01:22 stanrifkin left 02:01 apac joined 02:58 derpydoo joined 03:08 Aedil joined 03:24 apac left 03:25 Aedil left 03:29 Aedil joined 03:49 derpydoo left 04:47 derpydoo joined 05:43 abraxxa joined 06:34 Sgeo_ left 06:39 Some-body_ joined, DarthGandalf left 06:42 Some-body_ is now known as DarthGandalf 07:07 melezhik joined 08:03 dakkar joined 08:04 ACfromTX left 08:17 ACfromTX joined 08:29 lichtkind joined 08:59 apac joined 09:01 ACfromTX left 09:14 ACfromTX joined 09:16 melezhik left 10:07 justache joined
disbot2 <holmdunc> Homebrew's formula DSL has some bizarre stuff like ohai onoe opoo 🤔 10:16
10:38 justache left 10:44 justache joined 11:08 derpydoo left 11:20 apac left
disbot2 <simon_sibl> I apologize if thats not enough information but in my code, this line %profiles{$token}<rates_snapshot> = %LATEST_CURRENCY_RATES; binds instead of assign, I dont get the point of having = and := if sometimes = still binds 🥹 makes it so confusing, I simply wanna copy the value, but in my code after I modify %LATEST_CURRENCY_RATES it also changes in the %profile 11:30
<simon_sibl> how can i force the copy 11:31
<simon_sibl> %profiles{$token}<rates_snapshot> = {%LATEST_CURRENCY_RATES}; this worked adding {} but why ?? 11:37
ab5tract I agree that this is confusing. I think it’s designed this way to avoid excessive copying. Still, it feels wrong to me as well that you have to send the hash into a constructor in order get the assign semantics you are expecting 11:54
simon_sibl: do you have a GitHub account? It may be worth discussing in a problem solving ticket 11:55
disbot2 <librasteve> @simon_sibl good question, I was as suprised as you and tried to get a hint from ChatGPT regarding the why chatgpt.com/share/68da7480-5cd0-80...61f03b0b4e 11:59
<librasteve> that goes part way to explaning, but then I realised also that %profiles{$token}<rates_snapshot> is itself a Scalar container (ie the value of any Hash element is held in a Scalar container, so your example is different to assigning a Hash to another Hash %a = %b 12:01
<librasteve> so (yes I checked) %a = %b does do copy on assignment 12:03
<librasteve> but you are doing the same as $x = %b
<librasteve> and if you think of it that way, then it is natural for Raku (when you ask it to store a Hash in a Scalar) to store a reference to the Hash rather than to make a whoke new anonynous Hash 12:05
<librasteve> one more point - just as in perl, {a => 0, b => 4} is a literal Hash - same as writing %(a => 0, b => 4), so the curlies make a new Hash from their innards (a slightly weird use of them to as them to make a Hash from another Hash (the normal contents is a list of Pairs) but seems that Raku DWIMs that OK 12:07
<librasteve> ab5tract: don't know if you caught it but I made a reply to your point on HN - apologies 12:09
lizmat librasteve_: it's more that when the left-hand side of an assignment is an iterable (such as an array or a hash), that the RHS is being iterated over and assigned to the LHS 12:38
a scalar is not iterable, so assigning a hash to scalar just puts the Hash object into the scalar
12:48 apac joined 13:36 Sgeo joined 13:40 apac left
Voldenet m: my %x = :a(1); my %y; %y<x> = %x; %x<b> = 2; say (:%x, :%y); # this actually stores a scalar that points to %x 14:42
camelia (x => {a => 1, b => 2} y => {x => {a => 1, b => 2}})
Voldenet so it's akin to
m: my %x = :a(1); my %y; my $c = %x; %y<x> = $c; %x<b> = 2; say (:%x, :%y); 14:43
camelia (x => {a => 1, b => 2} y => {x => {a => 1, b => 2}})
