»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.
lambd0x timotimo: hm, so what would be a better option in your opinion? Parrot or JVM? 00:00
timotimo you can't use parrot for a rakudo any more at all
we dropped support over a year ago if i'm not mistaken
lambd0x Oh I see, it got limited to project at some point then? 00:01
00:02 stevieb9 left, yqt left
timotimo one reason we ported rakudo to the JVM was to rely on its mature multi-threading support so we could design proper multi-threading stuff for rakudo 00:02
after that we built the port to MoarVM, which also has multi-threading features based on what we learned
parrot also has multi-threading features, but nobody put the work in to make rakudo make use of it
00:03 tardisx left, stevieb9 joined 00:04 TEttinger joined
lambd0x Understood. Quite peculiar the way things end up :). That's nice really. Oh, so right now the optimizations you said that are being development, is to improve Raduko with MoarVM to the point where it gets stabvle 00:05
*stable?
timotimo these optimizations are for performance only. less memory usage, less CPU time usage, less contention on locks and resources when multi-threading and doing async I/O 00:06
stability improvements happen all the time, but moarvm has been considered stable for a long time now
there are still bugs, of course. you can still make it crash in some ways, of course 00:07
00:08 stevieb9 left
lambd0x :) Nice so even learning/clashing the language I might end up helping you guys somehow 00:09
00:11 cdg left, tardisx joined, jjido left, jjido_ joined 00:12 cdg joined
lambd0x So I think I'll get the compiler with MoarVM so that I can give feedback. Although I don't know it portage's version is current with the language dev. 00:12
timotimo it's easy to figure out, since we name our versions after year and month 00:16
so if you have 2015.03, you're far out of date
00:16 cdg left
timotimo if you get 2016.06, you're faster than everybody else! :) 00:16
just run perl6 --version and you'll see what your version is
00:20 jjido_ left
lambd0x almost timotimo, it is currently 2016/05 00:21
timotimo perfect
This is Rakudo version 2016.05-145-gac0dcdd built on MoarVM version 2016.05-34-gfbe9e24 <- you can't get newer than this at the moment 00:22
00:24 vendethiel left
lambd0x I see. here is "version 2016.05 built on MoarVM version 2016.05" 00:24
timotimo then you have a released version, and i have an in-development version 00:25
i'll go to bed soon, but i wish you the best of luck with perl6 :)
perl6.org/downloads/ - the "introductory material" on the right of this page is good 00:26
lambd0x timotimo: Man you helped me a lot really. Thanks for you time have a good night. o/
timotimo \o
lambd0x thanks
00:27 awwaiid left, awwaiid joined 00:32 MilkmanDan[Han] is now known as MilkmanDan, dfcarpenterak left 00:38 vendethiel joined 00:40 lambd0x left 00:43 tardisx left 00:47 tardisx joined 00:50 cpage_ left, cpage__ joined 00:51 b2gills joined 00:56 awwaiid left 00:57 awwaiid joined 01:01 vendethiel left 01:02 vendethiel joined 01:10 rindolf left
Xliff_ How can I trace execution in perl6? I am trying to debug and endless loop and am having no luck with the traditional debugging methods. 01:12
jdv79 throw an exception, catch it to print it? 01:13
i don't know if we have something like Carp::longmess from p5
01:14 molaf left
jdv79 maybe i'm wrong - rosettacode.org/wiki/Stack_traces#Perl_6 01:14
llfourn you can just say Backtrace.new.gist I think 01:15
m: say Backtrace.new.gist
camelia rakudo-moar ac0dcd: OUTPUT«Backtrace.new␤»
llfourn m: say Backtrace.new.Str 01:16
camelia rakudo-moar ac0dcd: OUTPUT« in block <unit> at <tmp> line 1␤␤»
llfourn Backtrace.new.Str*
jdv79 rosetta suggests .concise
whatever that is
llfourn m: say Backtrace.new.consise
camelia rakudo-moar ac0dcd: OUTPUT«Method 'consise' not found for invocant of class 'Backtrace'␤ in block <unit> at <tmp> line 1␤␤»
llfourn m: say Backtrace.new.concise
camelia rakudo-moar ac0dcd: OUTPUT«␤»
llfourn doesn't seem too useful :P 01:17
jdv79 i guess rosetta needs updating
llfourn maybe it's better if you have some frames
m: sub foo { say Backtrace.new.concise }; foo()
camelia rakudo-moar ac0dcd: OUTPUT« in sub foo at <tmp> line 1␤␤»
jdv79 yeah 01:18
(self.grep({ !.is-hidden && .is-routine && !.is-setting }) // "\n").join
llfourn m: sub foo { say Backtrace.new.concise }; sub bar { foo }; bar()
camelia rakudo-moar ac0dcd: OUTPUT« in sub foo at <tmp> line 1␤ in sub bar at <tmp> line 1␤␤»
jdv79 that's the method
Xliff_ Actually, isn't there a command line argument or an environment variable that gives you detailed output on what is happening under the hood?
