Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2016-02-07

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
07:16 ilbot3 joined #perl6
07:16 Topic for #perl6 is now »ö« Welcome to Perl 6! | https://perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:,  or /msg camelia p6: ... | irclog: http://irc.perl6.org or http://colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
07:27 kaare_ joined #perl6
07:35 moritz \o
07:38 masak o/
07:46 AlexDaniel I see that there were no commits to roast for 21 days. Is there any special reason for that or is it just that nobody contributed any tests?
08:01 darutoko joined #perl6
08:12 masak m: class A::B {}; class C { has A::B $.ab = A::B.new }; say "alive"
08:12 camelia rakudo-moar 58d339: OUTPUT«alive␤»
08:13 masak m: class A::B {}; class C { has A::B $.ab .= new }
08:13 camelia rakudo-moar 58d339: OUTPUT«===SORRY!===␤Could not locate compile-time value for symbol A::B␤»
08:13 masak (RT #126975)
08:14 musiKk joined #perl6
08:16 maybekoo5 joined #perl6
08:17 xinming_ joined #perl6
08:36 firstdayonthejob joined #perl6
08:39 Gentoolman joined #perl6
08:40 Gentoolman Is MoarVM realization better then ParrotVM?
08:48 perl6newbee joined #perl6
08:50 perl6newbee hi all
08:50 NZKindest joined #perl6
08:52 Gentoolman Hi
08:52 NZKindest Hello fellow traveller
08:52 perl6newbee hehe
08:52 Gentoolman Which VM is better? Parrot or Moar?
08:53 perl6newbee afaik is the spec support in MOar better
08:53 rindolf joined #perl6
08:53 bartolin MoarVM. support for Parrot has been suspended about a year ago
08:54 bartolin and it seems unlikely that Parrot will be resurrected in the future
08:54 Gentoolman For perl6 on MoarVM i need to install Rakudo perl and then MoarVM and configure rakudo perl tp use MoarVM ?
08:56 bartolin Gentoolman: best thing would probably be to use Rakudo Star (which includes MoarVM and some modules) http://rakudo.org/how-to-get-rakudo/
08:57 masak Gentoolman: hi. maybe you're a speaker of a slavic language? usually in English, it's "implementation", not "realization".
08:57 Gentoolman masak: yes, i am russian.
08:57 masak \o/
08:58 masak Gentoolman: I speak Esperanto, where the word is "realigo", so I'm aware of the false cognate ;)
08:58 Gentoolman masak: Esperanto is the language of the future. :)
08:58 masak ...and always will be :P
09:00 Gentoolman It seems i found rakudo port in Portage. Thanks.
09:01 masak just watch out for package managers sometimes carrying vary old versions
09:01 masak very*
09:01 masak (but yes, it'll vary, too) :)
09:03 perl6newbee masak: you surely know how can I adopt the IO::Prompt repo in mu/misc/shelter?
09:04 masak perl6newbee: hm. good question.
09:05 masak surely there must be instructions somewhere?
09:05 perl6newbee In the shelter Readme stands Aks us on IRC on how to adopt a module :-)
09:06 masak a good start though seems to be to clone the git repo, set up your own github repo for the module, and then aim for excellence with such intensity that people just accept your version as authoritative
09:06 masak :)
09:06 perl6newbee lol.
09:06 masak perl6newbee: I suspect it is that simple, actually
09:06 masak we just re-point the ecosystem files to your repo after some time
09:06 perl6newbee I already fork the repo and managent to get it running with 6.c. the tests passes
09:07 perl6newbee then I will add the META6.json raw link to the perl6 ecosystem like zoffix explained in his blog post
09:07 perl6newbee thx
09:08 masak yes, I think that's it
09:08 perl6newbee :-)
09:08 masak if you make a pull request to https://github.com/perl6/eco​system/blob/master/META.list
09:08 masak I will merge it
09:08 perl6newbee w00t
09:08 masak or if you have write access already, just go ahead and make the commit yerself
09:09 perl6newbee unfortunately, I am still a newbie no change to get a perl6 guru like you are
09:10 perl6newbee I will see what happens with IO::Promp in the wild.
09:12 Gentoolman Hi guys. I found that last perl6 from takudo has build problems
09:12 Gentoolman :(
09:13 perl6newbee http://rakudo.org/downloads/st​ar/rakudo-star-2016.01.tar.gz did a gread job for me
09:14 perl6newbee I mean, extract the tar write "perl Configure.pl --gen-moar" and make a make; make install; did the job
09:15 Gentoolman No, guy, not Gentoo-way. :)
09:15 perl6newbee haha
09:15 Gentoolman If there some sources i need to write a ebuild
09:15 perl6newbee sometimes, you have to take a different way ^^
09:16 Gentoolman Install sources with manual compiling instead automatic ports-like package manager is strange.
09:17 perl6newbee maybe. you can wait until the gentoo package maintainer updated the package or you do itself
09:17 Laurent_R joined #perl6
09:20 perl6newbee btw IMHO installing software with compiling is the linux way. Package manager is just a wrapper about it
09:20 Gentoolman perl6newbee: it seems the problem in sources, not i  package. Last update was at 3 feb
09:21 Gentoolman perl6newbee: portage like ports in FreeBSD. Portage in gentoo compile everything from build scripts -- ebuilds. It is awesome
09:22 RabidGravy joined #perl6
09:23 abaugher_ joined #perl6
09:24 perl6newbee francly, I never installed Gentoo. I only know Fedora and Arch
09:26 Gentoolman perl6newbee: gentoo is like "arch where yaourt is only way to install packages"
09:27 perl6newbee Gentoolman: I installed Perl6 with the sources without problems on Fedora as well as Arch :-)
09:27 Gentoolman perl6newbee: Dark side u r going to.
09:28 perl6newbee Gentoolman: Dark Side has cookies :-)
09:38 frederico hello, I'm learning some perl6 , I would like to call a sub by it full namespace name, 'Foo::bar::sub()'
09:38 frederico but my small example is not working
09:38 frederico http://pastebin.com/Du5DUJQ2
09:40 perl6newbee You use a '.'. That only works if you have a object of a class I think. So, FullName::qualy; should do the trick
09:41 cognominal joined #perl6
09:42 frederico perl6newbee: with FullName::Qualy the error moves to Could not find symbol '&qualy'
09:43 perl6newbee hmm
09:44 dolmen joined #perl6
09:50 perl6newbee frederico: sorry. With my limitied knowledge I cannot help you :-(
09:51 frederico perl6newbee: :), fine.
09:54 musiKk joined #perl6
09:55 RabidGravy frederico, by default subs are defined with 'my' scope. That is to say you can't see them from outside the scope they are defined in. (In this case your Fullname module)
09:55 perl6newbee off course: with our sub query it works
09:56 RabidGravy to be able to access them fully qualified you must define them as 'our' scoped:
09:56 perl6newbee facebalm
09:56 * flussence drops a friendly reminder that a Gentoo overlay for Rakudo already exists (https://gitlab.com/flussence/perl6-overlay)
09:56 RabidGravy m: module Foo { our sub foo() { say "foo" } }; Foo::foo()
09:56 camelia rakudo-moar 58d339: OUTPUT«foo␤»
09:56 frederico cool, our sub qualy worked
09:56 RabidGravy m: module Foo { sub foo() { say "foo" } }; Foo::foo()
09:56 camelia rakudo-moar 58d339: OUTPUT«Could not find symbol '&foo'␤  in block <unit> at /tmp/3MAntiSmId line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/3MAntiSmId line 1␤␤»
09:58 RabidGravy you can however "export" the 'my' subs with "is export" so they become symbols in the outer scope
09:58 perl6newbee RabidGravy: Why does a "sub foo() is exported" does not change the error.
09:58 RabidGravy m: module Foo { sub foo() is export { say "foo" } }; import foo;  foo()
09:58 camelia rakudo-moar 58d339: OUTPUT«5===SORRY!5=== Error while compiling /tmp/43UCK4Mczj␤Could not find module foo to import symbols from␤at /tmp/43UCK4Mczj:1␤------> 3() is export { say "foo" } }; import foo7⏏5;  foo()␤»
09:58 RabidGravy m: module Foo { sub foo() is export { say "foo" } }; import Foo;  foo()
09:58 camelia rakudo-moar 58d339: OUTPUT«foo␤»
09:59 RabidGravy perl6newbee, It's still a "my scoped" sub, it gets added to the outer scope with the unqualified name as it doesn't have a qualified name in that sense
10:00 JRaspass joined #perl6
10:00 perl6newbee the difference is you used "import Foo". with "use Foo;" it does not work.
10:00 perl6newbee That was my problem
10:01 RabidGravy the import Foo there is because it's "in the same file", of course if it was in a separate file you would use "use" (which implies an import)
10:02 perl6newbee cool. learned something new
10:05 perl6newbee RabidGravy: Foo::foo() does only works when foo is our, it does not work if foo is in the my-scope with the exported trait, right?
10:05 RabidGravy yes
10:05 perl6newbee thx
10:06 RabidGravy other methods of exporting are available
10:06 perlawhirl joined #perl6
10:07 perl6newbee :-) I am happy to have understand this one at the moment ggg
10:07 perl6newbee but maybe you can help me with another problem.
10:07 RabidGravy at it's simplest what happens when the compiler sees the "is export" it saves the name of the thing and the thing itself (in this case the subroutine) and then when you do "import Foo" it gets all those saved thingies and inserts them in the current scope
10:09 perl6newbee sounds easy and familiar. Read it already, but never need this information
10:11 abraxxa joined #perl6
10:15 perl6newbee anyway. I have problems to debug a script that imports a class from another script. The class in question is "unit class IO::Prompt". Start up the debbuger with a test.p6 script the IO::Prompt works but the code in the seperated file is never seen. If I create a bp with bp add Prompt.pm:13, the debugger says unknown file. I tried several strings: /lib/IO/Prompt.pm:14 or IO::Prompt:14 nothing worked
10:16 perl6newbee What is the problem here? or where is the problem
10:21 perlawhirl joined #perl6
10:22 RabidGravy this is a problem with the debugger, I'm guessing here that it hasn't kept up with the way that installed modules are stored.
10:23 perl6newbee Is there a walkthrough?
10:23 perl6newbee walkaround
10:23 RabidGravy of?
