🦋 Welcome to Raku! raku.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: colabti.org/irclogger/irclogger_log/raku
Set by ChanServ on 14 October 2019.
discord6 <theangryepicbanana> I've gotten a working example for my Raku LLVM library working here if anyone's interested 02:27
<theangryepicbanana> repl.it/@theangryepicbanana/llvm-p6-thing
<theangryepicbanana> lizmat: I think someone said that you might be interested in something like this. if you're not then just ignore the mention 🙂 02:31
<theangryepicbanana> also fyi I'm on the discord bridge 02:32
Geth doc: ac9517314f | Coke++ | doc/Type/Raku.pod6
remove doubled .
05:36
linkable6 Link: docs.raku.org/type/Raku
[Coke] lizmat: in doc repo, xt/examples-compilation.t doc/Type/Perl.pod6 is failing; I updated my rakudo, same failure. no idea why. Think it's one you touched today
tyil @theangryepicbanana messages from the bridge are easily distinguished from regular IRC messages :> 09:24
El_Che releasable6: status 10:09
releasable6 El_Che, Next release in ≈25 days and ≈8 hours. There are no known blockers. 0 out of 14 commits logged
El_Che, Details: gist.github.com/e733c40b6584a5a0f9...766bf08091
tellable6 hey El_Che, you have a message: gist.github.com/e9028f191c5371d23c...531b43c261
El_Che it's out?
El_Che .tell patrickb I am not territorials about packages. The more the merrier and even better if we can work together! 10:09
tellable6 El_Che, I'll pass your message to patrickb
fluca1978 I've a question about 'multi': why we should mark a sub/method with multi? If we don't the compiler will not allow it, so it seems to me Perl is smart enough to find out the method is duplicated. Why whould then explicitly say it's a multi method? 10:22
El_Che fluca1978: the compiler? sure 10:23
the programmer, less so
:)
El_Che what is you redefine the sub on an other file? 10:24
fluca1978 El_Che: therefore is just a "warning" to ask the developed he knows what he is doing, right?
El_Che the "use warnings" or "use strict" in Perl 5 10:25
I'd rather have a strict compiler, but opnions diverge on that I suppose
jnthn It can avoid quite a few surprises
El_Che Some people say compilers are really dumb
jnthn Especially if you declare an only sub in a scope, and their happens to be something in an outer scope with the same name 10:26
El_Che Still, a compiler is probably smarter than me :)
jnthn If that thing in the outer scope were to be refactored into a set of multis, then you might be in for a surprise.
fluca1978 El_Che: clear! It was a doubt because it reminds me the "virtual" of C++ methods...
jnthn You might thing "why'd I do that", but you might not; some module you import might
El_Che fluca1978: languages like go get away by not allowing it, or the classical dynamic langauges by not caring 10:27
multi is a nice middle ground, imho
jnthn Quite a few things in Raku are designed to preserve refactorability - whether you're doing it by hand, or to make it possible to implement more reliable tools to do it for you
fluca1978 El_Che jnthn: thanks
another question: I'd like to get a MAIN named argument act as a global variable, like MAIN( Bool :$verbose ), so that I can check it in whatever class I am into. Does that sounds dangerous? How can I do something right with that? 10:28
El_Che in my view, while staying dynamic in nature, raku is trying hard to fix some problems typical with dynamic languages, e.g. by moving failures to the compile phase instead of runtime
jnthn fluca1978: Declare it as Bool :$*verbose (a dynamically scoped variable) 10:29
It's the same danger as all dynamic scoping, I guess
Ah, and you access it as $*verbose too 10:31
fluca1978 jnthn: thanks, that is what I was looking for! shame on me I didn't think about $* 10:33
Lee66 Hello everyone. 11:23
Lee66 Does anyone have any suggestions about where to start looking for doing graphical programming in Raku? 11:24
sena_kun Lee66: graphical like GUI or like games? 11:30
lizmat Lee66: could you explain what you mean exactly by "graphical programming" ?
