»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
timotimo "such an old version" is a bit extreme for 2016.11, but still 00:00
Geth_ gtk-simple: b66db46864 | (Timo Paulssen)++ | .travis.yml
bump lowest rakudo version for zef compatibility
timotimo let's see how that goes
when i've found a working version i'll also bump appveyor 00:01
oh, huh. that's not clever ... 00:02
our travis script for rakudo does "make test" then "make install"
but if the core setting step explodes, it'll try to do the core setting again
which on the jvm version takes almost 400 seconds (on travis, that is) 00:03
Geth_ gtk-simple: a6938136dc | (Timo Paulssen)++ | .travis.yml
bump lowest rakudo version for zef compatibility
 00:08
timotimo .o( i hope it's .01 and not .1 )
benjikun timotimo: I appreciate you doing all of this just because I brought it up lol 00:09
timotimo well, the CI is hardly useful if nobody looks if it succeeds or fails to build
TEttinger I think it's also just important that stuff works :)
timotimo the tests in gtk-simple hardly do anything ;_; 00:11
00:12 El_Che left, El_Che_ joined
timotimo anyway, looks like 2017.01 is new enough for zef 00:12
i wonder if zef should warn for earlier versions?
ugexe its rakudo, not zef that is broke 00:13
timotimo of course
00:13 MilkmanDan left
timotimo but it'd be helpful if zef said "hey btw this rakudo is too old yo" 00:13
rather than the ominous and unhelpful "expected 2 parameters but got 1" 00:14
00:14 MilkmanDan joined, mr-foobar left, someuser left
timotimo wait, i said it's new enough but it hadn't even built rakudo yet, how was it supposed to maybe-fail 00:14
yeah, there it goes
ugexe: you happen to know which rakudo version fixed this problem?
00:15 someuser joined
Geth_ gtk-simple: d50c5dc437 | (Timo Paulssen)++ | .travis.yml
bump lowest rakudo version for zef compatibility
 00:15
ugexe clone zef for around the same time of your rakudo version if you want to use old rakudo
timotimo hm, right
actually, rakudobrew should do that
benjikun timotimo: How early did you have to claim timo on github for it not to be taken 00:16
timotimo (actually actually, maybe don't use rakudobrew on travis)
good question, let's see if my very first commit is there
ugexe rakudobrew wont do that
it could grep the git log or something to do it, but it doesnt 00:17
timotimo benjikun: i already had it in december 2008 it looks like
well, yeah
benjikun I call my cat timo sometimes
his name is Timothy
timotimo i bet it's a good cat
benjikun it sure is
timotimo there's commits on github from march 2008 but i can't actually tell when i created the github account 00:19
i.e. if i made these commits off of github and later (maybe much later) pushed them there
benjikun hmm 00:20
timotimo but i expect around that time to be correct 00:21
github wasn't very big when i got on
benjikun I'm honestly thinking about making and using a private repo thingy
and abandoning my github 00:22
ugexe any module that spawns procs in its Build.pm should require minimum 2017.07 imo
timotimo ah is that what causes the problem in gtk-simple?
ugexe probably not the one you are seeing, it usually is prone to deadlocks 00:23
timotimo oh, ouch 00:24
Geth_ gtk-simple: df15c903af | (Timo Paulssen)++ | 2 files
bump the "old rakudo" version by a lot

to benefit from zef fixes and such
00:30 Cabanossi left 00:32 mscha joined
mscha m: my %h{num}; %h{1e0} = 42; 00:32
camelia Type check failed in binding to parameter 'key'; expected num but got Num (1e0)
in block <unit> at <tmp> line 1
00:32 Cabanossi joined
mscha What's the point of allowing %hash{num} if you can't put anything in it? 00:33
ugexe m: my %h{num}; %h{num} = 1;
camelia ( no output )
mscha OK, you can put a type object in it. Very useful... :-) 00:34
ugexe killer language feature
geekosaur this is more a shortcoming in our handling of boxing/unboxing numbers 00:35
(I should say Num vs. num; things like Int vs. int work, mostly)
ugexe looks liek it gets coerced 00:36
m: use nqp; my %h{num}; %h{nqp::unbox_n(1e0)} = 1;
camelia Type check failed in binding to parameter 'key'; expected num but got Num (1e0)
in block <unit> at <tmp> line 1
00:36 itaipu left
timotimo yup, we're notg ood at that one yet 00:40
geekosaur actually that's also an LTA error on top of being wrong
well, on top of being LTA behavior
timotimo japhb: i just tried tris.p6 (tris di pasta?); fantastic! japhb++ 00:53
looking forward to being able to use arrow keys and such
00:58 Erroneous_ joined 01:03 BenGoldberg left
japhb timotimo: Yeah, now that I can get the individual codepoints reliably, I can start decoding key escape sequences. :-) 01:17
timotimo don't forget to kick out the comment about libuv and threads 01:25
and i expect you don't have to open /dev/tty
benjikun How hard would it be to implement an ssh client in perl6 01:31
01:33 Zor joined
timotimo well, there's already a libssh binding that does async i/o 01:38
01:45 ilbot3 left, Cabanossi left 01:48 Cabanossi joined
zengargoyle i seem to remember an article on RSA (or some other encryption) that had quite minimal p6 by using all of the expmod() and prime built-in mathy things. 01:49
TEttinger expmod? 01:50
exponent in a modular arithmetic field?
zengargoyle something like e**x mod n but yeah, but not the obvious simple math thing. 01:51
01:52 ilbot3 joined, ChanServ sets mode: +v ilbot3
timotimo yeah, there's algorithms for expmod that perform a whole lot better than exponentiating and then modding 01:54
both in terms of speed and memory usage
bedtime 02:00
02:07 vendethiel joined 02:15 bjz joined 02:29 araraloren joined 02:35 M-Illandan joined 02:45 Cabanossi left
SmokeMachine m: say "bla:ble".split: /:/ 02:45
camelia 5===SORRY!5=== Error while compiling <tmp>
Unrecognized regex metacharacter : (must be quoted to match literally)
at <tmp>:1
------> 3say "bla:ble".split: /:7⏏5/
expecting any of:
argument list
term
SmokeMachine m: say "bla:ble".split: /":"/ 02:46
camelia (bla ble)
SmokeMachine :)
02:48 Cabanossi joined 02:55 noganex_ joined 02:58 noganex left 03:05 Zor left
araraloren morning 03:17
String is faster than regex 03:18
03:18 Erroneous_ left 03:20 rubio_terra left
SmokeMachine araraloren: yes, but my real case is /\s* ":" \s*/ 03:21
araraloren SmokeMachine, I test it, split with ":" the trim is faster than split by regex 03:27
s/the/then/
03:28 Cabanossi left 03:31 Cabanossi joined 03:32 bjz left 03:36 skids joined 03:39 bjz joined
araraloren .tell andrzejku After we add the run configuration, it was working! 03:46
yoleaux araraloren: I'll pass your message to andrzejku.
araraloren .tell andrzejku Though it has a different with what we thought . 03:47
yoleaux araraloren: I'll pass your message to andrzejku.
04:00 itaipu joined 04:14 ryu0 joined 04:16 itaipu left 04:23 mscha left 04:29 Cabanossi left 04:31 Cabanossi joined
jeromelanteri hello all 04:44
with Git::Wrapper, how to get any unsucces/errors/failed things who should happen eventually ?
araraloren catch exception ? 04:50
jeromelanteri araraloren, there is no exception to catch... command is send, then... i hope to have with this module a kind of "success" variable... 04:56
an other one question, i have many modules <add remove modify install uninstall update> and some sub functgions had same name: <all_repos repo_from_group, etc...> 04:57
then there is conflicts the way i do declare things... our sub all_repos($things) is export(:MANDATORY) {...}
from there, i can call: add::all_repos(my_things); 04:58
i want to be able to call modules by there own name, and keep same names functions without have problems... it should be possible correct ? but how to ?
araraloren jeromelanteri, do you mean, call $whatmode:add_repos(...) ? 05:09
s/mode/module/
geekosaur I am not understanding the question here 05:11
geekosaur is guessing all those uses of "module" should be "method"? 05:12
araraloren jeromelanteri, I check Git::Wrapper source code, and I think it has no something like **success** , even not handle error message
geekosaur in which case: a method call has an invocant, a sub doesn't. 05:13
$foo.method_here(...)
vs. some_sub_here
er, some_sub_here(...)
