»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
devz3ro is there a perl6 equivalent to grep -A1 (match + line after) 02:15
Juerd m: my @foo = ^20; say @foo[ @foo.grep(* %% 4, :k).map: { $^i .. $i + 1 } ] 02:24
camelia ((0 1) (4 5) (8 9) (12 13) (16 17))
Juerd devz3ro: Something like that?
vrurg Juerd: did you mean $^i + 1? 02:29
vrurg I mean, yours works but it probably shouldn't. :) 02:30
Ah, no, it's correct. Sorry. 02:32
devz3ro perfect, thank you 02:35
tyil perl6.org/getting-started/involved shows "not found" 06:23
rba tyil: noted 06:32
Kaiepi camelia, help 07:05
camelia Kaiepi: Usage: <(nqp-js|nqp-jvm|debug-cat|rakudo-moar|prof-m|rakudo-jvm|nqp-moarvm|star-m|p5-to-p6|rm|rakudo|r-j|nqp-m|r|p6|nqp|j|rj|sm|p56|perl6|nqp-q|master|star|r-jvm|m|r-m|nqp-mvm)(?^::\s(?!OUTPUT)) $perl6_program>
Kaiepi rakudo-jvm: use NativeCall; say size_t 07:06
camelia (size_t)
lizmat weekly: blogs.perl.org/users/damian_conway/...-work.html 07:07
notable6 lizmat, Noted!
Kaiepi rakudo-jvm: say $*VM.version 07:08
camelia v1.8
Kaiepi 1.8...? i thought the target was 1.9 07:09
how long ago was perl6-j updated on camelia? 07:10
Kaiepi rakudo-jvm: say $*VM.properties 07:12
camelia {awt.toolkit => sun.awt.X11.XToolkit, file.encoding => UTF-8, file.encoding.pkg => sun.io, file.separator => /, java.awt.graphicsenv => sun.awt.X11GraphicsEnvironment, java.awt.printerjob => sun.print.PSPrinterJob, java.class.path => :/home/camelia/rak…
Kaiepi rakudo-jvm: say $*PERL.version 07:13
camelia v6.c
robertle releasable6: status 07:14
yoleaux 2 Jul 2019 14:02Z <AlexDaniel> robertle: colabti.org/irclogger/irclogger_lo...07-02#l273
releasable6 robertle, Next release will happen when it's ready. R6 is down. At least 1 blocker. 666 out of 681 commits logged
releasable6 robertle, Details: gist.github.com/3ed6da8a322d9845c0...3e416f31b0 07:14
robertle .tell dominix Dominique Dumont or I (robertle@semistable.com) would be good fols to talk to regarding debian packaging, there is some stuff to be found starting wiki.debian.org/Teams/DebianRakudoGroup and salsa.debian.org/perl6-team 07:28
yoleaux robertle: I'll pass your message to dominix.
lizmat weekly: news.ycombinator.com/item?id=20409663 08:50
notable6 lizmat, Noted!
antoniogamiz how do you pass a variable by reference in perl6? I mean, I want to pass an array to a function and I want all changes made to that array inside the function stay the same when the function finishs 09:40
jnthn When you pass an array you pass a reference to the array, so that'll happen anywy 09:44
If you want it on a Scalar container, add `is rw`
antoniogamiz mmmm 09:52
antoniogamiz then why @refs it's empty in this case? 09:55
gist.github.com/antoniogamiz/fbf0a...36ea5d7ee7
jnthn No idea without seeing your top-level call toit 10:04
*to it
antoniogamiz my @refs = [];find-references(:$pod, :$origin, url => $origin.url :@refs);
jnthn Missing , before the :@refs?