10:33 wamba joined #perl6
10:38 virtualsue joined #perl6
10:44 perl6newbee of the problem with the debugger
10:49 abraxxa is there a message bot around?
10:49 Hotkeys m: my &f = {sum(@_)/2 ∈ @_}; say f([5,3,2])
10:49 camelia rakudo-moar 58d339: OUTPUT«False␤»
10:49 Hotkeys m: my &f = {(sum(@_)/2) ∈ @_}; say f([5,3,2])
10:49 camelia rakudo-moar 58d339: OUTPUT«False␤»
10:50 Hotkeys m: say sum([5,3,2])/2
10:50 camelia rakudo-moar 58d339: OUTPUT«5␤»
10:50 Hotkeys m: say 5 (elem) [5,3,2]
10:50 camelia rakudo-moar 58d339: OUTPUT«True␤»
10:51 Hotkeys m: my &f = {sum(@_)/2 (elem) @_}; say f([5,3,2])
10:51 camelia rakudo-moar 58d339: OUTPUT«False␤»
10:51 Hotkeys Why isn't this working
10:51 Hotkeys Am I doing something dumb?
10:51 RabidGravy abraxxa, yeah yoleaux is alive and kicking
10:52 RabidGravy perl6newbee, I don't think so
10:52 abraxxa RabidGravy: found the syntax in my irc logs, thanks
10:53 Hotkeys m: my &f = {sum(@_)/2 ∈ @_}; say f((5,3,2))
10:53 camelia rakudo-moar 58d339: OUTPUT«False␤»
10:54 RabidGravy HotKeys, not quite sure what you're trying to do
10:54 Hotkeys RabidGravy: trying to figure out why the function doesn't work but outside the function this works
10:54 Hotkeys Eg
10:55 RabidGravy er quite, but what is the function trying to do
10:55 Hotkeys m: say sum([5,3,2])/2 (elem) [5,3,2]
10:55 camelia rakudo-moar 58d339: OUTPUT«False␤»
10:55 Hotkeys Well theb
10:55 espadrine joined #perl6
10:55 Hotkeys It's for a golf problem
10:56 Hotkeys Checks if the difference any two of the numbers in an array of 3 ints results in the third
10:56 Hotkeys eg in that one 5 - 3 = 2
10:56 Hotkeys So it is true
10:57 abraxxa .tell Skarsnik I've rebased my DBIish pull-request, the Travis setup seems to work now (finds the dependent modules), the fails are the same as on my dev box and subject to API discussion
10:57 yoleaux abraxxa: I'll pass your message to Skarsnik.
10:57 Hotkeys m: my &f = {say sum(@_)/2; say @_; sum(@_)/2 ∈ @_}; say f([5,3,2])
10:57 camelia rakudo-moar 58d339: OUTPUT«5␤[5 3 2]␤False␤»
10:58 Hotkeys Why isn't this resulting as true?
10:58 Hotkeys It seems like it should
10:59 Hotkeys m: say {@_ ∋ sum(@_)/2}([5,3,2])
10:59 camelia rakudo-moar 58d339: OUTPUT«False␤»
10:59 masak m: my &f = {say sum(@_)/2; say @_; (sum(@_)/2) ∈ @_}; say f([5,3,2])
10:59 camelia rakudo-moar 58d339: OUTPUT«5␤[5 3 2]␤False␤»
11:00 abraxxa have a nice Sunday guys!
11:01 Hotkeys m: say {any(@_) == sum(@_)/2}([5,3,2])
11:01 camelia rakudo-moar 58d339: OUTPUT«any(True, False, False)␤»
11:01 Hotkeys Hmm
11:02 Hotkeys So why isn't the element op working
11:06 rindolf joined #perl6
11:09 Hotkeys Figured it out
11:09 Hotkeys Division returns a rat
11:10 Hotkeys That list is integers
11:10 Hotkeys I'm a bit disappointed (elem) doesn't let you do that
11:11 ely-se joined #perl6
11:15 stmuk_ what terminal do the kewl kidz instead of cmd.exe?
11:15 stmuk_ ^ use
11:17 flussence .oO( probably one of those Mac ones that emulate CRT artifacts and make clicky sounds as you type... )
11:17 Hotkeys I use powershell
11:17 Hotkeys It's moderately better than cmd
11:18 Hotkeys And built in
11:18 stmuk_ I'll try it
11:18 ely-se I use Sakura on Gentoo
11:18 ely-se On Windows I use ConsoleZ
11:19 ely-se with PowerShell as shell
11:20 flussence I'm using `st` because it supports unicode and doesn't consume web-browser amounts of RAM, though I find its pedantic adherence to standards a little braindamaged (have to type ^D for delete)
11:21 flussence not a windows thing, though, so ignore me :)
11:22 Hotkeys When I run linoox I usually run urxvt
11:23 flussence the problem with urxvt is it advertises itself as unicode, but is really ucs-2
11:35 RabidGravy ucs-2 is almost ubiquitous on windows
11:35 pmurias joined #perl6
11:42 kjs_ joined #perl6
11:48 El_Che when on windows I use the cygwin shell. It's a fork of putty
11:48 El_Che utf8, copy paste, resizable fonts of the fly
11:48 El_Che you know, like linux well for ever :)
11:49 El_Che (it's mintty)
11:49 dalek ecosystem/jantz-patch-1: 4069725 | jantz++ | META.list:
11:49 dalek ecosystem/jantz-patch-1: hoelzro has switched to meta6.info
11:49 dalek ecosystem/jantz-patch-1:
11:49 dalek ecosystem/jantz-patch-1: Fix Task::Star build
11:49 dalek ecosystem/jantz-patch-1: https://github.com/hoelzro/p6-linenoise/comm​it/46f9534b1568e68a2a077ee80713914738b16e28
11:49 dalek ecosystem/jantz-patch-1: review: https://github.com/perl6/e​cosystem/commit/4069725434
11:53 maybekoo5 joined #perl6
11:58 dalek ecosystem: 4069725 | jantz++ | META.list:
11:58 dalek ecosystem: hoelzro has switched to meta6.info
11:58 dalek ecosystem:
11:58 dalek ecosystem: Fix Task::Star build
11:58 dalek ecosystem: https://github.com/hoelzro/p6-linenoise/comm​it/46f9534b1568e68a2a077ee80713914738b16e28
11:58 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/4069725434
11:58 dalek ecosystem: 31976cb | niner++ | META.list:
11:58 dalek ecosystem: Merge pull request #144 from perl6/jantz-patch-1
11:58 dalek ecosystem:
11:58 dalek ecosystem: hoelzro has switched linenoise to meta6.info
11:58 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/31976cbdc0
12:00 flussence .oO( maybe the commit bot shouldn't mention the same sha1 twice in a row )
12:03 dalek rakudo/nom: 06ed9cf | lizmat++ | src/core/ (3 files):
12:03 dalek rakudo/nom: Better fix for RT #126888
12:03 dalek rakudo/nom:
12:03 dalek rakudo/nom: This will actually allow binding into a list if the element is a container.
12:03 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/06ed9cf863
12:08 timotimo oh, interesting
12:11 El_Che what is it nowadays: META.info, META6.json, meta6.info?
12:12 lizmat forward looking, I think META6.json is the one
12:17 rindolf joined #perl6
12:22 dalek perl6.org: a43a96f | (Steve Mynott)++ | source/downloads/index.html:
12:22 dalek perl6.org: how to fix p6doc on windows corrected to follow tested process on 32 bit Windows 10 anyway
12:22 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/a43a96ff4d
12:22 stmuk_ virtualbox++
12:24 dalek ecosystem: 8584fda | okaoka++ | META.list:
12:24 dalek ecosystem: Add Algorithm::Treap to ecosystem
12:24 dalek ecosystem:
12:24 dalek ecosystem: See https://github.com/okaoka/p6-Algorithm-Treap
12:24 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/8584fda14f
12:24 dalek ecosystem: 4e294cd | (Shoichi Kaji)++ | META.list:
12:24 dalek ecosystem: Merge pull request #143 from okaoka/add-treap
12:24 dalek ecosystem:
12:24 dalek ecosystem: Add Algorithm::Treap to ecosystem
12:24 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/4e294cd279
12:28 dalek rakudo/nom: 8e967c8 | lizmat++ | src/core/Hash.pm:
12:28 dalek rakudo/nom: Make sure DELETE-KEY(:SINK) doesn't bomb on empty
12:28 dalek rakudo/nom:
12:28 dalek rakudo/nom: Even though this DELETE-KEY candidate isn't used yet, it had a flaw
12:28 dalek rakudo/nom: when being called on a Hash that didn't have any storage allocated yet
12:28 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8e967c8331
12:30 rpburkholder joined #perl6
12:35 dalek perl6.org: 7d8ed09 | (Steve Mynott)++ | source/downloads/index.html:
12:35 dalek perl6.org: close tag
12:35 dalek perl6.org: review: https://github.com/perl6/p​erl6.org/commit/7d8ed093a5
12:35 mr-foobar joined #perl6
12:41 Cabanossi joined #perl6
12:55 loic joined #perl6
12:56 RabidGravy joined #perl6
12:57 RabidGravy back again
13:01 dalek rakudo/nom: 62b5308 | lizmat++ | src/core/Hash.pm:
13:01 dalek rakudo/nom: Make Hash[Type].AT-KEY 25% faster
13:01 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/62b5308e1b
13:02 dalek Heuristic branch merge: pushed 103 commits to DBIish/oracle by jonathanstowe
13:06 dolmen joined #perl6
13:06 perlawhirl joined #perl6
13:13 Skarsnik joined #perl6
13:15 Skarsnik Hello
13:15 yoleaux 10:57Z <abraxxa> Skarsnik: I've rebased my DBIish pull-request, the Travis setup seems to work now (finds the dependent modules), the fails are the same as on my dev box and subject to API discussion
13:17 neilb joined #perl6
13:18 dalek Inline-Perl5: 6fb0c35 | (Stefan Seifert)++ | / (3 files):
13:18 dalek Inline-Perl5: Pass real code references to Perl 5 code
13:18 dalek Inline-Perl5:
13:18 dalek Inline-Perl5: When passing Callables to Perl 5 code, we turned them into Perl6::Callable
13:18 dalek Inline-Perl5: objects that overrode call dereferencing. While this allowed for calling
13:18 dalek Inline-Perl5: the referenced code, the passed references still pointed to objects instead
13:18 dalek Inline-Perl5: of code. Perl 5 code checking explicitly for CODE references would fail
13:18 dalek Inline-Perl5: even though just calling it anyway would have worked.