lizmat clickbaits rakudoweekly.blog/2020/01/27/2020-...pringtime/
Lee66 Sorry. GUI's and suchk. 11:31
such*
sena_kun Lee66: there are a couple of modules for GTK support, GTK::Simple is probably the most interesting 11:32
sena_kun Lee66: a sorta fun tutorial to look into is, for example, perl6advent.wordpress.com/2018/12/...ktop-apps/ 11:34
Lee66 Okay, I'll look into those. Thank you. 11:36
cpan-raku New module released to CPAN! LogP6-Writer-Journald (1.3.3) by 03ATROXAPER 11:45
rypervenche jnthn: Is there support for the Activate signal in GTK::Simple so that I can simply press Enter in an entry box and have it enter my text without needing to physically click a button? 12:02
sena_kun rypervenche: if there isn't, you can patch. :) 12:08
discord6 <theangryepicbanana> tyil: ah ok thanks 12:27
tyil ^_^
patrickb El_Che: Great! Then I'll give that a shot. 12:55
Voldenet m: sub x { "haha" }; sub y { { LEAVE x; return "not so haha" } }; y.say 13:01
camelia Cannot find method 'say' on 'BOOTCode': no method cache and no .^find_method
in block <unit> at <tmp> line 1
Voldenet uh, ok
dakkar Voldenet: more LEAVE weirdness! the gift that keeps on giving 13:16
lizmat worthy of an issue, I'd say 13:19
dakkar github.com/rakudo/rakudo/issues/2380
been there for a while ☺
lizmat hopefully as easily fixable as the recent double ENTER issue
ah, so I was right (about it being worthy of an issue :-) 13:20
dakkar that issue (openend 2018-10-16) has a bunch of weird interactions between `LEAVE`, nested blocks, and `return`
dakkar it also produces really undesirable behaviour in Log::Timeline (enabling logs changes the return value of some subs) 13:21
Voldenet it's very similar to the previous code, but I like that error message better 13:26
jnthn rypervenche: No idea, though if you don't see it implemented, probably not. I've not done any of the work on that module for years; I only wrote the initial small bit of code showing how to wire things up. 13:30
El_Che already? Found /opt/rakudo-pkg/bin/moar version 2020.01, which is too old. Wanted at least 2020.01.1 at /nqp/3rdparty/nqp-configure/lib/NQP/Config.pm line 192. =- 13:42
tellable6 2020-01-28T12:54:00Z #raku-dev <patrickb> El_Che: Great! Then I'll give that a shot.
El_Che :)
rypervenche jnthn: Got it. Thanks :) 14:09
Geth doc: 8cca87f95b | Coke++ | xt/examples-compilation.t
Track compiler name change
14:19
whateverable: 110ed6a356 | Altai-man++ (committed using GitHub Web editor) | lib/Whateverable.pm6
Fix operator
14:20
long_long_size hi 14:49
which is harder Perl or Raku ?
long_long_size hi Maylay 14:50
jnthn long_long_size: As with all language learning, it surely depends on which other languages you're familiar with. 14:53
[Coke] We're partial to Raku here, of course.
Raku has some very nice books available if that's how you learn. 14:54
El_Che long_long_size: they are both 4 characters 15:08
Maylay hi long_long_size 15:19
Geth doc: 2c083530bc | (Trey Harris)++ | doc/Type/Range.pod6
Fix Range.elems docs to match current behavior

This referred to old behavior which was incorrect by roast. Corrected.
15:48
linkable6 Link: docs.raku.org/type/Range
[Coke] doc - I have a .cache-doc folder -- is that a replacement of .pod-cache? 15:55
[Coke] .seen jjmerelo 16:05
tellable6 [Coke], I saw jjmerelo 2020-01-27T19:18:23Z in #raku: <jmerelo> antoniogamiz: in settings, same as AlexDaniel
[Coke] .tell jjmerelo - need to add .cache-doc to .gitignore for Raku/doc - guessing this is from the build change. might be able to remove .pod-cache and references to it. 16:06
tellable6 [Coke], I'll pass your message to jmerelo
tyil if anyone's feeling adventurous: there's now a rakudo-star-2020.01-rc1 on dist.tyil.nl/raku/rakudo-star/ 16:07
El_Che tyil++ 16:09
cpan-raku New module released to CPAN! Math::Libgsl::Elementary (0.0.1) by 03FRITH 16:12
New module released to CPAN! Math::Libgsl::Constants (0.0.1) by 03FRITH 16:13
cpan-raku New module released to CPAN! Gnome::N (0.15.4.1) by 03MARTIMM 17:00
melezhik new Sparrow plugin to handle zef distribution fetch, for someone needs to automate work with module distribution internals - github.com/melezhik/sparrow-plugin.../zef-fetch 17:12
cpan-raku New module released to CPAN! Math::Libgsl::Complex (0.0.1) by 03FRITH 17:16
Kaiepi . 18:27
jmerelo Hi, Kaiepi 18:27
tellable6 2020-01-27T21:06:45Z #raku <Kaiepi> jmerelo, saw your message, just trying to think of projects that i would be good for mentoring for and appealing for students. who's the target audience for these grants?
