Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2017-07-28

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:00 zoll i put up with hg not by choice. one thing i miss is gitk; hgk is barely usable in comparison.
00:00 timotimo but tortoisehg is very good
00:03 zoll is tortoise* family not confined to MS Windows? i may^Wwill update CentOS to v7 to get updated support software for other, better hg GUI.
00:04 timotimo not tortoisehg
00:04 timotimo there's also sourcetree, i think that has both git and hg, or maybe only hg
00:05 Cabanossi joined #perl6
00:07 zoll ok. i will see into this further. thanks timotimo
00:08 timotimo good luck!
00:08 timotimo merced: i often struggle with not finding projects when i want to play with something new, like a new programming language
00:08 timotimo merced: what's your style? tools, gui, web, games, maths, art, ...?
00:10 skids joined #perl6
00:11 jaldhar_ joined #perl6
00:14 R0b0t1 timotimo: Admittedly I think the submodule command renamed some things nonsensically
00:22 lookatme joined #perl6
00:22 lookatme morning
00:28 merced timotimo: i like automation, raw data processing, fast-scrolling terminals, systems engineering, shit like that
00:32 PerlJam "fast-scrolling terminals"?  heh
00:32 yoleaux 28 May 2017 09:33Z <Zoffix> PerlJam: IIRC it was you who lamented impossibility of Perl 5's `$.` for $*ARGFILES var in Perl 6. It's no longer impossible ($*ARGFILES is IO::CatHandle now): https://twitter.com/zoffix/status/868761719639871488
00:33 merced yeah slow scroll's so boring
00:35 Cabanossi joined #perl6
00:36 * PerlJam wonders if yoleaux can be made smart enough to deliver messages to any nicks that a person might use whether or not that particular nick is addressed
00:39 timotimo merced: hm, i don't think that intersects with a lot of my style, so i don't have any good suggestions
00:41 PerlJam merced: why not contribute to an existing project?
00:41 pharv_ joined #perl6
00:42 PerlJam merced: something I've thought of doing just because is implementing useful shell commands in Perl 6
00:43 PerlJam kind of like Perl Power Tools, but only a subset of the unix utilities that I find interesting or useful
00:43 coetry joined #perl6
00:44 coetry p6: my $to = 'all'; say "hello to $to :)";
00:44 camelia rakudo-moar 9658dd: OUTPUT: «hello to all :)?»
00:45 timotimo m: say "coetry".comb.pick(*) xx 5
00:45 camelia rakudo-moar 9658dd: OUTPUT: «((r y o c e t) (y c t e r o) (o r t e c y) (t o y c r e) (e y t r c o))?»
00:45 timotimo m: say "coetry".comb.pick(*).join xx 5
00:45 camelia rakudo-moar 9658dd: OUTPUT: «(teroyc oecryt oyetrc ytceor retyco)?»
00:48 lookatme m: say "coetry".flip;
00:48 camelia rakudo-moar 9658dd: OUTPUT: «yrteoc?»
00:49 zoll m: say split   "coetry"
00:49 camelia rakudo-moar 9658dd: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>?Calling split(Str) will never work with declared signature ($pat, Cool $target, |c is raw)?at <tmp>:1?------> 3say 7?5split   "coetry"?»
00:50 zoll that's actually good error message
00:50 timotimo m: say (do for "coetry".comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains($letter)).grep({ rand < 0.2 }).first }).join
00:50 camelia rakudo-moar 9658dd: OUTPUT: «243532015550141141414152863?»
00:50 timotimo m: say (do for "coetry".comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains($letter)).grep({ rand < 0.2 }).map(*.chr).first }).join
00:50 camelia rakudo-moar 9658dd: OUTPUT: «???????»
00:50 timotimo m: say (do for "coetry".comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains(" $letter ")).grep({ rand < 0.2 }).map(*.chr).first }).join
00:50 camelia rakudo-moar 9658dd: OUTPUT: «Use of Nil in string context?  in block <unit> at <tmp> line 1?Use of Nil in string context?  in block <unit> at <tmp> line 1?Use of Nil in string context?  in block <unit> at <tmp> line 1?Use of Nil in string context?  in block <unit> at <…»
00:50 timotimo eh.
00:51 timotimo ah, of course, upper case everything in unicode
00:51 timotimo m: say (do for "coetry".uc.comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains(" $letter ")).grep({ rand < 0.2 }).map(*.chr).first }).join
00:51 camelia rakudo-moar 9658dd: OUTPUT: «???????»
00:51 timotimo yeah boyyyyy that's what i'm talking about
00:51 PerlJam I did not think that would turn out so well
00:51 timotimo the unicode consortium has our backs
00:52 timotimo as well as our hacks
00:52 zoll what does "rand < 0.2" do?
00:52 timotimo rand gives you a random number between 0 and 1
00:52 timotimo actually, that's totally unnecessary
00:52 zoll no, besides the obvious
00:52 timotimo since i already randomize the sequence of numbers anyway
00:52 lookatme m: say rand for 1 .. 3;
00:52 camelia rakudo-moar 9658dd: OUTPUT: «0.808007052908069?0.675879308772485?0.975016660427953?»
00:52 PerlJam m: put [~] <c e r> Z~ <o t y>   # KISS
00:52 camelia rakudo-moar 9658dd: OUTPUT: «coetry?»
00:53 timotimo m: say [+] (rand xx 100).grep({ rand < 0.2 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «10.5056290971469?»
00:53 timotimo m: say [+] (rand xx 100).grep({ rand < 0.3 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «7.37854687110917?»
00:53 timotimo er, not what i was aiming for
00:53 timotimo m: say [+] (1 xx 100).grep({ rand < 0.3 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «25?»
00:53 timotimo m: say [+] (1 xx 100).grep({ rand < 0.5 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «49?»
00:53 timotimo m: say [+] (1 xx 100).grep({ rand < 0.2 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «26?»
00:53 timotimo m: say [+] (1 xx 100).grep({ rand < 0.2 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «20?»
00:53 timotimo m: say [+] (1 xx 100).grep({ rand < 0.2 })
00:53 camelia rakudo-moar 9658dd: OUTPUT: «16?»
00:53 timotimo when you grep by { rand < 0.2 } you'll get about 20% of all items
00:54 timotimo m: say (do for "coetry".uc.comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains(" $letter ")).map(*.chr).first }).join
00:54 camelia rakudo-moar 9658dd: OUTPUT: «???????»
00:54 timotimo anyway, it was useless anyway :)
00:54 * zoll hmms ...
00:54 timotimo m: say (do for "coetry".uc.comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains(" $letter ")).map(*.chr).first }).join
00:54 camelia rakudo-moar 9658dd: OUTPUT: «???????»
00:54 timotimo m: say (do for "coetry".uc.comb -> $letter { ^0x10000 .pick(*).grep((*.uniname ~ " ").contains(" $letter ")).map(*.chr).first }).join
00:54 camelia rakudo-moar 9658dd: OUTPUT: «???????»
00:54 timotimo m: say (do for "coetry".uc.comb -> $letter { ^0x10000 .pick(*).grep((*.uniname ~ " ").contains(" $letter ")).map(*.chr).first }).join
00:54 camelia rakudo-moar 9658dd: OUTPUT: «???????»
00:54 timotimo m: say (do for "coetry".uc.comb -> $letter { ^0x10000 .pick(*).grep((*.uniname ~ " ").contains(" $letter ")).map(*.chr).first }).join
00:54 camelia rakudo-moar 9658dd: OUTPUT: «????r??»
00:54 timotimo this variant also allows the letter to stand at the very end of the uniname
00:54 PerlJam those last two are more what I thought would happen  :)
00:55 timotimo yeah, it's because those that have " A " in them are usually "LATIN LETTER A WITH WEIRD SQUIGGLES"
00:55 jeromelanteri morning, someone see the last comment for me ? (to deep in the history for me)
00:55 timotimo whereas with accepting the " A" at the end of the uniname as well you get things like "MARTIAN LETTER KINDA LIKE A"
00:55 timotimo 184728*   Zoffix ? jeromelanteri: debugger was fixed like a few days after 2017.06 release
00:55 merced PerlJam: interesting idea
00:56 jeromelanteri timotimo, ok, thank you
00:57 jaldhar_ joined #perl6
01:01 timotimo R0b0t1: bad news: both dyncall and libffi use and actually rely on executable stack :\
01:02 timotimo (based on a very cursory google search)
01:10 timotimo coetry left :<
01:10 lookatme Haha
01:11 timotimo m: say (do for "goodbye".uc.comb -> $letter { ^0x10000 .pick(*).grep(*.uniname.contains(" $letter ")).map(*.chr).first }).join
01:11 camelia rakudo-moar 9658dd: OUTPUT: «??????ê?»
01:12 timotimo i'll go sleep now. have a good one!
01:12 * R0b0t1 sighs
01:13 R0b0t1 timotimo: Thanks for checking
01:14 timotimo https://github.com/Snaipe/dyncall/blob/master/dyncall/dyncall_call_x86_8a.s#L41
01:16 R0b0t1 So it doesn't really need it but changing it would be a contribution to dyncall
01:18 timotimo looks like
01:20 Cabanossi joined #perl6
01:23 coetry joined #perl6
01:41 raschipi joined #perl6
01:52 ilbot3 joined #perl6
01:52 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!
01:56 astj joined #perl6
01:59 dj_goku joined #perl6
02:03 astj joined #perl6
02:05 astj joined #perl6
02:06 astj joined #perl6
02:17 pilne joined #perl6
02:38 jaldhar Given two lists of characters of equal length how do I get a count of characters which are the identical and in the same position in both lists?  E.g. 'abcd' and 'adcb' should give 2.
02:39 jaldhar It's easy enough to do with a loop but is there a faster/more concise way?  I'm thinking something with reduction operators maybe?
02:51 colomon m: say <a b c d> >>eq<< <a d c b>
02:51 camelia rakudo-moar 9658dd: OUTPUT: «(True False True False)?»
02:52 colomon m: say (<a b c d> >>eq<< <a d c b>).map(*.Int)
02:52 camelia rakudo-moar 9658dd: OUTPUT: «(1 0 1 0)?»
02:52 colomon m: say [+] (<a b c d> >>eq<< <a d c b>).map(*.Int)
02:52 camelia rakudo-moar 9658dd: OUTPUT: «2?»
02:52 colomon m: say [+] (<a b c d> >>eq<< <a d c d>).map(*.Int)
02:52 camelia rakudo-moar 9658dd: OUTPUT: «3?»
