06 PM
only person who could probably say for sure would be Marshall
 
@ngn i mean, i haven't looked at the source code (because i don't have access to it of course), but there aren't many alternatives
 
@dzaima there is the clever mergesort but I don't suppose it is doing that
 
ngn
@Anush shakti uses hashing (i don't know much about how it works). ngn/k doesn't.
 
@ngn k
 
there are lots of different ks
 
12:08 PM
how many English words would you guess have a silent k?
I can think of 14
including one with two silent ks!
 
ngn
is the "k" in "ck" silent? :)
 
:) Lets say no
it's the c that is silent
 
which is silent in the word scent, the s or the c?
 
@rak1507 in reality the "sc" makes an s sound
so we can think of sc as a "letter" in this context
 
Which is silent in frick? The c or the k?
 
12:13 PM
@lyxal see above. Really it's the "ck" that makes the k sound
 
ngn
@Anush i'm sceptical about that claim :)
 
I was just having fun
@ngn :)
english spelling isn't really phonetic in any reasonable sense
 
english spelling isn't really phonetic in any reasonable sense
 
If "k" in "ck" is silent, then I know of an English word with 4 silent "k"s.
 
12:16 PM
Yes.
 
english is weird :D
 
@Anush I suppose either could be (but not both). Cf. "cent" and "sent".
 
According to caird and others, yx makes an ick sound
 
@lyxal which is true
onyx, asphyxiate, etc
 
12:23 PM
Uh, isn't it "icks"?
 
yeah
 
Huh, TIL that "Featherstonhaugh" is pronounced "FAN-shaw". That's a lot of silent letters.
 
what now lol
 
Is it just me or does Folkingham sound obscene?
 
if you prnounce it wrong it's fine
 
12:35 PM
Sure, but if you say it right…
 
1:33 PM
this is what the current draft for the first-time asker popup dialog would look like, on my screen (1080P approx. 24 inches)
this is a bit too large IMO, since my monitor is reasonably large and this will probably overflow on laptops and other smaller screens - I'd like some feedback on how we can shorten this
heh, this could go on the main site because it's asking for golfing :D
meta post here (also pinned on starboard for now)
 
2:18 PM
@Adám not to me
 
2:54 PM
@Bubbler I can't see any way of solving this in strictly less than O(n²) time though, because it takes that long to generate the nᵗʰ row of Pascal's triangle
 
there's got to be some reduceable mathematical relation that lets you do it in (i suspect) linear-log time, or possibly even linear time
step-by-step can't be less than N^2 but supposedly that's not intended
 
3:11 PM
 
ooh, this is cool.
it looks clear (to me) so +1
 
ngn
@Neil it's O(n) if computing a binomial coefficient is a primitive operation
 
@ngn you might needa CPU with 100000 bit numbers though
 
@pxeger would a multidimensional array with the correct spaces highlighted be a possible output method?
basically put a 1 at all the possible move locations
 
no
That's quite a different challenge I think
 
ngn
3:23 PM
@Neil this problem looks suspiciously similar to bubbler's recent "ijk" challenge (product of quaternions)
@dzaima the "numbers" there are just {R,G,B}
 
@ngn right, of course the idea is to compute the binomial mod 3, all i'm saying is binomial coefficient being a primitive operation doesn't really help
 
3:38 PM
Morning!
@Neil That's a good idea, I never thought of padding the start. I made sure _ was the least likely to be used at the end of an even length program, so it shouldn't ever be too much of an issue I guess. The docs aren't very good because there's something like 100 things overloaded onto those operators, so it'd take ages to write even a few sentences for each of them.
Basically I'm just lazy :p
If the collatz conjecture ever gets proven, can we undelete this? :p
 
4:02 PM
@Neil 05AB1E vibes lol
 
It's totally not how synesthesia works but it's hilarious
 
I'm aware lol
 
4:18 PM
@pxeger Is there an English version?
 
?
 
The auto translated subtitles were terrible
 
...it is in English??
 