13:18 dalek Inline-Perl5:
13:18 dalek Inline-Perl5: We now create a new wrapper sub for wrapped Callables that are
13:18 dalek Inline-Perl5: indistinguishable from ordinary Perl 5 code references.
13:18 dalek Inline-Perl5:
13:18 dalek Inline-Perl5: Fixes Perl RT #127454
13:18 dalek Inline-Perl5: review: https://github.com/niner/Inl​ine-Perl5/commit/6fb0c35bb0
13:18 dalek Inline-Perl5: b01a3b3 | (Stefan Seifert)++ | META.info:
13:18 dalek Inline-Perl5: Version 0.4
13:18 dalek Inline-Perl5: review: https://github.com/niner/Inl​ine-Perl5/commit/b01a3b379a
13:20 kid51 joined #perl6
13:22 loic left #perl6
13:23 nine I think this eliminated the last gotcha from my Leapfrogging the Bootstrap talk at last year's FOSDEM :)
13:23 ab5tract nine++
13:24 ab5tract thank you for all your work, man. Inline::Perl5 and otherwise :D
13:24 nine It's a pleasure to be a part of this community :)
13:25 ab5tract i hear that :)
13:25 Zero_Dogg joined #perl6
13:26 dalek rakudo/nom: 8588eeb | lizmat++ | src/core/Hash.pm:
13:26 dalek rakudo/nom: Don't create a named intermediary
13:26 dalek rakudo/nom:
13:26 dalek rakudo/nom: Not sure whether this will speed up things, but at least it won't leak
13:26 dalek rakudo/nom: a nonsense name in case of errors
13:26 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8588eeb846
13:30 dalek DBIish/oracle: 2202e81 | RabidGravy++ | t/35-pg-common.t:
13:30 dalek DBIish/oracle: Just skip the test if PGDATABASE not set
13:30 dalek DBIish/oracle: review: https://github.com/perl6/DBIish/commit/2202e81943
13:30 dalek DBIish/oracle: c3df49b | RabidGravy++ | .gitignore:
13:30 dalek DBIish/oracle: Add .precomp to .gitignore
13:30 dalek DBIish/oracle: review: https://github.com/perl6/DBIish/commit/c3df49bae3
13:35 lizmat m: my $m = Map.new("a",42); $m.Hash<b> = 666; dd $m   # this feels wrong
13:35 camelia rakudo-moar 8588ee: OUTPUT«Map $m = Map.new((:a(42),:b(666)))␤»
13:37 lizmat as a Map is considered to be immutable, right ?
13:37 lizmat well, like a List is immutable
13:38 kid511 joined #perl6
13:40 ab5tract lizmat: that's how I understood them, though the GHR (GMR?) went past with considerably less stress than the GLR
13:41 petercommand joined #perl6
13:41 timotimo great hash refactor? i think it was actually called GMR
13:42 nine Yes, it's GMR as the List refactor was also named after the immutable base type
13:42 * masak .oO( at some point, we started calling all refactors "Great", even the small ones )
13:43 nine masak: when I started out on the gmr branch, I thought it would be much more work than it actually was
13:43 ab5tract assignment to a coerced object could maybe be considered a 'useless use in sink' case
13:44 masak nine: well, that's good news in retrospect
13:44 ab5tract but as I understand it the result of the .Hash should be another object entirely, not the same container with a different object in it
13:44 lizmat yeah, working on a fix
13:44 ab5tract but maybe I've misunderstood that aspect of coercion
13:44 musiKk joined #perl6
13:45 lizmat well, a Hash is mutable, so by definition, it should not return the same immutable Map
13:46 ab5tract lizmat: so are you saying the output should be 'Hash $m = Hash.new((:a(42),:b(666)))' ?
13:48 lizmat $ 6 'my $m = Map.new("a",42); my $h = $m.Hash; $h<b> = 666; dd $m, $h'
13:48 lizmat Map $m = Map.new((:a(42)))
13:48 lizmat Hash $h = ${:a(42), :b(666)}
13:48 lizmat after I push my fix
13:48 lizmat m: my $m = Map.new("a",42); my $h = $m.Hash; $h<b> = 666; dd $m, $h   # now
13:48 camelia rakudo-moar 8588ee: OUTPUT«Map $m = Map.new((:a(42),:b(666)))␤Hash $h = ${:a(42), :b(666)}␤»
13:49 mwheat joined #perl6
13:50 ab5tract lizmat: perfecto :)
13:50 lizmat hmmm,,, alas, no:
13:50 lizmat $ 6 'my $m = Map.new(my $ = "a",42); my $h = $m.Hash; $h<a> = 666; dd $m, $h'
13:50 lizmat Map $m = Map.new((:a(666)))
13:50 lizmat Hash $h = ${:a(666)}
13:50 lizmat nqp::clone is too shallow  :-(
13:51 timotimo ugh :<
13:52 jnthn Well, no, nqp::clone is defined to clone precisely the thing you call it on :)
13:52 lizmat yeah, I guess
13:53 lizmat ok, looking at that again after some cycling&
13:53 moritz does IO::Notification work on the JVM?
13:55 ab5tract moritz: interesting question. is there any documentation on what's broken/divergent from moarvm? or just the results from roast?
13:56 timotimo moritz: i think it does, but it's probably not equivalent in behaviour
13:56 timotimo or perhaps nobody bothered to implement it yet?
13:59 BenGoldberg joined #perl6
14:01 jnthn throw new UnsupportedOperationException("watchfile is not yet implemented.");
14:02 jnthn So, no :(
14:05 timotimo damn.
14:06 rindolf joined #perl6
14:35 araujo joined #perl6
14:37 araujo joined #perl6
14:51 kurahaupo joined #perl6
14:53 AlexDaniel joined #perl6
14:59 lokien_ joined #perl6
15:01 timotimo hm.
15:01 timotimo http://friedo.com/blog/2016/02/ex​ploring-perl-6-signatures-part-2 - always using the variable name itself in the where clause rather than $_; that seems unidiomatic to me
15:03 BenGoldberg joined #perl6
15:07 gfldex m: sub is-positive($n){ $n > 0 }; sub f($x where .&is-positive){ say 'OK' }; f 1;
15:07 camelia rakudo-moar 8588ee: OUTPUT«OK␤»
15:08 gfldex i don't need no stinking variable while nitpicking :->
15:08 timotimo DrForr: in your article "from regular expressions to grammars", you seem to claim that the index method returns -1 on a failed match
15:08 lustslut joined #perl6
15:08 timotimo m: sub is-positive($n){ $n > 0 }; sub f($x where &is-positive){ say 'OK' }; f 1;
15:08 camelia rakudo-moar 8588ee: OUTPUT«OK␤»
15:08 lustslut left #perl6
15:09 sail0r joined #perl6
15:09 gfldex m: sub is-positive($n){ $n > 0 }; sub f($x where * > 0){ say 'OK' }; f 1;
15:09 camelia rakudo-moar 8588ee: OUTPUT«OK␤»
15:09 gfldex is-positive is not the best example possible
15:10 timotimo that's right
15:11 gfldex "Every Perl 6 subroutine can also specify its own return type as part of the signature." is plain wrong. There is a big difference between a type constraint and a type.
15:12 timotimo you mean if you supply Cool and return Str, that's just a type constraint on what it returns, not a return type?
15:13 AlexDaniel b2gills: hi! Do you know why your hyper tests are not merged yet? I reviewed your PR and it looks great, but I thought that I'd ask you before merging it in case you know something
15:14 timotimo fosdem's status.txt recently went up to 149, but still none in the 1124 room
15:14 timotimo or was it 2214 that we're interested in?
15:14 AlexDaniel 2214
15:14 timotimo anyway, will be AFK for a bit
15:14 timotimo (none in 2214 either)
15:15 Ben_Goldberg joined #perl6
15:19 musiKk joined #perl6
15:24 AlexDaniel j: (-1)⁴⁵⁵³⁵³⁵³⁴⁵³⁶⁴⁵³⁵³⁴⁵
15:24 camelia rakudo-jvm f99bdb: OUTPUT«5===SORRY!5=== Error while compiling /tmp/hS9Fz7O6KA␤Bogus postfix␤at /tmp/hS9Fz7O6KA:1␤------> 3(-1)7⏏5⁴⁵⁵³⁵³⁵³⁴⁵³⁶⁴⁵³⁵³⁴⁵␤    expecting any of:␤        infix␤        infix stopper␤        statement e…»
15:24 AlexDaniel j: say (-1)⁴⁵⁵³⁵³⁵³⁴⁵³⁶⁴⁵³⁵³⁴⁵
15:24 camelia rakudo-jvm f99bdb: OUTPUT«5===SORRY!5=== Error while compiling /tmp/KWfipmXbVa␤Bogus postfix␤at /tmp/KWfipmXbVa:1␤------> 3say (-1)7⏏5⁴⁵⁵³⁵³⁵³⁴⁵³⁶⁴⁵³⁵³⁴⁵␤    expecting any of:␤        infix␤        infix stopper␤        postfix…»
15:25 AlexDaniel j: 2²
15:25 camelia rakudo-jvm f99bdb: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Z9yl36IqoQ␤Bogus postfix␤at /tmp/Z9yl36IqoQ:1␤------> 27⏏5²␤    expecting any of:␤        infix␤        infix stopper␤        statement end␤        statement modifier␤        statement…»
15:25 AlexDaniel how old is that?
15:25 AlexDaniel hm that's pretty recent
15:27 jnthn m: 2²
15:27 camelia rakudo-moar 8588ee: ( no output )
15:28 jnthn Is it a regression or did it ever work on JVM?