Also, `= []` is pointless there; `my @refs` is already an empty array 10:05
antoniogamiz omg, the ,
............ Idk why I did not see it
thanks a lot!
antoniogamiz recomendations to debug this kind of error? => The iterator of this Seq is already in use/consumed by another Seq(you might solve this by adding .cache on usages of the Seq, orby assigning the Seq into an array) 10:55
sena_kun antoniogamiz, the error message already says what should be done, and points exact line where the used seq is. 11:04
antoniogamiz but it points to a function definition
sena_kun antoniogamiz, can you show the code? 11:05
antoniogamiz mmm code is quite long and I cannot simplify the situation x'd 11:17
I will try a little more and if I cannot solve it I will try to make a reduced example
lizmat are per chance binding a Seq to an array or so ? 11:21
antoniogamiz lizmat: yes! but I think I have found the problem 11:27
lizmat ok, good to hear
antoniogamiz the error is thrown when I consult a specific key of a hash
rfold p6: sub f(*@args) { @args.perl.say }; f: «foo bar»; 11:32
camelia WARNINGS for <tmp>:
Useless use of constant value foo bar in sink context (lines 1, 1, 1, 1)
lizmat rfold: you're creating a label called "f" there 11:33
rfold Ah, I see. 11:34
lizmat m: f: «foo bar»;
camelia WARNINGS for <tmp>:
Useless use of constant value foo bar in sink context (lines 1, 1, 1, 1)
lizmat : only works on methods
if the sub is known, you can omit parens
m: sub f(*@args) { @args.perl.say }; f «foo bar»;
camelia ["foo", "bar"]
rfold Aha, thanks, that makes sense!
El_Che this may intereset some folks here: medium.com/darklang/the-design-of-...f5d38e52d2 13:01
jnthn Read it yesterday; yes, it's an interesting perspective indeed. 13:04
moritz note that gitlab takes a similar approach 13:05
it started with just version control, then added issues / project management, CI/CD, and I think chat and editor/IDE support as well (or might be upcoming) 13:06
El_Che moritz: yeah, but is is ambition on an other level 13:12
moritz El_Che: agreed
El_Che wishful thinking of course, but a concurrent and flexible lang as p6 could come in handy autobuilding the backend glue code 13:13
Go's static nature makes is less flexible
python concurrency is a no-no
tadzik I wish people stopped using Medium :( 13:16
timotimo i'd love it if people brought back webrings with a vengeance 13:18
robertle I heard an interesting talk from Stroustrup, in the discussion afterwards the question about C++'s choice of a linker came up. it uses the C linker, which obviously has lot's of drawbacks for C++, like no namespaces and symbols do not understand types, and from there comes all the weird mangling and I guess something towards template complications.
strupstrup comment was that even if that was true in hindsight, it would have been impossible to build a linker *and* a new language at the time, so there you are. but also that if you do everything "batteries included", you end up with one monolithic thing that does all, and you probably do not have a community to support that. do you believe your compiler writers are also great at writing standard 13:20
libraries
pmurias tadzik: isn't having a site that consists almost entries of dubious quality blog posts convinient to keep them in one place?
the graalvm is the only decent thing I ever saw on medium 13:21
robertle I came away with the conclusion that it makes a lot more sense to do one thing and do it well, and then some other people who are better at it can compete, discuss and build a better make tool, library, repository system etc
tadzik pmurias: well, there is actually interesting stuff there, but it's confusingly mixed with attempts at viral marketing
robertle this dark language seems to be the exact opposite: reinvent the whole world in one big go
tadzik which is probably no different to most of the internet these days, but not all content publishing websites are hostile to your readers 13:22
El_Che tadzik: agree about medium 13:23
tadzik fortunately it looks mostly ok in Reader Mode :) 13:24
El_Che robertle: not reinvent, actually. It's pretty close to what Spring Boot does for Java People today
It abstract a lot of the moving pieces
s 13:25
lucs Hmm... 'zef install p6doc' wants to write to root owned "/usr/share/perl6..." :(
El_Che anyway, I think the monolith approach risk crashing against a basic assumption while the world moved on
e.g. concurrency and scripting languages
not easy to di
fix 13:26
lucs: is it a zef installed as root of a local one in your home?
lucs El_Che: I think it's a root-installed zef. Note that I've been installing other modules in $HOME/.zef without problem. 13:27
Can I fix this? 13:28
El_Che I have seen this before, that's why my pkgs provide a root zef and a script to install zef as a user
which zef
see which zef you're running 13:29
lucs Um, not sure how to find out
(I don't see a 'version' option or something)
renormalist Isn't Dark the redo of Smalltalk, just in a different century? 13:30
El_Che lucs: what OS? 13:31
renormalist: pretty much
lucs Oh, 'zef --version' works: v0.7.1 13:32
pmurias El_Che: is there an example of what the dark language is somewhere?