I got Korean subtitles initially
 
hmm yeah the subtitles are in Korean
I have no idea why
 
4:22 PM
Oh I didn't realize the audio's English
 
I think the subtitles are trying to interpret the English audio as Korean (so they're meaningless), and then they're translated to English (also meaningless)
 
I guess that makes sense, but why would they try to interpret it as Korean? There's nothing there to suggest that
 
it just mis-detected the language I guess
it's just a dumb AI
 
Bad bot :/
 
@user +1 for sponsorblock btw
 
4:29 PM
:)
 
ngn
4:54 PM
@dzaima why not? it reduces the complexity from the O(n^2) Neil mentioned to O(n)
 
@ngn You can reduce big O complexity of many things if n-bit arithmetic operations are O(1), but it's not useful to say that, say, sorting is O(n)
 
ngn
@dzaima well, i think for this problem there will be a considerable difference between O(n) and O(n^2)
 
but thinking a 100000-bit binomial coefficient result takes O(1) time to compute just because it's a "primitive operation" isn't gonna help practically getting to O(n)
 
ngn
@dzaima we only need to compute it mod3, as you said
 
@ngn but then you need to implement an optimized binomial coefficient mod 3 (unless you mean that if you had a primitive operation for specifically that)
 
ngn
5:06 PM
 
sierpinski, pretty cool
 
i mean, i know that there are good ways to compute binomial coefficient mod 3, but all i'm saying is that "if computing a binomial coefficient is a primitive operation" is a completely useless statement
 
I'm probably missing something, but I don't actually see a pattern there that helps with this
 
ngn
@hyper-neutrino first of all, did you figure out what the operation is if R=0,B=1,G=2?
 
i did not, unfortunately
 
ngn
5:11 PM
@dzaima when we talk about big O, we implicitly agree on what the primitive operations are
@hyper-neutrino it's negated addition mod3
 
ah i was trying stuff that was too complicated then
 
ngn
so, it's like an inverted pascal's triange in which every other row is negated
 
that makes sense
 
@ngn but I definitely don't agree that computing an n-bit number is a "primitive operation"
 
ngn
@hyper-neutrino in other words, the result is (-1)^n times some linear combination of the input with the binomial coefficients as weights
 
5:14 PM
oh, and each row of the %3 pascal triangle you showed is the coefficients?
 
ngn
@hyper-neutrino yes
(if i'm not mistaken)
 
yeah that sounds right
 
Oh cool, a dotcomma answer by a new user
 
5:57 PM
This is twelve years old and just as relevant today lol
 
6:19 PM
I would not be surprised by that
 
 
1 hour later…
7:26 PM
Anyone want to nominate ATS LotM? I'd do it myself, but I'd have to learn it first :/
 
Then it's probably not a good choice.
 
7:39 PM
Considering making an improved Risky, by changing up how the arity rules work
It'll be functional
 
Functional, as in working? Or actually functional?
 
Functional as in it will have functions
 
That's a new definition of functional :P
@Ausername What do you mean?
 
Functional as in FP
The existing definition :p
 
Oh ok
@RedwolfPrograms Having functions != FP (looking at you, Java)
 
7:44 PM
yeah lol
 
You know what I mean, idk :p
 
8:01 PM
Risky actually working? Surely not....
 
@RedwolfPrograms well, 0 does equal 0 . . .
 
8:25 PM
@Underslash Are you sure it equals zero? Maybe it equals infinity instead
:P
 
8:36 PM
@hyper-neutrino How often do mod flags occur in chat?
I'm not sure if it'd be worth adding support for detecting them in my chatbot library
 
27% of the 97 top flagged messages in order of recency of the message (not the flag) are mod flags
so they occur a decent amount but they are quite uncommon; i'd be surprised to get 2 in a day, maybe once every like 2-4 days or so
 
Actually now that I think about it, I doubt there's much that detecting flags of any sort will be useful for. I might not include that in the library at all.
 