15:28 AlexDaniel I don't know actually
15:29 bartolin locally it works fine (build from last night)
15:29 bartolin ./perl6-j -e 'say 2²'
15:29 bartolin 4
15:30 bartolin (this is on FreeBSD, 2015.12-262-g58d3398)
15:30 AlexDaniel j: say π
15:30 camelia rakudo-jvm f99bdb: OUTPUT«3.141592653589793␤»
15:33 AlexDaniel j: say (-1)**4553535345364535354
15:33 camelia rakudo-jvm f99bdb: OUTPUT«1␤»
15:34 AlexDaniel m: say (-1)**4553535345364535354
15:34 camelia rakudo-moar 8588ee: OUTPUT«Numeric overflow␤  in block <unit> at /tmp/zb1dCZ2T_Z line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/zb1dCZ2T_Z line 1␤␤»
15:34 AlexDaniel hm
15:35 sjoshi joined #perl6
15:38 raiph joined #perl6
15:44 s_kilk joined #perl6
15:50 khw joined #perl6
15:51 ELBeavers joined #perl6
15:57 bartolin ahh, the r-j camelia runs is actually rather old
15:58 bartolin r: say $*PERL.compiler
15:58 camelia rakudo-jvm f99bdb: OUTPUT«rakudo (2015.10.134.g.273.e.895)␤»
15:58 camelia ..rakudo-moar 8588ee: OUTPUT«rakudo (2015.12.266.g.8588.eeb)␤»
15:58 bartolin star-m: say 2² # same failure mode
15:58 camelia star-m 2015.09: OUTPUT«5===SORRY!5=== Error while compiling /tmp/G68OBeCWX0␤Bogus postfix␤at /tmp/G68OBeCWX0:1␤------> 3say 27⏏5² # same failure mode␤    expecting any of:␤        infix␤        infix stopper␤        postfix␤        statement end␤        …»
15:59 * bartolin doesn't know how to trigger a rebuild of r-j for camelia
16:14 maybekoo5 joined #perl6
16:15 ugexe jvm doesnt build
16:16 bartolin it does for me (donaldh++ for making it build again)
16:16 bartolin first successful build was two days ago: https://github.com/usev6/perl6-roast-​data/blob/master/perl6_pass_rates.csv
16:18 AlexDaniel meh, I can't build it
16:18 AlexDaniel “Unhandled exception: java.lang.StackOverflowError … bla-bla”
16:19 bartolin afaik this was the commit which unbusted the build: https://github.com/perl6/nqp/commit/175e8a7251
16:19 bartolin hmm, I just built it 40 min ago (on linux)
16:19 Guest45613 in case someone wants to weigh in on this discussion of Perl 6 parallelism, etc.  https://soylentnews.org/arti​cle.pl?sid=16/02/07/0145222
16:20 dalek ecosystem: 8c7b871 | (Jake Russo)++ | META.list:
16:20 dalek ecosystem: Fix for recent linenoise file change
16:20 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/8c7b871511
16:22 bartolin AlexDaniel: I built it from HEAD with 'perl Configure.pl --backends=all --gen-moar --gen-nqp'
16:22 bartolin (didn't 'make install', though)
16:26 isBEKaml joined #perl6
16:27 AlexDaniel bartolin: oh well, I hoped that rakudobrew will do it
16:28 * bartolin hasn't tried rakudobrew, yet
16:34 Ven joined #perl6
16:35 espadrine joined #perl6
16:37 Guest45613 the how to help page https://github.com/usev6/perl6-roast-data says to open a ticket for test failures - surely that is something that could be automated?  or am i misunderstanding something
16:39 Guest45613 another question i had is are there docker containers available for building the latest rakudo-star?
16:39 RabidGravy there is a docker image available
16:40 RabidGravy https://hub.docker.com/_/rakudo-star/
16:42 neilb joined #perl6
16:42 Guest45613 awesome - thanks RabidGravy
16:45 bartolin Guest45613: (please note, that README hasn't been updated in a while); automatically generated tickets won't be that good IMHO (at least not yet). in my experience it needs at least some investigation to generate useful tickets (for eventual test failures)
16:48 sufrostico joined #perl6
16:48 Guest45613 ok.  any suggested areas to look at with low hanging fruit?  i'm coming from a perl 5 background but am eager to get into perl 6
16:49 Juerd Documentation. Finding out what's missing or wrong.
16:50 geekosaur we need more cross-references too
16:50 Guest45613 ok
16:51 abaugher_ joined #perl6
16:57 ugexe bartolin: sorry, it builds but it doesnt pass its tests (so camelia wouldnt update the build exposed here)
16:58 bartolin ugexe: ah, that explains it. I didn't try 'make test' recently -- only 'make spectest'. thanks for clarifying
16:59 raiph 'noon all
16:59 yoleaux 6 Feb 2016 21:27Z <moritz> raiph: please mark rakudo-specific stuff as such in the docs, even if it's not always the case yet
16:59 ugexe the travis badge is also a false positive. it shows green for the jvm matrix items, but the tests say failure
17:02 Cheery joined #perl6
17:04 kjs_ joined #perl6
17:04 dnmfarrell joined #perl6
17:05 dnmfarrell hey folks I just raised a ticket (https://rt.perl.org/Ticket/Display.html?id=127483) but forgot to prepend "[BUG]" to it. Anyone know how I can updated it by email? I doing it from RT but I get "permission denied"
17:06 bartolin I don't think you can do that via email. I'll change the subject, mom
17:07 bartolin done
17:09 dnmfarrell bartolin: thank you sir!
17:09 bartolin yw :-)
17:10 DubiousPepper joined #perl6
17:10 autarch how would I get some debugging output from Perl6::Actions - something like dd?
17:11 moritz if it's a Perl 6 level object, you can nqp::say($theobject.perl())
17:11 autarch cool
17:11 llfourn autarch: if you mean QAST nodes .dump
17:11 autarch I have no idea what I mean - I'm trying to figure out what's getting passed to Perl6::Actions::bare_rat_number
17:12 autarch I'm guessing I want dump
17:12 moritz method bare_rat_number($/) {
17:12 moritz a match objet
17:12 llfourn autarch: $/ is a  NQPMatch
17:12 llfourn I think .dump might work on it too though?
17:13 moritz it's automatically called by the grammar token of the same name
17:13 autarch moritz: that much I understand
17:13 llfourn anyway what probably want is note $/.ast.dump
17:13 autarch I'm trying to figure out why in some cases calling @($<nu>.ast)[0].compile_time_value in that method throws an error that compile_time_value doesn't exist as a method
17:13 llfourn or rather $/<something>.dump
17:14 llfourn note $<nu>.ast.dump # try this
17:14 autarch $/.dump did work, btw
17:14 moritz except that you need parens for the sub call in nqp
17:15 llfourn cool
17:15 llfourn ^^ yes
17:16 autarch hmm, but that didn't help me figure out why it doesn't have a .compile_time_value method
17:16 llfourn well what type of QAST node at [0]?
17:16 moritz what type is its first child node?
17:16 autarch yeah, that's my next step
17:16 moritz the QAST node should dump its children too
17:17 moritz if it doesn't have one, it would explain why it has no .compile_time_value method :-)
17:17 FROGGS joined #perl6
17:17 autarch the .dump for cases where .compile_time_value worked and where it didn't looked fundamentally the same
17:18 autarch but we'll see what @($<nu>.ast) and @($<nu>.ast)[0] produce
17:18 autarch I'm trying to fix https://rt.perl.org/Ticket/Display.html?id=126873
17:19 llfourn (you should .dump them)
17:19 moritz m: say <1480211362/1021>
17:19 camelia rakudo-moar 8588ee: OUTPUT«1449766.270323␤»
17:19 moritz m: say <1480211362999/1021>
17:19 camelia rakudo-moar 8588ee: OUTPUT«===SORRY!===␤Cannot find method 'compile_time_value'␤»
17:19 moritz m: say <148021136299/1021>
17:19 camelia rakudo-moar 8588ee: OUTPUT«===SORRY!===␤Cannot find method 'compile_time_value'␤»
17:19 moritz m: say <14802113629/1021>
17:19 camelia rakudo-moar 8588ee: OUTPUT«===SORRY!===␤Cannot find method 'compile_time_value'␤»
17:19 moritz m: say <1480211362/1021>
17:19 camelia rakudo-moar 8588ee: OUTPUT«1449766.270323␤»
17:19 moritz m: say 14802113629.log(2)
17:19 camelia rakudo-moar 8588ee: OUTPUT«33.7850841452925␤»
17:19 moritz 32 bit boundary, eh?
17:21 autarch actually, it seems to be 2**31
17:21 autarch m: say 2**31/7
17:21 camelia rakudo-moar 8588ee: OUTPUT«306783378.285714␤»
17:21 autarch hmm, maybe not
17:21 autarch m: say 2**32/7
17:21 camelia rakudo-moar 8588ee: OUTPUT«613566756.571429␤»
17:21 autarch m: say (2**32)/7
17:21 camelia rakudo-moar 8588ee: OUTPUT«613566756.571429␤»
17:22 autarch m: say ((2**31)/3).perl.EVAL
17:22 camelia rakudo-moar 8588ee: OUTPUT«===SORRY!===␤Cannot find method 'compile_time_value'␤»
17:22 autarch m: say ((2**31 - 1)/3).perl.EVAL
17:22 camelia rakudo-moar 8588ee: OUTPUT«715827882.333333␤»
17:22 llfourn right so .compile_time_value will work on a **
17:22 dalek nqp/experiment-multi-line-repl: 09391cb | hoelzro++ | src/HLL/Compiler.nqp:
17:22 dalek nqp/experiment-multi-line-repl: Move continuation detection into eval
17:22 dalek nqp/experiment-multi-line-repl:
17:22 dalek nqp/experiment-multi-line-repl: With this change, eval can return the output of running the code, or
17:22 dalek nqp/experiment-multi-line-repl: a sentinel value that indicates that more input is needed to run a chunk
17:22 dalek nqp/experiment-multi-line-repl: of code.  This is so that more advanced forms of detecting when we need
17:22 dalek nqp/experiment-multi-line-repl: more input, such as if the user is entering input for a multi-line
17:22 dalek nqp/experiment-multi-line-repl: statement like a block.
17:22 dalek nqp/experiment-multi-line-repl:
17:22 dalek nqp/experiment-multi-line-repl: This is just a proof-of-concept change; I don't really like how eval
17:22 dalek nqp/experiment-multi-line-repl: indicates to its caller that it needs more input.  But at least this
17:22 dalek nqp/experiment-multi-line-repl: will get people to try and see if multi-line input could be feasible.
17:22 dalek nqp/experiment-multi-line-repl:
17:22 dalek nqp/experiment-multi-line-repl: We also add a new method to HLL::Compiler - needs-more-input, which is
17:22 dalek nqp/experiment-multi-line-repl: used in conjunction with the sentinel value returned by eval.