llfourn --ll-exception
Xliff_ Ahh!
jdv79 uh. there's --ll-exception for more verbose failures
llfourn they're more low level rahter than verbose in my experience 01:19
jdv79 idk about something like that for non-exceptional times except if you want to output a profile and dig
llfourn ie "ll"
jdv79 bug likely not fun
Xliff_ Hrm. --ll-exception didn't seem to do anything. 01:20
jdv79 well, verbose in one direction:)
are you throwing exceptions?
Xliff_ Nope
01:20 Zoffix joined
llfourn
.oO( what do you think the ll means in llfourn )
 01:20
jdv79 then... that won't be helpful
Xliff_ Endless loop. I'm trying to figure out why.
jdv79 could you show the code?
Xliff_ Not really. It' 01:21
jdv79 the Backtrace class seems like your friend here
Xliff_ It's test code and it's got several modules.
I'm suspecting it's Nativecall code, but I have no idea where.
jdv79 why is it called backtrace? isn't stacktrace more conventional.
llfourn Xliff_: are you sure it's a loop? There seems to be a bug I'm hitting where when rakudo throws an exception it hangs forever and never shows it.
Xliff_ So to show the code, I'd have to give you everything
llfourn I haven't golfed it yet
but it's something to do with typechecking parameters I think
Xliff_ llfourn, actually....that's a possibility.
llfourn it's quite annoying :( 01:22
Xliff_ I'm getting this a lot -- "Internal error: zeroed target thread ID in work pass" 01:23
01:23 Zoffix left
jdv79 hmm. interesting. wonder what %*ENV<RAKUDO_VERBOSE_STACKFRAME> does.... 01:23
llfourn :S that sounds bad
jdv79 Bastrace.Str seems to look at it
*Backtrace
llfourn it's probably -ll-exeception?
jdv79 ah
Xliff_ *sigh* 01:24
I finally isolated the call. It's in new code I just finished today. 01:25
Wheee
Pointer manipulation.
jdv79, THAT'S it!
%*ENV<RAKUDO_VERBOSE_STACKFRAME> <-- setting this will have rakudo send all of its operations to stdout.
jdv79 are you sure that helps? 01:26
llfourn oh cool?
01:26 molaf joined
Xliff_ Hrm. Dammit. I thought that was it. 01:27
Damn. No, that's not it. It's something close to that, though. 01:28
I will have to scour rakudo code to find that option now. It's bothering meee. 01:29
01:29 girafe3 left
jdv79 are you sure it exists? 01:30
Xliff_ I remember using it once. 01:31
It printed everything the compiler and VM was doing.
However "remember using it" !=:= "how to invoke it" 01:32
Which kills me.
jdv79 maybe its a moarvm thing
Xliff_ Probably.
At any rate, I finally have a place to start. It's probably in my converted XS code.
Which I am in no mood to debug, right now.