El_Che I mean which script 13:32
ugexe m: say first { .can-install() }, map { CompUnit::RepositoryRegistry.repository-for-name($_) }, <site home>;
camelia Nil
ugexe that will tell you which repository zef will try to install to with a default config
lucs ugexe: Um, what would <site home> be like? 13:33
pmurias El_Che: the medium post seems to be just a wall of marketing text
ugexe run it and see 13:34
lucs Oh, I though
Oh, I thought it was a place holder :)
inst#/home/lucs/.perl6 13:35
ugexe strange then that it is trying to install to the root owned directory
lucs More details then...
ugexe did you install zef from source or another way?
lucs Checking... 13:36
From a Slackbuild: slackbuilds.org/repository/14.2/perl/zef 13:37
ugexe do `zef --help`, towards the bottom it will tell you the full path of the configuration file. open that file and look at the ~5th entry called "DefaultCUR" -- what value is this field?
El_Che pmurias: my adblocker (ublock origin) + pihole blocks that shit for me
lucs "DefaultCUR" : ["auto"], 13:38
ugexe yeah that should be doing the right thing then :/
lucs :(
Let me get more details... 13:39
ugexe fwiw you should be able to force a workaround via `--install-to=home` 13:41
lucs ugexe: I will paste to you in /msg
I'll try that.
ugexe ok I understand now 13:42
p6doc Build.pm seems flawed
github.com/perl6/doc/blob/41c47393...m6#L11-L12 13:43
lucs (same result with `--install-to=home`)
Oho.
ugexe it assumes it can write to any repository that contains perl6 modules
El_Che code hidden in a private alpha it seems
ugexe lucs: you will want to file an issue with p6doc that they cannot assume they can create a /doc directory inside any perl6 repository in their Build.pm 13:45
lucs Could I mention where this directory should have been created, or will they figure that out? 13:46
ugexe its a little more complicated than just change the directory 13:50
lucs Yeah, I guess.
ugexe i.e. if you just install to $*HOME then the docs generated by root are not available 13:51
lucs :/
ugexe however it probably should not be installing docs under a perl6 repo directory 13:51
that would be ok if docs were installed along with the module, but after-the-fact 13:52
but not after-the-fact 13:53
lucs ugexe: I will paste parts of this conversation in the issue I will open. Are you okay with that?
ugexe sure 13:54
lucs Okay, thanks.
Geth doc: f5e2d6d917 | Coke++ | xt/words.pws
learn new word
 14:11
AlexDaniel El_Che: any idea how can I use github.com/perl6/rakudo-pkg-canary without the regular prerelease upload location? 14:54
El_Che AlexDaniel: I can change the wget by git clone of a tag? 15:11
AlexDaniel El_Che: but there's no tag :) 15:12
El_Che branch?
or better: what do you have
?
AlexDaniel El_Che: I'm not sure 15:13
:)
El_Che AlexDaniel: we need something :) 15:13
otherwide we need to add some AI code :) 15:14
El_Che AlexDaniel: in the meantime, I'll add it to my TODO list to add new distro releases and remove the EOL'ed 15:19
AlexDaniel El_Che: I mean, it almost sounds like we need a release branch :) 15:21
like, one that we wouldn't delete after the relese :) 15:22
a
El_Che or tag release candidates? 15:23
v2019.07.2-RC1
AlexDaniel El_Che: ye but then the link will be different all the time, no?