17:22 dalek nqp/experiment-multi-line-repl: review: https://github.com/perl6/nqp/commit/09391cb88e
17:22 dalek nqp/experiment-multi-line-repl: 7f07a9f | hoelzro++ | src/HLL/Compiler.nqp:
17:22 dalek nqp/experiment-multi-line-repl: Check if we need more input due to trailing backslash
17:22 dalek nqp/experiment-multi-line-repl:
17:22 dalek nqp/experiment-multi-line-repl: This is a bit of a hack, but it's good enough for this experiment
17:22 dalek nqp/experiment-multi-line-repl: review: https://github.com/perl6/nqp/commit/7f07a9f003
17:24 MadcapJake anyone ever used fish shell? I love how writing functions can happen right in the shell. (I've been thinking about writing a Perl 6 plugin for fish, not sure what's all possible though)
17:24 llfourn you can write function in sh though?
17:24 MadcapJake in fish you can
17:25 llfourn foo(){ echo "hello world"; } # /bin/sh
17:25 MadcapJake no it's a multiline kind of thing i mean
17:25 dupa joined #perl6
17:25 dalek rakudo/experiment-multi-line-repl: f9ff566 | hoelzro++ | tools/build/NQP_REVISION:
17:25 dalek rakudo/experiment-multi-line-repl: Require experimental NQP for multi-line REPL
17:25 dalek rakudo/experiment-multi-line-repl: review: https://github.com/rakudo/rakudo/commit/f9ff566cd9
17:25 dalek rakudo/experiment-multi-line-repl: 7a22486 | hoelzro++ | src/ (2 files):
17:25 dalek rakudo/experiment-multi-line-repl: Add pos attribute to X::Syntax
17:25 dalek rakudo/experiment-multi-line-repl: review: https://github.com/rakudo/rakudo/commit/7a22486c62
17:25 dalek rakudo/experiment-multi-line-repl: 2ecda56 | hoelzro++ | src/Perl6/Compiler.nqp:
17:25 MadcapJake like you type `function name` and then pressing enter actually indents the shell and only dedents when you type `end`
17:25 dalek rakudo/experiment-multi-line-repl: Override HLL::Compiler.eval for multi-line input
17:25 dalek rakudo/experiment-multi-line-repl:
17:25 llfourn I'm sure it's much better -- but you can do that in sh too :)
17:25 dalek rakudo/experiment-multi-line-repl: The way this works is pretty dodgy, but it's good enough for
17:25 dalek rakudo/experiment-multi-line-repl: my experiment.  How this works is that we catch any exceptions
17:25 dalek rakudo/experiment-multi-line-repl: from evaluating the code; if there's an exception, it's an
17:25 dalek rakudo/experiment-multi-line-repl: X::Syntax::Missing, and it occurs at the end of the input string,
17:25 dalek rakudo/experiment-multi-line-repl: that means that more input is required.  Instead of just throwing
17:25 dalek rakudo/experiment-multi-line-repl: the exception at the user, we signal HLL::Compiler that we need
17:25 dalek rakudo/experiment-multi-line-repl: more input via self.needs-more-input
17:25 dalek rakudo/experiment-multi-line-repl: review: https://github.com/rakudo/rakudo/commit/2ecda568f3
17:25 magiczna_dupa hey
17:26 magiczna_dupa could you fix documentation?
17:26 MadcapJake o_O
17:26 mspo hey nice
17:26 MadcapJake llfourn, i was speaking to the multi-line repl idea, it'd be good inspiration
17:27 ryan_ joined #perl6
17:27 MadcapJake fish shell functions: https://youtu.be/8pno3tTdywg?t=2m22s
17:27 llfourn MadcapJake: I was just mentioning that vanilla sh also does multiline repl.
17:27 MadcapJake llfourn, look at the video to see what i mean
17:28 llfourn (well on my mac at least)
17:29 llfourn sure does look nice :)
17:29 MadcapJake llfourn, that'd be cool to have for perl6 repl
17:30 llfourn yep. looks like hoelzro++ is getting us slightly closer to that.
17:30 MadcapJake sweet!
17:31 moritz m: say 2**31
17:31 camelia rakudo-moar 8588ee: OUTPUT«2147483648␤»
17:31 moritz m: say <2147483648/1>
17:31 camelia rakudo-moar 8588ee: OUTPUT«===SORRY!===␤Cannot find method 'compile_time_value'␤»
17:31 moritz m: say <2147483647/1>
17:31 camelia rakudo-moar 8588ee: OUTPUT«2147483647␤»
17:31 MadcapJake i might be able to write some functions for fish that'll give you some of the same power
17:31 hoelzro your wish is my command (line): http://hoelz.ro/files/multi-line-repl.gif
17:32 mspo hoelzro: https://asciinema.org/
17:32 hoelzro please try out the experiment-multi-line-repl branch of Rakudo and NQP, and let me know if I broke anything serious
17:32 Ven joined #perl6
17:32 MadcapJake hoelzro++
17:32 hoelzro mspo: thanks for the tip!
17:32 autarch so how can I get some insight into what's in "@($<nu>.ast)" - both .dump and .perl fail
17:32 llfourn hoelzro++ The people's champion
17:33 mspo hoelzro: it's pretty slick- you can copy and paste
17:33 hoelzro I'll work on getting multi-line history going too
17:33 llfourn autarch: eh, $<nu>.ast.dump doesn't show anything?
17:33 hoelzro gotta run and get coffee now =)
17:33 autarch llfourn: it blows up
17:34 llfourn autarch: erm does $/.dump show a 'nu' in the match?
17:34 autarch I'll check
17:34 autarch the build/test cycle is pretty slow :(
17:34 llfourn heh yep :\
17:35 llfourn autarch: also is it nu.ast that blows up or when you call a method?
17:35 autarch when I call the method
17:35 autarch cannot find method 'dump'
17:36 llfourn autarch: ok maybe try asking HOW for the type name -- my $tmp := $<nu>.ast; $tmp.HOW.name($tmp)
17:37 llfourn (I think that works)
17:37 jevin_ joined #perl6
17:37 llfourn (printing it out of course)
17:38 drrho joined #perl6
17:38 llfourn it should def have a nu in the match so nvm about that
17:40 autarch llfourn: so it looks like in the cases where it works $<nu>.ast is a QAST::Want and when it fails it's a QAST::WVal
17:41 autarch which makes sense to me, since the former has a compile_time_value method and the latter does not
17:41 autarch oh wait, no, they both have that method
17:41 llfourn yeah...
17:41 llfourn well Want doesn't always?
17:42 llfourn I'm not an expert of this stuff but I think WVal should always have compile time value
17:42 autarch no, it should since it does the QAST::CompileTimeValue role, right?
17:42 llfourn so it's Want when it works?
17:43 autarch yes
17:43 sevvie joined #perl6
17:43 llfourn well I think I'm at the end of my helpfulness then because I expected it to be the other way around!
17:43 autarch so if I get rid of the "make an array" bit in Perl6::Action and change it to "my $nu := $<nu>.ast.compile_time_value;" then it just works
17:44 autarch does anyone know _why_ it's making an array and getting the first element?
17:44 s_kilk joined #perl6
17:45 autarch because QAST::Want::compile_time_value is defined as "self[0].compile_time_value()" - which seems to do what's needed ;)
17:46 llfourn yeah I saw that too I guess give that a shot?
17:46 autarch I'll try running the entire spectest with this change but this might be the easy fix
17:46 TEttinger joined #perl6
17:46 llfourn autarch: although blame shows jnthn did that last august :P
17:47 autarch my guess is maybe something else changed in nqp-land since then to make this unnecessary, but that's very much a guess
17:47 llfourn Make <0x01/0x03> work reliably across platforms.
17:47 llfourn Fixes it on Windows, and probably on JVM too.
17:47 llfourn is the commit msg FYI
17:48 llfourn so it seems he's working around some low level stuff
17:48 autarch guess we'll see if this breaks that ;)
17:49 [Tux] test            22.092
17:49 [Tux] test-t          12.038
17:49 [Tux] csv-parser      50.114
17:51 autarch https://github.com/rakudo/rakudo/pull/707
17:52 llfourn autarch: oh did it fix the immediate issue?
17:53 autarch llfourn: yep
17:53 llfourn autarch++
17:53 autarch well, let's see what the spec test says in ... how many minutes/hours?
17:54 * lizmat waits for Travis before merging
17:54 autarch yeah, absolutely
17:54 lizmat but looks good!   autarch++
17:54 autarch I for one do not trust myself ;)
17:55 * lizmat always does a spectest before committing
17:56 NZKindest joined #perl6
17:58 autarch I was hoping travis might be quicker than me
17:58 autarch I'm all about just rebasing anyway
17:58 mst o/ all about rebase / no merges o/
18:00 lizmat [Tux]: again lower, right ?
18:00 autarch well, my change doesn't break the tests jnthn added when he changed Perl6::Actions
18:00 llfourn autarch: but he changed it because it didn't work on windows?
18:00 autarch but that's just on moar - I guess we'll see what travis says about the jvm
18:00 llfourn are you testing on windows?
18:01 autarch oh right, no
18:01 autarch and travis doesn't either :(
18:01 autarch someone should get rakudo running on appveyor too
18:05 autarch actually, looking at the earlier change, I think it's unlikely I broke windows
18:06 autarch I think the earlier problem had to do with making the numerator and denominator Int values explicitly in the action method
18:08 llfourn hmm yes it would be strange if your change was platform dependent
18:09 jevin joined #perl6
18:10 magiczna_dupa joined #perl6
18:13 woolfy joined #perl6
18:13 Xor_ joined #perl6
18:13 woolfy left #perl6
18:13 sjoshi joined #perl6
18:14 dalek rakudo/nom: 95e9655 | lizmat++ | src/core/Map.pm:
18:14 dalek rakudo/nom: Make sure Map.Hash works correctly
18:14 dalek rakudo/nom:
18:14 dalek rakudo/nom: Before, any subsequent changes made to the Hash, would be happening to
18:14 dalek rakudo/nom: the Map also.  This also makes Map.Hash 15x faster for an empty Map,
18:14 dalek rakudo/nom: to the same speed for a Map with 20 keys, to slower for larger Maps.
