« first day (4176 days earlier)      last day (664 days later) » 
00:00 - 13:0013:00 - 23:00

1:00 PM
Computing the hashes is only a small part of what that function does
 
In fact it implies to me that integer hashing is inefficient
@mousetail ah, go on
 
For the integers the time it takes to hash them is completly insignificnat
 
it also has to compare the equality of the values, to check there wasn't a hash collision
 
That should be much faster for two integers, right?
 
yes, that probably also contributes to the speed
 
1:02 PM
I'm looking at Flobnar, would it be better to ask about the examples in the documentation here or on main?
 
So why isn't the speedup at least a factor of two?
 
Because allocating memory is very expensive
Much more so than the actual hashing
 
@AncientSwordRage here in chat would be better
 
Hmm
 
| 85   5
| *p<
| 40+@
|   >  +
|      9
|      9
= Result: 18
@pxeger Struggling to see how that evaluates to 18
p sets (0,5) to 32, right?
 
1:10 PM
I don't know Flobnar, sorry
 
ah ok! no worries :D
 
not ok. worries D:
flobnar looks really cool
 
@thejonymyster Not worries, ah ok?
 
i should have stuck around for the lyal but it was super scary looking xd
 
@thejonymyster yes, I'm not bouncing off of it like my attempts to learn befunge
and it's making me realise what my proto-esolang needs
 
1:11 PM
LYAL is still going
it's a 24 hour event
 
well im at work now, so its over for me :P
 
:(
 
i'll def check it out on my own time though
 
@thejonymyster I'm checking it out from work 👀
I'll have to wait for Jo King to come back
 
In python, I have ten lists A, B, C, D etc each holding up to a million strings. I want to repeatedly compute the size of the set union of different prefixes of these lists. Currently I make a new list of sets of these prefixes each time and then take the union of the sets in the list and then do len (). Is there likely to be a faster way?
 
1:19 PM
@graffe Maybe, allocating such a large set then discarding it is going to be expensive. However, looping over them is also expensive since it's not a built in. I wouldn't be able to say for sure which is faster overall
 
@graffe What do you mean by "different prefixes of these lists"
as in if A is a, aa, aaa you're looking at a, aa and a etc
 
I assume you mean, you have a list of prefix lengths ls, and you want to compute the size of A[:l] | B[:l] | C[:l] | D[:l] for each l
@graffe something like this pseudocode: ato.pxeger.com/…?
 
1:34 PM
@pxeger yes.
@AncientSwordRage A[:100] for example
 
1:53 PM
@AncientSwordRage ahhh 32 is ASCII space
 
2:32 PM
 
 
1 hour later…
4:01 PM
@pxeger how do i tell pacman to upgrade everything but not npm?
 
If you want to do that, you probably have a problem that should be solved a different way
e.g., for npm, you may want to use nvm
 
actually npm throws errors when i upgrade it
 
when you upgrade it how?
using pacman?
 
Looks like you've used npm as root; don't do that
 
4:05 PM
@PyGamer0 found it online pacman -Syu --ignore npm
 
the proper way to fix it would be to remove /usr/lib/node_modules and then reinstall npm
 
ok
 
I really wish firefox didn't make me restart immediately after it's updated
 
@pxeger how do i reinstall npm not as root?
 
reinstall npm with pacman as normal, but then make sure you never use sudo npm for anything
 
4:09 PM
oh ok
when i run npm it says it can't find semver:
:
https://dzaima.github.io/paste#0rZVPb4JAEMXvfoq5IUkLglZx@@dS9VQlaeupaQ3CtMXCLt1Fa9P0u3dBPWiIrIW5bJawv8zOe3lLWYAkpCly6kVmzIJlhML0F8KMmBcgJ/1OF1SrAeAzKlJAzuEaKH7BkHPGmzEK4b2hfgknVGN/@wKVqgF1ViO/F4Fbj1KWwmtIA9gMDzSB8Qq5dgrtHj@XIUcQqed/kKq9nYO5FNyMwrlJpbyznao0ifOPfhQaC1EHbR7SbD1XJuYqeCmMc4gx4yhYtMJRGCH1YoQmVfAjsS70Qlr2iwKi5/SI3dNLeuNbTcp5VsuyidXXC2kKmHeMEuRCciTG0Y/NbXPMwHXCeCqgqSI0sToH2B3NnS/QT40rjzL6HbOluDlG3Beb2MTeyHCogs/iROqpMjgrG1xHP@YQXKdIRShjxcgspgLtSoe0Ch2iaJC@45C2rZe6VxHntKUAEvdTkJbyKGhjdzC9G84m7uNs5E4nA@3sX/m2tdrDJkSeKqelpuIulV7LaPvO0sppzzUm@W@t70K9r8xETipz/cpyjI7Rqkb7Aw
 
try a full uninstall pacman -Rn npm then reinstall it
 
doesnt work..
 
which npm?
 
/usr/sbin/npm
 
4:25 PM
Did you definitely do sudo rm -rf /usr/lib/node_modules; sudo pacman -Syu npm? Because I cannot see any reason that wouldn't work
 
yeah
alright i will try this tomorrow
 
4:58 PM
It's better to install npm using nvm so you can use multiple versions
 
 
1 hour later…
6:14 PM
CMP: should leading zeros be parsed as a separate number?
(in Fig)
 
If you want to sacrifice readability, then yes.
 
who cares about readability in golfing?
 
6:31 PM
@Seggan yes absolutely, its a golflang
in fact you should use bijective base for your numbers :P
 
6:57 PM
so im writing a little config dsl for a minecraft space plugin, how readable is this to yall?
 
the percent syntax is a bit confusing
 
how would you suggest it?
(this is a kotlin dsl btw, so im a bit limited)
 
it's confusing in the sense that it's not clear what motivates it or what it actually does
 
ah
 
LGTM
 
7:02 PM
its a definition of the composition of the atmosphere
maybe i should do like an addGas function
 
@Seggan lyxal
 
7:51 PM
Can someone tell me if this I/O format is valid?
 
 
1 hour later…
9:22 PM
@emanresuA wtf since when does desmos have ans
i wish theyd do a better job of letting their users know what features exist :P
challenge idea: opposite of a pristine program, errors unless you change it in which case it runs fine :P
 
Ooh I really like that idea
 
We've hit 66k total reviews :P
 
Definitely gonna try to do the review with id equal to my user id (66833) :P
 
@Zionmyceliaadamancy I'm 2.06% of those :D
 
9:37 PM
0
Q: extended euclidean without recursion or data structures. Just local variables

D. SikilaiExtended gcd in integer field is commonly programmed as a recursion. recently i found a non recursive derivation. I challenge you to program one. no libraries should be used.

 
@NoHaxJustRadvylf 3.46% here :P
Damn, I've done 2315 total reviews O.o
 
Do you have any Steward badges or are they spread evenly across the queues?
 
Spread pretty evenly
379, 140, 29, 542, 273, 341, 189 and 422 (542 is First Posts)
WW has over 5.5k total, for 8.45%
 
The review queue is always empty for me lol
I guess y'all just ninja me
 
It's almost always empty, if you want to get any reviews you basically have to have a userscript
Generic Review Tool is the one I and I think most of the other active reviewers here use, I made it a while back to replace Review Stalker Reloaded, which is another option (though not a very good one IMO :p)
 
9:47 PM
@NoHaxJustRadvylf Tho is being buggy for me, keeps opening multiple copies of a review task :/
 
Yeah. It's definitely not perfect, and I think Tampermonkey itself is part of the problem, though RSR just like...stopped working entirely for me lol
 
Yeah, RSR is basically broken
 
The issue with multiple being opened might have to do with tabs not running the scripts they're supposed to when they're not focused
 
That might be it, the tabs only load when I click on them
 
Yeah that makes sense. I'll try to think of a better way to track which tabs are open.
 
9:53 PM
@thejonymyster iirc such a thing exists?
@Steffan same
@NoHaxJustRadvylf how does that work
 
"Fun" fact: You can't include a file starting with .wh. in a Docker container (or OCI container)
 
... why ...
 
@Seggan You open codegolf.stackexchange.com/review and whenever there's a review task, it opens it in a new tab
I recommend pinning the tab
 
ah
does it have to be focused?
 
Nope
You can configure how often it auto-reloads, and set it to play a ping noise too
 
9:56 PM
noice
apparently it works with greasemonkey too
 
@emanresuA I'm pretty sure it is, it's been used many times and is also used in TI-Basic often
 
Most userscripts should, if they don't it's usually a bug
@emanresuA It's called a "whiteout file" and it indicates to delete that file in a previous layer
Because there's exactly zero less cursed ways you could indicate a file should be deleted
And having a way to escape the .wh. is totally unnecessary of course
Because why would you want your files to not randomly break
 
@Seggan ill search around for it then :)
 
@emanresuA Hmm. I suspect this would no longer be Turing-complete, but I'm not sure, it might be. (I assume there's still an accumulator.)
 
I think it actually might be
 
10:10 PM
(Technically, my goal for Acc!! wasn't Turing-completeness; my goal was that I could figure out how to write a program to find the third-largest of a series of positive integers.)
 
is it find the third largest of a series of positive integers complete
 
CMM: can I do a truthy/falsey value instead of light/dark in this question?
im having trouble generating white/black in piet :P
 
I don't think so. The challenge is very clear about the output requirements, even though those specific requirements would be frowned upon as too narrow if it were posted nowadays.
 
@Seggan you could add an alternative solution that does that but ^
 
@DLosc Why isn't this working?
Nvm I'm an idiot, I was passing it a single-line string
 
10:34 PM
Ah, lol. I started looking at it and couldn't see any problems either.
 
00:00 - 13:0013:00 - 23:00

« first day (4176 days earlier)      last day (664 days later) »