05:14 ilbot3 left
geekosaur if you actually mean modules, giving them names like add or remove is a really bad idea because yes, it conflicts. I'm not sure it even makes sense to split modules that way... 05:16
but if somehow it does, a namespace prefix is likely a good idea. use Foo::add; Foo::add::somesub(...); 05:17
also you should be aware that lowercase module names are reserved for pragmas
05:18 ilbot3 joined, ChanServ sets mode: +v ilbot3
geekosaur (as are all-uppercase names) 05:18
ryu0 I have experience mostly with traditional unix languages (C, Awk, Shell), with C being the one I have the most experience with. I have been looking for a language to supplement these, but I have no prior experience with Perl. Would Perl be a good fit for me? If so, why might it be? 05:19
geekosaur for example, "use fatal;" in a scope causes all Failures (delayed exceptions that throw if used without testing/defusing them first) to become normal exceptions
ryu0, with that background perl 5 is likely the better fit... but it's always good to expand your horizons 05:21
araraloren ryu0, here is Perl 6, Perl 5 is borrow many idea from C Awk Shell
ryu0 geekosaur: why? i was thinking perl6 might be more in line with what i'm wanting to do.
geekosaur and these days it's hard to get by without experience in *some* OO language. (perl 5 and python don't really count, their OO is baling wire and chewing gum, aside from p5's Moose which instead s painfully slow) 05:22
ryu0 I have also dabbled in functional languages, but wasn't too thrilled with them.
Mostly their syntax was weird.
araraloren Perl 6 is actually modern language
ryu0 Lisp in particular.
araraloren with multi programming form 05:23
geekosaur lisp certainly takes some getting used to. there are functional languages that aren't Lisp/Scheme or ML family, though; take a look at scala.
ryu0 I actually liked the SML variant of ML.
Was easier to relate to.
I also tried Go, so I do have some experience with OO constructs. 05:25
C++ is something i've also had to use at points.
jeromelanteri araraloren, geekosaur yes, i'm talking about modules, so, sub function, not method class object nor instance things. 05:26
unit module a; our sub blabla() is export(/MANDATORY) { } // use a; a:blabla(); works, but if i have: 05:27
unit module b; our sub blabla() is export(:MANDATORY) {} // use a; use b; a::blabla() (error... two functions has the same name) 05:28
geekosaur don't use all-lowercase module names; they are reserved for use as pragmas. (and a few internals modules that you proably don't want to mess with)
also, you are basically asking for that when you export the names; you are telling perl 6 to pull those names into your namespace. 05:29
although qualified *should* work.
since they're "our"
jeromelanteri geekosaur, ok 05:30
so actually, what can be the solution ?
if i remove "our", i can no more call A::blabla(); 05:31
05:31 benjikun left
geekosaur I am still not understanding 05:38
you want to force multiple declaration? what is supposed to happen?
I mean, you are basically creating an impossible situation and saying that you think it is the correct way to do things
you are exporting with forced import suns with the same name from multiple modules 05:39
the result will be Cannot import symbol &foo from Bb, because it already exists in this lexical scope
either don't export and use qualified names, or export with a tag so users can choose to import or not as needed 05:40
jeromelanteri geekosaur, each function is embeded inside a module...
geekosaur if I kill the is export(:MANDATORY) then I can call Aa::foo and Bb::foo with no problem
jeromelanteri it can have a function in module A same name that an other one function in module B 05:41
geekosaur if you force export with is export(:MANDATORY) then it cannot *import* two modules doing so with the same name into the same scope
jeromelanteri then call by A::bla(); B::bla(); for make this works ? why impossible ? it should be possible by the fact they are embeded in different modules...
geekosaur that does work
what does not work is "use"ing both of those, because they both try to define bla() *in the module that "use"s them* 05:42
because you said is export(:MANDATORY)
jeromelanteri geekosaur, so just remove is export(:MANDATORY) should do the job i want ?
geekosaur yes
you don't need to export them, you made them "our" 05:43
05:43 pilne left
jeromelanteri geekosaur, oh ok, fine... 05:43
so export is for put them in the scope with a kind of tag (:MANDATORY is the default one)
geekosaur, thank you 05:44
geekosaur export is to make them appear as if they were defined in the module that "use"s them
jeromelanteri ok
05:44 khw left
geekosaur you can see why this is a problem if you do that to functions with the same name in two different "use"d module 05:44
05:44 Cabanossi left
jeromelanteri so yes, definitly, by call module name, i absolutly no need is export 05:45
geekosaur you could put them on some other tag so people can choose to import them into their namespace, if they know it won't cause a problem, by specifying that tag
use Foo :IMPORTS; # or some such
jeromelanteri geekosaur, ho yes, that should be an other one solution
geekosaur then you can choose which one gets to use the unqualified names
jeromelanteri but then, is use A :Tag_A; use B :Tag_B with modules contain some same named functions... it should not works same or it should works ? 05:46
geekosaur that would fail
05:46 Cabanossi joined
jeromelanteri ok, well understand 05:46
it just call by a tag functions to export in the scop of the actual module use them... 05:47
geekosaur Haskell imports work the way you want, but it comes with other behaviors that surprise people
jeromelanteri yes
geekosaur there's no nice way to do things so it all somehow "just works"; *something* will be annoying
jeromelanteri geekosaur, you may be able to understand (if you want also) that this objection doesn't care about the fact i am not english and my vocabulary is limited... 05:48
geekosaur, but do what you want, and thank you for your help.
geekosaur I had noticed (you seem to be translating some French idioms that don't quite work in English) 05:49
sadly my French is not very good
jeromelanteri yes, i'm french 05:50
araraloren jeromelanteri, you can show him the module code
jeromelanteri i will not hurt you the day you will try to write french and not find the precise best words for explain your problem, don't worry.
araraloren, yes i can: bitbucket.org/jerome___/vimplugger/src 05:51
all modules actually... i'm going to just remove all export things
geekosaur I'm not complaining, just noting.
jeromelanteri geekosaur, nop
araraloren Actually , I think Module.method is not bad thing 05:52
jeromelanteri what is "module.method" ? i thank that method is not module, but class method
module.function and class.method.. isn't it ?
geekosaur I would expect module.method to look for methods defined on a Module 05:54
araraloren, I think you wanted :: not . 05:56
jeromelanteri geekosaur, i didn't no it was possible... i always thinki (so i just start learning perl6 one week an half ago) that method is not module but for class... or maybe all file finished by .pm6 is a module... also if there is class inside.
s/no/know
geekosaur a module (and other things that subclass Package, like Class) is an object like anything else, with methods. but introspection doesn't seem to work :/ 05:57
jeromelanteri well..; that makes me confuse actually.
geekosaur m: module A {}; .say for A.^methods 05:58
camelia No such method 'methods' for invocant of type 'Perl6::Metamodel::ModuleHOW'
in block <unit> at <tmp> line 1
geekosaur the MOP for Module is incomplete :/
module A {}; say A.WHAT
jeromelanteri :)
geekosaur er 05:59
m: module A {}; say A.WHAT
camelia (A)
geekosaur m: module A {}; say A.HOW
camelia Perl6::Metamodel::ModuleHOW.new
ryu0 geekosaur: So then why might I want to use Perl6? I don't have much of an opinion about either major version of Perl.
jeromelanteri so sub is for declaration of module.method in perl6 speaking "nice" ? 06:00
geekosaur ryu0, I'm not really the right prson to ask.. especially right no when the only reason I'm still up is someone set off the fire alarms (it's "all clear" but they're still making it impossible to sleep...) 06:02
also apparently I can't type... oy
ryu0 Ok.
jeromelanteri geekosaur, you taold something interesant about pragmas and modules names conventions... 06:03
what is a pragmas ?
do i need then to move files modules names in same convention (upper case for first letter) ? 06:04
geekosaur to the second, yes 06:06
jeromelanteri also, i have a files names datas.pm6 who doesn't contain a module, but a class... do i need to do same with this class name and the datas.pm6 file ?
ok, move modules files names to uv first letter, and declare them same convention... and for the class inside the datas.pm6 file ? same ? 06:07
geekosaur classes are the same way, yes.
jeromelanteri ok
thank you
and a link for explain pragmas ?
geekosaur as for pragmas, they're generally instructions to change the compiler / language's behavior in the current scope
docs.perl6.org/language/pragmas
so, for example, ``` use lib "/usr/lib/ourstuff"; ``` tells perl 6 to look for modules in the specified directory as well as the usual places 06:08
```use fatal;``` turns delayed exceptions ("Failure") into normal exceptions 06:09
jeromelanteri ok
it is kind of modules reserved then ?
araraloren geekosaur, yeah, class module, or class
zengargoyle hehe... 06:10
geekosaur so all-lowercase module names are reserved, all-uppercase module names are reserved (generally to enable dangerous things) 06:11
araraloren If I have a sample.pm6, do I have to declare module with: unit module sample; 06:14
or the `sample` can be other thing else ?
geekosaur "unit module" is just a way to save on braces
oh. they have to match, yes
araraloren So when I `use sample;` what the sample mean ?
mean the filename ? 06:15
geekosaur when it sees that, it has to go out and look for a file with that name.
it can't ask something if there is some file somewhere that contains a perl 6 module definition
(at least, not if you want it to actually run today...)