9:09 PM
@RedwolfPrograms Hi, Codidact rep here. I have some questions about rep and sunk costs and stuff. I totally get not wanting to start over somewhere else; I had a lot of rep and posts on Writing.SE when we moved over to Codidact. We imported a bunch of content, which we later realized was a mistake -- Google penalizes that heavily, so on top of the usual struggles of a new site getting exposure, we had that. [cont]
On Codidact, rep now doesn't lead to privileges like it does on SE. (It used to, but we've since reworked the abilities system. You get abilities by doing things successfully.) So, now, rep is very much just a number. A question for you and others who talked about the rep factor earlier: if you could bring over your SE rep number, without any other effects, would that be appealing?
What else would you want to see to feel more comfortable and less "starting over" there? You are, of course, welcome to re-post favorite challenges there to see what new answers you get, if you like. I'm asking more about the user/community aspects. Getting over the sunk-costs feeling is hard; what would help?
 
@MonicaCellio Personally, I don't really know. I'd like to say no, in that rep doesn't mean anything and is just a number, but I do like that number being larger rather than smaller. What I can say matters to me is the trust I believe I've earned here, reflected in privileges. I enjoy posting challenges and answers, but I can do that anywhere. Here, I've invested time into improving the site, and I've been rewarded for that with abilities such as voting to close, editing, voting to delete etc. cont.
I actually don't think it'd be fair for users to be able to "transfer" those powers across, but I'd have to put in just as much effort on Codidact to be able to reach the same position, and I don't have motivation to do that when SE is, more or less, a perfectly fine place for me. I still believe it needs a lot of work - but that's msotly the company, not this community which is still going strong and relatively independent from SE
 
@MonicaCellio I think one of the reservations I have about fully moving to Codidact is the time and energy I've committed here since picking up my diamond (and I've talked about this before but I really hope deciding to help keep this place alive hasn't come across as hurtful or insulting to you; I truly despise what SE did to you and I hope becoming a moderator hasn't sent the message that I sympathize with the corporation).
I am definitely willing to keep posting on Codidact; so far I've only answered a couple of times and whenever people like rak1507 bring up a Codidact question in TNB I'm always willing to go over and take a look and answer it. I feel like I'd only considered leaving SE entirely and solely being on Codidact if I feel like there is the same sense of community there, because I don't want to lose that over here for personal reasons.
I understand building a community from scratch is difficult so I'd be willing to devote some time to Codidact as a potential community member without sacrificing my position here to help build up that critical mass. But until I leave my position as moderator (for whatever reason) or I believe this community is too dead and I give up on it, I don't think I'll fully move to primarily be on Codidact. I'd be willing to contribute over there a bit though, but I can't really guarantee much.
I don't really want to just gain full privileges on Codidact just because of my SE permissions; I feel like that would be a bit unfair (like caird said, without putting any effort into Codidact, it doesn't feel fair to just have those powers that others have to work hard for), but starting over from the beginning is uncomfortable for me.
 
I think just copying reputation over wouldn't be a good solution, because (in my opinion) it does two things: showing the community that you've dedicated substantial amounts of time to it, and giving you the ability to help the community in more ways. Just copying the number over feels kind of fake, because I haven't put the work in there.
 
Anyway, I've gotta head out for a bit, sorry - I'd love to discuss this more later tonight when I might have some more time though.
 
Yeah, I can't speak for others, but my main argument for staying here is that the negative things that SE have done have not been impactful enough on me to significantly make me want to leave - rather, I want to try to encourage change. I know you've struggled with the same position before you left @MonicaCellio, and I absolutely do not want to trivialise the way SE treated you. That incident, combined with the way SE treated us during the leadup to our last election was the closest I came to leaving.
 
9:26 PM
wait we have elections?
 
Yeah, we've had 3. 2016, 2018 and 2020
 
@hyper-neutrino No worries; we're all caught in a bind, caring deeply about our communities despite the actions of the landlord. I don't fault anyone for gritting one's teeth and staying for the community. I'm just asking how another landlord could make the community feel more welcome. :-)
 
what do they elect?
 
9:27 PM
@Underslash The site moderators - the users with diamonds by their names
 
I appreciate that you participate on Codidact as well as here. It doesn't have to be either/or!
 
who was the mod that got banned?
(or deleted their account)
 
@MonicaCellio I'm looking at the Code Golf Codidact Meta rn - do you know if they've had a discussion about challenge vs answer reputation?
 
Not banned, deleted their account
 
9:28 PM
ah
 
@hyper-neutrino makes sense, and I wasn't suggesting you abandon your position of responsibility here. I'm trying to get a sense of what would make people comfortable there so we can all work together.
 