El_Che or Tag "canary" 15:24
AlexDaniel El_Che: and with the release branch, we can configure the same travis thingy you have on github.com/perl6/rakudo-pkg-canary, but right in the rakudo repo
El_Che and move the tag around (not done :), I know)
but lighter than branching
AlexDaniel ok let's try it… hold on… 15:24
El_Che rakudo-pkg is building every release, maybe a cron job with a extra param is an option? 15:25
(just thinking out loud) 15:26
uzl Hello, everyone.
yoleaux 4 Jul 2019 15:23Z <jmerelo> uzl: you're probably right...
antoniogamiz uzl: o/
El_Che uzl: and you're probably right about something
uzl El_Che: 😃
El_Che :)
uzl antoniogamiz: hey!
m: my @a = letter => 'a', letter => 'b'; @a.hash.put 15:27
camelia letter b
uzl How would I create a hash from @a without using a for loop?
sena_kun uzl, how are elements stored? 15:29
m: my @a = 'a', 2; @a.Hash.say;
camelia {a => 2}
sena_kun m: my @a = 'a' => 2, 'b' => 3; @a.Hash.say;
camelia {a => 2, b => 3}
sena_kun m: my @a = 'a', 2, 'b', 3; @a.Hash.say;
camelia {a => 2, b => 3}
sena_kun uzl ^ ?
uzl sena_kun: hold on a sec. 15:31
AlexDaniel there's also classify and categorize, it could be that you need that
AlexDaniel El_Che: ok, so if we have this github.com/rakudo/rakudo/tree/prerelease 15:33
El_Che: there's no way to configure travis specifically for that branch?
uzl m: my @a = a = ["test-walk/contents" => "file1.md", "test-walk/contents" => "file3.md", "test-walk/contents/blog" => "blog4.md", "test-walk/contents/blog" => "blog2.md"]; 15:34
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix = instead.
Did you make a mistake in Pod syntax?
at <tmp>:1
------> 3my @a = a =7⏏5 ["test-walk/contents" => "file1.md", "t
uzl sena_kun: that's a simplified version. 15:34
sena_kun `@a = a = [...]`?
uzl m: my @a = ["test-walk/contents" => "file1.md", "test-walk/contents" => "file3.md", "test-walk/contents/blog" => "blog4.md", "test-walk/contents/blog" => "blog2.md"]; 15:35
camelia ( no output )
sena_kun just .Hash it
uzl That one.
sena_kun ah, stop
.classify?
uzl Yeah, but then I loss some of the values except one.
m: my @a = ["test-walk/contents" => "file1.md", "test-walk/contents" => "file3.md", "test-walk/contents/blog" => "blog4.md", "test-walk/contents/blog" => "blog2.md"]; @a.hash.say 15:36
camelia {test-walk/contents => file3.md, test-walk/contents/blog => blog2.md}
»
uzl I'd like the directory key to map to an array of the files on it.
sena_kun I remember a trick to collect those, but not exact lines. /afk
uzl m: my @a = ["test-walk/contents" => "file1.md", "test-walk/contents" => "file3.md", "test-walk/contents/blog" => "blog4.md", "test-walk/contents/blog" => "blog2.md"]; my %h; for @a { %h{$_.key}.push: $_.value }; say %h 15:37
camelia {test-walk/contents => [file1.md file3.md], test-walk/contents/blog => [blog4.md blog2.md]}
»
uzl I can do that but I was wondering if there was possibly a more straightforward way.
uzl looks into classify. 15:39
antoniogamiz uzl: in perl6 there's always more than one way :DDDD 15:40
jnthn m: my @a = ["test-walk/contents" => "file1.md", "test-walk/contents" => "file3.md", "test-walk/contents/blog" => "blog4.md", "test-walk/contents/blog" => "blog2.md"]; my %h.push(@a); say %h; 15:42
camelia {test-walk/contents => [file1.md file3.md], test-walk/contents/blog => [blog4.md blog2.md]}
»
uzl jnthn: That will do the trick ;-)! Didn't thought it'd be so simple. Thanks! 15:44
antoniogamiz: I know. That's why I like P6! 15:45
TreyHarris .tell grondilu looking back in logs, that was sent in March--you were looking for syntax on how to specify a constraint clause where the Whatevers referred to a list's contents being numeric in various ways instead of the list itself as a unit (which is obviously always numeric, but in very different ways to its possibly-numeric contents). Someone suggested .ACCEPTS(Array) as a decent way of getting at list-y 16:46
yoleaux TreyHarris: I'll pass your message to grondilu.