Thanks for the help! 01:33
01:33 ssotka left 01:43 awwaiid left, awwaiid joined
timotimo Xliff_: you can run perl6-gdb-m, hit ctrl-c at a random point, go "up" until you have a frame where the "tc" variable is available (almost always the first parameter to a function) and then "print MVM_dump_backtrace(tc)" 01:44
and then "c" to continue running and ctrl-c again
01:46 ilbot3 left, mohae joined
jdv79 where are the "commandline_options" actually used? 01:46
01:47 ilbot3 joined
timotimo also, there's "use trace", which isn't working 100% great 01:48
01:49 vendethiel left, mohae_ left, Sgeo joined
jdv79 is it 100% good? 01:49
which translates to much great then? 01:50
s/much/how much/
timotimo it's sometimes very strange 01:51
01:57 vendethiel joined 01:59 BenGoldberg joined 02:01 BenGoldberg left 02:16 kurahaupo_ joined 02:18 kurahaupo left 02:20 kid51 left 02:23 kurahaupo_ left 02:25 inokenty left 02:29 noganex_ joined, colomon left, awwaiid left, awwaiid joined 02:31 noganex left 02:39 wamba joined, vendethiel left 02:40 inokenty joined 02:47 grassass left
tony-o it's watermelons good 02:49
mature fruit, also known as a vine-like (scrambler and trailer) flowering the crop. The fruit, 02:50
02:52 vendethiel joined 02:53 kaare_ joined 03:12 mohae_ joined
timeless .tell AlexDaniel sounds like the right problem and the right solution 03:14
yoleaux timeless: I'll pass your message to AlexDaniel.
03:14 vendethiel left, mohae left, grassass joined 03:16 ssotka joined
dalek c/eval-desc-grammar: 1b3676a | (Josh Soref)++ | doc/Type/Cool.pod:
fix EVAL desc grammar
 03:17
c: 1b3676a | (Josh Soref)++ | doc/Type/Cool.pod:
fix EVAL desc grammar
 03:18
c: 26365b9 | (Josh Soref)++ | type-graph.txt:
Add order to typegraph
 03:19
03:23 aries_liuxueyang left 03:25 aries_liuxueyang joined 03:28 telex left, telex joined 03:31 travis-ci joined
travis-ci Doc build passed. Josh Soref 'fix EVAL desc grammar' 03:31
travis-ci.org/perl6/doc/builds/137692297 github.com/perl6/doc/commit/1b3676af1dee
03:31 travis-ci left 03:32 mosh left 03:34 beatdown joined 03:35 travis-ci joined
travis-ci Doc build passed. Josh Soref 'Add order to typegraph' 03:35
travis-ci.org/perl6/doc/builds/137692478 github.com/perl6/doc/compare/typegraph-order
03:35 travis-ci left
sammers_ good afternoon from Japan 03:39
03:51 cpage__ left, MadcapJake left, jrusso joined 03:52 cpage_ joined, cpage_ left, jrusso is now known as MadcapJake 04:00 colomon joined
grondilu kon icha wa 04:00
grondilu has no idea if that's how one writes it but whatev 04:01
"ichi" instead of "icha" for sure though
04:05 skids left 04:09 jevin_ joined 04:11 Cabanossi left, jevin left 04:13 Cabanossi joined 04:15 tardisx left 04:18 khw left 04:22 awwaiid left, Zero_Dogg left, Zero_Dogg joined, awwaiid joined, baest_ is now known as baest, baest left 04:23 baest joined 04:26 snarkyboojum joined, snarkyboojum left 04:32 awwaiid left 04:33 awwaiid joined 04:34 lizmat left 04:35 tardisx joined 04:37 vendethiel joined 04:45 sortiz left, cpage_ joined 04:48 tardisx left
timeless こんにちは 04:50
今日は
04:54 Possum left 04:58 vendethiel left 05:02 awwaiid left 05:03 tardisx joined, awwaiid joined 05:05 sno left 05:08 wamba left 05:20 zzzzzzzzz left, zzzzzzzzz joined 05:28 rurban joined 05:37 rurban1 joined 05:39 rurban2 joined 05:40 rurban left 05:42 rurban1 left 05:50 vendethiel joined
timeless m: $a = '/a/b#c'; say "hi" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!; 06:03
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '$a' is not declared␤at <tmp>:1␤------> 3<BOL>7⏏5$a = '/a/b#c'; say "hi" if $a ~~ m!(\/<-␤»