Maybe showing the user's SE reputation as a separate number (if they moved from there) could be an option
 
The +5 vs +10 rep makes sense for answer-forward sites, but +10 vs +10 (or even +10 vs +5) makes more sense for the kind of site that Code Golf orients itself towards, and Codidact has definitely said they're open to making changes specific to certain sites
 
@RedwolfPrograms That would indicate that they've put a lot of work into a similar community, without the problems that just copying it over could introduce
 
9:31 PM
@hyper-neutrino Codidact has abilities that are earned through activity. A community can customize the thresholds (make it easier or harder to earn an ability), and moderators can grant specific abilities to specific users. That last can cause unrest if done poorly, but if the community wanted to set some "immigration criteria", some sort of "if you've done X on SE you can do Y on CD", that would be possible. Up to the community; we don't dictate.
Wow I'm being laggy in responding, sorry. :-) Still have stuff to read.
 
If linking your account allowed users to show their CGCC rep as well, that'd be a good way to implement it
 
so is codidact basically a separate site made to be similar to stackexchange in concept, but different in how the community is built? because it seems like theres sentiment that the moderation or the community on this site can be overbearing, but I wouldn't know the entireity of that
 
@MonicaCellio No worries, we're quick at typing and you are few and we are many :)
 
@Underslash Have you read about the things that occured in 2019?
 
nope
 
9:32 PM
@RedwolfPrograms something I was wondering about (and this is just me talking; I haven't discussed this with anybody there) is if we could show two numbers, the "native" CD one and the associated SE one, as a way of saying "yo, super-experienced user from over there" without affecting the Codidact number directly.
 
@Underslash tldr se bad some people want an alternative site that isn't se, codidact is one of those
 
@MonicaCellio That'd be an awesome way to implement it, in my opinion
 
what is making it bad?
 
@hyper-neutrino happy to, either here or on the Codidact server -- just ping me so I notice, please. :-)
 
@Underslash Monica will probably write a better response to this, as a founding member, but: in 2019, SE fired Monica from her position as a moderator on 6 SE sites. There was a lot of backlash, which eventually resulted in Monica and a large number of others users wanting to find an alternative Q&A format that was run in a more fair way than SE (a for-profit company). They (a group of specific users) then built Codidact, an open source Q&A format network of sites
I tried to frame that in the most neutral way possible - there was a lot of bad blood involved with the various things that took place in late 2019, and I'm not trying to pass judgement here
 
9:35 PM
@cairdcoinheringaahing understood. I'm not trying to get people to leave, just asking how we can make things more comfortable since it looks like people are frustrated to varying degrees. I appreciate the people from here who are also participating on Codidact! We're small but I hope growing. Starting new communities is hard -- getting traction is hard.
 
ah
that sounds like a hassle
 
I'd recommend reading some posts about it on MM to get a better idea of what happened, it definitely had a massive impact on the site (including CGCC)
 
@cairdcoinheringaahing I don't think so. We've mentioned that varying rep grants for different post types is possible, but I don't think anybody's run with that idea yet. (Similarly, we could make sandbox vs challenge posts behave differently if the community wanted.)
@RedwolfPrograms heh, I had not yet seen this when I suggested basically the same thing.
@cairdcoinheringaahing it's just a configuration setting; that's easy to change.
 
@MonicaCellio I think I'll draw up a meta post then - as someone who posts a substantial number of challenges, I've always disliked how long it took SE to make the change, and even then I think it should've been specific to a couple of sites, not network wide
 
9:40 PM
@Underslash The Codidact Vision (blog post).
 
I'll go take a look at Codidact later, might try answering some questions there.
2
 
@MonicaCellio While you're here, what are your thoughts on the non-compete clause of the new Community Ads? It seems to be specifically targeting Codidact. There are a couple of comments from members of the team (Mith, Art) but I'd be interested in knowing what you make of it
 
@MonicaCellio is there a way to see questions by a user, not just all questions + answers?
 
@cairdcoinheringaahing sounds good!
@rak1507 alas, our profile filtering and search both need work -- we have issues open, but no, you can see all the posts from a user but can't filter it down by type or category yet. Working on it.
 