zengargoyle jeromelanteri: remember i said that you try too hard with export and easy single name modules like 'check', 'install', 'add'? it would go easier if you did VimPlugin::Check, VimPlugin::Install, VimPlugin::Add and then made it easy after it works. 06:16
geekosaur I suggested that earlier, actually
not sure if it was understood
araraloren, so when it sees 'use module Foo;' it checks a number of standard places (which you can add to with a ```use lib 'wherever'```; pragma) for files named Foo.pm6 or Foo.pm 06:17
zengargoyle geekosaur: i made a repo with mostly decent distro standards but jeromelanteri didn't wan't to even look ATM.
jeromelanteri zengargoyle, yes, i miss understood this part. 06:18
zengargoyle it is convention that most people do not break, so nobody knows how to not do it that way.. :)
araraloren geekosaur, oh, it make sense, thanks
jeromelanteri zengargoyle, so not VimPlugin, but Add::function_name ...
zengargoyle better, but what if you ever write another program with 'Add' modules? 06:19
geekosaur someday it might be interesting to have object filesystems where e.g. perl 6 could register a handler with it to let it respond to "find a perl 6 module named Foo". but currently nothing knows how to use such things, unless you count the simplified forms built into various GUIs (windows explorer, os x finder, gnome/kde/whatever, etc.)
jeromelanteri zengargoyle, an other one programm with Add module should not be mixed in the same lib... but ok, i understand what you mean. 06:20
zengargoyle you can not easily have two different Add or Check modules, but you can easily have VimPlugin::Add and VimPlugin::Check.
jeromelanteri zengargoyle, look at my code, i do modify it... i think there is no VimPlugger::things... 06:21
geekosaur jeromelanteri, maybe more to the point: someday you may want to use a program that wants to use Add from both
zengargoyle yeah, if you don't realy want to make a module you can install, or ever mix parts of different projects together... short Add will do. :P
geekosaur let's say you have a git library with Add::whatever, and a vim library with Add::whatever, and at some point you want to use them to add git integration to your vim setup
"oops"
06:22 statisfiable6 left
zengargoyle jeromelanteri: i know is feels a bit bad to have SomeLong::Name::For::A::Module, but that almost always works, and there are ways to do short name after module is finished. 06:24
jeromelanteri zengargoyle, that also was my idea... or i missunderstand you. Could you please look at my code for be sure ? 06:25
geekosaur m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; constant RLN = ReallyLongAnnoyingName; RLN::foo; 06:26
camelia foo!
06:26 darutoko joined
zengargoyle m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; my \foo = ReallyLongAnnoyingName::foo; foo; 06:27
camelia WARNINGS for <tmp>:
Useless use of foo symbol in sink context (line 1)
foo!
geekosaur (this, btw, is one reason why modules are objects)
zengargoyle heh
geekosaur hrm. that's LTA
zengargoyle m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; my &foo =&ReallyLongAnnoyingName::foo; foo;
camelia foo!
zengargoyle m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; my \foo =&ReallyLongAnnoyingName::foo; foo;
camelia WARNINGS for <tmp>:
Useless use of foo symbol in sink context (line 1)
zengargoyle guess sigil matters in assignment. 06:28
06:29 wamba joined
geekosaur actually... 06:29
m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; my \foo = ReallyLongAnnoyingName::foo; say "bar"; foo;
camelia WARNINGS for <tmp>:
Useless use of foo symbol in sink context (line 1)
foo!
bar
geekosaur see what it did there?
jeromelanteri zengargoyle, geekosaur the idea is not to be annoying, but to give a name that make understand what it is supposed to do... then other coming (and me to) should be able to understand what the function want to do. 06:30
geekosaur oh, I know.
jeromelanteri but sure... i can change names by LongAnnoyingName => A. It should quickly become unreadable. 06:31
geekosaur er. that wasn't a response to you
zengargoyle call vs name for hot using &Really....
geekosaur m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; my \foo = &ReallyLongAnnoyingName::foo; say "bar"; foo();
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '&foo' is not declared
at <tmp>:1
------> 3ReallyLongAnnoyingName::foo; say "bar"; 7⏏5foo();
zengargoyle because no &
geekosaur yeh
zengargoyle but i sorta thought my \foo = &R... would maybe work.
geekosaur m: module ReallyLongAnnoyingName { our sub foo { say "foo!" } }; my \foo = &ReallyLongAnnoyingName::foo; say "bar"; foo.();
camelia bar
foo!
zengargoyle thought that was basically a bind.
geekosaur it is. but a bare simple term like "foo" is special cased 06:32
zengargoyle ah. *nods*
geekosaur so you need the parens, but then it explodes because that too is a special case (specifically looks for & sigil), so have to use .() 06:33
jeromelanteri wel.. all is updated on bitbucket. And now it works fine.
thank you
geekosaur every DWIM has a WAT, etc.
zengargoyle jeromelanteri: i will look later...
my brain still hurts from Git::Wrapper :P 06:34
jeromelanteri zengargoyle, it is not an obligation. But i think we were speaking about the same idea but your example seems to talk about my code without read it (so it makes me strange i xas not sure to well understand what you mean). Git::Wrapper was a real good fix for me... 06:35
Gt::Wrapper doesn't have any success "getter"... so i have to do something for read if clone things worked or not... it will be complicate i think. 06:36
it oesn't have any error getter to...
in perl5 RPC things has scucces catch...
success
zengargoyle does p6 have the same thing as some p6 Exporter modules where use Foo bar => { -as => 'baz' }; or similar makes &baz = &Foo::bar ? 06:38
araraloren The namespace is not a easy thing can better control, like c++, you must not define a function already exists in **std** 06:39
zengargoyle jeromelanteri: yes, i need to pull code again before looking.
araraloren EXPORT ?
zengargoyle jeromelanteri: and yeah, Git::Wrapper i don't think is a *good* module. captures stderr and throws it away, doesn't check system() return codes, just assumes it works and returns stdout. 06:41
jeromelanteri zengargoyle, but then use "run" do the same job... i not well understand the interest of Git::Wrapper if it can not do that. Maybe it will grow on this way ? 06:42
zengargoyle somebody needed it and did the thing that worked and kept going with original problem. :P
06:42 raiph left
zengargoyle i have done worse myself. :) 06:43
jeromelanteri zengargoyle, if i wasn't so nOObs i would add some features...
zengargoyle i have done better myself. :)
jeromelanteri actually, i can not i think.
zengargoyle, please... do it now... 06:44
06:44 Cabanossi left
jeromelanteri i promise i will put your name in the title of my poor nOObs bitbucket repo... 06:44
zengargoyle i did a lot of Wrapper at last $WORK. but don't know p6 run/shell stuff that well yet. 06:45
jeromelanteri (crying) 06:46
zengargoyle (sometimes the CLI program that is Wrapped doesn't even have decent error or return code.... )
jeromelanteri i failed to find some situations for git...
06:46 Cabanossi joined
jeromelanteri the idea is to check repo exist (that is ok... Validate), and also to tell if clone/pull requests success or not 06:47
(i said that is ok... but i'm not so sure in all situation of url that has been given) 06:48
06:49 espadrine joined
zengargoyle i have not used git in an automatic way. always more manual and seeing errors and 'not working' was an OK thing for me. 06:51
so i do not know git error code or failure mode that well. usually a typo or internet is broken :)
jeromelanteri same 06:52
06:53 skids left
araraloren That Git::Wrapper has a little problem in META6.json 07:02
07:03 abraxxa joined 07:04 rindolf joined
zengargoyle araraloren: what is the problem? i have another pull-request for not working when precompiled. (so jeromelanteri is using a patched Git::Wrapper anyway). 07:05
araraloren the provides 07:06
jeromelanteri zengargoyle, merged with master
used also the perlpilot one
(but this should not make problems) 07:07
wich branch are you in ?
zengargoyle oh, cool. i didn't see a github email. :)
jeromelanteri because it is not github... it is bitbucket
07:08 abraxxa left
jeromelanteri zengargoyle, but... if you want to push on it, you are welcome 07:08
zengargoyle jeromelanteri: we are missing communication. :)
araraloren and i were talking about source: git://github.com/nicqrocks/p6-Git-Wrapper.git 07:09
jeromelanteri yes
zengargoyle i think. that is not bitbucket....
jeromelanteri i do speak about this one in my README to
araraloren, normally, if you read all the content of my readme file, you should find the solution by do things i explain. (but if not, tell me and i will correct my README file for be more clear) 07:11
07:11 setty1 joined
jeromelanteri zengargoyle, ho well ... we cross speaking different things. 07:11
i do coding in the same time than... 07:12
zengargoyle when i pull or clone i only get master branch.
of your bitbucket.
araraloren zengargoyle, yeah github.com/nicqrocks/p6-Git-Wrappe...META6.json
zengargoyle guess i need to maybe look at bitbucket api to learn branch names to pull. 07:13
jeromelanteri zengargoyle, perfect... actually, there is only master branch
zengargoyle araraloren: what is the problem with the provides section.