18:14 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/95e9655ab2
18:17 s_kilk joined #perl6
18:21 skids joined #perl6
18:23 Relsak joined #perl6
18:27 magiczna_dupa please fix documentation
18:30 dnmfarrell joined #perl6
18:32 timotimo can you be more specific, magiczna_dupa?
18:33 magiczna_dupa try to find docs for .=
18:33 magiczna_dupa ;D
18:33 magiczna_dupa the site will error
18:34 dnmfarrell m: @*INC.say;
18:34 camelia rakudo-moar 95e965: OUTPUT«Dynamic variable @*INC not found␤  in block <unit> at /tmp/wOsT_q0iuU line 1␤␤Actually thrown at:␤  in block <unit> at /tmp/wOsT_q0iuU line 1␤␤»
18:34 dnmfarrell Does anyone know how to inspect @*INC?
18:34 llfourn magiczna_dupa: https://github.com/perl6/doc/issues/155 # known issue
18:34 MadcapJake m: $*REPO.say
18:34 camelia rakudo-moar 95e965: OUTPUT«inst#/home/camelia/.p​erl6/2015.12-267-g95e9655␤»
18:35 MadcapJake dnmfarrell, @*INC is gone
18:35 magiczna_dupa oh ok
18:35 dnmfarrell Meh, it's still in S28
18:36 dnmfarrell MadcapJake: is there an alternative?
18:36 MadcapJake m: $*REPO.^methods.say
18:36 llfourn dnmfarrell: you can look at $*REPO
18:36 camelia rakudo-moar 95e965: OUTPUT«(BUILD writeable-path can-install install files need resource id short-id loaded precomp-repository path-spec prefix load repo-chain new installed perl gist Str WHICH <anon>)␤»
18:36 MadcapJake dnmfarrell, ^
18:36 dnmfarrell MadcapJake: thanks :)
18:37 llfourn m: note $*REPO.repo-chain.perl
18:37 camelia rakudo-moar 95e965: OUTPUT«(CompUnit::Repository::Installation.new​("/home/camelia/.perl6/2015.12-267-g95e9655"), CompUnit::Repository::Installation.new("/home​/camelia/rakudo-m-inst-2/share/perl6/site"), CompUnit::Repository::Installation.new("/​home/camelia/rakudo-m-inst-2/share/perl…»
18:39 maybekoo5 joined #perl6
18:42 Sqirrel joined #perl6
18:43 ryan_ joined #perl6
18:43 autarch magiczna_dupa: patches for that would be very welcome, but I think it's a somewhat challenging issue to fix properly
18:44 autarch but maybe I'm overestimating the work involved?
18:44 MadcapJake magiczna_dupa, if you ever have questions on operators, the best resource available is http://doc.perl6.org/language/operators
18:44 autarch jnthn: are you around today?
18:45 MadcapJake autarch, the current status on operators is that we should encode them in some custom way
18:45 MadcapJake wrt doc search
18:45 autarch right, I know that
18:46 MadcapJake oh ok, sorry!
18:46 autarch I'm saying that might be a fair bit of work
18:46 MadcapJake where's the search code in perl6/doc?
18:47 autarch MadcapJake: it's some JS in the perl6/doc repo, I believe
18:48 autarch https://github.com/perl6/doc/blob/​master/template/search_template.js
18:48 MadcapJake thanks!
18:49 autarch so I think the trick is to substitute the operators with words when generating the file that template is for - and then update the autocomplete JS to perform the same substitution
18:49 autarch hmm, after describing it it doesn't sound so hard ;)
18:50 MadcapJake maybe a json file could contain the translations used by both sides
18:51 sjn \o
18:51 sjn is there a good way to search for rakudo bugs?
18:51 * sjn looks at rt.perl.org
18:52 ZoffixWin joined #perl6
18:53 llfourn sjn: rt.perl.org is the best we got atm
18:54 sjn hm. no free text search?
18:55 llfourn nope
18:55 dalek rakudo/nom: 2775261 | autarch++ | src/Perl6/Actions.nqp:
18:55 dalek rakudo/nom: Just call $<nu>.ast.compile_time_value in Perl6::Actions.bare_rat_number
18:55 dalek rakudo/nom:
18:55 dalek rakudo/nom: The value of $<nu>.ast can either be a QAST::Want or a QAST::WVal depending on
18:55 dalek rakudo/nom: the value of the numerator. Any value of 2**31 or greater gives us a
18:55 dalek rakudo/nom: QAST::WVal.
18:55 dalek rakudo/nom:
18:55 dalek rakudo/nom: Previously the code called "@($<nu>.ast)[0].compile_time_value" - this works
18:55 dalek rakudo/nom: fine for a QAST::Want, but a QAST::WVal does not have any children, and so
18:55 dalek rakudo/nom: this would lead to an error of "Cannot find method 'compile_time_value'".
18:55 dalek rakudo/nom:
18:55 dalek rakudo/nom: Internally QAST::Want.compile_time_value is implemented as
18:55 dalek rakudo/nom: "self[0].compile_time_value()" so it seems like there's no good reason for us
18:55 dalek rakudo/nom: to do same thing in Perl6::Actions. Yay polymorphism!
18:55 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/2775261502
18:55 dalek rakudo/nom: 09181a4 | lizmat++ | src/Perl6/Actions.nqp:
18:55 dalek rakudo/nom: Merge pull request #707 from autarch/autarch/fix-large-numerator-Rat-literal
18:55 dalek rakudo/nom:
18:55 dalek rakudo/nom: Just call $<nu>.ast.compile_time_value in Perl6::Actions.bare_rat_number
18:55 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/09181a4105
18:56 dalek roast: c785caa | autarch++ | S32-num/rat.t:
18:56 dalek roast: Fix incorrect plan in rat.t
18:56 dalek roast: review: https://github.com/perl6/roast/commit/c785caacab
18:56 dalek roast: 1b4f796 | autarch++ | S32-num/rat.t:
18:56 dalek roast: Add a test for Rat.perl.EVAL round tripping with arbitrary numerator & denomator
18:56 dalek roast:
18:56 dalek roast: As of the latest Rakudo this leads to an error of the form:
18:56 dalek roast:
18:56 dalek roast:   ===SORRY!===
18:56 dalek roast:   Cannot find method 'compile_time_value'
18:56 dalek roast:
18:56 dalek roast: See https://rt.perl.org/Ticket/Display.html?id=126873 for the original report.
18:56 dalek roast: review: https://github.com/perl6/roast/commit/1b4f796e10
18:56 dalek roast: 8be23bb | lizmat++ | S32-num/rat.t:
18:56 dalek roast: Merge pull request #101 from perl6/autarch/fix-Rat.perl
18:56 dalek roast:
18:56 dalek roast: Tests that Rat.perl to actually generates valid code
18:56 dalek roast: review: https://github.com/perl6/roast/commit/8be23bb124
19:08 autarch MadcapJake: re: json file - yeah, I was thinking the same thing
19:24 rindolf joined #perl6
19:26 virtualsue joined #perl6
19:36 JRaspass joined #perl6
19:40 hankache joined #perl6
19:43 * hoelzro is back from coffee
19:44 hoelzro what's the best way to announce a branch for public testing? I really want people to try out the experiment multi-line REPL branch, to make sure it behaves well, doesn't break anything, and help me improve the code
19:44 hoelzro I admit that the change to introduce multi-line statements is more than a little hacky =/
19:44 moritz ask for testing here and on p6u
19:46 hoelzro thanks for the advice, moritz; I'll promote it again tomorrow morning when more people are on, and I'll follow up on p6u!
19:46 * hankache puts testing hat
19:59 lizmat m: dd Map.new.hash
19:59 camelia rakudo-moar 09181a: OUTPUT«Map.new(())␤»
20:00 lizmat that feels wrong, seems to me it should be the same as:
20:00 lizmat m: dd Map.new.Hash
20:00 camelia rakudo-moar 09181a: OUTPUT«{}␤»
20:00 lizmat m: dd Map.new("a",42).Hash
20:00 camelia rakudo-moar 09181a: OUTPUT«{:a(42)}␤»
20:10 autarch http://blog.urth.org/2016/02/07/ha​cking-on-perl-6-rakudo-and-roast/
20:18 dalek rakudo/nom: e387a76 | lizmat++ | src/core/Map.pm:
20:18 dalek rakudo/nom: Several Map optimizations
20:18 dalek rakudo/nom:
20:18 dalek rakudo/nom: - directly setting $!storage instead of using bindattr
20:18 dalek rakudo/nom: - using postfix unless instead of ||
20:18 dalek rakudo/nom: - using ternaries if possible
20:18 dalek rakudo/nom: - don't attempt to copy if there's nothing to copy
20:18 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/e387a76da3
20:19 mcmillhj joined #perl6
20:20 lizmat autarch++  # nice blog posts pointing to one of rakudo's USP's
20:21 BenGoldberg joined #perl6
20:24 * lizmat is starting to wonder what the exact semantics of .hash are
20:35 pmurias autarch: re why 2**31 I think that such numbers are just treated as big integers
20:35 magiczna_dupa hey I find smh in PSGI code that I can pass array into submodule
20:35 magiczna_dupa and it will autoextracto to parameters
20:35 autarch pmurias: that'd make sense
20:35 magiczna_dupa like in python
20:35 timotimo what are USPs?
20:35 lizmat Unique Selling Points
20:35 lizmat that what differentiate you from other suppliers
20:35 lizmat *differentiates
20:37 magiczna_dupa oh okay that not true
20:38 timotimo oh
20:39 RabidGravy right that's update functions basically implemented, now for attachments
20:42 geraud joined #perl6
20:43 jnthn lizmat: .hash is like .list, that is "given me something I can use in a hash-ish kind of way"
20:43 jnthn lizmat: I'd expect it to be identify for Map
20:43 jnthn .hash also de-itemizes
20:44 jnthn Since it's the thing used by %$foo
20:44 jnthn autarch: I'm only about a little today...
20:44 jnthn autarch: Anything I can answer quickly?