timeless m: my $a = '/a/b#c'; say "hi" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!; 06:04
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at <tmp>:1␤------> 3$a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!;7⏏5<EOL>␤Regex not terminated.␤at <tmp>:1␤------> 3$a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!;7⏏5<EOL>␤Unable to parse expression in…»
timeless m: my $a = '/a/b#c'; say "hi" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!; 06:05
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at <tmp>:1␤------> 3$a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!;7⏏5<EOL>␤Regex not terminated.␤at <tmp>:1␤------> 3$a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[#(.*)|]!;7⏏5<EOL>␤Unable to parse expression in…»
timeless realizes that the blue is from the irc client channel linkifier
m: my $a = '/a/b#c'; say "hi" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(#(.*)|)!; 06:06
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at <tmp>:1␤------> 3$a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(#(.*)|)!;7⏏5<EOL>␤Regex not terminated.␤at <tmp>:1␤------> 3$a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(#(.*)|)!;7⏏5<EOL>␤Unable to parse expression in…»
timeless m: my $a = '/a/b#c'; say "hi" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)!;
camelia rakudo-moar ac0dcd: OUTPUT«hi␤»
06:06 sno joined
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)!; 06:06
camelia rakudo-moar ac0dcd: OUTPUT«Use of Nil in string context in block <unit> at <tmp> line 1␤hi </a/><b><>␤»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)()!;
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3<$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(7⏏5)!;␤ expecting any of:␤ infix stopper␤»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(.*)!;
camelia rakudo-moar ac0dcd: OUTPUT«hi </a/><b><#c>␤»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(#(.*))!; 06:07
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at <tmp>:1␤------> 3 $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(#(.*))!;7⏏5<EOL>␤Regex not terminated.␤at <tmp>:1␤------> 3 $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(#(.*))!;7⏏5<EOL>␤Unable to parse expression in…»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)([#](.*))!;
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Regex not terminated.␤at <tmp>:1␤------> 3a ~~ m!(\/<-[/]>*\/)(<-[#]>*)([#](.*))!;7⏏5<EOL>␤Regex not terminated.␤at <tmp>:1␤------> 3a ~~ m!(\/<-[/]>*\/)(<-[#]>*)([#](.*))!;7⏏5<EOL>␤Unable to parse expression in…»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(<#>(.*))!;
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter < (must be quoted to match literally)␤at <tmp>:1␤------> 3$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(<7⏏5#>(.*))!;␤Unrecognized regex metacharacter # (must be quoted to match literally)␤at <tm…»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)(<[#]>(.*))!; 06:08
camelia rakudo-moar ac0dcd: OUTPUT«hi </a/><b><#c>␤»
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*)]!;
camelia rakudo-moar ac0dcd: OUTPUT«hi </a/><b><c>␤»
timeless is that really how i'm supposed to write this?
06:14 awwaiid left, awwaiid joined, vendethiel left 06:15 domidumont joined
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || ]!; 06:16
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || 7⏏5]!;␤ expecting any of:␤ infix stopper␤»
timeless that's unhelpful
06:20 domidumont left, domidumont joined, vendethiel joined 06:22 huggable joined 06:26 firstdayonthejob joined 06:33 mr-foobar left 06:34 ssotka left 06:40 dwarring left 06:42 rurban_ joined 06:44 firstdayonthejob left