jeromelanteri zengargoyle, you no need to pull Git::Wrapper branch, this branch is merged back in master now. 07:14
and bitbucket is same cxommand as git (so github same).
araraloren so the module is the file name. Can a file provide three module /
?
zengargoyle a file can provide 3 classes. 07:15
jeromelanteri araraloren, now you touch why i don't like to speak about module for a class...
class/module... not the same.
then why for me, method is about class (not module). 07:16
araraloren I don't think it's class
In the provides section, include all the namespaces provided by your distribution and that you wish to be installed, only module files that are explicitly included here will be installed and available with use or require in other programs. 07:17
This is what the document said
jeromelanteri a class can be instanciate (i don't know if this english word is correct also), but module can not.
zengargoyle i *think* once it's compiled, a class is a CompUnit. it makes little difference where on the filesystem they were before being compiled.
araraloren **namespaces **
jeromelanteri araraloren, so why and when use class instead of module ? class is definitly for create objects to be instanciate, not modules. 07:19
concept is definitly not the same
araraloren you use a module, import name
not use a name
07:20 holyghost joined
jeromelanteri ??? 07:20
araraloren an exported class is a name, a variable is name. 07:21
this is something in symbol table
m: module A { our $e is export; }; say ::.keys; import A; say ::.keys; 07:23
camelia ($=pod $_ $/ A !UNIT_MARKER $e $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
($=pod $_ $/ A !UNIT_MARKER $e $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
araraloren m: module A { our $e is export; }; say ::.keys;
camelia ($=pod $_ $/ A !UNIT_MARKER $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
araraloren say the $e in ::.keys
evalable6 (exit code 1) 04===SORRY!04=== Error while compiling /tmp/n3NtbyZ2L7
Variable '$e' is not declared
at /tmp/n3NtbyZ2L7:1
------> 03say the 08⏏04$e in ::.keys
araraloren see
m: module A { class Want {} is export; }; say ::.keys; 07:24
camelia 5===SORRY!5=== Error while compiling <tmp>
Strange text after block (missing semicolon or comma?)
at <tmp>:1
------> 3module A { class Want {}7⏏5 is export; }; say ::.keys;
expecting any of:
infix
infix stopper…
jeromelanteri araraloren, a variable is an object... defined by a class object.
araraloren m: module A { class Want is export {}; }; say ::.keys; 07:25
camelia ($=pod $_ $/ A !UNIT_MARKER $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
araraloren m: module A { class Want is export {}; }; import A; say ::.keys;
camelia ($=pod Want $_ $/ A !UNIT_MARKER $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
jeromelanteri and again, a module is not a class, a class is not a module. 07:26
they have some name yes... everything has a name. Do they are the same ?
araraloren I mean the name in symbol table. When you reference a class or variable, it was search that **name** in the symbol table.
so the provides is not correct
zengargoyle m: module A { class Want is export {}; }; say ::.keys; say ::A::Want.WHAT;
camelia ($=pod $_ $/ A !UNIT_MARKER $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
(Want)
jeromelanteri araraloren, yes sure, but what i mean is that call moduyle everything is not clear. 07:27
araraloren you mean you don't know what the module provides ?
jeromelanteri same, a variable is define has a specific object class.. but you call it a variable, not a class. my Str $x; is a variable defined by class Str... 07:28
araraloren, no, nevermind.
zengargoyle m: module A { class Want is export {}; }; say ::.keys; say ::A::Want.new.WHAT;
camelia ($=pod $_ $/ A !UNIT_MARKER $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
(Want)
zengargoyle m: module A { class Want is export {}; }; say ::.keys; say ::A::Want.new.WHAT; say Want.new.WHAT;
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
Want used at line 1
zengargoyle m: module A { class Want is export {}; }; say ::.keys; say ::A::Want.new.WHAT; say ::Want.new.WHAT; 07:29
camelia 5===SORRY!5=== Error while compiling <tmp>
No such symbol 'Want'
at <tmp>:1
------> 3keys; say ::A::Want.new.WHAT; say ::Want7⏏5.new.WHAT;
jeromelanteri zengargoyle, ...oO...
zengargoyle m: module A { class Want is export {}; }; import A; say ::.keys; say ::A::Want.new.WHAT; say ::Want.new.WHAT; 07:31
camelia ($=pod Want $_ $/ A !UNIT_MARKER $=finish EXPORT $! ::?PACKAGE GLOBALish $¢ $?PACKAGE)
(Want)
(Want)
zengargoyle missed the import part.
jeromelanteri sure, a module can contain a class... a class can not contain a module... and you not use module and class for do same things. It is not the design pattern idea of "what is a class" and "what is a module".
that is also why a file can not contain more than one module (due to fact that it needs to follox name of his file also), but can contain many class... 07:32
zengargoyle i usually don't export class. 07:33
jeromelanteri there is conceptual idea about what a class is for and what a module is for... but sure, after, you can use class or module for other things, but it will not be a powerfull design pattern.
07:33 kaare__ joined
zengargoyle m: module A { class Want is export {}; class Dont {}; }; import A; say A::Dont.new.WHAT; 07:34
camelia (Dont)
jeromelanteri and that is not specific for perl6 or other OO language... it is same for all OO langage.
zengargoyle class in module is to save typing -ish
jeromelanteri yes
zengargoyle m: class A::Want is export {}; class A::Dont {}; }; say A::Dont.new.WHAT;
camelia 5===SORRY!5=== Error while compiling <tmp>
Unexpected closing bracket
at <tmp>:1
------> 3A::Want is export {}; class A::Dont {}; 7⏏5}; say A::Dont.new.WHAT;
jeromelanteri save an object you use specifically with this module
zengargoyle m: class A::Want {}; class A::Dont {}; }; say A::Dont.new.WHAT;
camelia 5===SORRY!5=== Error while compiling <tmp>
Unexpected closing bracket
at <tmp>:1
------> 3class A::Want {}; class A::Dont {}; 7⏏5}; say A::Dont.new.WHAT;
zengargoyle m: class A::Want {}; class A::Dont {}; say A::Dont.new.WHAT; 07:35
camelia (Dont)
jeromelanteri araraloren, you get it ?
zengargoyle m: module A { class A::Want is export {}; class A::Dont {}; }; import A; say A::Dont.new.WHAT; 07:36
camelia (Dont)
zengargoyle and it's smart....
m: module A { class A::Want is export {}; class A::Dont {}; }; import A; say A::Dont.new.WHAT; say A::A::Dont.WHAT; 07:37
camelia (Dont)
Could not find symbol '&Dont'
in block <unit> at <tmp> line 1
07:37 kaare_ left
araraloren no I don't mean the module you declare in a .pm6 or whatever file 07:39
I mean the module what we `use module;` 07:40
use
07:40 kaare__ left
jeromelanteri there is also this problem about declare or define...i used to make things with C++... 07:41
07:41 kaare__ joined
jeromelanteri araraloren, i understand what you mean... do you understand what i mean to ? 07:42
zengargoyle if i understood correctly, when using a filesystem CUR like with -Ilib or 'use "lib"' then the name of the file (minus extension and :: ized) is used to find the module.
araraloren a little..
jeromelanteri talk about module for a class make things confuse more than clear (this resume what i mean)
zengargoyle unless there is a META that has that information like when using -I. in a distribution directory, then it used information from META. 07:43
whan looking in other system CUR, it is just the name of the thing. there's no filesystem file name information left once it has been compiled and in CUR. 07:44
anyways, it's really complicated compared to perl 5 :)
07:45 holyghost_ joined, Cabanossi left
jeromelanteri so in perl6, if i well understand, "use" is a kind of concept macro command for ("import" +" load")... ? And it can depend of a context (use lib // use Something => what is inside ?//). 07:45
araraloren it's like c++, use a namespace, import some class or variable you declare
zengargoyle maybe module all get loaded at once so you don't use each of them individually?
araraloren jeromelanteri, you have to read document carefully 07:46
jeromelanteri ok
araraloren docs.perl6.org/language/modules#use
jeromelanteri araraloren, actually, i have many problem to well understand document sometimes... it needs maybe some more examples
zengargoyle jeromelanteri: yes, at lest use == need + import
07:46 Cabanossi joined
jeromelanteri zengargoyle, that what i understood to. 07:46
araraloren, also, when you asked if a file can contain many modules, i'm not sure you read that doc so carrefully (or maybe same me, sometime you not reeally understand this doc). 07:48
07:48 holyghost left
araraloren Hmm, don't use **module** is better solution, it's a CompUnit 07:49
but, everyone call them module
zengargoyle araraloren: did you look that up?
jeromelanteri i like to use module when his concept is good for what i need... and then, i call it a module.
zengargoyle i just guessed that might be true, i don't really know. 07:50
but it makes sense if true.