TreyHarris structures, so I cobbled that together 16:47
Geth doc: 7925341365 | (Tim Van den Langenbergh)++ | doc/Language/regexes.pod6
Add the `S///` section from operators to regexes
 17:20
doc: 84a44ceb2a | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | doc/Language/regexes.pod6
Merge pull request #2894 from tmtvl/2892-add-non-destructive-substitution-to-regex-documentation

Add the `S///` section from operators to regexes
synopsebot Link: doc.perl6.org/language/regexes
doc: JJ assigned to noisegul Issue Can't install p6doc as a non-privileged user, github.com/perl6/doc/issues/2896
0b0a01fc4e | (Tom Browder)++ (committed using GitHub Web editor) | doc/Language/unicode.pod6
antoniogamiz What do you think of this? => github.com/antoniogamiz/Perl6-Documentable 18:08
antoniogamiz recommendations? suggestions? any feedback is welcomed 18:08
it's a fix for github.com/perl6/doc/issues/1823 18:09
ugexe improve CI testing 18:12
look at how zef tests the various cli commands -- github.com/ugexe/zef/blob/master/....is.yml#L60
which also act as examples of use 18:13
antoniogamiz mmm currently it does not have any CI tests so I will add them :D 18:15
so you execute every command to see if it fails, right?
ugexe right
antoniogamiz good, thanks for that :D 18:16
ugexe you probably should write subtests like this 18:17
subtest 'Foo bar' => { }
instead of subtest { }, 'Foo bar';
antoniogamiz oh that way is more readable isn't it? 18:18
ugexe exactly
lib/Perl6/Documentable.pm6 uses Pod::Utilites, but it is only listed under test-depends, not depends 18:21
antoniogamiz oh right, I will fix it right now
antoniogamiz do you think I should shorten the modules names stripping out Perl6::? 18:24
ugexe i think its fine with Perl6
you list `resources/language-order-control.json` in resources section of META6.json, but it should just be `language-order-control.json`
i'm guessing you havent tried to install this yet? 18:25
antoniogamiz eemm you are right
ugexe there is a `unit class Perl6::Documentable:ver<0.0.1>;` but the META6.json says it is version 1.0.0 18:26
as part of your CI test you will want to actually install your module (not just run the tests)
antoniogamiz I will take that in mind! 18:27
antoniogamiz ugexe: after fix a lot of problems now I have learned that I really have to add installing the module to CI 18:48
thanks a lot for taking the time and help me! :D 18:49
ugexe have fun progressing
El_Che there was less perl drama when Twitter was down. 20:45
El_Che closes Twitter
synopsebot Link: doc.perl6.org/language/unicode
wildtrees how can I erase or redefine a class, say at the repl? 21:14
Xliff \o 21:34
m: %a = ( a => 1, b =>2, c => 3)
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '%a' is not declared
at <tmp>:1
------> 3<BOL>7⏏5%a = ( a => 1, b =>2, c => 3)
Xliff m: my %a = ( a => 1, b =>2, c => 3 ); %a.values »*=« 5; %a.gist.say
camelia Lists on either side of non-dwimmy hyperop of infix:<*> + {assigning} are not of the same length while recursing
left: 3 elements, right: 1 elements
in block <unit> at <tmp> line 1
Xliff m: my %a = ( a => 1, b =>2, c => 3 ); %a.values «*=» 5; %a.gist.say 21:35
camelia {a => 5, b => 10, c => 15}
Xliff I love hyper-ops
m: my %a = ( a => 1, b =>2, c => 3 ); %a.values.map( * * 5 ).say 21:36
camelia (15 5 10)
Xliff m: my %a = ( a => 1, b =>2, c => 3 ); %a.values.map( * * 5 ).sum.say 21:37
camelia 30
Xliff Is there a more readable way to do this: ^^
m: my %a = ( a => 1, b =>2, c => 3 ); %a.values.map({ $_ * 5 }).sum.say # may have to do 21:38
camelia 30
Xliff Yeah.