02:52 colomon dunno that that’s the best approach
02:53 zengargoyle m: say ('abcd'.comb Zeq 'adcb'.comb).grep(*.so, :k)
02:53 camelia rakudo-moar 9658dd: OUTPUT: «(0 2)?»
02:53 yoleaux 27 Jul 2017 21:43Z <thundergnat> zengargoyle: The evolution one isn't one I wrote. I think that one was mostly grondilu and TimToady. Not too many people say my code is formatted too nicely so that was a big clue :-)
02:53 lookatme m: say <a b c d> Zeq <a d c b>
02:53 camelia rakudo-moar 9658dd: OUTPUT: «(True False True False)?»
02:53 colomon m: say [+] (<a b c d> >>eq<< <a d c d>)
02:53 camelia rakudo-moar 9658dd: OUTPUT: «3?»
02:53 colomon m: say [+] (<a b c d> >>eq<< <a d c b>)
02:53 camelia rakudo-moar 9658dd: OUTPUT: «2?»
02:54 jaldhar colomon: cool and very interesting!  Thanks.
02:55 colomon Zeq might be a better choice than >>eq<< … I’m afraid I don’t really know which is more efficient at the moment.
02:55 colomon m: say [+] (<a b c d> Zeq <a d c b>)
02:55 camelia rakudo-moar 9658dd: OUTPUT: «2?»
02:55 bwisti joined #perl6
02:55 zengargoyle m: say ('abcd'.comb Zeq 'adcb'.comb).grep(*.so).elems
02:55 camelia rakudo-moar 9658dd: OUTPUT: «2?»
02:58 noganex_ joined #perl6
02:58 jaldhar colomon: it seems both >>+<< and Zeq perform about equally in rakudo 2017.07 atleast for my arrays which aren't very large.
02:59 colomon jaldhar++ # testing performance is good
02:59 zengargoyle m: say [Zeq] ('abcd','adcb')>>.comb
02:59 camelia rakudo-moar 9658dd: OUTPUT: «(True False True False)?»
03:05 zengargoyle m: say ([~^] ('abcd','adcb')).comb>>.ord.grep(!*,:k)
03:05 camelia rakudo-moar 9658dd: OUTPUT: «(0 2)?»
03:06 zengargoyle depending on data, string xor and counting 0's might work. :)
03:10 jaldhar to clarify these are arrays of characters not strings though I suppose I could join them and then do the xor.
03:12 zengargoyle i'm just goofing/golfing.  if you already have array of chars, eq is probably best.  guess i didn't read the original question closely enough. :)
03:14 zengargoyle buggable: eco leven
03:14 buggable zengargoyle, Found 2 results: Text::Levenshtein, Text::Levenshtein::Damerau. See https://modules.perl6.org/#q=leven
03:15 zengargoyle there's a module that does text 'difference' calculations that i haven't used...
03:16 jaldhar zengargoyle: btw I'm finally looking at that rosettacode example you mentioned earlier.  I don't understand everything it is doing but I'm pleased to see I came up with the same general approach.
03:18 zengargoyle jaldhar: i understand after i figured out the loop thing and it's pretty sweet code written by somebody with some p6 skills.  it's quite nice.
03:19 jaldhar agreed
03:26 pompomcrab joined #perl6
03:29 zengargoyle jaldhar: oh, somebody else mentioned the code originally, i just looked out of curiosity.
03:34 Cabanossi joined #perl6
03:40 coetry joined #perl6
03:42 pharv joined #perl6
03:47 coetry anyone know of anyone working on a rust backend / implementation of perl6?
03:47 coetry timotimo: haha i checked the irc logs and see that ya'll had some fun with my name :), made me smile!
03:56 pharv1 joined #perl6
03:56 pharv1 left #perl6
04:00 jeromelanteri someone use Git::Wrapper ? if i just want to send command git ls-remote https://url from Git::Wrapper, how to do that ? and then how to get back outputs ?
04:00 jeromelanteri https://github.com/nicqrocks/p6-Git-Wrapper
04:00 jeromelanteri doc and example are minimalist.
04:01 MasterDuke joined #perl6
04:09 zengargoyle i'd think $git.run('ls-remote', $url);
04:11 zengargoyle returns @out of lines of the command given.
04:14 zengargoyle perl6 -e 'use Git::Wrapper; .say for Git::Wrapper.new(gitdir=>"p6-Search-Dict").run: "ls-remote", "https://github.com/zengargoyle/p6-Search-Dict.git"; '
04:15 zengargoyle does what i'd expect.  seems you must supply a gitdir argument which is a bit lame but understandable.  guess you need a git repository somewhere to actually use the wrapper.
04:17 jaldhar Before I submit my module for inclusion in the ecosystem I just want to clarify one thing.  If my module is dual-licensed, Artistic or GPL2 and above, I include this in META6.json
04:17 jaldhar "license" : [
04:17 jaldhar "Artistic-2.0",
04:17 jaldhar "GPL-2.0+"
04:17 jaldhar ],
04:17 jaldhar is that correct?
04:17 zengargoyle but... gitdir=>'.' also works in this case even when '.' is not a gitdir.... :)
04:20 astj joined #perl6
04:32 Todd_ joined #perl6
04:33 Todd_ Hi!  I am trying to convert a p5 program to p6.  what do I use in place of `LWP::UserAgent`.  I use it for downloading files from the web
04:34 Todd_ I need to be able to pass the following to the web page: Caller, Host, UserAgent, Referer, Cookies
04:34 raiph joined #perl6
04:37 jaldhar Todd: there's a perl6 port of LWP::Simple you can install with zef.  Haven't tried it myself.
04:38 Todd_ I fpund it but could not find the directions
04:40 jaldhar Todd: then there is WWW and HTTP::UserAgent which seem to be better documented.  I haven't tried those either but the latter looks like a port of LWP.
04:46 Todd_ someone wrote a how to over on the mailing list.  I will dig it out as I think it has the directions.  Thank you for www and useragent!  bye bye
04:46 skids .oO(aww no Perl6 page on http://en.illogicopedia.org)
04:46 jeromelanteri zengargoyle, thank you
04:49 zengargoyle jeromelanteri: just noticed that the nicqrocks url for Git::Wrapper you gave is a fork of perlpilot's code.  don't know if this matters or there are any differences... just FYI
04:52 jeromelanteri zengargoyle, ho ok... but i install module from zef, so it should instal the one you was talking about.
04:54 zengargoyle looks like zef returns the niqrocks version... maybe niqrocks took over development or something... no big deal.
04:56 zengargoyle i just noticed that one can't create an issue on the niqrocks fork so i can't complain about the gitdir thing. :P
04:58 pharv joined #perl6
04:58 perigrin m: my @a = <a a c>; my @b = <a b>; say (@a (&) @b) / (@a | @b);
04:58 camelia rakudo-moar 9658dd: OUTPUT: «any(0.333333, 0.5)?»
04:59 perigrin m: my @a = <a a c>; my @b = <a b>; say (@a (&) @b) / (@a (|) @b);
04:59 camelia rakudo-moar 9658dd: OUTPUT: «0.333333?»
05:05 xtreak joined #perl6
05:09 ufobat joined #perl6
05:11 xtreak joined #perl6
05:16 xtreak_ joined #perl6
05:19 xtreak joined #perl6
05:34 coetry joined #perl6
05:43 coetry joined #perl6
05:47 jeromelanteri zengargoyle, from Git::Wrapper.new(gitdir => '.'); it said SORRY! STable conflict detected during deserialization. (so the good news is that the conflict is stable... :) )
05:48 jeromelanteri no.. it is directly at use Git::Wrapper time...
05:49 Cabanossi joined #perl6
05:49 jeromelanteri it also said that: Probable attempt to load two modules that can not be loaded together.
05:50 jeromelanteri i have use Prompt::Gruff::Export and use Term::Choose loaded.
05:51 cpage_ joined #perl6
05:52 coetry joined #perl6
05:56 jeromelanteri if i do a little file just for add modules and see.. it run.
05:57 jeromelanteri but if i comment use Git::Wrapper from my programm... it works. So it is strange, how is it possible to not be able to use two modules together ? and how to know why precisely there is a STable conflict ?
05:57 jeromelanteri if i know that, maybe i should add en issue for the faulted module who can not have a life with other.
06:01 coetry_ joined #perl6
06:02 jeromelanteri conflict seems to be (maybe) with Prompt::Gruff::Export
06:03 Miro_ joined #perl6
06:03 zengargoyle perl6 -e 'use Term::Choose; use Prompt::Gruff::Export; use Git::Wrapper;  .say for Git::Wrapper.new(gitdir=>".").run: "ls-remote", "https://github.com/zengargoyle/p6-Search-Dict.git";'
06:03 zengargoyle still works for me....
06:05 zengargoyle so all three can be use'd and at least Git::Wrapper works in that case.
06:11 zengargoyle maybe a version/os thing, i'd suspect Term::Choose because of ncursed and nativall type stuff.  the Prompt and Git::Wrapper seem mostly generic....
06:12 * zengargoyle lol at s/ncursed/ncurses/
06:13 zengargoyle it will probably take somebody who knows what STable could mean and looking at your code.
06:16 zengargoyle perl6 -e 'use Term::Choose; use Prompt::Gruff::Export; use Git::Wrapper;  .say for Git::Wrapper.new(gitdir=>".").run: "ls-remote", "https://github.com/zengargoyle/p6-Search-Dict.git"; say my $chosen = choose( <1 2 3>, :layout(2) );say prompt-for("Enter name (required): ");'
06:17 andrzejku joined #perl6
06:17 zengargoyle simple test of all 3 modules works for me...
06:20 coetry_ joined #perl6
06:20 jeromelanteri zengargoyle, same for me
06:21 jeromelanteri only my code failed. but use lib is call from a module in do. who are introduced from own lib (use lib 'my path")
06:23 jeromelanteri an other one different question, i use vim and TagBar plugin, but still no have tags list... i do install universal-ctags in place of ctags for try to resolve, but not works to. How to resolve this problem (if someone use vim and TagBar and know) ?
06:24 jeromelanteri zengargoyle, i see that if i add this Git::Wrapper lib from main file (.pl6, not module file .pm6) no error... but i need it in the module file...