07:54 wamba left
araraloren yeah, I read it 07:57
zengargoyle i usually make a distribution and write tests so i don't think about many details. if test works, i'm right!
araraloren I often read document, I was not familiar with Perl's thinking
zengargoyle araraloren: cool, now i know. :)
aka think like Larry. :) 07:58
araraloren yeah 07:59
jeromelanteri module/class is not specific to perl6. It is specific to OO languages. Like MVC is not specific to one GUI or web coding... it is a large concept of design pattern to use. 08:00
zengargoyle i do not know many other languages well enough, and fear thinking perl 6 uses the same word to mean the same thing. 08:02
they might be using 'module' wrong! :P 08:03
jeromelanteri zengargoyle, i think you will be interested by read some book about algorythm and design pattern things. That is large concept of programmation. I mean... that is very funny also to read that kind of books. 08:04
zengargoyle i have, but mostly back in C days... not really module/class as a keyword in development. 08:05
jeromelanteri the first time i read that... OMG! it is like i realize the interest of "concept" things of algorythms.
and then... rewrite most of my codes again. For me, it was like... open the light inside my dirty room... after that, i wanted to clean my room. 08:06
i nbever see that was so dirty. 08:07
zengargoyle i guess a module is one .o file with many symbols in it. but a .c of many classes without a 'module' would be many .o files with only a single symbol.
jeromelanteri yes
araraloren I know what you mean the module. But perl6 also has a keyword module, it's confused anyway 08:10
jeromelanteri that's it. 08:11
zengargoyle now i'm maybe remembering when it was 'module Foo;' like perl 5 and was changed to 'unit module Foo;' to make a distintion that it was the whole file..... 08:14
where you could have multiple 'package Foo;' 'package Bar;' in same file without the 'package Foo { } package Bar { }' scope syntax. 08:16
08:22 setty1 left 08:23 nadim joined 08:28 ChoHag left
jeromelanteri zengargoyle, actually, the machine langage (and you know that if you like asm) talk only about registers for memories, or events, or CPU unit, or mounted interfaces, adress, then command jump, branch, set, ref (some MISC proc have only 35 commands...). But conceptuels langages (and that is the power of OO and functionnal langages) is for make sesn relatively of what we know allready in our life. It is like a human interface 08:29
langage for make code readable and also for make it DRY and KISS.
i wascurious of perl because of the conceptor is a linguist. 08:30
08:44 Cabanossi left 08:45 holyghost_ left 08:46 Cabanossi joined
zengargoyle jeromelanteri: continue later, i must sleep. :) bon nuit. 08:47
je pense mon francais est tres mal. :P 08:48
08:54 Cabanossi left 08:55 Cabanossi joined
jeromelanteri zengargoyle, il est compréhensible, bonne nuit l'homme. 09:24
09:24 brrt joined 09:29 andrzejku joined
andrzejku ls 09:30
yoleaux 03:46Z <araraloren> andrzejku: After we add the run configuration, it was working!
03:47Z <araraloren> andrzejku: Though it has a different with what we thought .
09:44 Cabanossi left 09:47 Cabanossi joined 09:57 brrt left 09:58 andrzejku left 10:03 wamba joined
jeromelanteri an other one module who print nothing (and also... from README file copy/pasted...). No error, but also no nothing. module is Text::Table::List github.com/supernovus/perl6-text-table-list 10:24
10:25 TEttinger left 10:44 bjz left 10:48 bjz joined 10:53 Aaronepower left 10:57 espadrine left 11:06 kaare__ left
gfldex travis fails to build rakudo for me, could anyone check please? 11:20
11:27 Aaronepower joined 11:30 pharv left 11:33 espadrine joined 11:34 andrzejku joined 11:40 Rawriful joined
gfldex rakudo fails to build locally too 11:41
11:41 _cronus joined
_cronus hello, I am having trouble installing bailador 11:42
gfldex samcv: could you unbreak rakudo please?
_cronus when I run zef install Bailador it stops after failing some tests 11:43
gfldex _cronus: rakudo itself is broken right now
_cronus gfldex: does the previous version still work? 11:44
gfldex _cronus: 51e59eeb5e48917ed3e825947989f91a219ce2cc should work 11:45
jeromelanteri gfldex, what is travis ? a kind of docker ?
gfldex jeromelanteri: it's an automatic build system that is triggered by github commits
jeromelanteri ok
11:46 bjz_ joined
timotimo i'll look into the rakudo breakage right now 11:46
jeromelanteri _cronus, i see that there is a specific channel on IRC for bailador 11:47
11:48 bjz left
jeromelanteri bailador.net/ they are talking about "bailador stack channel (maybe not IRC) 11:48
_cronus gfldex: i'm using rakudo-star-2017.07 which is older
jeromelanteri: thanks 11:51
timotimo i do have a fix, but i'll test it first 11:54
tadzik yeah, it's a Slack channel
11:56 HoboWithShotgun joined
HoboWithShotgun can i tell the .perl method not to recurse into objects? 11:56
you know, like $Data::Dumper::Maxdepth in P5 11:58
timotimo no, .perl is a bit dumber than that 11:59
what's your use case?
perl has detection for circular references, for example
12:00 Cabanossi left
timotimo m: my $a = []; $a.push($a); $a.push($a); say $a.perl 12:00
camelia ((my @Array_77166080) = $[@Array_77166080, @Array_77166080])
timotimo so if that's the problem, no problem :)
HoboWithShotgun debugging. i am not sure what is in the array of xml::element objects, so i call .perl on it. but that produces thousands of lines of output cause these things are all interconnected by their parent properies
12:01 brrt joined
HoboWithShotgun and .perl happily follows each reference 12:01
12:01 Cabanossi joined
timotimo also 12:01
perl6 -e 'my @foo; @foo[0][0][0][0][0] = "hi"; @foo[0][1] = "hi"; @foo[1] = "bye"; say @foo.tree(*.self, *.self, *.self, { [] }).perl;'
er
m: my @foo; @foo[0][0][0][0][0] = "hi"; @foo[0][1] = "hi"; @foo[1] = "bye"; say @foo.tree(*.self, *.self, *.self, { [] }).perl; 12:02
camelia ((([],).Seq, "hi").Seq, "bye").Seq
timotimo m: my @foo; @foo[0][0][0][0][0] = "hi"; @foo[0][1] = "hi"; @foo[1] = "bye"; say @foo.perl;
camelia [[[[["hi"],],], "hi"], "bye"]
timotimo .tree can help you here
12:07 nattefrost joined 12:08 Rawriful left, Rawriful joined, Rawriful left 12:13 avalenn left
jeromelanteri is there any ncurses kind of perl6 module(s) for forme questions tags ? 12:15
i see module use ncurses for table print, menu choose (very nice one)... but if i want to ask wuestion with ncurse ? 12:16
questions
timotimo honestly, i'd just use zenity or any of the similar programs 12:18
or Text::Print, that's powerful and cool
though tbh i've only looked at the examples, i haven't used it myself yet 12:19
and without the raw input branch it might not yet have great input, come to think of it
okay forget i said anything until in a week or so
12:26 andrzejku left
HoboWithShotgun m: (1,2,3).map(*.WHAT) 12:30
camelia Cannot resolve caller map(List: Whatever); none of these signatures match:
($: Hash \h, *%_)
(\SELF: &block;; :$label, :$item, *%_)
(HyperIterable:D $: &block;; :$label, *%_)
in block <unit> at <tmp> line 1
HoboWithShotgun why is that?
12:31 kaare__ joined 12:32 itaipu joined, andrzejku joined 12:40 someuser left
araraloren m:say (1,2,3).map(*.&WHAT) 12:43
evalable6 ((Int) (Int) (Int))
12:44 brrt left, brrt joined 12:46 brrt left 12:49 itaipu left 12:53 andrzejku left 12:54 clarkema joined
timotimo .WHAT is a little macro-ish 12:57
when you do *.WHAT you get Whatever instead of a whatever code curried to call .WHAT on something
m: say *.WHAT
camelia (Whatever)
timotimo m: say *.foo
camelia { ... }
timotimo that's the difference
12:59 Cabanossi left
HoboWithShotgun excuses. WHAT should give me the type object as advertised. 13:00
this language is far from production ready
13:00 mr-foobar joined
HoboWithShotgun i am trying to get a simple script working and i constantly hit walls like this 13:01
timotimo i don't think *.WHAT will be "fixed"
13:01 Cabanossi joined
timotimo and .WHAT does give you the type object 13:01
*.WHAT gives you the type object for Whatever 13:02
m: say (* - 1).WHAT
camelia (WhateverCode)
timotimo here the * - 1 gets curried as you expect, and .WHAT doesn't, so this time you get WhateverCode instead
HoboWithShotgun but thats not what it should do. if *.foo executes foo on the current list element
timotimo upper-case things are all special 13:03
HoboWithShotgun then *.WHAT should execute WHAT on the fucking list element and not the *
fine, then how would i do what i want to do? 13:04
13:07 _cronus left 13:08 stmuk_ joined 13:10 stmuk left, Aaronepower left, nadim left, nadim joined
jeromelanteri timotimo, text::print ? where ? 13:11
zenity ?