Xliff m: my $i; sub wtf { if $i++ > 10 { say 'H!'; exit }; for ^100 { wtf }; say "WTF?" 21:51
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing block
at <tmp>:1
------> 3!'; exit }; for ^100 { wtf }; say "WTF?"7⏏5<EOL>
expecting any of:
postfix
statement end
statement modifier
stat…
Xliff m: my $i; sub wtf { if $i++ > 10 { say 'H!'; exit }; }; for ^100 { wtf }; say "WTF?"; 21:52
camelia H!
Xliff m: my $isItAMatterOfLength; sub wtf { if $isItMatterOfLength++ > 10 { say 'H!'; exit }; }; for ^100 { wtf }; say "WTF?";
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$isItMatterOfLength' is not declared. Did you mean '$isItAMatterOfLength'?
at <tmp>:1
------> 3my $isItAMatterOfLength; sub wtf { if 7⏏5$isItMatterOfLength++ > 10 { say 'H!'; e
Xliff m: my $isItAMatterOfLength; sub wtf { if $isItAMatterOfLength++ > 10 { say 'H!'; exit }; }; for ^100 { wtf }; say "WTF?"; 21:53
camelia H!
Xliff m: $*PERL.compiler.version.say
camelia v2019.03.1.683.g.20.e.74837.a
sena_kun was rakudo released? I see some failures on HEAD with Cro::LDAP, now checking it again, might be a blocker. 21:54
sena_kun can someone on release branch test it out? 21:56
Geth ¦ doc: tbrowder self-assigned Need section on bracketing constructs github.com/perl6/doc/issues/2897 21:57
wildtrees how can I erase or redefine a class, say at the repl? (reposting ) 22:02
jnthn Declare it lexically (my class Foo { }) 22:04
wildtrees cool, thanks jnthn 22:18
Xliff m: 1000.exp(5).say 22:20
camelia 93326361850321887899008954472381716961709144637170802462171433979596691097577563445444032709788110235959498993032424262421548752135403239484152081720393075623441066613832515027399507598590183151110049079626511311824051251479593379080517827112541510381…
Xliff m: 1000.log.say 22:21
camelia 6.907755278982137
Xliff m: 1000.log(5).say
camelia 4.29202967422018
Xliff m: 5 =~= 5.1 22:23
camelia ( no output )
Xliff m: say 5 =~= 5.1
camelia False
Xliff m: say 5 =~= 5 + 1e-10
camelia False
Xliff m: say 5 =~= 5 + 1e-15
camelia True
Xliff m: say 5 =~= 5 + 1e-14
camelia False
ugexe sena_kun: I see regression between 2019.03.1 and blead. although i dont know about the release itself 22:40
'===SORRY!=== Missing serialize REPR function for REPR NativeRef (StrLexRef)' for each test 22:41
sena_kun ugexe, sigh, I saw the same error on travis. 22:42
I wonder if I get killed after creating a blocker ticket... 22:43
ugexe which module uses NativeCall in that dependency chain?
sena_kun ugexe, only OpenSSL, I believe.
also SSL::Async.
ugexe er, i dont know if NativeRef is nativecall or not
its not 22:44
nqp/ops.markdown shows it alot for the atomics/cas 22:45
sena_kun: `perl6 -I. -e 'use Cro::LDAP::Client;'` 22:55
thats enough to trigger it 22:56
sena_kun ugexe, on bleed?
Xliff nfin!sole
ugexe yes 22:57
use Cro::LDAP::Types
thats better, as that module only uses 3 other modules (all of which can be used fine) 22:58
sena_kun nooooooooo
Cro::LDAP::Types has ASN::META behind it and it contains the most black magic.
ugexe looks like `use ASN::META <file docs/ldap.asn>;` is the line that does it (but only if used inside a module, not when dong via -e 'use use ASN::META <file docs/ldap.asn>;') 22:59
sena_kun github.com/rakudo/rakudo/issues/3045 23:01
ugexe im not sure how that will work 23:04
that docs/ldap.asn
that means you have to run things from a specific directory 23:05
sena_kun yes
from cro-ldap
apparently
sena_kun it relies on zef being run from there and it does, as it worked before just fine 23:06
though it is bad
and ugly too
ugexe yeah, and its unrelated to this. but i mean when its installed
that wont work
sena_kun ugexe, can module be installed and then re-precompiled?
without sources?
during compilation I basically parse it, build AST, then generate types, export them, finita la comedia. 23:07
so it only happens once.