20:44 autarch jnthn: I think I figured out my question, actually
20:44 jnthn OK :)
20:45 autarch but you might want to take a look at my PR that lizmat merged to make sure I didn't break something you fixed a few years ago
20:45 autarch I'm fairly sure I didn't, but we (for some value of we) really need to get rakudo running via appveyor
20:46 ugexe https://github.com/ugexe/ze​f/blob/master/.appveyor.yml
20:48 jnthn lizmat: oops
20:48 jnthn if nqp::defined('$!storage') {
20:48 jnthn That's checking if the string is defined
20:49 jnthn autarch: It looks reasonable
20:49 autarch cool, thanks for checking
20:50 jnthn lizmat: (That was in your latest commit)
20:50 jnthn Rest looks fine
20:52 DubiousPepper joined #perl6
20:53 Hotkeys codegolf is v fun
21:04 magiczna_dupa should modules in perl be wrote as OOP design
21:05 magiczna_dupa I mean user should only see what he need no more
21:06 revhippie joined #perl6
21:08 gregf___ joined #perl6
21:11 SmokeMachine____ joined #perl6
21:12 raiph magiczna_dupa: "user should only see what he need no more" <-- That's how P6 is designed
21:12 raiph doesn't have to be OO
21:12 raiph (re: modules)
21:13 ely-se I have a few modules in the module index, but I want to move their repos to bitbucket
21:13 ely-se IIRC they have to be on GitHub
21:13 musiKk joined #perl6
21:17 autarch how can I build everything HEAD with rakudobrew?
21:18 lokien_ joined #perl6
21:21 Sqirrel_ joined #perl6
21:24 yoleaux joined #perl6
21:24 AlexDaniel joined #perl6
21:24 rpburkholder joined #perl6
21:25 ely-se joined #perl6
21:25 autarch I don't think I understand what "perl6 --version" is telling me - This is Rakudo version 2015.12-259-g5605d5f
21:26 autarch what is that commit from? I can't find it in the logs for rakudo or MoarVM
21:26 Sqirrel joined #perl6
21:27 ugexe 5606d5f, rakudo
21:28 MadcapJake autarch, it's «$*PERL.compiler.version()»
21:28 SCHAAP137 joined #perl6
21:28 autarch but that commit doesn't show up in the rakudo log
21:29 ugexe --configure_opts="--gen-moar=HEAD --gen-nqp=HEAD" or something like it should get your rakudobrew setup
21:29 dolmen joined #perl6
21:29 autarch the problem is the rakudo version, not the others "rakudobrew build moar HEAD" doesn't give me rakudo HEAD
21:30 RabidGravy I think rakudobrew triple HEAD HEAD HEAD
21:30 autarch cool, that seems to be working
21:31 ugexe i thought rakudobrew `moar` by itself was supposed to get HEAD
21:31 autarch I think it somehow figured out a tag and gets that - it's all a bit confusing
21:35 ugexe maybe it gets HEAD from master instead of nom? (or the other way around)
21:39 virtualsue joined #perl6
21:41 Hotkeys is there a method like 'min()' but for strings
21:41 Hotkeys ie choose the string with shortest length
21:42 Zero_Dogg joined #perl6
21:42 Zero_Dogg joined #perl6
21:42 flussence m: say <zzzz foo foobar>.min(:by(*.chars))
21:42 camelia rakudo-moar e387a7: OUTPUT«foo␤»
21:42 Hotkeys oh dang
21:42 Hotkeys I was going way too complicated
21:43 Hotkeys I had min("foo", "bar", :by({$^a.chars cmp $^b.chars}))
21:43 Hotkeys thanks flussence
21:43 kjs_ joined #perl6
21:46 magiczna_dupa joined #perl6
21:47 autarch it seems to get HEAD the first time you build, but on future build calls it does something else
21:51 Hotkeys flussence: http://i.imgur.com/eIh9rz6.png (for reference the returned value is $_)
21:52 Hotkeys it's treating r as greater even though it isn't
21:52 Hotkeys regardless of the value of $_
21:53 Amendil joined #perl6
21:53 yoleaux joined #perl6
21:54 flussence huh, no idea...
21:54 Hotkeys http://i.imgur.com/wjoOYw4.png
21:54 Hotkeys seems strange
21:55 dalek roast: 2212bf2 | (Brad Gilbert)++ | S07-hyperrace/ (2 files):
21:55 dalek roast: Test hyper and race in reverse
21:55 dalek roast: review: https://github.com/perl6/roast/commit/2212bf2335
21:55 dalek roast: 5016fe1 | (Aleks-Daniel Jakimenko-Aleksejev)++ | S07-hyperrace/ (2 files):
21:55 dalek roast: Merge pull request #99 from b2gills/reversed_hyper_race
21:55 dalek roast:
21:55 dalek roast: Test hyper and race in reverse
21:55 dalek roast: review: https://github.com/perl6/roast/commit/5016fe1ef2
21:55 Hotkeys m: ("foobar", 3000000).min(:by(*.chars)).say
21:55 camelia rakudo-moar e387a7: OUTPUT«3000000␤»
21:55 Hotkeys m: ("foobar", "foobarbaz").min(:by(*.chars)).say
21:55 camelia rakudo-moar e387a7: OUTPUT«foobar␤»
21:55 Hotkeys m: ("foobar", "3000000").min(:by(*.chars)).say
21:55 camelia rakudo-moar e387a7: OUTPUT«3000000␤»
21:55 Hotkeys m: ("foobar", "3000000".Str).min(:by(*.chars)).say
21:55 camelia rakudo-moar e387a7: OUTPUT«3000000␤»
21:55 * Hotkeys is very confused
21:57 Hotkeys m: ("300", "3000000").min(:by(*.chars)).say
21:57 AlexDaniel m: ("foobar", "30000000000000000".Str).min(:by(*.chars)).say
21:57 camelia rakudo-moar e387a7: OUTPUT«300␤»
21:57 camelia rakudo-moar e387a7: OUTPUT«30000000000000000␤»
21:57 Hotkeys m: ("3000000", "foobar").min(:by(*.chars)).say
21:57 AlexDaniel Hotkeys: honestly that's the first time I see :by :)
21:57 camelia rakudo-moar e387a7: OUTPUT«3000000␤»
21:58 Hotkeys AlexDaniel: it's undocumented for min except in supply
21:58 AlexDaniel Hotkeys: great!
21:58 Kristien joined #perl6
21:58 Hotkeys m: min("foo", "3000000", :by({$^a.chars cmp $^b.chars}))
21:58 camelia rakudo-moar e387a7: ( no output )
21:58 Hotkeys m: min("foo", "3000000", :by({$^a.chars cmp $^b.chars})).say
21:58 camelia rakudo-moar e387a7: OUTPUT«foo␤»
21:59 Hotkeys m: min("foo", "3000000", :by(*.chars})).say
21:59 camelia rakudo-moar e387a7: OUTPUT«5===SORRY!5=== Error while compiling /tmp/s_arJLzP3s␤Unable to parse expression in parenthesized expression; couldn't find final ')' ␤at /tmp/s_arJLzP3s:1␤------> 3min("foo", "3000000", :by(*.chars7⏏5})).say␤    expecting any of:␤      …»
21:59 Hotkeys m: min("foo", "3000000", :by(*.chars)).say
21:59 camelia rakudo-moar e387a7: OUTPUT«foo␤»
21:59 Hotkeys ohg
21:59 Hotkeys weird
21:59 Hotkeys m: min("foo", 3000000, :by(*.chars)).say
21:59 camelia rakudo-moar e387a7: OUTPUT«foo␤»
21:59 Hotkeys well that works for me
21:59 Hotkeys the dotty method not working is strange though
21:59 AlexDaniel m: ("foobar", "3000000".Str).min(:by({say $_.chars}))
21:59 camelia rakudo-moar e387a7: ( no output )
22:00 AlexDaniel does it even work?
22:00 AlexDaniel m: ("foobar", "3000000".Str).min(:byhoaesuhoae({say $_.chars}))
22:00 camelia rakudo-moar e387a7: ( no output )
22:00 AlexDaniel right
22:00 AlexDaniel Hotkeys: look, it does nothing
22:00 Hotkeys weird
22:00 Hotkeys I feel like it should be there
22:00 Hotkeys if it's there in other places
22:01 Hotkeys m: min("foo", 3000000, :byagdkjl(*.chars)).say
22:01 camelia rakudo-moar e387a7: OUTPUT«Unexpected named parameter 'byagdkjl' passed␤  in block <unit> at /tmp/se17MuQ3Xh line 1␤␤»
22:01 Hotkeys just checkin
22:01 AlexDaniel m:  min("foo", "3000000", :by({say .chars}})).say
22:01 camelia rakudo-moar e387a7: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tnpVTFvFr7␤Unable to parse expression in parenthesized expression; couldn't find final ')' ␤at /tmp/tnpVTFvFr7:1␤------> 3min("foo", "3000000", :by({say .chars}7⏏5})).say␤    expecting any of:␤ …»
22:01 AlexDaniel m:  min("foo", "3000000", :by({say .chars}))
22:01 camelia rakudo-moar e387a7: OUTPUT«7␤3␤»
22:02 AlexDaniel this one works, yeah
22:03 Hotkeys strange
22:07 autarch m: my $n = now; say $n.perl; say $n.perl.EVAL.perl
22:07 camelia rakudo-moar e387a7: OUTPUT«Instant.from-posix((<1744404512151/1199>, 0))␤Instant.from-posix((2.0, Bool::False))␤»
22:07 autarch another round tripping bug - the extra parens are screwing things up
22:10 Cabanossi joined #perl6
22:17 lizmat jnthn: good catch
22:19 perlawhirl joined #perl6
22:19 yqt joined #perl6
22:19 dalek rakudo/nom: 93a3479 | lizmat++ | src/core/Map.pm:
22:19 dalek rakudo/nom: Fix copy-pasto, spotted by jnthn++
22:19 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/93a34799e1
22:19 jnthn lizmat: :)
22:22 ely-se well then, I think all my Perl 6 modules broke
22:23 lizmat ely-se: recently?
22:24 ely-se A few moments ago. Because I deleted all my GitHub repositories.
22:24 ely-se I should remove them from the module list until non-GitHub hosts are supported and before people start depending on them
22:24 lizmat autarch: can we expecte a PR ?
22:25 autarch lizmat: I'm so sleepy right now, but probably at some point - I think this one is a much easier fix, actually
22:26 * lizmat wonders what autarch's timezone is
22:26 lizmat it being pretty late in CET
22:26 autarch heh, it's 16:30 here I've just been feeling lazy
22:26 lizmat ah...  :-)
22:26 lizmat siesta  :-)
22:32 bowtie joined #perl6
22:33 sotka joined #perl6
22:37 pmurias ely-se: why did you remove all your GitHub repos?