13:13 mr-foobar left
timotimo not text::print, sorry 13:16
jeromelanteri ok
i really like ncurses lib use, it make nice terminal print
timotimo, instead, which command could add 1000 lines in the terminal ? 13:18
13:18 Aaronepower joined
timotimo "add 1000 lines"? 13:18
jeromelanteri for (1..1000)-> { run: "echo"}; ?? 13:19
timotimo, yes, for wash the prompt
timotimo m: <foo bar baz 100>.map(*.^name).perl.say
camelia ("Str", "Str", "Str", "IntStr").Seq
timotimo HoboWithShotgun: how do you like this?
jeromelanteri because if i use mix of ncurse and actual prompter, it is very dirty
timotimo or do you need the actual type object to call methods on it?
13:20 mr-foobar joined
timotimo why would you "run echo" instead of "say"? 13:20
jeromelanteri timotimo, i don't know... because i'm stupid sometimes...
13:23 Skarsnik joined 13:25 nadim left 13:30 zakharyas joined
Skarsnik Hello 13:31
timotimo don't worry jeromelanteri :)
araraloren HoboWithShotgun, I show the code to you
m:say (1,2,3).map(*.&WHAT)
evalable6 ((Int) (Int) (Int))
jeromelanteri timotimo, i'm not. We can all be stupid sometimes. this is not a big problem when we know that.
araraloren Skarsnik, hi 13:32
HoboWithShotgun araraloren: if i do that, i get no output 13:36
as i said. walls everywhere. i gonna do this in perl5 instead 13:37
araraloren HoboWithShotgun, you have to print them use say or print
HoboWithShotgun tx for listening
13:37 HoboWithShotgun left
jeromelanteri he looks like not happy... 13:38
:)
araraloren haha
yeah
araraloren hear nothing, I hear nothing 13:39
jeromelanteri he choose an adapted pseudo also... very funny.
13:45 zakharyas left 13:49 zakharyas joined, mr-foobar left 14:14 Cabanossi left 14:16 Cabanossi joined 14:18 bjz_ left, BenGoldberg joined 14:22 bjz joined 14:31 cdg joined 14:41 bjz left, holyghost_ joined
ugexe someone putting (1,2,3).map(*.WHAT) in production code, since it was brought up, would not speak highly of that person 14:45
14:45 holyghost_ left
ugexe if you don't want to bother learning the less basic parts of the language just use block form of map: (1,2,3).map({ .WHAT }) 14:47
BenGoldberg m: say *.WHAT 14:52
camelia (Whatever)
BenGoldberg m: say *+1 14:53
camelia { ... }
BenGoldberg m: say *.HOW 14:54
camelia Perl6::Metamodel::ClassHOW.new
BenGoldberg m: say *.WHY
camelia { ... }
15:00 nadim joined 15:02 smls joined
smls m: grammar A { has $.a; token TOP { {say $.a} } }; A.new(a => 42).parse(""); 15:04
camelia (Any)
smls ^^ Is using *instances* of grammar classes with class attributes meant to work? 15:05
bisectable6: grammar A { has $.a; token TOP { {say $.a} } }; A.new(a => 42).parse(""); 15:08
bisectable6 smls, On both starting points (old=2015.12 new=e051dd2) the exit code is 0 and the output is identical as well
smls, Output on both points: «(Any)»
15:13 zakharyas left
ugexe m: role Foo { has $.a; method a { $!a } }; grammar A does Foo { token TOP { {say $.a} } }; A.new(a => 42).parse(""); # seems broke, this should work at the very least 15:13
camelia (Any)
15:13 nadim left 15:14 Cabanossi left, nadim joined 15:15 khw joined 15:16 Cabanossi joined 15:25 Aaronepower left, Rawriful joined 15:26 hoelzro_ left, huggable left, hobbs left, buggable left, NeuralAnomaly left, ZofBot left, plicease_ left, Geth_ left, hobbs joined, hobbs left, hobbs joined 15:27 leego joined, Geth___ joined, ChanServ sets mode: +v Geth___, buggable joined, ChanServ sets mode: +v buggable, plicease joined 15:36 cdg left
jnthn smls: They work in so far as that being how a Cursor stores its own internal state. However, each rule gets a cursor of its own, meaning a fresh instance is made per token that is entered 15:36
So the instance inside of token TOP is not the one that .parse was called on 15:37
15:37 popov_ joined
smls ok 15:37
15:50 Aaronepower joined 15:51 Aaronepower left 15:53 BenGoldberg left 15:56 BenGoldberg joined 15:57 clarkema left, BenGoldberg left 16:03 mr-foobar joined, clarkema joined 16:06 Aaronepower joined 16:07 BenGoldberg joined 16:09 andrzejku joined 16:12 Aaronepower left 16:14 Cabanossi left 16:16 Cabanossi joined 16:21 Aaronepower joined 16:25 pharv joined, Aaronepower left 16:30 bjz joined 16:33 araraloren left 16:37 Cabanossi left 16:38 bjz left, khw left 16:41 Cabanossi joined 16:46 nattefrost left 16:52 leego is now known as leedo, Cabanossi left, bjz joined 16:54 Cabanossi joined 16:55 skids joined 17:01 popov_ left
japhb jnthn, timotimo: Not only is raw tty input seeming stable, as timotimo guessed, it doesn't need the fd reopen hack to force the input stream onto the processing thread anymore. Thank you, jnthn++! 17:02
Of course, now this means I need to wade back into the battle with decoding escape codes for arrow and function keys .... 17:03
17:09 bjz left 17:12 zakharyas joined 17:15 JJ_ joined
JJ_ Hi there 17:15
japhb JJ_: Hello! Did you have a question, or just saying hi? 17:17
JJ_ For the time being, just hi and lurking for a while...
japhb It tends to be a little slower on the weekends, so don't be surprised if you don't hear much at times. 17:18
JJ_ :-)
I'm getting kind of crazy, of all things, looping over an array 17:19
Got this array in JSON
github.com/JJ/perl6em/blob/master/...short.json
Gobble it up: my $json = $urls-file.IO.slurp(); my $decoded = from-json $json; 17:20
$decoded.elems returns 39. So far so good.
but then for $decoded -> $u assigns the whole array to $u in the first iteration.
Any idea? This is probably pretty stupid, I know, but I can't figure it out. 17:21
I've worked around it by using .keys, which for some reason works OK.
timotimo you want to for @$decoded or for $decoded.list 17:22
it's in a scalar container, the semantics of scalar containers are "a single thing"
alternatively you could assign to @decoded instead of $decoded 17:23
JJ_ I tried that, didn't work. Let me check again.
Yep, does not work. It @decoded.elems returns 1
timotimo oh, heh 17:25
JJ_ OK, I think I got it
timotimo sorry about that, you'd have to either bind with := or put @(...)
because there's still the scalar container
it treats it the same way as
JJ_ my @urls = @(from-json $json);
timotimo m: my @foo = $(1, 2, 3); say @foo.elems; say @foo[0].elems 17:26
camelia 1
3
timotimo m: my @foo := $(1, 2, 3); say @foo.elems; say @foo[0].elems
camelia 3
1
JJ_ The problem seems to be from-json returns in scalar context. So you have to change it to Array context.
timotimo aye
17:26 bjz joined
JJ_ Thanks! 17:26
timotimo though context isn't the right word
in perl6 context works the other way around from perl5 and earlier
JJ_ Yep, it's perl5ish 17:27
sorry.
timotimo no problem
m: use LWP::Simple; LWP::Simple.get: raw.githubusercontent.com/JJ/perl6...short.json
camelia ===SORRY!===
Could not find LWP::Simple at line 1 in:
/home/camelia/.perl6
/home/camelia/rakudo-m-inst-2/share/perl6/site
/home/camelia/rakudo-m-inst-2/share/perl6/vendor
/home/camelia/rakudo-m-inst-2/share/perl6
CompUnit:…
timotimo er
star: use LWP::Simple; LWP::Simple.get: raw.githubusercontent.com/JJ/perl6...short.json
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3use LWP::Simple; LWP::Simple.get: https:7⏏5//raw.githubusercontent.com/JJ/perl6em/m
expecting any of:
colon pair
timotimo star: use LWP::Simple; LWP::Simple.get: "raw.githubusercontent.com/JJ/perl6...hort.json" 17:28
camelia 501 Protocol scheme 'https' is only supported if IO::Socket::SSL is installed <URL:raw.githubusercontent.com/JJ/perl6...json>

 in method request_shell at /home/camelia/star-2016.10/share/perl6/site/sources/2A99795C7…
timotimo oh wow that's 2016.10?
nine: can we get that up to date?