2020-01-28T16:06:04Z #raku <[Coke]> jjmerelo - need to add .cache-doc to .gitignore for Raku/doc - guessing this is from the build change. might be able to remove .pod-cache and references to it.
jmerelo Answering directly
Kaiepi: in general, people whose first contact with Raku will be this one.
Kaiepi: but other than that, they're generally enthusiastic and hard-working 18:28
.tell [Coke] OK
tellable6 jmerelo, I'll pass your message to [Coke]
Kaiepi aight
jmerelo hey, stoned75 18:34
stoned75 hi jmerelo 18:35
stoned75 jmerelo: I recall what I wanted to ask you this other day. is there a logic in the individual operator sections in operators.pod6 ? 18:36
jmerelo stoned75: there's history. Not sure if there's any logic at all. 18:38
stoned75: you mean grouping them by precedence
stoned75: or in general?
stoned75 jmerelo: I mean the order of the operators inside a precedence section 18:39
jmerelo stoned75: I can't figure out any.
stoned75 jmerelo: neither can I :) 18:39
TreyHarris Can anyone think of a core routine name that's used for two entirely different things depending on type? (semantically entirely different, not just multis) 18:40
jmerelo stoned75: it will probably have grown by accretion, with new operators added as soon as they were created or simply being noticed.
TreyHarris like, the old example Dog.bark() and Tree.bark() 18:41
stoned75 jmereolo: sure but as I noticed some operators are missing (well at least one), I was wondering where to insert the section I wanted to add :)
jmerelo TreyHarris: hm. No.
stoned75: I guess that you can add it wherever and increase the entropy :-)
TreyHarris wrong channel anyway
stoned75 jmerelo: eheh 18:42
TreyHarris: .e ? 18:44
TreyHarris stoned75: indeed! thanks 18:44
stoned75 np 18:45
also .from 18:47
TreyHarris thanks! 18:49
cpan-raku New module released to CPAN! IP::Addr (0.0.3) by 03VRURG 19:22
Geth ¦ problem-solving: treyharris assigned to JJ Issue Routine doc pages need front matter github.com/Raku/problem-solving/issues/155 19:40
El_Che weekly: github.com/nxadm/rakudo-pkg/releas...g/v2020.01
notable6 El_Che, Noted! (weekly)
guifa is seriously considering a GSoC application after talking with Kaipei and jmerelo for pushing Intl stuff even further 19:42
But would need a mentor who’s got the time (jmerelo’s got too much on his plate to add it). Anyone potentially interested?
samcv: I know you did GSoC as a student and I’ve been meaning to play around with your Unicode sort work, but no idea if you’d have time 19:43
sjn El_Che: suggestion: produce sha256 checksums too? 19:52
El_Che sjn: sure, it's changing 1 line :) 19:53
or maybe gpg sigining
melezhik I _probably_ found the issue with p6doc install under user, where can I put a ticket? - repo.westus.cloudapp.azure.com/raku...241616.txt 20:07
cpan-raku New module released to CPAN! Gnome::Glib (0.15.5.3) by 03MARTIMM 20:09
[Coke] what is "cpan-p6doc-alpine" ? 20:11
tellable6 2020-01-28T18:28:43Z #raku <jmerelo> [Coke] OK
lucasb 'raku -v' says "This is Rakudo version (...) implementing *Perl* 6.d" 20:12
shouldn't it be implementing the new language Raku instead? :) 20:13
sena_kun well... 20:13
melezhik Coke: it's local user under which zef install gets run for p6doc module 20:14
[Coke]: it's local user under which zef install gets run for p6doc module
[Coke] would you expect to be able to install into the system directory with that user? 20:15
melezhik no, but the problem is when others modules use the p6doc their installs fails for the same reasons 20:16
for example - repo.westus.cloudapp.azure.com/raku...239499.txt
I mean user install should be successful
antoniogamiz o/ 20:17
Voldenet isn't there some support for local manpath?
antoniogamiz do you know some docker container with raku and zef installed?
