00:25 ecocode___ left 01:05 hulk joined, kylese left 02:15 hulk left, kylese joined 02:22 facetious joined 03:56 topnep left 03:57 topnep joined 04:38 gordonfish- joined 04:39 sivoais left, gordonfish left, ilogger2 left, phogg` joined, _________ left, phogg left, avuserow left 04:40 sivoais joined 04:41 _________ joined, gordonfish- is now known as gordonfish, avuserow joined 04:51 ilogger2 joined 05:18 Aedil joined 06:03 abraxxa joined 06:06 abraxxa left 06:09 abraxxa joined 06:49 Sgeo left
disbot8 <librasteve> it makes me weep to see so many dist.ini files sprinkled around raku.land ... a couple of years back I filed an issue to allow .yaml (or toml?) but there was no response 07:20
<librasteve> it's just so Windows 3.11
07:36 skaji__ left, skaji__ joined 08:05 wayland joined 08:07 topnep left 08:08 topnep joined 08:15 lichtkind joined 08:51 Aedil left
lizmat fwiw, /me hates yaml, but that might just be personal, because it feels like programming in Perl without "use strict" 08:53
08:54 Aedil joined 09:24 nine left, nine joined 09:28 apac joined 10:12 topnep left 10:13 topnep joined 10:36 apac left, apa_c joined
disbot8 <holmdunc> @tbrowder That sounds similar to how some people don't like masses of ~/.something files in their home directory and ask various program maintainers to use ~/.config/.something instead (i.e. $XDG_CONFIG_HOME) 11:08
<holmdunc> ~/.config/something, rather 11:09
11:17 itaipu joined 11:36 apa_c left 12:17 topnep left 12:21 topnep joined 12:38 tejr_ joined, tejr left, tejr_ is now known as tejr 13:26 apa_c joined 13:47 zostay joined 14:04 ab5tract left 14:05 ab5tract joined
disbot8 <librasteve> lizmat: there is obviously a wide range of config file types (json, yaml, toml, etc) - please give it a 2nd look because (i) yaml is a very clean way to present large amounts of config - such as github.com/librasteve/raku-Physics...neral.yaml and (ii) the raku YAMLish module (by Leon Timermans) is role model behaviour for "how to make a grammar that's extensible - such as 14:19
github.com/librasteve/raku-Physics...rakumod#L6 ;-)
lizmat librasteve I'm glad you like it, but as I said, it reminds me too much of Perl and that just brings back bad memories 14:21
disbot8 <antononcube> I don't hate YAML, but only use it if I have to. Especially, for the Raku LLM packages, using JSON is always preferable because: (1) LLMs work with JSON, (2) LLMs produce reliably JSON results, and (3) the easy conversions to/from Raku data structures. 14:26
<librasteve> we're going to agree to differ on that then! (i fully accept what you say on LLM side ... but all those quotes are a bit heavy for my needs) 14:36
14:40 apac joined 14:41 apa_c left
lizmat I'd rather type some extra quotes than have the entire meaning change because of some sloppy whitespacing 14:43
15:10 itaipu left 15:13 itaipu joined 15:14 librasteve_ joined
tbrowder hi, once more routine "split" is confounding me. given a csv header or record with empty internal or trailing empty fields, how do i run split to capture all the fields' contents into a list--where the list will include the empty fields? 15:20
i should have said all the header fields *will* be filled 15:22
^^^ no empty fields in the header 15:23
lizmat use Text::CSV 15:36
disbot8 <antononcube> @tbrowder Do you mean the sum of the roles have fewer fields than the number of columns? 15:42
[Coke] m: dd "this,that,other".split(',') 15:49
camelia ("this", "that", "other").Seq
[Coke] "no empty fields" ^^
m: dd "this,that,,".split(',') 15:50
camelia ("this", "that", "", "").Seq
[Coke] m: dd "this,that,,".split(',', :skip-empty)
camelia ("this", "that").Seq
[Coke] but, yes, if you're doing anything with CSV, do NOT do it by hand.
you'll only regret it later.
15:58 apac left 16:15 apac joined 16:31 topnep left 16:32 topnep joined
tbrowder thnx all. it's a subtle string problem it seems, i shouldn't have cried wolf 16:34
and [Coke] do you mean using say vi to fiddle with it? 16:35
16:54 apac left, apac joined 17:11 apac left
[Coke] I mean don't write code to do CSV - it's a solved problem, and you'll miss so many non-simple cases that are valid in CSV 17:29
disbot8 <librasteve> Coke++ CSV decode can be quite misleading - I always use Text::CSV 17:32
17:47 topnep left 17:49 topnep joined 18:26 itaipu left 18:28 itaipu joined
[Coke] I've fallen into the trap myself a few times, and almost immediately need to handle commas inside quotes 18:36
librasteve_ notable6: weekly
notable6 librasteve_, 1 note: 2025-08-19T18:21:56Z <librasteve_>: imgur.com/EP3ZuYh
18:38 thowe left, thowe joined 18:45 itaipu left 19:00 itaipu joined
librasteve_ notable6: weekly reset 19:01
notable6 librasteve_, Moved existing notes to “weekly_2025-08-25T19:01:41Z”
librasteve_ rakudoweekly.blog/2025/08/25/2025-...reducible/ 19:10
disbot8 <antononcube> @librasteve I think "Antons’s Corner" should be corrected to "Anton’s Corner". 19:14
19:53 topnep left 19:54 topnep joined
disbot8 <librasteve> @antononcube oops - thanks!! 19:59
20:04 nort joined 20:23 itaipu left 20:40 itaipu joined
disbot8 <cultoftheglenda> Hello everyone, apologies for using the bridge, but I have a rather simple question. I am interested in getting the version of the current package. I have a method in a .rakumod file that returns the following reference: $?DISTRIBUTION.meta<ver>. This evaluates to "*" rather than what I would expect "2.5.0" which is defined in the META6.json file. What meaning does "*" have here? 20:59
21:12 ecocode___ joined
lizmat whatever 21:56
librasteve++
22:21 librasteve_ left 22:24 topnep_ joined
ugexe cultoftheglenda: use $?DISTRIBUTION.meta<version>. <ver> is an old leaked implementation detail 22:25
in other words: write $?DISTRIBUTION.meta<$key> where $key matches the key as it is literally written in the META6.json file
oh also 22:26
you'd get * as a version when using an include path that does not contain a META6.json
-Ilib would return *, -I. will return what is in the META6.json
22:27 nort left, Aedil left, topnep left 22:29 Aedil joined, lichtkind left 22:41 Aedil left 22:57 wayland left 23:14 Sgeo joined 23:22 ecocode___ left