17:28 andrzejku left 17:29 smls left 17:30 clarkema left 17:37 andrzejku joined 17:41 zakharyas left 17:48 andrzejku left 18:03 nhywyll joined 18:18 jaldhar_ left 18:19 jaldhar_ joined 18:25 setty1 joined
moritz I'm on it 18:38
18:46 klapperl joined 18:54 pharv left 18:59 domidumont joined 19:01 domidumont left 19:02 ZofBot joined, ChanServ sets mode: +v ZofBot 19:03 huggable joined, ChanServ sets mode: +v huggable 19:04 NeuralAnomaly joined, ChanServ sets mode: +v NeuralAnomaly 19:05 darutoko left
JJ_ What do you people use for debugging? 19:05
I'm stuck with the old say this dd that
19:06 dwarring joined
moritz too 19:08
which I tend to do in all languages I program in
JJ_ but still 19:09
breakpoints help.
the old "x" in the perl5 debugger was a nice way of having formatted variables printed. 19:10
So there's no better alternative in perl6, right?
moritz you can install Debugger::UI::CommandLine (if it works right now), and then use perl6-debug-m instead perl6-m to start your code
JJ_ That's a big if 19:11
19:11 kaare__ is now known as kaare_
moritz star: say 42 # for timotimo 19:11
camelia 42
JJ_ @moritz it apparently works, thanks for the suggestion 19:12
moritz \o/ 19:13
19:14 Cabanossi left 19:16 Cabanossi joined 19:21 hankache joined 19:26 andrzejku joined 19:27 bjz left
JJ_ Now regex stuff 19:33
$text = "[captura](una.captura) y [otra](capt.ura)" 19:34
$text ~~ /\] \( (http <-[ ) ]> +?) \) / returns una.captura
Now 19:35
$text ~~ m:global/\] \( (http <-[ ) ]> +?) \) /
returns ](una.captura)
So the global switch makes it forget the grouping parentheses?
timotimo moritz: perfect! <3
JJ_ um 19:36
moritz JJ_: I don't see any grouping parenthesis in your regex
JJ_ right before http and after ?
moritz oh right 19:37
JJ_ they actually work... without the global switch
moritz m: my $text = "[captura](una.captura) y [otra](capt.ura)"; for $text ~~ /\] \( (http <-[ ) ]> +?) \) / -> $/ { say $0 }
camelia Nil
JJ_ m: my $text = "[captura](una.captura) y [otra](capt.ura)"; for $text ~~ m:g/\] \( (http <-[ ) ]> +?) \) / -> $/ { say $0 } 19:38
camelia ｢una.captura｣
｢https://capt.ura｣
»
JJ_ um
moritz that looks right, now? 19:39
s/now/no/
JJ_ looks totally right
moritz it returns a list of matches, and in each match, the [0] element matched the URL 19:40
m: my $text = "[captura](una.captura) y [otra](capt.ura)"; say $text ~~ m:g/\] \( (http <-[ ) ]> +?) \) /
camelia (｢](una.captura)｣
 0 => ｢una.captura｣ ｢](capt.ura)｣
 0 => ｢capt.ura｣)
»
moritz same thing, but since the list delimiter is not explicit, it's easier to lose track of what's going on
geekosaur looks ot me like they stringified the match object, losing the information they actually wanted 19:41
moritz btw I'm currently working on leanpub.com/perl6regex :-)
explaining "token" and "rule" right now
JJ_ Good luck! 19:42
moritz thanks 19:46
19:47 JJ_ left
zengargoyle m: my $text = "[captura](una.captura) y [otra](capt.ura)"; $text ~~ m:g/\] \( (http <-[ ) ]> +?) \) /; say ~$_.[0] for $/[]; 19:51
camelia una.captura
https://capt.ura
»
zengargoyle man $/ is so bizzare to get my head around... :)
moritz read my book! 19:52
I hope it explains it well
if not, tell me, and I'll do my best to improve the descirption
zengargoyle hehe, i do need to re-start reading.
19:54 pharv joined
zengargoyle moritz: i sorta started waiting for chapter contents to settle down because going back and tracking additions gave me too much deja vu. :) 20:00
zengargoyle needs to up my game on viewing diff of prose. 20:03
moritz zengargoyle: I understand, no problem 20:05
20:07 TEttinger joined 20:19 rindolf left 20:24 Skarsnik left, skids1 joined, itaipu joined 20:26 skids left 20:33 andrzejku left, dwarring left 20:36 skids joined 20:39 skids1 left 20:50 deep-book-gk_ joined, bjz joined 20:53 deep-book-gk_ left 20:55 itaylor57 left, itaylor57 joined 20:59 dct joined 21:01 Cabanossi left, cog_ left 21:02 Cabanossi joined 21:06 lizmat left 21:09 lizmat joined 21:15 dct left 21:23 hankache left 21:26 hythm_ joined 21:29 rindolf joined 21:32 wamba left
hythm_ hello perl6, I would like to start learning web development in perl6, is there a resource, documentation or book I can read? 21:32
timotimo gabor szabo is writing a book about bailador (and since he started, development around bailador picked up a few more volunteers) 21:33
japhb: did you know that with Terminal::Print $p.indices -> [$x, $y] is almost twice as slow as flat $p.indices -> $x, $y? 21:38
though of course the API is fixed so can't just change what it returns 21:41
hythm_ just checked it, I would like to crowdfund to help, since the book will be released in December, are there other resources to learn from now (as I want to start soon), or may be the book offer a draft version if helped with crowndfunding it!
timotimo you already know a fair bit about perl6 itself, right? 21:42
i think i've seen you in here a bunch of times
hythm_ yes I know the basics, but I'm tatally new to web development 21:43
ryu0 Not even HTML/CSS experience?
hythm_ s/totaly/totally/
yes I know HTML and CSS 21:44
21:45 Cabanossi left 21:46 Cabanossi joined
hythm_ but I don't have experience creating web app with API and nice UI, and that is my next goal, 21:47
timotimo i somehow managed to segfault the tris example :| 21:49
21:50 bjz left
zengargoyle .ask jeromelanteri could you add an example my_pl6.yml file to your repo? 21:51
yoleaux zengargoyle: I'll pass your message to jeromelanteri.
hythm_ Actually, I just got a draft version from the book with a few chapters to start with. Awesome! 21:55
zengargoyle .ask jeromelanteri or you could email me zengargoyle at gmail.com if my_pl6.yml is sensitive data.... 21:56
yoleaux zengargoyle: I'll pass your message to jeromelanteri.
22:00 rindolf left 22:01 andrzejku joined
nadim japhb: did you get my message? 22:02
22:03 andrzejku left 22:06 Hotkeys joined
nadim hi, I am wondering (there may be an example I misssed) if it is possible to pass arguments to a pointy block that does not declare parameters. something like @a = { ;}, {;} ; and then iterate through the array an call the blocks with arguments. 22:06
timotimo blocks, as opposed to methods, don't slurp up all nameds by default 22:08
m: my &foo = {;}; say &foo.signature
camelia (;; $_? is raw)
timotimo blocks without signatures take a single optional argument
nadim hmm, which one? hand how are they called? 22:09
timotimo a single optional positional argument
i.e. you can call the blocks with &block() or with &block(42) 22:10
nadim I want to avoid writing sub ($a, $b){ ;} if I can write {;} only
timotimo then you can use $^a, $^b inside the { }
nadim adnd 42 gets in $_, right?
timotimo yes
if you want to avoid writing "sub", you can give a pointy block a signature 22:11
-> $a, $b { }
22:11 hythm_ left
nadim and call it like: &foo(1, 2) ? 22:11
or five the parameters after?
-> $s, $b {}(1, 2) 22:12
s/five/give/
timotimo m: -> $a, $b { say "hi $a, $b!" }(99, 42)
camelia hi 99, 42!
timotimo you can, yes
nadim can a reference be used later to do that or does it have to be inline as the example above? 22:13
timotimo of course you can put a pointy block into a variable
ilmari m: my $s = -> $a, $b { say $^a + $^b }; $s->(1,2)
camelia 5===SORRY!5=== Error while compiling <tmp>
Redeclaration of symbol '$^a' as a placeholder parameter
at <tmp>:1
------> 3my $s = -> $a, $b { say $^a7⏏5 + $^b }; $s->(1,2)
ilmari m: my $s = -> $a, $b { say $a + $b }; $s->(1,2)
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of ->(), ->{} or ->[] as postfix dereferencer; in Perl 6 please use .(), .[] or .{} to deref, or whitespace to delimit a pointy block
at <tmp>:1
------> 3my $s = -> $a, $b { say $a + $…
timotimo or pass a pointy block as a parameter to something else without a variable in the middle
ilmari m: my $s = -> $a, $b { say $a + $b }; $s(1,2) 22:14
camelia 3
ilmari m: my $s = { $^a + $^b }; $s(1,2) # look ma, no pointy!
camelia ( no output )
ilmari m: my $s = { say $^a + $^b }; $s(1,2) # look ma, no pointy!
camelia 3
22:14 |oLa| joined
nadim I guess that's the closes one can come 22:16
In P5 I made variables "appear in context", practical to make thing look magic.