Voldenet m: my %x = :a(1); my %y; %y<x> = %x; %x<b> = 2; say (%x.WHERE, %y<x>.WHERE, %x.WHERE == %y<x>.WHERE); 14:46
camelia (2609782467152 2609782467152 True)
Voldenet (the example to what was said above) 14:52
not cloning anything is actually common strategy in languages, usually it's difficult to make deep clone of nested collections 14:53
well, not too difficult, just that you have to ask for that explicitly 14:54
e.g. you have to recursively clone everything
using your own functions
m: my %x = :a(1); my %y; %y<x> = Hash.new(%x); %x<b> = 2; say (%x.WHERE, %y<x>.WHERE, %x.WHERE eq %y<x>.WHERE); say (:%x, :%y); # Hash.new is being used to explicitly do the shallow copy 14:57
camelia (3168631941920 3168632024872 False)
(x => {a => 1, b => 2} y => {x => {a => 1}})
Voldenet what I'm a bit surprised about it the following 14:58
m: my %x = :a(1); my %y; %y<x> = %x.Hash; %x<b> = 2; say (%x.WHERE, %y<x>.WHERE, %x.WHERE eq %y<x>.WHERE); say (:%x, :%y); # .Hash on hash doesn't do anything?
camelia (2659745247152 2659745247152 True)
(x => {a => 1, b => 2} y => {x => {a => 1, b => 2}})
Voldenet However 15:00
m: my @n = [1]; my @m; @m[0] = @n.Array; @n[1] = 2; say (:@n, :@m) 15:01
camelia (n => [1 2] m => [[1]])
Voldenet soooo… .Hash is not copying the hash, but .Array is copying the array - I'm not entirely sure if that's sane
15:19 coleman left 15:21 coleman joined 15:26 apac joined
disbot2 <simon_sibl> Hey yes I have a Github account, sorry for the late reply, busy day T-T 15:41
<simon_sibl> well as long as I got it to work with {} just makes no sense to me if Raku does the distinction between = and := = should be deep copy and := bind (ref) But I am far from being an expert so maybe what I am saying is a terrible idea, just thats what makes sense to me, or then to add an explicit .copy method ? idk 15:43
16:00 Aedil left 16:05 Aedil joined 16:08 marcprux joined 16:19 arkiuat joined
arkiuat Voldenet: huh, that is weird. You'd think that .Hash in that context would do the same thing as .Array 16:20
disbot2 <antononcube> Please vote: news.ycombinator.com/item?id=45415790 16:38
16:38 marcprux left 16:41 apac left 16:42 dakkar left 16:47 simcop2387 left 16:53 abraxxa left
ab5tract antononcube: I only see ‘[dead]’ on that post, plus your commment 16:57
disbot2 <antononcube> Ha! Strange.
<antononcube> This is what I see: 16:58
<antononcube> cdn.discordapp.com/attachments/633...f3730&
<antononcube> Maybe, someone at HW actively hates Raku or me. (I hope it is just me.) 16:59
librasteve_ notable6: weekly 17:04
notable6 librasteve_, 17 notes: gist.github.com/fa37e372f4855d30bd...47d3c21f15
[Coke] librasteve_: There is no blog post, but a one liner about Snyk in there. let me know if you need/want more. 17:09
17:22 simcop2387 joined
disbot2 <librasteve> Coke: I could use some more please ... is there a url to link to some context maybe? 17:24
<librasteve> @antononcube I see dead also and the article does not appear in the new article sorted by time at 47 mins 17:26
<antononcube> Damn! 🙂 17:30
[Coke] again, there is no blog post. 17:35
don't have time to write one today, apologies.
disbot2 <librasteve> m: no worries; 17:43
evalable6
17:44 marcprux joined
librasteve_ rakudoweekly.blog/2025/09/29/2025-...f-control/ 17:54
disbot2 <antononcube> - Don't call us, we'll call you. - Leave the driving to us. 17:59
18:26 Sgeo_ joined 18:29 Sgeo left 19:34 marcprux left 19:37 Aedil left, arkiuat left 19:50 arkiuat joined 19:54 arkiuat left

Powered by Cro and the Raku® Programming Language.

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

Thank you!