02:15 jgaz left 02:45 topnep left 02:46 topnep joined 03:59 stanrifkin joined 04:00 stanrifkin_ left
disbot8 <rcmlz> the „old“ definition builds rakudo 2022 without problems? 05:07
<simon_sibl> yeah the "old" definition worked good 05:43
<simon_sibl> although they have a patch part where they disable some tests 05:44
<simon_sibl> they say it works when ran manually
06:56 topnep left 06:58 topnep joined 07:14 Guest85 joined 07:15 Guest85 left
disbot8 <rcmlz> That is the advantage/challenge with Nix and Gnuix - they apparently reveal (m)any impurities and force developers to be crystal clear about any dependencies and versions of dependencies and sources for each and every dependency … and make incompatibilities appear that otherwise might have stay hidden („but it works on my system …“ sort of mentality) However, in the light of that „Cyber Resilience Act“ Liz was 09:06
talking about in the weekly and all related efforts to make supply chain attacks less likely: being able to reproduce any binary software package bit-by-bit is the foundation - IMHO. If it is not easy to compile a software like Rakudo, how would you know (as a company or private person) you are not running the „patched“-version with some „extra-features“ supplied by your local or foreign Survilance Agency?
<rcmlz> So one approach to have a BOM for Raku would be to create and constantly use a Nix or Guix definition (in Nix it is called „flake“ these days) that works with the „—pure“ flag - basically assuming an empty system with nothing installed. Just a thought … maybe I am misunderstanding things. 09:11
lizmat I think that's roughly the direction we're going to, yes 09:21
disbot8 <rcmlz> @simon_sibl The good people maintainng Rakudo on Nix use icu zlib gmp perl as build inputs and removing some "dynext" links (L42). I have no idea what that is ... ;-) github.com/NixOS/nixpkgs/blob/nixo...efault.nix 10:12
10:49 librasteve_ joined 11:07 topnep left 11:09 topnep joined 13:49 librasteve_ left
disbot8 <simon_sibl> with Guix they use channels and commits to get the exact same version 13:49
<simon_sibl> I will take a look, I will try to help but I am really not used to make packages definition yet 😆 I just modified some easy lines to get all to compile except rakudo
<antononcube> related to the last comment by Simon ( @simon_sibl ): 1. Which packages from the Raku ecosystem help making new packages? 2. Which one is your favorite? 13:53
<antononcube> I know that @tbrowder mentions (-ed) "App::Mi6". Also, CommaIDE -- or the corresponding plug-in -- have a wizard for creating a new Raku package. (Or Cro app, and others.) 13:57
<simon_sibl> oh thats true I dont have experience making Raku packages but I meant Guix packages 😆 🙏 14:00
16:19 ntv left 17:25 topnep left 17:27 topnep joined
disbot8 <winfredraj> Hello all, When I use the character class <print> in a grammar, it seems to think it is a method or sub and tries to resolve it 18:42
<winfredraj> Anyway around it??
<winfredraj> I get the error Cannot resolve caller print(ladder_actions:U: ladder:D) 18:44
<winfredraj> Where ladder is the grammar and the ladder_actions is the action
<winfredraj> <graph> doesn't through an error, but I need to match "=" so I need print 18:45
<winfredraj> If I use <:S> I can match but would prefer <print> to include text as well 18:46
lizmat winfredraj which version of Rakudo are you using ? 19:02
19:54 habere-et-disper joined
habere-et-disper Why are the following not equivalent... 19:55
m: map *.key, grep { .value == 2 }, ( 0 => 1, 1 => 2 )
camelia ( no output )
habere-et-disper m: say map *.key, grep { .value == 2 }, ( 0 => 1, 1 => 2 )
camelia (1)
habere-et-disper and:
m: say grep { .value == 2, :k }, ( 0 => 1, 1 => 2 )
camelia (0 => 1 1 => 2)
20:23 arkiuat joined
arkiuat habere-et-disper, the :k adverb belongs outside the block in your second form 20:24
m: say grep { .value == 2 }, :k, ( 0 => 1, 1 => 2 )
camelia (1)
arkiuat inside the block, the :k is just ignored 20:25
20:26 habere-et-disper left
disbot8 <wambash> m: say grep {$_ == 2}, (1, 2), :k 20:30
<Raku eval> (1)
20:43 DarthGandalf left, DarthGandalf joined 20:59 arkiuat left
disbot8 <nahita3882> it's ignored in the sense that it doesn't do what's desired, but what is happening is that the block always evaluates a 2-length List (Bool, Pair), which, since has length > 0, is considered truthful and &grep behaves like an all-pass filter, i.e., no item will ever be discarded 21:04
21:10 arkiuat joined 21:15 arkiuat left 21:22 arkiuat joined 21:51 arkiuat left 22:04 arkiuat joined 22:12 arkiuat left 22:28 arkiuat joined 22:33 arkiuat left 22:47 arkiuat joined 22:49 stanrifkin left 22:51 stanrifkin joined, stanrifkin left 23:38 habere-et-disper joined 23:54 habere-et-disper left