@MonicaCellio If changed, would it be retroactive or not?
 
9:46 PM
@MonicaCellio no worries
 
@cairdcoinheringaahing the original version of that announcement seemed like it was targeting Codidact and TopAnswers. The newer version restricts it to tech sites, if I understand correctly, so, for example Mi Yodeya (SE) and Judaism Codidact can continue to work together and the former can have an ad for the latter. The revised language sounds like they care about the SO audience but not cooking or religion or physics (or code golf?).
My reaction to the first wording was basically "wow, the 800-pound gorilla is afraid of little us?". We're not trying to displace SO. We're building something new, taking advantage of lessons we learned here. SO is still SO; not trying to touch that. But they seemed to be afraid of us anyway, which was puzzling.
 
@MonicaCellio Well, it's by no means the first time that SE has only cared about the SO audience and nothing else
 
@cairdcoinheringaahing it would be retroactive, yes. I don't think we could make it not be.
 
@MonicaCellio I find it quite telling tbh. That was my reaction as well, and I definitely got the impression that SE understands just how alienating it is being to non-SO sites, and that Codidact genuinely poses a threat to its smaller sites
 
@cairdcoinheringaahing right, they really don't seem to care about the other communities. The rest are dust particles in the asteroid field around the SO sun.
@cairdcoinheringaahing well, now apparently the smaller communities can run ads for Codidact communities that aren't Software Development. And maybe they don't like our brand new Linux community either; not sure where their boundaries are. But most of our others should be fine to advertise, if SE communities want to run those ads.
What SE seems to have forgotten (originally and even in the revised draft) is that community ads are supposed to be chosen by the communities, so if a community doesn't want an ad for something they won't vote it up, and if they do then SE should leave them alone. Last year employees with 101 rep (i.e. could only downvote because they were employees) were downvoting ads. Bad form -- they should let the communities decide.
 
9:54 PM
@MonicaCellio Well, I expect a Codidact ad to be run here once they set them up for us :)
 
@cairdcoinheringaahing I've been hoping. :-)
 
@MonicaCellio That seems very much not ok, and border-line (minor) abuse of power
 
Also, if y'all are using our sandbox to support challenges on both Codidact and SE, that's cool -- and something you might want to make an ad for.
 
@cairdcoinheringaahing In this case, at least, it's a good thing that they're leaving Code Golf Codidact and the like alone
 
@MonicaCellio I doubt you can speak to this, as I don't think you're especially active on Code Golf, but I have doubts that the Codidact sandbox can be effective. Our sandbox relies on having multiple experienced challenge writers reviewing posts, and it doesn't seem as tho Codidact has that experience base yet
 
9:57 PM
@MonicaCellio Thanks! There was a discussion in chat and on Meta just the other day about what should be done with the Sandbox
@cairdcoinheringaahing You could draw those users by announcing posts here or in the Guild of Reviewers chatroom
 
How does sandboxing work on Codidact? Is it just a meta post like it is here, or is there a feature specifically for that?
 
@cairdcoinheringaahing getting the people is the challenge, yes. I'm not sure how we can improve that, other than by making it as easy as we can to participate there. Yeah it's another account you need to make, but beyond that, if there's friction, we'd like to know.
 
@RedwolfPrograms It's a different category entirely: codegolf.codidact.com/categories/50
 
@MonicaCellio Well, I did actually have quite a bit of "friction" setting up my codidact account :P
 
@RedwolfPrograms there's a special place for them, and an RSS feed if you want to collect them somewhere.
@cairdcoinheringaahing oh, was it you who had the email-confirmation problems? I remember somebody running into that. I'm sorry. I believe that was not typical.
 
10:01 PM
@MonicaCellio It might be a nice feature to be able to click a "Sign Up from Stack Exchange" button, a la the ability to create an account via gmail/facebook
 
I'm working on a bot that posts new sandbox posts in a room here, I suppose I could use the RSS feed to post Codidact ones as well.
 
@MonicaCellio Yep, that was me. No harm done tho :)
 
ngn
@MonicaCellio why do you require emails at all?
 