Voldenet alpine-perl6 20:18
github.com/JJ/alpine-perl6
[Coke] I don't understand. you're able to install *other* modules to your system folder with zef, just not p6doc?
antoniogamiz Voldenet: thanks :D 20:19
Voldenet Question: Is there any way to defer statements for the block? I want to do conditional "close after opened", but I'm not sure how to implement this 20:22
[Coke] basically, if that user can't create files or dirs in /opt/rakudo-pkg ... then I'd expect any zef install to die with a similar error. P6doc may be "special" in that it's copying things in addition to raku lib files. (more) 20:23
if others are working, maybe you have permissions setup properly for *part* of /opt/...
stoned75 Voldenet: may be docs.raku.org/language/phasers#LEAVE
Voldenet LEAVE is always being called, even if there's return before it
[Coke] It's hard to tell if you're doing anything else given the log format here (rather than, e.g. a typescript of an install) 20:24
stoned75 IC
Doc_Holliwood A comment needs likes: www.youtube.com/watch?v=6gddK-cOxYc&t=0s It's the one linking to raku
melezhik Coke: the install is dead simple. It's 20:25
zef install p6doc under $user 20:26
[Coke] do any zef installs work for you in this env?
rypervenche I also ran into this p6doc issue because it tries to install things in /usr, I think. 20:27
[Coke] I see Test::When in the pre-reqs, if you need one to test.
ok, but it's not complaining about /usr in the logs you show.
melezhik and then _for some_ reasons `zef install` tries to mkdir /opt/rakudo-pkg/share/perl6/site/doc which is obviously is forbidden for none privileged user
[Coke] Just trying to see what the root cause is.
rypervenche My installation is different. Let me get my own logs, one sec. 20:28
[Coke] (/usr) oh, two different people. :)
melezhik Coke: the test/installation script is written on Sparrow and could be found here github.com/melezhik/RakuDist/blob/...arrowfile. Again nothing complicated here 20:29
[Coke] (create dir): github.com/Raku/doc/blob/master/Build.pm6#L12 20:30
rypervenche [Coke]: gist.github.com/rypervenche/76e0c2...d5cb84dede
melezhik it just create a `$user`, then fetch zef distribution , then `cd to $distribution_dir` and then `zef install .` , the last command run _under_ $user
[Coke] So you've got a restricted permission area and you're trying to copy files into it with a non-restricted user. 20:31
sorry s/non-restricted user/user who doesn't have privs/ ?
melezhik well then thing _I_ am not doing this. It's done by `zef` or other piece of code not controlled by test scenario ( well me ) 20:32
[Coke] Let me back up. Why are you trying to run this as this specific user? 20:33
melezhik because this I test _user_ install 20:34
[Coke] ... ok. but if the user doesn't have permissions, why do you expect it to work?
melezhik because I test user install
[Coke] I wouldn't expect a random test user to be able to install into /opt 20:35
melezhik I do not. IMHO it's bug. Install process should not try to create files in /opt/rakudo-pkg/ when gets run under user.
[Coke] then look at the line I linked earlier.
it's not a problem with zef, perse.
er, per se. 20:36
melezhik so, you mean this _should be_ overridden somewhere else?
rypervenche I came into this late. So it's not possible to install p6doc using zef as a simple user?
melezhik looks like that
[Coke] I'm saying the Build step for p6doc is explicitly creating a directory and copying things.
melezhik technically it results in zef install fails for p6doc under $user 20:37
[Coke] I am not saying whether that's right or wrong. But it explains the issue with the non-priveledged user you're seeing.
it's not a zef issue, though. zef is just doing what the lib says to.
p6doc switched to this Build *years* ago when we moved the pod from 'lib' to 'doc' folder inside the module. 20:38
It was intended to be a transparent way to get the pod to end up in the target lib directory. 20:39
melezhik yes. I knew that. another question If someone's module _use_ p6doc as a dependency will it fails when gets installed under $user?
[Coke] if they're using p6doc, it has to be installed.
so then it's still p6doc's install failing. 20:40
Would be my guess.
cpan-raku New module released to CPAN! Gnome::Gtk3 (0.22.0.2) by 03MARTIMM
melezhik yeah, here is example of fail - repo.westus.cloudapp.azure.com/raku...239499.txt 20:40
[Coke] I am still surprised you're able to install ANY modules with an unpriveledged user, but I don't know how your system is setup.
melezhik there is no surprise , because modules get installed into $user paths not system 20:41
[Coke] Looks like chunks of Build.pm go back to 2016 20:42
How are you passing those paths to zef?
melezhik I don't pass $users paths to zef it happens out of the box when you run zef under $user 20:43
[Coke] I can't read what the sparrow file is doing entirely.