06:25 xtreak joined #perl6
06:26 rindolf joined #perl6
06:28 wamba joined #perl6
06:29 xtreak_ joined #perl6
06:30 mr-foobar joined #perl6
06:31 coetry joined #perl6
06:35 cpage_ joined #perl6
06:39 jeromelanteri same problem if i do use Git::Wrapper from a file .pm6 who contain a class
06:39 zengargoyle i'd say put your code on github or somewhere so somebody smarter than me can look.  if it is really a problem with 'use lib' and use from module and conflict errors.... it might be something weird.
06:40 jeromelanteri ok
06:41 coetry_ joined #perl6
06:41 zengargoyle STable conflict detected during deserialization -- is a weird internally looking error that is less than helpful :)
06:42 zengargoyle at least i've never seen it before and have no clue what it could mean.
06:42 jeromelanteri IDK, my level is to bad (start to learn perl6 not one week ago)
06:43 jeromelanteri so it should also be something i done wrong. I 'm going to create a public bitbucket repo and git push this
06:44 andrzejku joined #perl6
06:49 Cabanossi joined #perl6
06:49 lookatme STable, I remember this is a thing in NQP
07:03 darutoko joined #perl6
07:04 parv joined #perl6
07:05 domidumont joined #perl6
07:12 domidumont joined #perl6
07:16 abraxxa joined #perl6
07:19 nadim joined #perl6
07:20 miro_ joined #perl6
07:27 rindolf joined #perl6
07:30 jeromelanteri https://bitbucket.org/jerome___/vimplugger/src
07:30 jeromelanteri my bitbucket (git) public repo for this little code
07:32 lowbro joined #perl6
07:32 lowbro joined #perl6
07:33 jeromelanteri NQP ==> nOOb Questions Protocol,correct ?
07:33 eviltwin_b joined #perl6
07:34 lookatme nqp: nqp::say("123");
07:34 camelia nqp-moarvm: OUTPUT: «123?»
07:35 geekosaur joined #perl6
07:35 lookatme Not Quite Perl
07:37 lookatme nqp: nqp::if(1 == 3, nqp::say("123"), nqp::say("453"));
07:37 camelia nqp-moarvm: OUTPUT: «453?»
07:39 pecastro joined #perl6
07:39 lookatme jeromelanteri, How to use your script ?
07:40 jeromelanteri lookatme, actually, just start ./VimPlugger.pl6
07:40 lookatme ok
07:41 jeromelanteri if you have libs i use installed, then it should start by a menu when this error gone
07:41 jeromelanteri if you comment use Git::Wrapper from check.pm6, it should start well
07:42 geekosaur joined #perl6
07:42 lookatme OK, I will try
07:42 jeromelanteri thank you
07:44 geekosaur joined #perl6
07:47 lookatme There is some error, undeclared routine: yaml ?
07:47 lookatme jeromelanteri
07:48 zengargoyle jeromelanteri: first peek.  it will get much easier if you use App::Mi6 or something and create a standard module/distribution structure.  for example, you're code won't work for anybody because 'use lib "/home/jerome/PROGRAMATION/....";
07:49 Cabanossi joined #perl6
07:49 lookatme I change it to  "."
07:51 jeromelanteri yes, i didn't know that, thank you
07:51 jeromelanteri i have to read/learn about App::Mi6
07:52 lookatme What is yaml ?
07:53 jeromelanteri a kind of yaml parser
07:55 lookatme The code in the bitbucket is not correct
07:56 zengargoyle lookatme: it's dump/load not yaml.dump/yaml.load.
07:56 lookatme I delete all `yaml.` in the pm file, and all thing working expect that error you mentioned
07:56 lookatme yeah
07:57 jeromelanteri lookatme, ho really... let me check this.
07:57 lookatme s/expect/except/
07:58 zakharyas joined #perl6
07:59 zengargoyle there are some 'has method' in datas
08:00 samcv huggable, deb
08:00 huggable samcv, CentOS, Debian, Fedora and Ubuntu Rakudo packages: https://github.com/nxadm/rakudo-pkg/releases
08:00 espadrine joined #perl6
08:01 jeromelanteri zengargoyle, it should not in a class ?
08:01 jeromelanteri oh yes...
08:02 zengargoyle m: check
08:02 camelia rakudo-moar 9658dd: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>?Undeclared routine:?    check used at line 1. Did you mean 'CHECK'???»
08:03 jeromelanteri lookatme, https://github.com/yaml/yaml-perl6/blob/master/README.md
08:06 zengargoyle jeromelanteri: on first look, you're trying too hard. :)  lots of my sub and our sub and is export(:MANDATORY) and when "$answer" that looks like more than you probably need to actually do. :)
08:06 zakharyas joined #perl6
08:07 lookatme Yeah, I got the error
08:07 lookatme STable conflict detected during deserialization.
08:07 lookatme It always report this error, though I comment that use Git::Wrapper.
08:09 jeromelanteri zengargoyle, wow but... i done it.
08:10 jeromelanteri zengargoyle, maybe it seems to hard because the script code is not shiny... i used to first do something, then rewrite the code (and sometimes redesign all the design pattern after).
08:11 jeromelanteri lookatme, not for me, if Git::Wrapper is commented, no more error on that.
08:13 jeromelanteri do retry it three times, i confirm, if i comment Git::Wrapper from my side, it works good.
08:13 zengargoyle jeromelanteri: yeah, i see that in places.. :)   how far did you get before you got the error?  like... which parts did work.
08:14 zengargoyle m: add
08:14 camelia rakudo-moar 9658dd: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>?Undeclared routine:?    add used at line 1. Did you mean 'dd'???»
08:14 zengargoyle m: install
08:14 camelia rakudo-moar 9658dd: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>?Undeclared routine:?    install used at line 1??»
08:14 jeromelanteri zengargoyle, just when i try to add Git wrapper from check.pm6 file in git_repos_exist, and in same time in datas method install_repos_from
08:16 jeromelanteri zengargoyle, add and install are modules shared by "our" i choosed to do like this for clarity and be able to have functions with same names in different modules... then because of "our" in fucntion declaration modules, i can call them by the first module name :: function
08:17 jeromelanteri fomr a kind_of.pm6 file, unit module kind_off; our sub blabla is export(:MANDATORY) {...} from an other file: use kind_of; kind_of::blabla;
08:18 TimToady joined #perl6
08:18 jeromelanteri normaly no need to put use kind_of :MAN?DATORY; the doc said :MANDATORY is automatically exported.
08:18 jeromelanteri by defqault
08:18 jeromelanteri default
08:18 jeromelanteri and that's it.
08:18 zengargoyle jeromelanteri: assuming most code works, i'm thinking our and symbols and same name stuff causes collisions somewhere.
08:18 lookatme What is $datas, I found this in some module reference module datas
08:19 jeromelanteri zengargoyle, ho yes, that's should be a reason, why not... IDK
08:19 lookatme k
08:19 Cabanossi joined #perl6
08:20 jeromelanteri from vim /$datas ==> not find. I think you mean $.datas ?
08:20 zengargoyle for me.... it should be VimPlugins::datas, VimPlugins::install, VimPlugins::add, etc.  and maybe VimPlugins::add::somefunc()  before trying to make it look easy.
08:21 zengargoyle once it works, you can make it easy to use. :)
08:21 jeromelanteri this is a class public variable with attributes. It contain $yaml datas parsed from yaml file
08:21 lookatme I don't see where $datas is create
08:21 lookatme OH..
08:21 jeromelanteri zengargoyle, i not well understand. Do you mean at import time ?
08:22 jeromelanteri lookatme, ligne 8
08:23 jeromelanteri and ligne 30 show you the yaml parse
08:23 jeromelanteri i do initialize things at method TWEAK
08:23 jeromelanteri i not used submethod BUILD
08:24 lookatme Oh, that's my mistake
08:25 nadim joined #perl6
08:25 jeromelanteri I'm sorry also, i not add coments... i didn't know this to be shared also... i can add some comments if you want.
08:26 jeromelanteri (but with my poor english... not sure it will not be more confuse... :)
08:26 jeromelanteri or more clear
08:27 lookatme Oh, it said YAML.load not implement ?
08:27 jeromelanteri lookatme, yes sorry, my fault to not precise at first commit in thge README file to NOT use the zef install YAML
08:28 lookatme OH, so how to ?
08:28 jeromelanteri but instead use zef install from git yaml-perl6 ... i do change the README actually
08:28 jeromelanteri waiting, i copy paste
08:28 jeromelanteri lookatme, zef install https://github.com/yaml/yaml-perl6.git
08:29 jeromelanteri (remove the other one first)
08:29 cpage joined #perl6
08:31 jeromelanteri then also, if Git::Wrapper is use from main file... no error ! very strange
08:32 jeromelanteri nut maybe zengargoyle find the idea... it should be my modules who make conflmict because of "our"
08:32 jeromelanteri i also not well understand how to try to fix it.
08:32 jeromelanteri and keep the separat call for module: name_of_module::function
08:34 jeromelanteri also, i tryed to use the perl6-debug-m for rack deeply the problem... serve nothing, the debuger doesn't works on this case (or i miss the point to make it work on that error)
08:34 jeromelanteri for track
08:42 eythian joined #perl6
08:44 lookatme Cannot locate native library 'libyaml.so': /lib64/libyaml.so: file too short
08:45 zengargoyle lookatme: on arch?
08:45 lookatme fedora
08:45 zengargoyle cat /lib64/libyaml.so
08:45 lookatme INPUT(libyaml-0.so.2)
08:46 zengargoyle like arch, fedora evidently uses `ld` magick which doesn't work with nativecall.....
08:47 lookatme Oh, How to fix that ? create a symbol link ?
08:47 zengargoyle you could rm /lib64/libyaml.so and (cd /lib64; ln -s libyaml-0.so.2 libyaml.so) or something....
08:48 Doe joined #perl6
08:48 lookatme Oh, I wonder why they not using a symbol link
08:48 zengargoyle or edit look at module and edit to use a libyaml-0.so.2....
08:49 zengargoyle it's like a known issue that has no good solution yet i think...
08:49 zengargoyle why fedora/arch does this insane thing i have no idea. :P
08:49 lookatme Hmm
08:52 jeromelanteri lookatme, at one point, i had this problem and created a environment variable (but the module asked for that) PERL6_NCURSEs_LIB
08:52 lookatme OH
08:53 jeromelanteri zengargoyle, look if there is the way of variable env export
08:53 jeromelanteri i'm from archlinux
08:53 robertle joined #perl6
08:53 zengargoyle native call.  is native('libyaml.so')....  some librarys make that a sub or environment variable and do magic.