masak morning, #perl6 06:47
dalek c: 5fe50f7 | parabolize++ | doc/Type/Capture.pod:
Corrections in Capture doc (#598)

  * Change delimiters on formatted code
  * Parenthesize single numeric value
example code generates warnings of potential difficulties
  * Update Capture.pod		 06:52
06:53 melezhik joined 06:54 tardisx left
melezhik Hi all! I have a perl6 HASH ; my %h1 = 'apples' => 3, 'oranges' => 7; my $s = 'apples'; how can I get %h apples value ? seems like %a<$s> does not work? 06:55
gfldex m: my %h = 'apples' => 3, 'oranges' => 7; my $s = 'apples'; dd %h<<$s>>; 06:56
camelia rakudo-moar ac0dcd: OUTPUT«(3,)␤»
gfldex m: my %h = 'apples' => 3, 'oranges' => 7; my $s = 'apples'; dd %h{$s}; 06:59
camelia rakudo-moar ac0dcd: OUTPUT«Int %h = 3␤»
gfldex melezhik: see doc.perl6.org/language/subscripts 07:00
07:01 _mg_ joined 07:02 vendethiel left
moritz melezhik: {} is for subscripting with expressions, <> for subscripting with literal strings 07:05
gfldex m: my %h = 'apples' => 3, 'oranges' => 7; my $s = 'apples'; dd %h<apples oranges>; 07:06
camelia rakudo-moar ac0dcd: OUTPUT«(3, 7)␤»
gfldex lists of literal strings
07:06 _mg_ left 07:08 tardisx joined
melezhik gfldex, mortiz - sounds good!, thanks 07:09
07:12 RabidGravy joined, abraxxa joined
RabidGravy MARNIN! 07:13
grondilu moinz
07:14 krakan left, krakan joined 07:19 nebuchadnezzar left, nebuchadnezzar joined 07:22 xinming left 07:23 xinming joined
dalek c: 0feafb2 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Hash.pod:
improve Hash, provide link to subscripts
 07:33
gfldex melezhik++ # for asking the right question
07:43 vendethiel joined 07:45 zakharyas joined 07:51 _mg_ joined 07:57 tardisx left 07:58 ocbtec joined 08:03 CIAvash joined 08:04 vendethiel left 08:09 _mg_ left
masak gradual insight about language creation: it's not always clear when you're arguing a great feature for your language, or when you have an axe to grind. from inside your head, the latter often looks like the former. 08:09
08:11 _mg_ joined 08:15 dakkar joined 08:16 mohae_ left 08:19 zakharyas left
melezhik gfldex: :)) 08:21
08:26 brrt joined, [particle]1 left, melezhik left, [particle] joined 08:32 araujo joined 08:41 vendethiel joined 08:43 donaldh joined 08:48 Actualeyes left 08:51 g4 joined 08:52 rurban2 left 08:53 brrt left 08:59 jack_rabbit_ joined 09:00 jack_rabbit left 09:03 bjz joined 09:04 jack_rabbit_ is now known as jack_rabbit 09:14 Actualeyes joined 09:16 TEttinger left 09:19 bjz left, rindolf joined 09:24 vendethiel left
dalek c: e5d4cfb | (Wenzel P. P. Peppmeyer)++ | doc/Type/Hash.pod:
forgot to remove old example
 09:29
09:33 lizmat joined 09:35 zakharyas joined 09:37 azawawi joined
azawawi hi 09:38
RabidGravy: pong
RabidGravy: $pong.flip
09:39 _mg_ left 09:42 Possum joined, tadzik left, M-Illandan left, M-matthew left 09:47 bjz joined 09:51 melezhik joined
melezhik Hi guys! Does anybody know a proper perl6 YAML parser? I need to convert perl6 hashes into YAML ... Thanks ... 09:52
azawawi modules.perl6.org/#q=YAML ? 09:53
melezhik: modules.perl6.org is your friend. Type and it will autocomplete your query :) 09:54
melezhik unfortunatelly a candidates found quite raw ... not stable 09:55
09:58 lizmat left 09:59 lizmat joined, vendethiel joined
gregf_ melezhik: you can also use Inline::Perl5. but yeah, pure Perl6 would be better 09:59
10:02 andreoss joined
gregf_ m: gist.github.com/anonymous/018007d5...072e91ee33 10:02
camelia rakudo-moar ac0dcd: OUTPUT«5===SORRY!5===␤Unrecognized regex metacharacter ; (must be quoted to match literally)␤at <tmp>:1␤------> 3m = Module.new7⏏5; p m.methods; m.send(:define_method, :h␤Couldn't find terminator = (corresponding = was at line 1)␤at <tmp>:1␤…»
