🦋 Welcome to the former MAIN() IRC channel of the Raku Programming Language (raku.org). This channel has moved to Libera (irc.libera.chat #raku)
Set by lizmat on 23 May 2021.
00:02 reportable6 left 00:03 reportable6 joined
moon-child the same applies to \n 00:05
in a string literal, it's an lf, but in a regex, it matches a logical newline
and \v seems like an analogous case; in regex, it matches logical vertical whitespace, so why shouldn't it be the prototypical vertically white character in a string literal? 00:06
mykhal good counterarg. 00:09
00:11 Garlandg joined
mykhal i think it's rather uncommon char 00:13
00:18 neshpion joined 00:36 Xliff joined 00:40 squashable6 left
Xliff \o 00:43
tellable6 2021-08-04T22:34:34Z #raku <dwarring> Xliff 'use LibXML::Raw imports xmlNs as class native xmlNs
2021-08-04T22:38:12Z #raku <dwarring> Xliff constructor is xmlNs.new: :$URI, :$prefix
Xliff .tell dwarring Thanks!
tellable6 Xliff, I'll pass your message to dwarring
Xliff Is there a way to get available and free disk space through Raku without parsing something from the command line? 00:44
japhb Xliff: raku.land/github:ramiroencinas/Sys...:DISKUsage ? 00:47
Xliff: Looks like that author did a few different system stats modules (some of which work on Windows also): github.com/ramiroencinas 00:50
Xliff japhb: Thanks. Was looking at that. Was hoping to NOT need a supplementary module. 01:00
Maybe I should write the spec for %*MOUNTS or something?
01:40 evalable6 left, linkable6 left 01:41 linkable6 joined 01:42 evalable6 joined 02:08 neshpion left 02:42 squashable6 joined
japhb Why avoid the external module? I have over 60 that I always keep installed at this point. 03:10
03:16 Doc_Holliwood left, Doc_Holliwood joined 03:30 squashable6 left 03:32 squashable6 joined 04:26 justsomeguy joined 04:38 dwarring left
Xliff japhb: One less thing to "zef install". 05:02
I don't think something like %*MOUNTS is a bad idea 05:03
05:15 Sgeo left 05:17 Sgeo joined 05:32 casaca left, justsomeguy left 05:35 Sgeo_ joined 05:36 Sgeo left 05:38 casaca joined 05:39 casaca left 05:42 casaca joined 06:02 Sgeo_ left, reportable6 left 06:03 reportable6 joined 06:22 Doc_Holliwood left, Doc_Holliwood joined 06:30 Doc_Holliwood left 06:38 ilogger2 left 06:39 ilogger2 joined 06:40 patrickb joined 07:00 patrickb left 07:16 Xliff left 07:29 ufobat joined 07:33 frost joined 07:35 MoC joined 07:36 Garlandg left 08:04 linkable6 left 09:04 evalable6 left 09:05 evalable6 joined 09:07 linkable6 joined
Altreus hmm, raku has already rewritten so many things to avoid cross-discipline ambiguity that I'm kind of surprised backslash escapes exist at all in raku regex 09:27
logical newline didn't have to be spelled the same in regex as LF is in strings 09:28
lizmat Altreus: well, you could think of them as leftovers from the time that Raku was supposed to be the next Perl 09:32
Altreus Not to be unduly critical, but it kinda feels bad that Raku already has historical reasons, and it's not even dry yet 09:33
Altai-man Altreus, one cannot escape historical reasons since day 0 when creating anything, nothing can be created out of "nothing", without some history before that. :) 09:36
09:38 Xliff joined
lizmat Altreus: the paint may not be completely dry yet, but there's 20 year of history there 09:38
also: this implementation of Raku could be considered the 3rd attempt from scratch, basically
Altreus Yeah, I mean you can't replace everything, and it's only 20 years later when someone in IRC goes "This seems weird" that you get the benefit of hindsight 09:51
Xliff checks his scrollback 09:53
One cannot account for Raku without the history of its Perl roots. 09:54
10:10 sono left 10:46 Doc_Holliwood joined 10:47 sena_kun joined 11:00 Scotteh_ is now known as Scotteh 11:02 Xliff left 12:02 reportable6 left 12:05 Xliff joined
Xliff lizmat: What's the progress on log inspection for channels like #cro. 12:07
lizmat you mean: logs.liz.nl/cro/2021-08-05.html
12:14 stoned75_ left 12:16 stoned75 joined
Xliff Thanks! 12:39
12:48 Doc_Holliwood left 13:05 reportable6 joined 13:19 Doc_Holliwood joined
[Coke] irssi inside tmux on an ssh session from my iPhone surprisingly usable 13:59
jdv s/tmux/screen/ && ditto 14:07
Altreus what are you using for ios ssh? 14:08
ugexe i switched from irssi to weechat years ago since it seemed similar enough while allowing me to use glowing-bear.org as a web based irc client that connects to that weechat instance 14:10
jdv oh, right. android here. connectbot iirc.