When you say "under $user" can you be more specific? do you mean *as* the user?
melezhik su $user
then - zef install $stuff
yes . as the $user
[Coke] Looks like you're not using your own zef. 20:44
Are you copying it in from /opt/rakudo-pkg in that file? wonder if that's related.
melezhik no it does not. it's only the case for alpine 20:45
the same error is for debian
let me find a report for debian
rypervenche Note: Please note that, due to changes in the parsing of Pod6, this will fail in versions of Raku older than 2018.06. Please upgrade to that version, or install using --force. 20:46
melezhik repo.westus.cloudapp.azure.com/raku...244331.txt 20:47
rypervenche p6doc finally installed for me when I added --force to my zef command. 20:48
See: github.com/Raku/doc#install-p6doc
[Coke] Guessing that $*REPO in your install is still pointing at /opt/rakudo-pkg for some reason
rypervenche: I imagine the force is just ignoring the error, at which point, do you have the docs installed?
melezhik I enabled bash debug in this test, so all underlying commands are seen now
"$*REPO in your install is still pointing at /opt/rakudo-pkg for some reason" - this was my though two 20:49
rypervenche [Coke]: I have it installed now, yes. It wasn't getting past that error until I ran it with --force and it just ignored it after that and continued with the installation.
melezhik it's run under docker, not sure if it's relevant anyway ...
rypervenche Hmmm, p6doc doesn't seem to be working though. 20:54
[Coke] rypervenche: that's what I'd expect, if you force away the error, you'll get an install with no pod 20:55
Voldenet m: my \x = class { method CALL-ME { say "allright" } }.new; x()
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '&x' is not declared
at <tmp>:1
------> 3ethod CALL-ME { say "allright" } }.new; 7⏏5x()
melezhik "that's what I'd expect, if you force away the error, you'll get an install with no pod" - yeah 20:56
rypervenche That seems odd to me that there's no way to install this module as a user. I don't install things into root-owned directories outside of my package manager, so for my Raku modules I'd like to install them as my user (or via my package manager).
[Coke] trying a local install now. 20:59
melezhik under docker ideally ((((=;
Voldenet uh, there should be scopes for project/user/system 21:00
raku-modules is a cheerful name
[Coke] Do you have a dockerfile?
[Coke] under a local git checkout, did "$ zef install ." - it got to ===> Testing: p6doc:ver<1.003>:auth<perl6> and is stalled 21:02
melezhik so no luck with local install?
you can test on docker using Sparrowdo - github.com/melezhik/RakuDist#low-level-api 21:03
the scenario will be dead simple
`user $user; zef "p6doc", %( user => $user )` 21:04
rypervenche Let me update my zef and try again.
[Coke] finds github.com/Raku/doc/issues/2896 21:05
I am probably not going to install sparrow.
antoniogamiz is there a way to install a module without testing its dependencies?
(using zef) 21:06
discord6 <theangryepicbanana> add --/test
<theangryepicbanana> to the end of the command
antoniogamiz but --/test also avoids the module tests i think
[Coke] if you want to run all tests for your module but not the deps, you can install the deps first.
antoniogamiz i want to test the module, but not its dependencies
discord6 <theangryepicbanana> ah 21:07
[Coke] zef --deps-only --/test
melezhik Coke: it's up to you. But this is how one can easy repeat the test on docker
[Coke] then 'zef install <thing>'
antoniogamiz mmmm thanks coke :)
[Coke] the issue is in the Build file; someone else will have to figure out the right code to put into Build.pm, unfortunately.
melezhik: 21:08
16:06 < discord6> <theangryepicbanana> add --/test
melezhik: github.com/Raku/doc/issues/2896 - I'd follow that. I added my notes about Build and why it is that way, hopefully that helps someone figure out the fix.
rypervenche [Coke]: Thanks for that. I'll be watching it as well. 21:10
melezhik Thanks Coke: I've added my comment 21:13
rypervenche tyil: I had a failure when installing rakudo star with moarvm: gist.github.com/rypervenche/d44bba...17a8065bb2 . I can get the full logs if you need them. 21:17
uzl[m] lizmat++ for this Rakudo weekly. 21:28
.seen jmerelo
tellable6 uzl[m], I saw jmerelo 2020-01-28T18:41:48Z in #raku: <jmerelo> stoned75: I guess that you can add it wherever and increase the entropy :-)
TreyHarris lizmat++ for weekly too
uzl[m] Totally. I always look forward to reading it every week :-)! 21:29
lizmat thanks, glad to hear :-) 21:34
uzl[m] .tell jmerelo I've generated the documentation and hosted it on a Github repo (uzluisf.github.io/rakudoc/). However, all the links are relative so I'm wondering if there's way to set up a base url before generating the HTML like you could do with some static site generators (ar.al/2018/06/20/override-baseurl-...-server/). BTW, kudos for all the work with Documentable; generating the documentation couldn't be 21:36
easier now.
tellable6 uzl[m], I'll pass your message to jmerelo
uzl[m] Np, lizmat! 21:37
rypervenche tyil: Well..., I did a make clean and then make install and then it installed correctly... 21:39