gregf_ bah :/
m: gist.github.com/anonymous/6101f317...4da2919477
camelia rakudo-moar ac0dcd: OUTPUT«===SORRY!===␤Could not find Inline::Perl5 at line 3 in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-1/share/perl6/site␤ /home/camelia/rakudo-m-inst-1/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-1/share/perl6␤ CompUnit:…»
gregf_ m: use Inline::Perl5;use YAML::XS:from<Perl5>;my $h = { "foo" => [1,2,3,4], "bar" => 100 };say Dump($h); 10:03
camelia rakudo-moar ac0dcd: OUTPUT«===SORRY!===␤Could not find Inline::Perl5 at line 1 in:␤ /home/camelia/.perl6␤ /home/camelia/rakudo-m-inst-1/share/perl6/site␤ /home/camelia/rakudo-m-inst-1/share/perl6/vendor␤ /home/camelia/rakudo-m-inst-1/share/perl6␤ CompUnit:…»
10:03 silug left
RabidGravy boo! 10:04
lizmat hiss! 10:05
gregf_ *runs*
RabidGravy melezhik, I think you will find that a large proportion of the ecosystem is fairly raw and unstable due to the relative immaturity of the whole ecosystem 10:07
melezhik, as a rule of thumb it's either use what you find and help improve it or write something better yourself 10:08
azawawi, WAKEY WAKEY
azawawi RabidGravy: im sleepy lol 10:09
10:09 eyck left, eyck joined
azawawi RabidGravy: new rig 'i7 6700' eats 'rakudobrew build moar' parse stage in 42 secs :) 10:09
RabidGravy coolio 10:10
azawawi RabidGravy: vs ~70 sec for amd fx-8150
RabidGravy: 35% on multi thread, 25% on single thread performance difference 10:11
RabidGravy: github.com/rakudo/rakudo/commit/ac...09116bd602 improved GTK::Simple performance considerably
RabidGravy: FROGGS++ 10:12
RabidGravy coo, I saw that one go in a wondered about that 10:14
FROGGS++ indeed
10:14 silug joined
melezhik RabidGravy : Ok , sure )) 10:15
RabidGravy may have to re-run "test all the things" to see if I see an improvement with a broad set of NC modules
the burden of being an "early adopter" 10:16
azawawi RabidGravy: .precomp was like 42 MB 10:18
gfldex the docs say that %-sigiled containers got the Hash as a default type. That could be understood as if the default could be changed. Is that possible (without roles)?
azawawi RabidGravy: now it is 4MB 10:19
jnthn m: my %h is SetHash; %h<a>++; say %h.perl
camelia rakudo-moar ac0dcd: OUTPUT«SetHash.new("a")␤»
gfldex i shall doc that 10:20
m: my %h is SetHash = oranges => 'round'; %h<a>++; say %h.perl 10:21
camelia rakudo-moar ac0dcd: OUTPUT«Cannot modify an immutable SetHash␤ in block <unit> at <tmp> line 1␤␤»
10:22 tadzik joined, huggable left
gfldex m: my %h is SetHash = { oranges => 'round' }; 10:23
camelia rakudo-moar ac0dcd: OUTPUT«Cannot modify an immutable SetHash␤ in block <unit> at <tmp> line 1␤␤»
jnthn SetHash/BagHash/MixHash don't implement STORE, afaik, so you can't assign them. I think we probably could implement that... 10:24
moritz aren't some of those intentially read-only? 10:25
10:26 abruanese joined
masak moritz: not the Hash types 10:26
psch m: my $h = SetHash.new(<a b>); $h<c>++; say $h.perl
camelia rakudo-moar ac0dcd: OUTPUT«SetHash.new("a","c","b")␤»
masak moritz: Set/Bag/Mix are all immutable
jnthn Right, those certainly shouldn't get a STORE
moritz ok, good
10:26 bjz left 10:28 azawawi left
RabidGravy while we're on the subject of "is"ing containers, I'm actually surprised that no-one has freaked it about "my @a is Hash" (and friends) 10:30
if nothing else it makes it more complicated to explain aggregate containers and their respective sigils 10:31
jnthn That probably should be a type error
lizmat well, that's part of the malleability of Perl 6, no ?