Altreus I switched to weechat because of that very same relay thing, and there's a weechat app for android that connects directly to it 14:12
(Although I seem to have screwed mine up somehow and it's very reluctant to output diagnostics)
But there's no such client for ios
I do like that glowingbear works on any weechat instance so you don't have to actually run your own
14:19 evalable6 left, linkable6 left, evalable6 joined
[Coke] Altreus: Termius 14:28
Altreus looks swish 14:31
14:33 jrjsmrtn_ joined 14:35 jrjsmrtn left, frost left 14:55 Doc_Holliwood left 15:02 Sgeo joined
tbrowder .tell CIAvash i used yr emacs config line and, so far, no problems with unicode fonts! thnx! 15:09
tellable6 tbrowder, I'll pass your message to CIAvash
15:20 linkable6 joined 15:21 Garlandg joined 15:50 ufobat left
tbrowder anyone heard of c static analysis tool frama-c (frama-c.com)? 16:22
moon-child verifast is cool 16:27
[Coke] using HTTP::UserAgent... getting a hard "Could not receive data from socket 16:41
exception that isn't caught; I can put in a wrapper here, but wondering if this is something that should be caught by H:UA 16:42
if we were java, the get method would declare it could throw an IOException, or somesuch. 16:43
16:50 linkable6 left, evalable6 left 16:58 ufobat joined 17:02 Sgeo left 17:08 Sgeo joined 17:16 sena_kun left
Xliff Can you get the current invocant from a Method? 17:29
17:52 linkable6 joined
moritz the current invocant is a property of the method invocation, not of the method 17:52
what are you trying to achieve?
17:53 evalable6 joined
moritz m: my $method = Num.^find_method('sqrt'); say $method # that method has no invocation going on currently 17:54
camelia sqrt
moritz so no way to extract any invocant
17:55 Guest12 joined 18:01 ufobat left, Sgeo left 18:02 reportable6 left 18:04 Sgeo joined 18:05 reportable6 joined 18:07 Sgeo_ joined 18:11 Sgeo left 18:20 Altai-man left, Altai-man joined 18:30 Guest12 left 18:38 Doc_Holliwood joined 18:46 Sgeo_ left, Sgeo joined 19:08 sono joined 19:27 sono left
Util . 19:58
tellable6 2021-07-29T10:03:37Z #raku <lizmat> Util ttps://github.com/rakudo/rakudo/pull/4484
Util lizmat: thanks for writing pull #4484. I will revisit after the conference. 20:07
lizmat sure... :-)
the PR is not going away :-)
20:15 SmokeMachine_ left 20:16 skaji left
JRaspass m: require ::($_) for "JSON::Fast" 20:25
camelia ===SORRY!===
Cannot invoke this object (REPR: Null; VMNull)
tellable6 2021-07-08T12:57:00Z #raku-dev <tbrowder> JRaspass ^^^
JRaspass m: $_ = "JSON::Fast"; require ::($_)
camelia Could not find JSON::Fast in:
JRaspass is require not meant to work inside a for loop/
if so, "Cannot invoke this object (REPR: Null; VMNull)" aint a great message 20:26
20:30 sono joined 20:42 skaji joined, SmokeMachine_ joined
Xliff If a compunuit is spending alomost 4 seconds in stag mbc, is there a possibility that it is not being precompiled? 20:49
ugexe i think you are misunderstanding your example JRaspass 21:11
m: for <Test NativeCall> { say try require ::($_) }
camelia (Test)
JRaspass It's just postfix it doesn't like, interesting 21:12
ugexe oh wait i misunderstood... i didnt even see the `for`
JRaspass m: for "Test" { require ::($_) } 21:15
camelia ( no output )
JRaspass yeah block for is fine, postfix dies, this smells like a bug
ugexe yeah unfortunately i think `require` escapes lexical scope 21:16
or maybe fortunately if you want to look at it like that
unfortunately because `use` statements are lexical
require leaks out 21:17
hmmm thats not strictly true either 21:18