22:16 El_Che_ left
timotimo well, you can use dynamic variables 22:17
geekosaur suppose you could do somethng "clever" with $*vars
that
nadim I have an interfacethat is going to take a lot of subs, I need to make it as typing-less as possible
geekosaur: do tell, even if it is just to learn something and not use it 22:18
timotimo like you can declare $*foo, $*bar, $*baz, $*quux where you call these blocks and every block can decide which of those they want to use by just refering to them like that
geekosaur I basically said what timotimo did
nadim ok, what's the mechanism behind them? 22:19
22:19 espadrine left
timotimo m: my @blocks = { say "nothing" }, { say "foo is $*foo" }, { say "foo and quux: $*foo, $*quux" }; for @blocks -> &block { my $*foo = "this is foo"; my $*bar = "and bar"; my $*quux = "quux here"; &block() } 22:19
camelia nothing
foo is this is foo
foo and quux: this is foo, quux here
nadim their reason to exist, except for being cool
22:19 |oLa| left
geekosaur because sometimes you need a variable that depends on the dynamic call stack instead of lexical 22:19
timotimo dynamic variables have "dynamic scope", which means whatever is called by the function that defines it (or a function that is called by that, and so on) has access to it
geekosaur in perl 5 you do this with local()
lisp variables are generally dynamic 22:20
timotimo and you can "override" or "shadow" dynamic variables from further up the stack by defining them in your function
geekosaur and in Haskell this is MonadReader's local
nadim so I now have 4 ways to do what I want!, land of plenty. 22:21
is it possible to type a block?
timotimo yes 22:22
you can have a restriction on the return type or the signature
geekosaur that's what timotimo was saying about -> with a signature
nadim and what can be done with a typed block? parameters wise
I mean typing like assigning a class to it
timotimo docs.perl6.org/syntax/Constraining..._Signature on this very unwieldily-named page there's examples for what to put into your sub signatures
if i understand right what you mean by "typed" 22:23
nadim thank yo both for the tips and the link
nadim saves this QA session 22:24
timotimo can you elaborate a bit more what you want to achieve?
nadim sure
timotimo with the typing, i mean
i can sort of guess what you want in general
nadim I have lines that look like this :flat(Hash, sub ($s, $depth){$s ~~ Array, 5}, sub ($s, $depth){$s ~~ Int}, ....) 22:25
the subs are user defined, and a pain to type 22:26
I can provide some sugar for common cases but when the user wants to add something special then he needs to type all the boilerplate, I'd like to avoid that, as I am the user too 22:27
22:27 nhywyll left
nadim :flat(Hash, {$_ ~~ Array, 5}, {$_ ~~ Int}, ....) looks a bit better 22:27
timotimo i'm not sure how these blocks work exactly 22:28
it checks the thing against a type and returns a True or False and also optionally a number?
nadim yes
I's like to pass more than one variable but for the sake of the dicussion let's say I pass only one 22:29
timotimo another thing you could do is have like (Array) => 5 meaning "if it smart matches against Array, expand it if it's below 5 levels down" or whatever
nadim I pass a type and the sub tells me if it should be flattened or not
timotimo OK, and the number? 22:30
nadim let me paste an example
22:31 pharv left
nadim nopaste.linux-dev.org/?1160205 22:31
:flat flattens, if you have an Array then all it's elements are put on the same line or multiple lines depending on the teminal depth 22:32
timotimo what's the 10 doing there off to the side?
nadim the number says, puts 5 entries in a colums, followed by another colum of 5 entries, ...
timotimo oh
that's the 11th thing in the one to its left
22:33 gregf_ left
nadim the 11th entry contains an array of 10 entries which is itself flattened and columnized 22:34
timotimo right
nadim data is $ds = $([1, 2, 3, 4, 5, 6, 7, 8, 9, 10], [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, [1, 2, 3, 4, 5, 6, 7, 8, 9, 10]], {:a(IntStr.new(1, "1")), :b(IntStr.new(2, "2"))}) 22:36
which is pretty easy to read in dd because it is short and all number are consecutive, with mixed number it becomes impossible 22:37
22:37 rindolf joined
timotimo yeah 22:39
nadim imgur.com/Hfhiouw 22:40
timotimo still don't understand what the flag parameters do exactly
nadim Data::Dump::Tree renders all the data vertically 22:41
:flat tell ddt to take a specific entry or type, or depth, and render it horizontally 22:42
timotimo and how exactly does that work? 22:43
nadim the arguments to :flat tells it what to do. and if, on top of telling it if an entry should be flattened or not, it can also say that it wants columns
timotimo that's the number you return? 22:44
nadim the details? :flat installs a sub_elements post filter, then an element is going to be displayed the filter is called and rather that return a list of sub elements it returns a Horizontal object that is the representatio of the sub elements, that is obtained by running a sub renderer. the sub rendere returns the rendering lines, the Horizontal object re-arranges the lines horizontally and in columns too. 22:46
the number that is returned is the amount of lines in a colum.
timotimo OK 22:47
nadim so it would be, for example, be possible to say "a specific element needs to be rendered in columns" something like :flat({$_ === $my_big_array, 10}) 22:48
timotimo ah, sure 22:49
nadim actually one can already say :flat(($my_big_array, 5)). but I want a generic mechanism via subs
timotimo right
nadim the nice things with subs is that it can decide dynamically to leave it in vertical rendering, horizontal or columnar based on the element itself 22:50
timotimo of course
22:52 rindolf left
nadim what I like is that it is not even 100 lines of code, and I have a very fluffy style 22:53
timotimo japhb: you know how i told you flat and -> $x, $y makes things 2x faster? no, actually, it makes things completely wrong :) 22:54
japhb: however, if i take @alphabet.roll(*), put it into my @alphseq, and use @alphseq.shift in place of @alphabet.roll, it gets about 1.8x faster 22:59
m: say 2.9 / 5.7
camelia 0.508772
timotimo m: say 2.9 R/ 5.7 23:00
camelia 1.965517
timotimo wow, much closer to 2x than i thought
23:02 BenGoldberg left
nadim is there a shortcut to set a Bool variable to True or False, ie not writting $var = True 23:04
timotimo if the bool variable has an "is default", you can assign Nil 23:05
that's 1 shorter than True or False
m: my $boo = False; $foo or= 1; say $foo.perl 23:06
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$foo' is not declared. Did you mean '$boo'?
at <tmp>:1
------> 3my $boo = False; 7⏏5$foo or= 1; say $foo.perl
timotimo m: my $boo = False; $boo or= 1; say $boo.perl
camelia 1
timotimo nah, that doesn't turn it into a bool for you
skids m: my Bool $a; $a++; $a.say; $a++; $a.say; $a--; $a.say; $a--; $a.say; 23:09
camelia True
True
False
False
timotimo japhb: but changing (1..5).roll and (1..6).roll to a shift on an infinite roll doesn't make it any faster
ooh, good thinking, skids
23:09 CIAvash[m] left, mack[m] left, tadzik left, mienaikage left, Matthew[m] left, tyil[m] left 23:10 ugjka left, unclechu left, M-Illandan left, xui_nya[m] left, wictory[m] left 23:12 ugjka joined
nadim indeed shorte and it works well with the $a++ use to make things true 23:13
23:14 ilmari[m] left 23:15 Cabanossi left 23:16 Cabanossi joined
timotimo you could have prefix:<⊥> (or suffix) and the one in the other direction 23:19
to truthify and falsify variables
u: ⊥
unicodable6 timotimo, U+22A5 UP TACK [Sm] (⊥)
timotimo u: DOWN TACK
unicodable6 timotimo, U+02D5 MODIFIER LETTER DOWN TACK [Sk] (˕)
timotimo, U+031E COMBINING DOWN TACK BELOW [Mn] ( ̞)
timotimo, 12 characters in total: gist.github.com/66d3f6d27f28205ccd...6fcca19259
23:19 samb1 left
timotimo ⊤ is the other one 23:19
23:20 itaipu left 23:31 samb1 joined 23:34 setty1 left 23:37 mienaikage joined 23:40 tzekid left 23:41 cog_ joined
Geth doc: aa912b4b19 | (Will "Coke" Coleda)++ | 2 files
fixup PseudoStash docs

  (whitespace, spelling, grammar)		 23:45
23:45 ilmari[m] joined, tyil[m] joined, unclechu joined, Matthew[m] joined, M-Illandan joined, mack[m] joined, tadzik joined, wictory[m] joined, Cabanossi left, xui_nya[m] joined, CIAvash[m] joined 23:46 Cabanossi joined 23:57 El_Che joined 23:58 Rawriful left

Powered by Cro and the Raku® Programming Language.

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

Thank you!