AlexDaniel *heavy breathing* 23:08
sena_kun OTOH I am not aware of guts, so there is probably a more robust way of passing the file instead of this relative filename. :/
AlexDaniel, o/
AlexDaniel, feel free to pretend my ticket doesn't exist 23:09
AlexDaniel sena_kun: sooo… `git bisect` ?
sena_kun AlexDaniel, it fails on release branch too, not just master?
ugexe bisect `use Cro::LDAP::Types`
AlexDaniel sena_kun: I don't know! But any reason why you can't run `git bisect` to figure it out? :) 23:10
sena_kun AlexDaniel, I don't have rakudo development stuff to git bisect it. :/ 23:11
inb4: so just get it
AlexDaniel let's see if I can get the answer reeeally quick with blin 23:15
sena_kun AlexDaniel, can Blin work with modules not from ecosystem? 23:16
AlexDaniel sena_kun: it can work with any arbitrary script, yeah
in this case I'm doing something like: use lib ‘/home/alex/git/Blin/cro-ldap’; use Cro::LDAP::Types; 23:17
it should work I guess
and these are the deps: OO::Monitors Text::LDIF Cro::TLS Cro::Core IO::Socket::Async::SSL ASN::BER ASN::META ASN::Grammar OpenSSL Type::EnumHOW
so it's like 23:18
sena_kun the deps are likely safe, I think it's a regression for ASN::META.
AlexDaniel bin/blin.p6 --old=2019.03.1 --new=HEAD --custom-script=doodle.p6 OO::Monitors Text::LDIF Cro::TLS Cro::Core IO::Socket::Async::SSL ASN::BER ASN::META ASN::Grammar OpenSSL Type::EnumHOW
Xliff AlexDaniel: see ugexe's comment.
sena_kun you need to use it with a file
AlexDaniel ooo… oh? 23:19
sena_kun AlexDaniel, `use Cro::LDAP::Types <file /home/alex/git/Blin/cro-ldap/docs/ldap.asn>;`, I think
ugexe no, use Cro::LDAP::Types is enough 23:20
that will call ASN::META <file ...> 23:21
which needs to be called from a module to get the same error
sena_kun ah, sorry
it seems I am brain dead from lack of sleep already...
AlexDaniel OK I had a path problem, trying again… 23:25
(to ldap.asn file that is) 23:26
sena_kun I think you need cd, to do it from `cro-lda` directory 23:27
AlexDaniel 🥞🥞🥞 Bisecting doodle.p6 23:29
okay!
gist.github.com/AlexDaniel/602d4af...d336533172 23:34
sena_kun, ugexe: ↑ 23:35
nine: :)
Blin++ 23:36
sena_kun AlexDaniel++ 23:39
AlexDaniel and yeah, it's on the release branch too, so please take a look while I'm doing the release stuff at turtle speeds 23:40
sena_kun well, I know there is a lot of hard work around this release, so feel very, ahem, not too nice reporting this. :( I'll present Cro::LDAP at Perlcon, so we won't be able to cut a new release before that. Though I still can say that there was a regression and 03 is ok. 23:42
anyway, I am going to sleep to keep some sanity. o/
Xliff Yikes! 23:45
That was done to support my use of Method::Also on various projects. 23:46
ugexe i'd prefer to not mess up the talk if possible 23:51
AlexDaniel ah re “the deps are likely safe”, you *have* to tell Blin about these modules, otherwise they won't be installed 23:56
and I believe Blin won't use anything installed on your system

Powered by Cro and the Raku® Programming Language.

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

Thank you!