08:55 zengargoyle so if module author thought to add env override maybe you can change it, otherwise edit code or make symlink.
08:55 jeromelanteri zengargoyle, if you make symlink, it should works, but also you shoumd break your distro...
08:55 jeromelanteri should
08:55 cpage joined #perl6
08:56 jeromelanteri at package time call from packet manager, it should cry
08:56 lookatme Oh, it's working now
08:57 zengargoyle well, i've moved and edited a bunch of files, and even moved the 'use Git::Wrapper;' in datas down to inside the single function that uses it....  it bonks if it's there, and works otherwise (i guess not knowing exactly what it's supped to do but getting an ncurses menu thing.... :) )
08:57 lookatme Which module conflict with Git::Wrapper ?
08:57 jeromelanteri mine maybe...
08:58 zengargoyle jeromelanteri: i've used unix for 30 years... :) i can *make* it work.
08:58 jeromelanteri in fact, if i import Git::Wrapper from main file, it works... it doesn't works only when import from pm6 files
08:59 jeromelanteri zengargoyle, you better than me, i make use it from 32 years, and sometimes i loose again some hairs...
08:59 zengargoyle perl6 on the other hand...
09:00 jeromelanteri did you try to run debuger ?
09:01 jeromelanteri maybe you can see some datas make sens for you (not for me, but maybe for you)
09:01 jeromelanteri or strace ?
09:01 zengargoyle anyway, it's 2am here and i'm going to bed for now.  i'll revisit tomorrow.
09:02 jeromelanteri zengargoyle, thank you for try to help me and have a good night man.
09:02 zengargoyle hopefully somebody figures it out...
09:02 astj joined #perl6
09:03 astj joined #perl6
09:03 lookatme good night! zengargoyle
09:04 Cabanossi joined #perl6
09:08 jeromelanteri lookatme, i make a break pause... will be AFK... and i will take time for reflection then, i will be back with fresh ideas. It should be my module callas who make problems... it's possible i think.
09:13 lookatme oik
09:13 lookatme ok
09:35 jeromelanteri by the way, remove our declaration of functions from modules and cal by &group instead of add::group (and same for all other) doesn't fix the problem... so it should be a module Git::Wrapper code problem (maybe).
09:37 jeromelanteri i created a branch git_wrapper also...
09:38 jeromelanteri a commit show that not fix problem... now i go back next commit to decclaration with our and use module name for call functions
09:51 jeromelanteri i also asked on archlinux channel why do use magic links instead of symlink... approximative answer was that it get more control, but no one actually know exactly what this control is and why is it better...
10:04 Cabanossi joined #perl6
10:06 andreoss joined #perl6
10:07 Geth joined #perl6
10:21 nadim hi, I am requiring a module which is not installed, its part of my development, I tried use lib '.' but I get this error 'use lib' may not be pre-compiled
10:25 nadim I am trying a (try require ::Data::Dump::Tree::LayFlat) but that gives me run time error: No such symbol 'Data::Dump::Tree::LayFlat'
10:26 jeromelanteri nadim, could you call a simple module lib from your code ? does it works ?
10:27 jeromelanteri did you also declare your module from your module file by unit module name_of_your_module (same as file_name without .pm6)
10:28 nadim Yes, but let me nopaste the code, it's simpler
10:28 jeromelanteri yes
10:35 nadim jeromelanteri: https://nopaste.linux-dev.org/?1160079
10:39 jeromelanteri nadim, you not use is export for sub function, and then i'm not sure the way you declare the module can works..  you can use lib from main file (it should share libs for other), then call unit module LayFlat (name of your file). Not sure for the last reqquest... also, not sure but files extentions convention is .pm6
10:39 jeromelanteri also, i do use is export(:MANDATORY): sub lay_flat is export(:MANDATORY) { ... }
10:40 nadim I believe the problem is finding the module, the error doe not seem to indicate we got that far.
10:40 jeromelanteri if u declare use lib 'lib', it goes for search lib from ./lib/ ...
10:40 nadim that's where the module is
10:41 jeromelanteri nadim, good. you should wait for an other one know better... i'm a noob actualy...
10:41 nadim also, using .pm works fine, although I should change that some day
10:42 jeromelanteri ok
10:42 nadim there are 3 Gods, 10 titans, 300 noobs and a whole bunch clueless, you still are in quite ok cathegory ;)
10:43 Aaronepower joined #perl6
10:43 nadim I am requiring another module but that one is installed, I just want to test without having to install a, potentially broken, module all the time
10:43 jeromelanteri nadim :) not sure... my frustration is to not be able to help you more by the fact...
10:44 jeromelanteri i'm waiting to see what should be yyour solution for learn more around that to.
10:44 nadim well, we'll both eventually learn something when one of the titans wakes up
10:49 jeromelanteri from sources of module, they add "use v6;"
10:49 nadim ok
10:50 Zoffix joined #perl6
10:50 nadim hmm, it was a good idea to try to installll too it seem as I found another install error!
10:50 Zoffix nadim: you can't precompile `use lib`; so you eaither need a `no precompile` or put it into your script instead of the module
10:51 Zoffix nadim: as for require. Is that sub exported?
10:51 Zoffix Use ::('Data::Dump::Tree::LayFlat::EXPORT::DEFAULT::&lay_flat')
10:52 nadim no it was not exported, adding that
10:52 Zoffix Right without an export or `our` it's not accessible from the outside
10:55 nadim I run p6 via an alias, I thought it would take care of things. alias p6='perl6 -Ilib,t'
10:55 timotimo setting lib to t? that's ... an interesting decision
10:56 nadim old habit when test libs end up in t
10:57 nadim they should probably be somewhere else, maybe
10:57 Zoffix nadim: it takes care of adding lib and t to module search paths, not to what symbols are visible from outside the module
10:59 nadim ok, bu the error, No such symbol 'Data::Dump::Tree::LayFlat', happens oin the require line, not when I am using a symbol from that module. https://nopaste.linux-dev.org/?1160079
10:59 Zoffix left #perl6
11:04 Cabanossi joined #perl6
11:04 llfourn nadim: why's there a :: before Data?
11:05 llfourn try require Data::Dump::Tree::LayFlat # <-- should be?
11:08 nadim llfourn: if I recall properly, it was from an example I saw. I require Terminal::ANSIColor like that and it works. I tried without the :: and got no better result
11:08 llfourn hmm weird I've never seen that before
11:09 llfourn is there some reason you're not doing try require Data::Dump::Tree::LayFlat <&lay_flat>; nand the njust pushing &lay_flat?
11:09 llfourn *and then
11:10 llfourn hmm but you say the error happens on the require line...
11:10 nadim no special reason, that certainly looks neater. but right now the problem is not with the sub, even if I do not use lay_flat, i still get an error loading the module
11:10 llfourn it could easily be a bug
11:10 nadim I hope not :)
11:10 llfourn the require implementation is pretty dodgey in that area
11:11 llfourn it does some dynamic lookup to get the class it's meant to be returning
11:11 llfourn and that sounds like where the error is coming from
11:11 llfourn I've been meaning to fix it...
11:11 nadim Do it, do it ;)
11:12 nadim right now I use this: (try require ::Data::Dump::Tree::LayFlat:file('Data/Dump/Tree/LayFlat.pm')) !=== Nil
11:12 llfourn can you give the exact reproduction steps and send to rakudobug@perl.org?
11:13 nadim so I am giving the path too. I do not "use lib" as it is not necessary. I found out by giving a wrong name and I got an error message showing me the libs paths, they were fine
11:13 nadim llfourn: of course I can do that :)
11:13 llfourn does it happen if you just do: require Data::Dump::Tree::LayFlat
11:13 nadim that's how I started, same error
11:14 nadim require works for installed modules
11:14 llfourn well, if everything is set out like you show it's probably a bug I *might* have a fix for
11:14 nadim this one is part of the development not installed yet
11:14 nadim let's see if I can golf this
11:15 llfourn nadim: that would help :)
11:19 cog_ joined #perl6
11:21 nadim llfourn: golfed it works :) let me break it again ;)
11:22 llfourn it might be something with the many :: deep name
11:24 jeromelanteri well... let's learn why STable conflict happen when i use Git::Wrapper module then ? (if you want)
11:24 jeromelanteri https://bitbucket.org/jerome___/vimplugger/src/258cb34cefe6b4d394c00ff586bb52ee277a1220?at=git_wrapper
11:26 jeromelanteri i'm burning sage and make incantations at this time...
11:36 nadim llfourn: no golfed broken version yet but the working code from a script, when called from a method, generates an error.
11:36 mr-fooba_ joined #perl6
11:41 llfourn nadim: so it sounds really tricky to reproduce which supports my conclusion it's related to the problem I suspected
11:42 llfourn it would be very handy to have a golfed version so we can turn it into a test though
11:46 mr-foobar joined #perl6
11:47 jaldhar This is strange
11:48 jaldhar Enabled extracting backends [git path tar p5tar unzip] don't understand /home/jaldhar/.zef/tmp/Algorithm-DawkinsWeasel
11:48 jaldhar You may need to configure one of the following backends, or install its underlying software - [psunzip]
11:48 jaldhar but I have unzip in my path.  What gives?
11:49 jaldhar Is my module (Algorithm::DawkinsWeasel) set up wrong?
11:49 jaldhar or is zef not configured properly?  If so where do I configure it?  (homemade install of rakudo 2017.07 in ~/.perl6)
11:51 llfourn jaldhar: I'm pretty sure it's meant to find those automatically
11:52 llfourn I've never used zef with anything other than git
11:52 nadim llfourn: trying to wrap things up in the golfed version but the darn things continues to work!
11:53 llfourn ugexe: what do you think wrt jaldhar's problem?
11:53 llfourn nadim: damn that's the worst
11:53 llfourn I hate it when you can't golf a rakudo bug
11:59 committable6 joined #perl6
11:59 quotable6 joined #perl6
11:59 bloatable6 joined #perl6
11:59 bisectable6 joined #perl6
11:59 greppable6 joined #perl6
11:59 evalable6 joined #perl6
11:59 coverable6 joined #perl6
11:59 unicodable6 joined #perl6
11:59 benchable6 joined #perl6
11:59 statisfiable6 joined #perl6
12:04 Cabanossi joined #perl6
12:08 amalia___ joined #perl6
12:17 bdmatatu joined #perl6
12:21 bdmatatu I am trying to eval code and save the context, like REPL::repl-loop.  It almost works but not quite: https://git.io/v7c6L
12:35 asdaddqwqe joined #perl6
12:35 asdaddqwqe hello for all
12:35 asdaddqwqe I am newcomer in Perl
12:35 nadim welcome
12:36 asdaddqwqe thanks
12:36 asdaddqwqe what for this irc channel?
12:36 moritz general questions around Perl 6
12:37 moritz (and specific questions as well :-)
12:37 asdaddqwqe mmm
12:37 asdaddqwqe good
12:37 asdaddqwqe will be know
12:39 zakharyas joined #perl6
12:44 Swey_Hey joined #perl6
12:44 mcmillhj joined #perl6
12:46 asdaddqwqe guys, are you like football?
12:47 asdaddqwqe fantasy premier league
12:52 perlpilot joined #perl6
12:52 perlpilot blah
12:56 perlpilot Zoffix++
13:04 itaipu joined #perl6
13:09 awwaiid bdmatatu: I would like it if what you have in your gist (or a variant) worked, but the REPL is not fully de-magic'd in ways I don't quite understand
13:09 awwaiid bdmatatu: However, I have constructed a devious way of _simulating_ the same thing through constructing nested scopes for subsequent evals
13:09 bdmatatu I think I just got it actually -- adding :interactive(1) seems to fix it
13:10 awwaiid bdmatatu: https://github.com/awwaiid/p6-lrep/blob/master/lib/LREP.pm6#L35
13:10 awwaiid bdmatatu: ah very good!
13:10 awwaiid On my very-deep-todo-stack is to come back around to some rakudo repl work :)
13:14 awwaiid I would rather the setup you had to be only creating an instance of REPL and then invoking .repl-eval, without any of the other things you had to do
13:14 awwaiid and _especially_ it shouldn't depend on nqp
13:15 bdmatatu Sounds good -- this is the application btw, have been trying to write a jupyter kernel: https://github.com/bduggan/p6-jupyter-kernel
13:16 awwaiid bdmatatu++ # !!!
13:16 awwaiid that is also somewhere on my infinite todo list :)
13:16 bdmatatu :)
13:16 andreoss` joined #perl6
13:16 awwaiid bdmatatu: you found the previous attempt someone worked on?
13:16 bdmatatu Yes
13:16 awwaiid very good
13:17 andreoss` m: class Foo {}; sub foo(Foo:D $x?) { 1 }; say foo();
13:17 camelia rakudo-moar c1e41f: OUTPUT: «Parameter '$x' of routine 'foo' must be an object instance of type 'Foo', not a type object of type 'Foo'.  Did you forget a '.new'??  in sub foo at <tmp> line 1?  in block <unit> at <tmp> line 1??»
13:17 awwaiid I would be delighted to hear of your pgoress on this and will try to find time to beta-test
13:17 bdmatatu awwaiid++ great!
13:18 andreoss` can I use optional parametrs with :D objects?
13:19 awwaiid andreoss`: maybe give it a default (defined) value
13:21 jnthn Only if you provide a default. Maybe we should catch that at compile time.
13:31 mspo stmuk: so benny is saying to package all of the "star" packages individually and just track them like that; which I think is what you were also recommending
13:31 mspo guess that's the way to go
13:32 mspo just need to figure out all of the right INC's to use + the incantation with zef (maybe)
13:33 mspo I always wasn't super clear on how other packaging systems are doing it
13:35 [Coke] timotimo: see also my unicode mangler if you want rando combiners.
13:35 skids joined #perl6
13:35 perlpilot joined #perl6
13:40 [Coke] timotimo: (g?????ö??????????d?????b?????y?????é????)
13:41 mspo [Coke]: that one works better than the one yesterday for me
13:41 lizmat joined #perl6
14:09 itaipu joined #perl6
14:10 Voldenet joined #perl6
14:10 Voldenet joined #perl6
14:20 zakharyas joined #perl6
14:25 nadim joined #perl6
14:27 popo__ joined #perl6
14:54 grumble joined #perl6
14:59 khw joined #perl6
15:03 domidumont joined #perl6
15:08 mspo so if I install rakudo star and want to update one of the included modules can zef do it?
15:08 mspo since there is no more @INC I'm not sure how to order stuff
15:08 timotimo yeah, zef can update stuff for you
15:12 mspo would it replace what star installed?
15:12 timotimo if the :ver is the same, yes
15:12 timotimo if it's a new :ver, it'll be installed alongside the older one so you can still get it with a more specific use statement
15:13 mspo okay
15:13 geekosaur iirc it will install alongside, since unlike perl 5 you can distinguish by :ver and possibly :auth
15:13 geekosaur heh
15:13 * geekosaur slow
15:13 timotimo like a little part of my zef list --installed looks like this:
15:13 timotimo JSON::Fast:ver('0.4')
15:13 timotimo JSON::Fast:ver('0.8.5')
15:13 timotimo JSON::Fast:ver('0.9')
15:13 timotimo JSON::Fast:ver('0.9.1')
15:14 mspo are those in site/ or vendor/ ?
15:14 timotimo i think vendor is for rakudo star and site is for what you install
15:14 timotimo or something
15:15 mspo timotimo: yeah I'm asking because I'm going to try (again) to get rakudo-star into pkgsrc
15:15 mspo timotimo: but right now we're headed down the path of just combining stand-alone nqp, moar, rakudo
15:15 timotimo did you look at nine's work on getting it packaged for suse?
15:16 mspo timotimo: and it's up in the air about how to handle the 30+ packages inside of star
15:16 timotimo problem with that is that the latest star release uses an unreleased nqp
15:16 mspo timotimo: people seem to be leaning towards just packing them up on their own
15:16 mspo timotimo: we can apply patchsets into nqp.  it's not a problem
15:17 mspo but why wouldn't nqp just do a point release to fix ? :)
15:18 mspo https://en.opensuse.org/Rakudo  <- looks out of date
15:19 timotimo huh i don't know where to find that stuff
15:19 mspo :)
15:19 mspo timotimo: pkgsrc has a bunch of stale attempts too
15:20 mspo timotimo: I'm hoping to get it cleaned up and modernized.  One of our guys is already most of the way there
15:20 timotimo nine built something called "CompUnitRepo::Staging" or similar
15:20 timotimo it allows you to install precompiled stuff by building it once and distributing the binaries
15:20 timotimo IIUC
15:22 mspo we also need a sanctioned way to intsall modules that might get packaged up
15:23 timotimo how do you mean?
15:24 mspo integrate zef into pkgsrc as a usable "tool" for building packages
15:24 mspo or some other method
15:24 mspo right now perl5 stuff uses makemaker directly I think
15:24 timotimo kind of like arch has alacryd?
15:24 timotimo not that i know anything about that tool
15:25 mspo I don't know it :)
15:25 mspo timotimo: pkgsrc is a big collection of software for netbsd, illumos, osx, and others
15:25 mspo timotimo: it's a giant collection of Makefile's
15:25 timotimo i mean i don't know anything about alacryd
15:25 mspo oh me either
15:25 timotimo i've heard of pkgsrc
15:26 mspo cool :)
15:26 mspo much nicer than brew for osx ;)
15:26 HoboWithShotgun joined #perl6
15:26 mspo just use pkgin to get all the binary packages; including perl6 if I can get some time to work on it!
15:26 HoboWithShotgun hi. I just installed rakudo and some modules with this zef tool.
15:27 HoboWithShotgun where do the modules get installed to?
15:27 mspo .say for $*REPO.repo-chain;
15:27 mspo run that I think
15:27 timotimo m: .say for $*REPO.repo-chain
15:27 camelia rakudo-moar c1e41f: OUTPUT: «inst#/home/camelia/.perl6?inst#/home/camelia/rakudo-m-inst-2/share/perl6/site?inst#/home/camelia/rakudo-m-inst-2/share/perl6/vendor?inst#/home/camelia/rakudo-m-inst-2/share/perl6?CompUnit::Repository::AbsolutePath.new(next-repo => CompUnit::Repos…»
15:28 mspo timotimo: oh I see; we could use compunit::repoisotry to package up star and just stick it on the install
15:28 mspo interesting but probably not what Benny wants
15:29 mspo also that means you can use compunit::repository to make FatPack-style stuff?
15:30 mspo ugexe++
15:32 HoboWithShotgun interesting. why do the file names of modules appear to be some checksum?
15:32 jdv79 joined #perl6
15:34 geekosaur because you can have multiple versions of a module installed
15:35 geekosaur so they get named by a hash of their identifying information (version, author, etc.) and I think maybe ABI?
15:35 moritz and the module name may contain Unicode characters that your file system may not like
15:35 geekosaur use Foo.Bar:ver(1.16); # for a specific version; default should be the latest installed
15:36 andrzejku_ moritz, hey!
15:36 HoboWithShotgun that's a bummer. i wanted to look at xml::node code
15:37 rumble joined #perl6
15:37 HoboWithShotgun actually, where is the documentation for the XML class?
15:37 stmuk is it possible to have nqp c0abee7953ac tagged as 2017.07.1?
15:38 geekosaur that wouldn't work anyway, I think --- iirc the CUR has only the compiled version?
15:38 HoboWithShotgun nvm, found it
15:39 nine The vendor repo is for packaged modules (think rpm, dpkg, ...). They are shipped by your operating system vendor.
15:39 nine The site repo is for modules installed by the user.
15:39 epony joined #perl6
15:39 nine Well user with sufficient access priviledges that is. There's still the home repo for unpriviledged users.
15:41 nine timotimo: CompUnit::Repository::Staging is essential for building precomp files with an unpriviledged user before packaging. It lets you install modules to a different location than where the will end up in, i.e. /tmp/BUILDROOT/usr/share/perl6/vendor.
15:41 perlpilot mi6++  (tried it for the first time today and it was doing something weird ... turned out it was a bug in my code!)
15:47 lucs I had the impression that mi6 was deprecated in favor of ddt.
15:47 lucs Is it?
15:49 timotimo nine: thank you for clarifying, i'll try to remember it
15:49 perlpilot I don't see that ddt handles CPAN distributions.  (that's the only reason I was using mi6)
15:50 nhywyll joined #perl6
15:51 lucs perlpilot: Oh, important difference, eh.
15:51 ugexe HoboWithShotgun: you can find the source for an installed module with `zef locate My::Module`
15:52 perlpilot but also ... mi6-- for changing my README and META6.json without me asking it to
15:52 pecastro joined #perl6
15:53 ugexe or `zef locate lib/My/Module.pm6` or `zef locate $some-sha1 --sha1`
16:04 ugexe llfourn: their source-url does not identify its type (https://github.com/jaldhar/Algorithm-DawkinsWeasel)
16:05 nadim llfourn: I thnk I will give up on trying to golf something to reproduce error: No such symbol 'Data::Dump::Tree::LayFlat', On the other hand I can try to debug the problem I have, if there is some sata I can extract from this case that can help, I'd be happy to do it
16:06 nadim llfourn: and maybe I give golfing another try but right now it has frustrated me enough.
16:08 mr-foobar joined #perl6
16:20 eroux joined #perl6
16:22 HoboWithShotgun perlbot: 'md5=c1a031bd227df4f7a4dae94e0135dd09' ~~ /md5\=(<[0..9a..z]>+)/; say qq[$0];
16:23 HoboWithShotgun rakudo: 'md5=c1a031bd227df4f7a4dae94e0135dd09' ~~ /md5\=(<[0..9a..z]>+)/; say qq[$0];
16:23 camelia rakudo-moar f083cf: OUTPUT: «c1a031bd227df4f7a4dae94e0135dd09?»
16:23 HoboWithShotgun rakudo: 'md5=c1a031bd227df4f7a4dae94e0135dd09' ~~ /md5\=(<[0..9a..z]>+)/; say $0;
16:23 camelia rakudo-moar f083cf: OUTPUT: «?c1a031bd227df4f7a4dae94e0135dd09??»
16:23 HoboWithShotgun where do these two extra characters come from?
16:25 zengargoyle m: 'md5=c1a031bd227df4f7a4dae94e0135dd09' ~~ /md5\=(<[0..9a..z]>+)/; say ~$0;
16:25 camelia rakudo-moar f083cf: OUTPUT: «c1a031bd227df4f7a4dae94e0135dd09?»
16:25 HoboWithShotgun oh, it's a match object.
16:25 AlexDaniel HoboWithShotgun: oh, you get a Match object. “say” is using .gist on whatever it gets, which is a human-readable output.
16:26 perlpilot Having to .Str the results of a regex capture is annoying
16:26 skids rakudo: 'md5=c1a031bd227df4f7a4dae94e0135dd09' ~~ /md5\=(<[0..9a..z]>+)/; print $0;
16:26 camelia rakudo-moar f083cf: OUTPUT: «c1a031bd227df4f7a4dae94e0135dd09»
16:27 pharv joined #perl6
16:27 AlexDaniel rakudo: 'md5=c1a031bd227df4f7a4dae94e0135dd09' ~~ /md5\=(<[0..9a..z]>+)/; put $0;
16:27 camelia rakudo-moar f083cf: OUTPUT: «c1a031bd227df4f7a4dae94e0135dd09?»
16:27 * zengargoyle notices nobody tried explaining the STable thing from last night. :(
16:27 cdg joined #perl6
16:32 Geth ¦ doc: 25806762ef | (JJ Merelo)++ | doc/Type/PseudoStash.pod6
16:32 Geth ¦ doc: First version of this missing page
16:32 Geth ¦ doc:
16:32 Geth ¦ doc: Refs #1341. Please check it out to see if the paraphrashing of @jnthn
16:32 Geth ¦ doc: and @lizmat is basically correct.
16:32 Geth ¦ doc:
16:32 Geth ¦ doc: I also looked up class inheritance in the "Any" diagram, so that
16:32 Geth ¦ doc: should be correct.
16:32 Geth ¦ doc: review: https://github.com/perl6/doc/commit/25806762ef
16:38 robertle joined #perl6
16:45 raschipi joined #perl6
16:46 andreoss` left #perl6
16:48 jnthn zengargoyle: I don't backlog this channel so closely, but I can probably answer most things about STable given I coined the term. :) What was the question?
16:48 andreoss` joined #perl6
16:50 zengargoyle jnthn: i think it boiled down to adding 'use GitWrapper;' in a module caused the STable conflict "maybe using two modules you can't use together" error from the reposess code.  (i just looked).
16:53 jnthn zengargoyle: Usually because a module does augment or other meta-programming without marking itself `no precompilation`
16:53 jnthn We should make augment imply that, most probably
16:53 jnthn Modules that use the MOP to change other module's objections will have to make the call themselves
16:54 zengargoyle jnthn: i'm not sure it's that complicated.   nothing fancy like that.  (it's sombody elses code)
16:54 nadim .tell llfourn I founf what the problem was in my code; I have unit module X ; use Y ; use Z ; sub want_this_sub{} ; When unit module X is after the 'use's it works, as if require did not put it in the right unit.
16:54 yoleaux nadim: I'll pass your message to llfourn.
16:55 jnthn Could be because of some dependency also
16:55 jnthn Anyway, it'll probably be something along those lines
16:58 zengargoyle i'm thinking it's not namespacing things, the code uses modules directly from . named like 'datas', 'check', 'add', and might have some circular thing going on.  i'm going to try and re-write it in a bit more standard fashion.
16:58 cdg_ joined #perl6
16:59 andreoss` m: my @x := 1, 1, | @x Z+ @x[1..*]
16:59 camelia rakudo-moar f083cf: OUTPUT: «5===SORRY!5=== Error while compiling <tmp>?Cannot use variable @x in declaration to initialize itself?at <tmp>:1?------> 3my @x := 1, 1, | @7?5x Z+ @x[1..*]?    expecting any of:?        prefix?»
16:59 andreoss` i recall self-referencing in declaration was possible before, wasn't it?
17:01 jnthn Don't think so. Even if the compiler didn't check it and complain, the @x on the right would be some empty Array unrelated to the result that gets bound into @x
17:01 jnthn Since the RHS is evaluated before the bind
17:01 jnthn home time; bbl &
17:05 mspo what is this stuff? 9AD391BB028FBF86169E02934A6181BA21049736.1497985253.01802/07/071E327F6BCF9C81A7F69FDD6CD4B79413A1AAD5
17:07 linuksz joined #perl6
17:09 linuksz i get the following error during the test when i try to install Email::Simple
17:09 linuksz http://sprunge.us/QdLe
17:09 linuksz did i fucked up something, or the module is wrong?
17:09 AlexDaniel andreoss`: it is, but not with binding
17:10 linuksz i use git version of rakudo and git version of module
17:10 geekosaur ummmmm
17:10 AlexDaniel andreoss: although… I'm no longer sure. Trying to get it working
17:11 geekosaur samcv ^ did something break with newline handling in your recent changes? ( http://sprunge.us/QdLe )
17:12 geekosaur (I *think* the missing backslashes are the test harness's fault; it's hard to get that right)
17:13 geekosaur (note that the network-format newlines are "backwards", \n\r instead of \r\n)
17:15 AlexDaniel well yeah, in this particular case it's much easier to write it as a sequence
17:15 geekosaur linuksz, this wouldn't be the first time something has gone sideways with trying to deal with variant forms of newline
17:16 linuksz so can i do something to get it working?
17:16 AlexDaniel m: my @x = 1, 1, * + * … ?; say @x[^10]
17:16 camelia rakudo-moar f083cf: OUTPUT: «(1 1 2 3 5 8 13 21 34 55)?»
17:16 geekosaur and I am reading those expected/got as "Subject: test\n\rTo: foo\n\r\n\rfoo\n\r" which would be a valid SMTP message aside from the reversed newline sequences
17:17 geekosaur linuksz, probably try an older not-from-git rakudo
17:17 geekosaur m: say "\n\r" ~~ "\n\r"
17:17 camelia rakudo-moar f083cf: OUTPUT: «True?»
17:17 geekosaur huh...
17:18 itaipu joined #perl6
17:18 geekosaur m: say "\n\r" eq "\n\r"
17:18 camelia rakudo-moar f083cf: OUTPUT: «True?»
17:18 linuksz it is annoying, because i use my distro package manager for installing packages to my system, and the program i use for packaging (alacryd) only runs on git version
17:18 geekosaur maybe it's not rakudo
17:19 geekosaur buggable: eco Email::Simple
17:19 buggable geekosaur, Email::Simple 'Simple email parsing module': https://github.com/retupmoca/p6-Email-Simple
17:20 mspo okay I'm reading https://github.com/ugexe/zef/issues/117
17:21 geekosaur hrm, suspicious, last commit is about newline issues
17:22 mspo "simple email parsing" lol
17:22 geekosaur m: use Test; my $nr = "\x0a\x0d"; my $nasty = "Subject: test{$nr}To: foo{$nr}{$nr}foo{$nr}"; is ~$nasty, $nasty, "fii"
17:22 camelia rakudo-moar f083cf: OUTPUT: «ok 1 - fii?»
17:24 * geekosaur trying to install locally but suspect this is a bug in Email::Simple or possibly an incompatibility with recent rakudo
17:24 geekosaur ...the latter apparently as it passes here
17:25 linuksz if i now compile nqp with an older version of moarvm (git compiled today morning) installed on my system, then compile rakudo with the today morning version of nqp and moarvm, will it work with the now compiled nqp and moarvm?
17:25 geekosaur suggest you file a bug at https://github.com/retupmoca/p6-Email-Simple/issues
17:26 geekosaur I have no idea when it changed, sorry
17:26 linuksz isn't there any way to make it working? take a look at the developer's github activity statistics.
17:26 geekosaur This is Rakudo version 2017.06-133-ga7b2304 built on MoarVM version 2017.06-29-ga51ba62
17:26 geekosaur is what I tested on
17:27 geekosaur you could try forcing the install as it's onl;y likely to be a problem with weird newlines
17:27 espadrine joined #perl6
17:27 geekosaur stick to the RFC (\r\n) and you'll be fine
17:28 linuksz geekosaur, did it work for you?
17:28 geekosaur by forcing the install I mean zef install --/test Email::Simple
17:28 cptAngry joined #perl6
17:29 linuksz so without tests?
17:29 geekosaur and I guess I have to figure out how to rephrase what I already said because it didn't register
17:30 geekosaur it will work fine if you don't feed it noncompliant messages with weird newlines
17:31 geekosaur it may work fine if you *do* feed it such messages, it's not clear that there is a real problem aside from a strange comparison somewhere since the expected and got strings are identical
17:32 geekosaur which is why I pinged the person working on the unicode stuff, since newlines have special unicode handling and it's been a problem before
17:32 geekosaur and for now, since the only test that failed is the one with the non-RFC-complaint newlines, yes, just skip the tests
17:33 geekosaur *compliant
17:35 mcmillhj joined #perl6
17:38 linuksz i get the "Malformed UTF-8 at line 1 col 1" message with several modules i trying to install
17:38 geekosaur erm, or the possibility I didn't think of which is that maybe the version I installed predates the change to try to support weird newlines...
17:39 geekosaur actually looks like it got thay version
17:40 geekosaur "at line 1 col 1" is a bit suspicious. what modules?
17:40 linuksz developing programs for humans is NOT possible in perl6, because installing moarvm, then nqp, then rakudo, then the module manager, then the module, and get them working requires at least a power user
17:40 geekosaur (I suspect Windows byte order marks)
17:40 linuksz Email::MIME
17:41 geekosaur linuksz, humans are supposed to use rakudo star, not git
17:41 bartolin joined #perl6
17:41 geekosaur buggable6, eco alacryd
17:41 geekosaur buggable: eco alacryd
17:41 buggable geekosaur, Nothing found
17:43 linuksz humans can't use zef. humans can use the graphical package manager, and click on the 'install' button of the program, and the package manager automatically downloads the packages, extract them, and installs them
17:44 AlexDaniel ZofBot: humans?
17:44 timotimo linuksz: if the end user is a windows user, you can build a little msi for them with just a little config file and a command
17:44 linuksz but it is not possible to make packages that are can give a working module with an extraction and installation
17:44 geekosaur linuksz, that train of thought goes nowhere. come back in a few years when there's been time to build up the kind of ecosystem you seem to think is mandatory
17:45 geekosaur or go join the schmuck who insisted perl 6 can't be a real language unless its install is curl|sh
17:45 linuksz i want to develop applications for linux
17:45 geekosaur because your attitude and theirs are (a) both nonsense (b) antithetical, so you're set for a cage match
17:46 geekosaur so you're using arch? and you tak about humans requiring gui package managers?
17:47 geekosaur you're on arch, you're using what looks like an arch specific or at least heavily arch oriented tool, I don't think you get to complain about complexity
17:47 mcmillhj joined #perl6
17:47 linuksz yes, because i want to develop not for myself
17:47 geekosaur rright, that point went flying verhead
17:48 geekosaur why. are. you. using. a. power. user. distro. if. you. think. humans. can't. cope. with. power. user. tools./
17:48 linuksz so is it possible to build packages and write programs that can be installed with a click?
17:49 geekosaur pretty sure the ubuntu packaging is a lot friendlier than arch's, yes
17:50 geekosaur you are using Arch. you have already excluded what you think "humans" are,
17:51 geekosaur heck, you are targeting Linux, you have already alienated 90% of what you think "humans" are
17:51 mst linuksz: of course it is
17:51 linuksz Manjaro has an one-click package manager.
17:52 mst yes, and?
17:52 mst look, the fact you don't understand how to package software
17:52 mst does not mean that software cannot be packaged
17:53 mst the fact that you're installing everything the hard way does not mean users are required to do so :)
17:54 zengargoyle welp... i cleaned up dist layout of jeromelanteri's VimPlugger, wrote some use-ok tests and got it working... then just add 'use Git::Wrapper;' in VimPlugger::install module and boom STable conflict... :(
17:54 linuksz so can you define the way perl6 software can be packaged for one click install?
17:55 mst linuksz: you build a package for the distro. the user installs the package.
17:55 geekosaur use a platform not based on a rolling release, so it has tools that don't require you to live on the bleeding edge. manjaro is based on arch, it inherits arch's bleeding edge tendencies
17:56 mst geekosaur: I don't honestly see how that relates
17:56 geekosaur the packager they found for arch only works with git rakudo and git moarvm
17:56 geekosaur not the packaged versions
17:56 linuksz ok, and what is the tool with i can build packages?
17:56 mst 'the packager they found' <- eh?
17:56 geekosaur "alacryd"
17:57 mst linuksz: that's a manjaro question, nothing to do with us.
17:57 geekosaur which generate arch pkgbuilds from perl 6 metadata
17:57 mst oh, right, but that's a developer-only tool
17:57 mst nothing to do with packaging for users
17:57 mst so nothing to do with linuksz problem
17:59 mst linuksz: anyway, "developing programs for humans is NOT possible in perl6" is a lie. please learn how to build packages for your target distro if you want to distribute packages for users.
18:00 mst meanwhile, 'zef' is the correct thing to use for development
18:00 linuksz arch is far away before other distros in packaging perl6. for example, fedora doesn't have any perl6 packages.
18:01 linuksz mst, and what is the correct thing to use for packaging?
18:03 mst linuksz: dpkg-deb --build
18:03 mst you need to figure out your own answer for your own distro
18:03 mst you don't have a perl6 problem
18:03 mst you have a "not understanding manjaro" problem
18:03 linuksz i am not familiar with debian.
18:04 mst and since manjaro uses a fork of an obsolete version of arch's package manager
18:04 linuksz but as i see, it isn't an efficient tool, because https://packages.debian.org/search?keywords=perl6
18:04 mst I have no idea
18:04 mst I have no idea what you mean by "efficient tool"
18:05 mst debian has 2017.06 just the same as arch
18:06 linuksz a tool with packagers can package several packages (there is 855) perl6 module currently
18:06 mcmillhj joined #perl6
18:06 linuksz and how many modules in binary form?
18:06 mst I don't actually care, I install to a local CUR and ship that
18:09 linuksz i will take a try with zef.
18:09 mst for shipping apps to end-users I tend to make fat packages that have the dependencies in the package so that you don't run into problems when distros upgrade stuff
18:09 mst this works fine with most languages
18:10 mst the problem here isn't "perl6 is missing a feature", the problem is "linuksz doesn't know how to use manjaro's build system"
18:10 raschipi Yes, then sysadmins have to upgrade all packages by hand when there's a problem, yay!
18:10 mst and I'm afraid I can't help you with that
18:10 mst raschipi: eh?
18:10 cdg joined #perl6
18:10 raschipi If you bundle, it won't be upgraded when the base package is upgraded.
18:11 raschipi It's a serious problem, instead of doing it once, doing it for every bundle. But I'm just ranting, sorry.
18:12 linuksz i know how to use arch's build system.
18:12 raschipi I wish distros somehow forbade bundling, so that devs didn't have that option and were forced to do it properly.
18:12 mst linuksz: then you should have no problems packaging perl6 applications
18:13 mcmillhj joined #perl6
18:13 tokomer_ joined #perl6
18:13 mst raschipi: for a package for a centralised distro repository you should work with the main packages, sure
18:13 mst but if I'm supplying a 3rd party package I'd rather have to rebuild myself than risk a distro changing a version of something and breaking deployed code
18:13 perlpilot raschipi: It's a problem that my software continues to work as expected because it bundles the specific version of a module and doesn't break because someone decides to "upgrade" a dependency?
18:14 linuksz i have no problems packaging perl 5 applications, but it's not perl6
18:14 linuksz why does it work for perl5?
18:14 mst I don't understand
18:14 raschipi perlpilot: The appropriate way to do it is to use stable distros.
18:14 mst there are plenty of alacryd-built packages in AUR
18:14 mst https://aur.archlinux.org/packages/?O=0&amp;SeB=nd&amp;K=perl6&amp;outdated=&amp;SB=n&amp;SO=a&amp;PP=50&amp;do_Search=Go
18:15 perlpilot raschipi: sure ... in a perfect world, that's true.   We don't live in a perfect world  :)
18:15 mst I would suggest that if you want to use arch, you actually use arch
18:15 linuksz all packages are alacryd-built
18:15 mst so what in that is "doesn't work" ?
18:16 linuksz perlpilot, what happens if two applications wants to use the same module?
18:18 mst with bundling, you get two copies
18:18 mst without, you need centralised QA of revdeps before upgrade
18:18 perlpilot linuksz: um ... then they do so?  I'm not saying this is the one true way to do things.  I'm saying bundling specific versions is perfectly valid way to solve some problems.
18:18 mst right
18:18 mst I really don't understand what the problem's supposed to be here
18:18 perlpilot me either
18:19 linuksz but it uses more space, and file conflicts are possible
18:19 mst file conflicts are not possible
18:19 mst you keep the deps by the app
18:23 pmurias joined #perl6
18:23 pmurias geekosaur: why the hate for curl|sh? :)
18:23 mst geekosaur: note, I suspect the arch version of alacryd and the arch version of rakudo work fine together
18:23 mst geekosaur: and actually what we have here is a "manjaro is a shitshow" problem
18:24 raschipi The problem is having to upgrade something more than once using non-standard tools.
18:24 mst raschipi: eh?
18:24 raschipi Well, only someone supporting a multitude of applications would be annoyed by that problem, of course.
18:24 raschipi Anything that upgrades using something other than apt is a PITA.
18:25 mst ok, so, you appear to still be ranting at me, and I was talking about building .deb packages
18:25 mst so I'm rather confused now
18:25 mspo would it be reasonable for install-dist.pl to have an option to install the source files or create symlinks with meaningful names or something? :)
18:25 mspo everyone is packaging!
18:26 perlpilot mspo: why do you want meaningful names?
18:26 mspo perlpilot: I don't know
18:26 raschipi I'm just thinking aloud, please don't be bothered.
18:26 perlpilot mspo: contrapuntally, I think that the lack of meaningful names discourages people from trying to edit them, which is a very good thing  :-)
18:27 raschipi I can make my own packages if it comes to it, :).
18:27 mspo also is this install-perl6-dist a script?
18:27 mspo perlpilot: I think it's more like a question of "what do I have installed" and "how can I use it?"
18:27 mst raschipi: so, basically, I often have more than one app on the same server, and sometimes they have incompatible module requirements, and perl6's CUR system is about the only thing that can handle that itself
18:28 mst raschipi: so I prefer bundling into a 'fat deb' and then rebuilding that if there's a problem
18:28 perlpilot mspo: zef can answer the first one and presumably, p6doc can answer the second.
18:29 raschipi I understand, I just wish people would do it properly and don't require specific versions. Every developer thinks they are justified in doing it, yet it multiplies my  workload thousandfold.
18:30 cdg joined #perl6
18:30 kristoft joined #perl6
18:30 kristoft Hi everyone
18:31 mst raschipi: how does this increase your workload? done right, it just means that after a nasty bug is fixed, apt-get upgrade upgrades both the system copy of the module and the copy in the fat deb
18:32 geekosaur this sounds like the "eaiser for all to not bound versions" <sandbagged by abi change> "whee glad that'll never happen again easier for all to not bound versions"
18:34 raschipi geekosaur: Everything is stable.
18:35 raschipi mst: Would the package have new features added?
18:36 pmurias mst: re figure out your own answer for your own distro, wouldn't having a "package-this-up debian App::SomeRandomApp" script that does stuff sanely be something that the widely understood Perl 6 community should aim for?
18:37 geekosaur until yet another linux distribution comes up with yet another linux package format
18:37 nadim hi, I have sub my_sub(*@i, :$n, :$n2) ; my $i1 = 1, my $i2 = 2; my $n1 = :n1(100) ; my $n2 = :n2(200) ; my  capture $c =\($i1, $i2, $n1, $n2) ; calling  my_sub $c puts a Capture object in @i when I want the capture to be "expanded"
18:37 mst pmurias: there's already assorted tooling for this, but IME *debugging* such tooling requires learning the underlying stuff anyway
18:38 perlpilot nadim: my_sub |$c;  # If I read you right
18:38 geekosaur nadim, |$c
18:38 geekosaur iirc
18:38 mst raschipi: depends - but "this may force me to upgrade to a version with additional features" is something that can happen in unbundled packages too - often *more* often rather thanless
18:38 nadim wouldn't that just flatten the .list part of the capture?
18:40 zengargoyle .tell jnth ok so Git::Wrapper does some Git::Wrapper.HOW.add_method(Git::Wrapper, $method, anon method (*@p, *%n) { MOP magic that fails when it's inside a 'unit module Foo;'
18:40 yoleaux zengargoyle: I'll pass your message to jnth.
18:45 * zengargoyle wonders if it would work inside a class vs a module...
18:46 zengargoyle or if something MOPish changed that makes it not work....
18:50 zengargoyle ah, i guess Git::Wrapper should maybe do 'no precompilation' or does that have to go in the thing that's using Git::Wrapper (which works).
18:51 zakharyas joined #perl6
18:54 R0b0t1 joined #perl6
18:55 zengargoyle woot!
18:56 jnthn zengargoyle: Does it do that .add_method in a BEGIN block?
18:58 zengargoyle jnthn: no, in a for loop at the end of the class.   class { methods.... for <blah> .add_method }
18:58 jnthn Ah
18:58 jnthn That'd be a bad idea :)
18:58 jnthn 'cus it changes the object at runtime
18:58 zengargoyle https://github.com/nicqrocks/p6-Git-Wrapper/blob/master/lib/Git/Wrapper.pm6
18:59 jnthn And those changes will get serialized into the module that uses this one
18:59 jnthn Then if two modules both use this one, then they'll end up with independent copies of those
18:59 jnthn And boom, two pre-comps you can't use together
18:59 zengargoyle yeah, it sorta makes sense now...  not sure of best fix though..
18:59 jnthn Wrap the foor loop in a BEGIN block, preferably before the closing } of the class :)
19:00 jnthn Oh, sounds like the for already is in the class body
19:00 itaipu joined #perl6
19:00 jnthn In fact BEGIN takes a statement
19:00 zengargoyle yes.  in the body.
19:00 jnthn So you can just s/for/BEGIN for/ and see how that helps :)
19:00 zengargoyle putting no precompilation worked, i'll try the begin thing.
19:03 zengargoyle jnthn: yep, BEGIN for works (or at least doesn't fail).  jnthn++
19:05 zengargoyle sadly the nicqrocks fork that's in the eco doesn't have an issues option... :(
19:08 mspo is there a reason star isn't using install-dist.pl instead of zef?
19:10 mst star bundles zef already
19:10 mst the point of install-dist was as a low level option when you didn't have zef or an equivalent yet
19:10 mspo I'm following along with https://build.opensuse.org/package/view_file/devel:languages:perl6/perl6-Inline-Perl5/perl6-Inline-Perl5.spec?expand=1
19:11 mspo star uses perl6 zef  --/build-depends --/test-depends --/depends --/p6c
19:11 mspo --/metacpan --/cpan --force install ./modules$path_sep$module;
19:11 mspo right?
19:14 zengargoyle .tell jeromelanteri the Git::Wrapper thing has been solved!  i made pull-request on Git::Wrapper with the fix.
19:14 yoleaux zengargoyle: I'll pass your message to jeromelanteri.
19:15 mspo how predictable are precomp names?  I was just trying to install zef using someone's package and got different precomp names
19:16 mspo his : precomp/9AD391BB028FBF86169E02934A6181BA21049736.1497985253.01802/FE/FE48D13883ACBA2CF35E1746EEFE956A173BE8AE.repo-id
19:16 mspo mine: precomp/6C3B1056A15F23B58ED8A7017D2C24B8F7524EB2.1501209729.51376/FE/FE48D13883ACBA2CF35E1746EEFE956A173BE8AE.repo-id
19:16 mspo it's just the first part!
19:18 mspo that middle part is a timestamp so that's not going to be good :)
19:23 mspo nine: any hints?
19:31 wamba joined #perl6
19:33 nhywyll joined #perl6
19:36 mr-foobar joined #perl6
19:37 mspo or maybe I should try -toolchain?
19:45 revere Is there a camel book for 6 yet? Or some WIP?
19:45 itaipu joined #perl6
19:49 hobbs joined #perl6
19:49 hobbs joined #perl6
19:50 timotimo not a camel book, but other books
19:50 timotimo https://perl6book.com/
20:00 tzekid joined #perl6
20:00 TEttinger joined #perl6
20:03 revere That chart looks like what I'm looking for, thanks timotimo
20:04 timotimo cool
20:08 coetry_ joined #perl6
20:09 Kyo91 joined #perl6
20:17 nine mspo: yes, -toolchain is the best place as I follow that in full. Sort of answered the question there :)
20:22 |oLa| joined #perl6
20:32 timotimo why did nobody force me to look at the Terminal::Print examples
20:32 timotimo there's fantastic stuff in there
20:32 timotimo ab5tract++
20:34 Cabanossi joined #perl6
20:44 |oLa| joined #perl6
20:45 fantazo joined #perl6
20:46 fantazo left #perl6
20:54 nadim I am going to force you to look at Data::Dump::Tree then!
20:55 timotimo that's a different kind fo fantastic stuff
20:56 nadim next version as :flat control, here is an example of a dump done by dd, ddt, and ddt in normal vertical mode https://nopaste.linux-dev.org/?1160097
20:57 * nadim will look at Termina::Print examples again
20:59 grondilu joined #perl6
21:00 timotimo that looks neat how it puts things into columns
21:01 nadim took most of this week to implement
21:03 nadim it's neater than it looks like, you can say things like ddt $my_stuf, :flat( Hash, sub($s){ $s ~~ Array && $s.elems > 30), $something_want_flat)
21:03 timotimo ah so a decider function so to say?
21:03 nadim and it will flatten all hashes, all your arrays that have 30 elements or more, and the stuff you want flattened
21:04 nadim yes
21:04 nadim put you can say :flat(0) and first level get flattened
21:04 timotimo what's the ... for?
21:04 timotimo oh because it's a seq
21:05 coetry_ joined #perl6
21:06 jameslenz joined #perl6
21:06 nadim you can do multi colums also like this https://nopaste.linux-dev.org/?1160098
21:06 nadim Seq of 11 elements in that case, you can control the amount displayed
21:07 zengargoyle .ask jeromelanteri was it you who wanted TagBar config for perl6?  https://gist.github.com/77f253d6d721e4b6287b8e5321b31e9f
21:07 yoleaux zengargoyle: I'll pass your message to jeromelanteri.
21:07 nadim timotimo: its also does references, something I miss in dd
21:07 brrt joined #perl6
21:07 zengargoyle or if anybody else has a better tagbar config... :)
21:09 nadim timotimo: maybe I shoud have "... +n" to show how many more elements, although how many elements are in the header
21:10 coetry_ Interesting discussion on HN regarding concurrency primitives in Perl6: https://news.ycombinator.com/item?id=14854418&amp;utm_term=comment
21:10 timotimo both would be fine
21:11 nadim the number of element is already in the header. .Seq(32) or .Seq(*)
21:13 timotimo yes
21:13 timotimo i mean either keep it only in the header or put it on the bottom as well
21:13 timotimo both choices are fine with me
21:13 timotimo coetry_: seen it, thanks :)
21:14 nadim I like Hashes even more https://nopaste.linux-dev.org/?1160099
21:15 timotimo yeah, not bad
21:15 coetry_ timotimo++
21:15 nadim but it must be seen in color, black and white takes half the legibility away
21:15 coetry_ ++timotimo
21:15 timotimo coetry_: i believe one or two of our channel inhabitants commented, too
21:16 coetry_ oh sorry, havn't seen the logs
21:16 timotimo i mean commented on hackernews itself
21:16 coetry_ ahh i see
21:16 timotimo at least raiph has 11 comments
21:16 coetry_ looking up right now
21:17 timotimo i also see b2gills in there
21:17 nadim http://imgur.com/Hfhiouw  and then I stop spamming
21:18 coetry_ what does TIMTOWTDIBSCINABTE stand for?
21:19 nadim Honestly, one has to be a smi God to read dd's output if it is more thana a few lines.  Although dd is invaluable.
21:19 geekosaur There Is More Than One Way To Do It
21:19 timotimo more than one way to do it, but sometimes consistency is not a bad thing either
21:19 coetry_ timotimo, yeah i was looking for that last part. cool.
21:24 itaipu joined #perl6
21:44 mcmillhj joined #perl6
21:50 coetry joined #perl6
21:52 zengargoyle joined #perl6
22:00 mcmillhj joined #perl6
22:08 nadim timotimo: I mixed Terminal::Print and something else :), yes Terminal::Print is cool, maybe Ii should use that instead for Curses.
22:13 cog__ joined #perl6
22:13 timotimo perhaps
22:14 timotimo i think i'm going to use it for p6profiler-tui
22:35 pharv_ joined #perl6
22:37 nadim timotimo: looking forward to that! I need a profiler and a debugger. Btw, is the debugger working now? it was broken some times ago.
22:38 timotimo it works right now
22:38 timotimo not necessarily if you have multiple threads active
22:40 coetry___ joined #perl6
22:43 sufrostico joined #perl6
22:49 Cabanossi joined #perl6
23:09 pilne joined #perl6
23:18 nadim I will check it up
23:22 nadim when an object is but'ed with a role that has an attribute, where is the attribute found?
23:26 nadim found them!
23:34 Cabanossi joined #perl6
23:57 kyan joined #perl6

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

Perl 6 | Reference Documentation | Rakudo