it is true in one case or another... i just forget which one :) 21:19
21:19 jess is now known as j
JRaspass So is it a bug? For now I'll use a block loop 21:22
ugexe if its a bug its probably more a compile time thing with postfix instead of specifically e.g. require 21:24
er no that wouldnt make sense, its require no compile time heh 21:28
JRaspass I don't really grok why the variable has to be so dynamic either 21:30
The docs say "The compunit name can be in a runtime variable if you put it inside an indirect lookup.", but not really why
dynamic is the wrong word but you know what i mean 21:31
21:32 MoC left, discord-raku-bot left, discord-raku-bot joined
ugexe if you mean why cant you just use a string, probably because its just too ambiguous as to what it could represent 21:39
i.e. module name, a path, some other spec for theoretical curs like http urls 21:41
thats not the greatest reason though so maybe thats less likely than i implied
JRaspass require by path would actually be pretty useful for my purposes, im dynamically trying to load all the modules in a dir 21:43
gfldex Maybe require tries to turn it's argument into Package unless it already is one?
ugexe you can require by path
JRaspass I can!? ooo 21:45
ugexe raku -I. -e 'require "lib/Zef.rakumod"; say "ok"'
JRaspass Hah, I really went the long way around then 21:46
I have: for dir 'lib/Local/Routes', :test(*.ends-with: '.rakumod') { require ::("Local::Routes::{.extension('').basename}"); }
ugexe that dir code isnt for an installable distribution though right? 21:47
gfldex m: require :file<foo.rakumod>; 21:48
camelia No such method 'IO' for invocant of type 'Pair'. Did you mean 'so'?
in block <unit> at <tmp> line 1
JRaspass Nah this is just a cro app trying to load some libs, i am being a bit evil and assuming lib is in the cwd 21:50
I didn't know what the raku equiv of @INC was
ugexe something like `$?DISTRIBUTION.meta<provides>.grep({ $_.key.starts-with("lib/Local/Routes") }).map({ require $_.value; ::($_.key) })` 21:52
might work 21:53
JRaspass Ah none of this is in the meta
this is an app rather than a dist
ugexe yeah but you still are putting your app files in a lib/
JRaspass sure
What does `-I` and `RAKUDOLIB` influence? There must be an array of lib paths somewhere? 21:54
ugexe m: say $*REPO.repo-chain
camelia (inst#/home/camelia/.raku inst#/home/camelia/rakudo-m-inst-2/share/perl6/site inst#/home/camelia/rakudo-m-inst-2/share/perl6/vendor inst#/home/camelia/rakudo-m-inst-2/share/perl6/core ap# nqp# perl5#)
JRaspass Ah ta 21:55
ugexe where inst# file# ap# tell it what type of CUR to load it with
s/load it with/it is/
gfldex m: use lib <./>; require Foo;
camelia Could not find Foo in:
gfldex `use lib <./>` points to the $CWD . 21:56
JRaspass ah require by file is slower right? looks like its not being pre-compiled? 21:59
gfldex aye
ugexe github.com/tony-o/perl6-pluggable/...#L123-L135 the comment i left here is what i was talking about earlier about it leaking out 22:05
22:11 juanfra__ left
ugexe should probably make CURI work with relative paths... no reason it shouldnt afaict 22:12
i.e. `require "lib/Test.rakumod";` might as well just work
22:13 juanfra__ joined
ugexe at the very least it would require a slow path though since we wouldnt have the sha1 of the module name itself 22:14
i dunno maybe not
JRaspass Thanks for all the help gfldex and ugexe , I shipped my raku.land change with dynamic loading of all routes :-) 22:45
22:57 gordonfish- joined 22:59 gordonfish left 23:11 j is now known as jess 23:38 Doc_Holliwood left