22:38 ely-se pmurias: political reasons
22:38 ely-se I'm not using GH anymore
22:39 pmurias did GH do anything outrageous recently?
22:42 autarch some people seem to think so
22:43 pmurias autarch: what did they do?
22:43 dalek rakudo/nom: f59c769 | lizmat++ | src/core/Hash.pm:
22:43 dalek rakudo/nom: Scrape a few % off of Hash's introspection methods
22:43 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f59c76939f
22:44 autarch http://www.businessinsider.com/gi​thub-the-full-inside-story-2016-2
22:44 ely-se pmurias: http://hintjens.com/blog:111
22:46 AlexDaniel ely-se: so have you moved everything somewhere else?
22:46 ely-se I moved everything to Bitbucket and I'm soon evaluating gitgud.io
22:46 ely-se luckily Git is a DVCS so I can keep it on both BB and GG
22:46 AlexDaniel ely-se: did you have any perl 6 modules? If yes, then how is it going?
22:47 ely-se I did, they aren't there anymore. If you try to install them with Panda now you get an error
22:47 * flussence was half-tempted to add my last thing to META.list via a gitlab url just to see what exploded, but it was in mid-december so I held off :)
22:47 AlexDaniel ely-se: yes, that's what I've been doing for some other projects too. There's no reason not to have several mirrors
22:47 ely-se I'm trying to figure out how best to remove them from META.info without using GH pull requests
22:48 ely-se until Panda supports non-GH repositories
22:48 AlexDaniel ely-se: yeah, we should have more people like you
22:48 lizmat I thought the URL was pretty agnostic about where things lived ?
22:48 AlexDaniel that's what I thought too, but who knows
22:49 ely-se It's late so I'll check it out tomorrow
22:49 AlexDaniel ely-se: if you can contribute anything to support non-GH repos please do
22:49 AlexDaniel ely-se: it will make lots of people happy (me)
22:49 flussence ely-se: I can remove 'em if you want, what username?
22:49 ely-se ok :p
22:49 ely-se flussence: rightfold
22:49 ely-se I think it's only the three topmost in META.info
22:50 hoelzro MadcapJake: btw, thanks for fixing my screwup with linenoise earlier =)
22:50 hoelzro MadcapJake++
22:50 ely-se they're now on BB, e.g. https://bitbucket.org/rightfold/nativecall-errno
22:50 crux joined #perl6
22:50 ely-se problem is that you can't have version-agnostic links to source files on BB, so you can't use it with META.info that way
22:51 ely-se META.list*, unless you want to update META.list for each revision
22:52 dalek ecosystem: f54969a | (Anthony Parsons)++ | META.list:
22:52 dalek ecosystem: g/github.*\/rightfold/d, per request in #perl6
22:52 dalek ecosystem:
22:52 dalek ecosystem: http://irclog.perlgeek.de/p​erl6/2016-02-07#i_12003067
22:52 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/f54969a810
22:53 flussence that url thing sounds kinda awkward, gitlab has a similar problem where it's impossible to get tarballs with a consistent md5sum
22:54 geekosaur ? github does that too, seems to gc tarballs if unused for a while an make new ones with different md5sums on demand
22:55 flussence huh... maybe they're stable if you use the "releases" feature
22:55 flussence I know a few gentoo packages pull straight from the site, so there has to be some way to do it
22:56 BenGoldberg joined #perl6
22:58 AlexDaniel ely-se: what's the difference between gitgud.io and gitlab itself?
22:58 ely-se gitlab is software, gitgud.io is a running instance of that software
22:59 ely-se you can host gitlab yourself if you want
22:59 pmurias ely-se: so it's mostly a fear of github trying to rake in profits by turning into an evil empire and taking over everything to rake in profits?
22:59 ugexe zef should handle bitbucket repos (and local/remote tar.gz archives)
22:59 AlexDaniel ely-se: no-no I mean if you hit “Explore” button on gitlab.com you can see some projects hosted there
23:00 ely-se Oh, no idea.
23:00 flussence gitlab.com runs an instance of their commercial version, so you get a few more buttons to click
23:01 ely-se the thing I like about gitgud is that it claims to be "100% politically neutral"
23:01 flussence (it's also *full* of automated spam, it seems)
23:01 AlexDaniel so if I want to host something, should I prefer gitlab.com or gitgud.io?
23:01 jast claims are cheap...
23:01 AlexDaniel ely-se: yeah I kinda liked that too
23:02 AlexDaniel jast: well claims are cheap but at least the source of gitlab is released while github is proprietary. You might say that it doesn't really matter because you're still using git, but in my opinion one is definitely better than the other in that respect
23:03 jast there's also notabug.org
23:03 jast I was talking about the claim of political neutrality mostly, which is mostly unrelated to how closed the source is
23:03 jast and in this particular case I don't personally care about whether the source code is available, I only do if I host it myself
23:04 AlexDaniel jast: but the difference is that you can move to self-hosted solution as soon as you don't like anything
23:04 AlexDaniel jast: or move to another gitlab hoster which looks exactly the same
23:04 geekosaur I am somewhat biased about gitlab because they handled buying and shutting down gitorious very badly
23:04 AlexDaniel /anything/something/
23:05 jast geekosaur: what do you think they could/should have done better?
23:05 jast short of not shutting it down at all :)
23:06 AlexDaniel jast: e.g. moving from github to something else is going to painful for sure (but still possible, yeah). Moving between different gitlab hosters is supposed to be easy (at least in theory it is so). I think that that's why the source code matters in that case
23:06 geekosaur well, for one, gitorious suffered some failure just before the start of the migration period. they didn't extend the migration period and gitorious was pretty broken/unable to migrate stuff for half of the period
23:06 jast AlexDaniel: the difference is that gitlab actually started out as open source and didn't offer hosting for the longest time
23:06 geekosaur which made it really painful to move stuff away in time
23:07 jast and, well, personally I wouldn't consider hosting gitlab myself, it's a bit much
23:07 AlexDaniel jast: I know. I was really disappointed when they started commercializing it
23:07 jast it was enough having to set it up at work, despite my misgivings
23:08 * flussence isn't willing to self-host anything more complex than gitweb/cgit, and those are horrible to hack on :(
23:08 jast we had a perfectly good gitolite setup (basically a headless git repo manager), and we're not even using gitlab's pull requests and issues and wiki and such
23:08 AlexDaniel anyway, we should be more TIMTOWTDISH with that. I've already added support for gitlab and bitbucket on camelia, but we should definitely do more.
23:08 geekosaur so I keep thinking "is that how they're going to handle gitlab outages too?"
23:08 AlexDaniel m: https://gitlab.com/snippets/14172
23:08 camelia rakudo-moar f59c76: OUTPUT«((a a a) (b b) c)␤»
23:09 jast geekosaur: fair enough. in fact they're having scaling issues already, although they say they're working on fixing that
23:10 jast yeah, supporting more platforms is definitely good
23:13 pmurias joined #perl6
23:13 lizmat .tell jnthn do you have an idea as to when object hashes would make it to Moar ?
23:13 yoleaux lizmat: I'll pass your message to jnthn.
23:15 dalek rakudo/nom: 63fdd19 | lizmat++ | src/core/Hash.pm:
23:15 dalek rakudo/nom: Scrape a few % off of initializing a Hash
23:15 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/63fdd19e6d
23:19 perlawhirl joined #perl6
23:20 captain-adequate joined #perl6
23:24 dalek rakudo/nom: f0a4504 | lizmat++ | src/core/Hash.pm:
23:24 dalek rakudo/nom: Slice off one frame for each Hash.ASSIGN-KEY
23:24 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/f0a45046d1
23:28 wamba joined #perl6
23:28 dalek roast/autarch/fix-Instant.perl: ade9e00 | autarch++ | S02-types/instants-and-durations.t:
23:28 dalek roast/autarch/fix-Instant.perl: Add a test for the round trip of Instant via .perl.EVAL
23:28 dalek roast/autarch/fix-Instant.perl: review: https://github.com/perl6/roast/commit/ade9e00fe1
23:29 AlexDaniel lizmat++ # I love that stuff
23:29 AlexDaniel progress!
23:30 autarch if someone wants to give me access I could close the RT bugs I'm fixing
23:31 AlexDaniel autarch: access to what?
23:31 AlexDaniel RT?
23:31 autarch access to edit tickets in rt
23:31 AlexDaniel autarch: ok, so you have to create bitcard account
23:31 autarch I have a bitcard account
23:32 autarch I just don't have rights to the queue
23:32 AlexDaniel autarch: what do you see when you log in?
23:32 autarch I see a bunch of RT stuff ;)
23:32 AlexDaniel hmmm
23:32 autarch I don't have perms to manage the queue, basically
23:32 AlexDaniel ah!
23:33 AlexDaniel right…
23:34 autarch although that begs the question of whether bugs should be marked resolved when a fix is pushed or when that fix is in a release
23:34 AlexDaniel right, so it seems like you can close your own stuff but can't close other tickets, hmm
23:35 kjs_ joined #perl6
23:36 KotH joined #perl6
23:40 dalek roast/autarch/fix-Instant.perl: 8650ef2 | autarch++ | S02-types/instants-and-durations.t:
23:40 dalek roast/autarch/fix-Instant.perl: Add a test for the round trip of Instant via .perl.EVAL
23:40 dalek roast/autarch/fix-Instant.perl: review: https://github.com/perl6/roast/commit/8650ef2979
23:42 autarch lizmat: PRs made
23:43 azawawi joined #perl6
23:43 azawawi hi
23:44 azawawi https://github.com/azawawi/atom-perl6​-editor-tools#context-sensitive-help   # animated gif screenshot to explain what it does... ctrl-f5 to refresh if needed
23:45 dalek rakudo-star-daily: 0e0c7c8 | coke++ | log/ (9 files):
23:45 dalek rakudo-star-daily: today (automated commit)
23:45 dalek rakudo-star-daily: review: https://github.com/coke/rakudo​-star-daily/commit/0e0c7c87e8
23:50 Hotkeys wait so what did github do?
23:56 atweiden joined #perl6
23:56 lizmat autarch: commented on PR

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo