🦋 Welcome to the MAIN() IRC channel of the Raku Programming Language (raku.org). Log available at irclogs.raku.org/raku/live.html . If you're a beginner, you can also check out the #raku-beginner channel!
Set by lizmat on 6 September 2022.
01:28 hulk joined, kylese left
tbrowder no need now, i see Voldenet already mentioned the oopsie 01:48
01:58 LainIwakura joined 02:04 LainIwakura left 02:15 hulk left, kylese joined 02:16 LainIwakura joined 03:01 LainIwakura left
perryprog > how can we on "normal" irc pm with the growing number of ppl on discord and other strange places? 04:10
by peer pressuring them to come back to irc
04:37 Aedil joined 04:42 [Coke] left 05:43 Aedil left 05:47 Pixi joined 05:52 derpydoo joined 06:34 [Coke] joined 06:55 Aedil joined 07:02 Sgeo left 07:05 abraxxa joined 07:07 abraxxa1 joined 07:08 abraxxa2 joined 07:10 abraxxa left 07:11 abraxxa1 left
disbot <librasteve> lol … i am trying hard to make the journey from Discord to IRC. Discord has the convenience of multiple server tabs to eg the HTMX and Polars servers, works well on various devices, supports media sharing and keeps the message history. otoh IRC doesn’t mess with text such as code for camelia and the bots are more reliable. IRCcloud has helped … it is always logged in in ios so i get the message history there … my 07:24
macbook habit is to log off a few times a day so i need something like screen or tmux to keep the history live and then would run them on my aws instance which I haven’t got around to yet…
07:33 abraxxa joined 07:37 abraxxa2 left 07:51 melezhik joined 08:02 abraxxa1 joined, crnlskn joined 08:06 abraxxa left, dakkar joined 08:09 abraxxa joined 08:12 abraxxa1 left, abraxxa1 joined 08:15 abraxxa left 09:00 abraxxa joined 09:04 abraxxa1 left 09:57 crnlskn left 09:58 crnlskn joined 10:40 melezhik left 10:47 abraxxa1 joined
tbrowder yes, Irccloud has been Godsend for me who has not figured out the overarching msg stuff. and their suport folks can fix it when i do something stupid 10:48
10:48 disbot left, disbot1 joined 10:50 abraxxa left 11:07 melezhik joined 11:10 Aedil left 12:27 derpydoo left 12:35 Aedil joined
Geth ¦ raku.org: librasteve self-assigned Pin Deps github.com/Raku/raku.org/issues/256 12:37
12:43 apac joined 12:44 apac left 13:28 crnlskn left 14:22 apac joined 15:30 melezhik left 15:57 abraxxa1 left 16:05 human-blip left 16:07 human-blip joined 16:18 apac left 16:19 apac joined 16:36 dakkar left 16:47 arkiuat joined
arkiuat I'm on MacOS as well, but I just keep a browser tab open to irclogs.raku.org/raku/live.html 16:48
tellable6 2025-09-02T07:44:47Z #raku <librasteve> arkiuat: thanks - now fixed
arkiuat not familiar with IRCcloud 16:49
so I was wondering about my lack of luck in ever constraining return values of subs that return a List or Seq 16:50
has anyone else ever tried to do this and run into troubles with it?
[Coke] Are you trying to use a list all of a type? or just a List?
arkiuat [Coke], I've tried both I think 16:51
certainly the former, since I don't see much point in constraining a return value to any ole List 16:52
[Coke] m: sub b returns List { return 1,2,3}; b;
camelia ( no output )
[Coke] m: sub a returns List { return 3 }; a
camelia Type check failed for return value; expected List but got Int
in sub a at <tmp> line 1
in block <unit> at <tmp> line 1
arkiuat I guess I'll have to settle for restraining the return value to any ole List, then 16:54
there's no way to constrain it to a List[Int] or suchlike?
[Coke] m: sub b returns Array[Int] { my @a of Int = 1,2,3 } ; b 16:55
camelia ( no output )
[Coke] Don't think you can do that with list - but you can with Array, for sure. 16:56
arkiuat Okay, so List and Seq "cannot be parametrized" (that's in the error message), and I need to containerize the return value if I want to constrain the type of its component items. Is that right? 16:58
[Coke]++ 16:59
[Coke] I believe that is correct. (need the container to "hold" the parameterization) 17:00
arkiuat Gotcha, thanks! 17:01
[Coke] m: sub b returns Array[Int] { my @a =1,2,3}; b
camelia Type check failed for return value; expected Array[Int] but got Array. You have to pass an explicitly typed
array, not one that just might happen to contain elements of the correct
type.
in sub b at <tmp> line 1
in block <unit> at <tmp> li…
[Coke] ^^
arkiuat Right
[Coke] (and you can't "guess" at the type, has to be explicit)
Good luck. :)
arkiuat Well, it works. I mean I think I'll get more helpful error messages this way when I break stuff. But having the return value of the sub begin with "my @r of Capture =" is pretty funny-looking if you ask me. 17:07
(it was returning a list of 4 Captures originally) 17:08
librasteve you may want to try coercion in the return type 18:07
m: sub fn(--> Hash()) { :a(0), :b(1) }; dd fn() 18:11
evalable6 {:a(0), :b(1)}
Raku eval {:a(0), :b(1)}
librasteve m: sub fn(--> Hash()) { :a(0), :b(1) }; say fn().WHAT
evalable6 (Hash)
Raku eval (Hash)
librasteve not sure if/how that interacts with the Array[Int] types - would need to experiment a bit 18:12
18:28 arkiuat left 18:29 arkiuat joined
arkiuat hmm! 18:46
19:15 Aedil left
librasteve m: 1.6605402e-27.FatRat.nude 21:07
evalable6
Raku eval
librasteve m: 1.6605402e-27.FatRat.nude.say
evalable6 (0 1)
Raku eval (0 1)
librasteve this feels wrong to me ... since the Num (IEEE 1754 64-bit FP) has finite information, it can be expressed as a FatRat without too much trouble, seems that Raku holds the opinion that an irrational number cannot be converted to a rational one 21:11
m: 1e-1.FatRat.nude.say 21:12
evalable6 (1 10)
Raku eval (1 10)
perryprog "fatrat nude say" is a crazy thing to come up in a programming language 100% unintentionally
librasteve hmm sometimes it works, I wonder if Num => FatRat is implemented as Num => Rat => FatRat (which should be the other way around) 21:13
lol - but this small Larry jokettes stick in the mind when it comes to having to remenber stuff 21:14
21:43 arkiuat left 21:48 arkiuat joined 21:52 arkiuat left 22:05 arkiuat joined
Voldenet m: sub tiniest-fat-rat($n = 2) { with $n / 2 { return $n if .FatRat == 0; samewith($n / 2); } }; say tiniest-fat-rat.FatRat.nude; 22:29
camelia (1 9223372036854775808)
Voldenet one might see a very specific number in here…
(roughly maximum value of int64) 22:31
FatRat internally can hold large values though 22:37
m: FatRat.new(1, 16 ** 32).nude.say
camelia (1 340282366920938463463374607431768211456)
Voldenet hah, of course, "reat the docs" 22:40
read 22:42
m: 1.6605402e-27.FatRat(1e-28).nude.say
camelia (1 602213665167515970973990912)
Voldenet by default FatRat cuts Num's precision to 1e-6 22:43
m: 1.6605402e-200.FatRat(1e-200).say # it might seem perfectly sane to set this high 22:54
camelia 0.0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000166054
Voldenet m: sub n64 { my num64 $n = $^x.Num }; (n64(0.1) + n64(0.2)).FatRat(1e-30).say; # but then you get this 22:55
camelia 0.3000000000000000266454
22:58 Sgeo joined
perryprog Hmm, does DateTime not have a way to initialize a date from a string given some format specifier? 23:19
(using a strftime-like format string)
(To be more specific, I have a yyyymmddHHMMSS string, and I want to parse it) 23:23
23:31 apac left
arkiuat perryprog, I don't believe so. It's basically the .new multi methods listed in docs.raku.org/type/DateTime#method_new 23:51
I've rolled my own plenty of times
perryprog Thought so, that's too bad
arkiuat Although I haven't checked raku.land -- maybe someone has already written it? 23:52

Powered by Cro and the Raku® Programming Language.

Please report any issues / comments / feature requests as an issue on App::Raku::Log.

Thank you!