@cairdcoinheringaahing hmm, I wonder if that's technically possible (in a secure way). You've told (say) Gmail that SE can use your identity, but does SE have permission to pass it on to us? I would think people would be concerned about that. We do currently have logins based on email/password; I'm not sure what prevents us from doing Gmail/FB/others -- technical concern or haven't gotten to it?
 
@MonicaCellio I can't speak to any of that unfortunately, I'm probably less of a lawyer than you :) Just a suggestion for a way to make it slightly easier to "steal" SE users ;)
 
10:06 PM
@ngn currently, the account sign-in is an email address (known to be unique) and we send a confirmation link (so people can't sign you up for stuff). We don't use the email address for anything else now, though would like to be able to send important notices and allow subscriptions (opt-in). I'm afraid I don't know a lot about this part of the system, so I don't know how much is that it was easy vs. being a specific choice.
 
I think Top Answers just sets a cookie. Is it possible for Codidact to do something like that initially and let users add an email later if we want to subscribe to something?
 
@cairdcoinheringaahing thanks. I'm going to ask the devs for more info about why and how we do login, for my own edification. (And if somebody asks a question on our main meta about it, we can record it there too.)
 
Email is also nice for sign-in stuff because (unless you own a domain name) you only have one or two email addresses so making tons of sockpuppets is difficult
 
I have like 5 different email addresses, not counting the sock puppet accounts I've created to test low-rep user bugs :P
 
Is having sockpuppets such a bad thing? Codidact could do with the activity :P
 
10:08 PM
@cairdcoinheringaahing I just make up a random name and prepend @redwolfprograms.com, they all just forward to my personal one lol
 
@user I'd have to ask. The TA approach is nice in that they store no personal info, not even an email address. The downside is that if you lose that token you need manual intervention from an admin to get back in, vs. sending a password-reset link to the address on file. (Oh yeah, I guess we use the email address for that too.)
 
I guess there is that
 
@MonicaCellio in the true spirit of code golf, signing up to code golf on codidact should require you to 'golf' your passwords. A single character should be golfed enough. Thoughts? I think this is a very good idea.
 
Why use more than 0 bytes?
 
@MonicaCellio Does Codidact have some form of reputation cap (e.g. SE's 200/day)? I'm finding it difficult to find any help pages on the reputation system
Interesting that "Improve Edit" in Suggested Edit reviews is an insta-approve of the original edit (and "Reject and Edit" is an insta-reject), but "Approve" and "Reject" both require 2 votes to take effect
@MonicaCellio Posted
 
ngn
10:28 PM
@Bubbler @hyper-neutrino @dzaima 50 bytes
 
10:42 PM
0
Q: Golf this MMIX constant!

NoLongerBreathedInHeavily inspired by Golf this Thumb-2 Constant, but with a different encoding scheme. As MMIX has only 32-bit instructions, it's obviously impossible to encode every 64-bit value into a single instruction. Up to four instructions might be needed, in fact, due to the instruction format. The challe...

 
@rak1507 rak is out here asking the real questions
 
I know of at least one user who will dislike this
 
Make that two
 
ngn
"Standard loopholes apply." - hm.. they are usually forbidden :)
 
10:49 PM
Time to create a new programming language for the question maniacal laughter
 
@ngn Nice, it's definitely below O(n^2)
 
I think I'm going to read through all of the KotHs
 
There are currently 160
@cairdcoinheringaahing A few reasons
 
Training a machine learning AI to write your Koths for you?
 
10:51 PM
Well I wasn't going to do that, but...
 
I would love to get a solid machine learning program (a la whatever Lyxal uses but heavy-duty) and get it to write challenges :p
 
If redwolf is reading them, then I doubt it is for machine learning
Maybe a koth guide?
To avoid older mistakes on koths
 
That's the main reason
I'm going to write a short summary of every KotH, and tag them (on a txt file) with things like what languages they use
 
Damn, WW is only 81 CV reviews away from another Steward badge
 
I also want to analyze what sorts of things add strategy to KotHs
 
10:55 PM
@cairdcoinheringaahing And I'm officially over 50% of the way towards my first :D
 
Only got 497 more First Posts reviews to go :P
 

« first day (3788 days earlier)      last day (137 days later) »