jnthn (Hash isn't Positional)
m: my @a is Hash 10:32
camelia ( no output )
jnthn Hm, yeah, that's busted
RabidGravy and also
jnthn m: my @a is Hash; say @a ~~ Positional
camelia rakudo-moar b3e9f5: OUTPUT«False␤»
RabidGravy m: my %a is Array
camelia ( no output )
masak nine: was it you who fixed the precomp trouble?
jnthn Yeah, it'll be the same place that wants fixing.
masak nine: I just want to say that it's much better now. thank you. <3
10:33 awwaiid left, awwaiid joined
lizmat glfdex: in a Set we only have keys basically, what should happen to the values ??? 10:33
10:34 rurban_ left
lizmat m: my %h is SetHash; %h<a> = 5; dd %h # jnthn: you *can* assign to SetHashes 10:34
camelia rakudo-moar b3e9f5: OUTPUT«SetHash.new("a")␤»
jnthn lizmat: No 10:35
lizmat: You can assign to *elements* of SetHashes
That's different.
lizmat ah, you mean:
jnthn (The assignment is actually to a Scalar container bound into a SetHash)
lizmat m: my %h is SetHash = <a b c d>
camelia rakudo-moar b3e9f5: OUTPUT«Cannot modify an immutable SetHash␤ in block <unit> at <tmp> line 1␤␤»
jnthn Right
Though if that one works it'll discard b and d presumably 10:36
psch m: my %h is SetHash .= new: <a b c d>
camelia rakudo-moar b3e9f5: OUTPUT«Cannot modify an immutable SetHash␤ in block <unit> at <tmp> line 1␤␤»
lizmat well, there you go: that's not what I expecred
$ted
jnthn Though would be good for TimToady to weigh in on that.
lizmat yeah
10:36 bjz joined
jnthn It'd be the consistent thing for assigning to something with a % sigil...as usual, we have to pick the consistency we want :) 10:37
masak .oO( we're good at picking consistencies, we're just not very consistent about it ) 10:38
10:39 araujo_ joined 10:41 araujo_ left
nine masak: FROGGS++ fixed the part that probably hurt you. I fixed issues on archaic file systems like FAT32 and HFS+ 10:41
10:41 araujo_ joined 10:42 _mg_ joined 10:43 araujo left 10:44 araujo_ left 10:46 AlexDaniel joined
dalek c: de1690d | (Zoffix Znet)++ | doc/Type/Capture.pod:
Fix typo mentioned in #598
 10:48
masak ok, in that case, FROGGS++
oh right, now I remember. FROGGS++ even announced his fix on the #6macros channel. 10:49
TIL that there are two numbers.Integral types in Python 3: int, and bool 10:52
10:54 _mg_ left 10:57 lizmat left 11:00 vibha joined 11:07 rurban joined 11:09 M-matthew joined, M-Illandan joined, labster left, kid51 joined
dalek c: 0ba409e | (Wenzel P. P. Peppmeyer)++ | doc/Language/variables.pod:
show how to change default type for %-sigiled containers
 11:12
11:15 rurban left 11:20 zakharyas left
dalek c: 0e7dd37 | (Wenzel P. P. Peppmeyer)++ | doc/Language/variables.pod:
add is for container type to index
 11:22
c: c251707 | (Wenzel P. P. Peppmeyer)++ | doc/Language/testing.pod:
better index entry for is (testing)
11:27 nowan_ joined 11:28 bjz_ joined, vendethiel left 11:30 bjz left, nowan left 11:34 vendethiel joined
AlexDaniel . 11:44
yoleaux 03:14Z <timeless> AlexDaniel: sounds like the right problem and the right solution
AlexDaniel timeless: let's just hope that there is a right way to implement it, liz has some doubts about this one :) 11:45
11:46 kid51 left 11:48 brrt joined 11:56 _mg_ joined, donaldh left 12:03 rurban joined, _mg_ left 12:05 Emeric joined 12:06 rurban_ joined, Emeric left, Emeric joined 12:09 _mg_ joined
literal if I want [<foo bar>, <baz quux>] without flattening, is there an easier way than adding .List to the end of each list? 12:13
psch literal: what's the context? are you return from some sub or..? 12:14
literal no, I just want to pass an array of lists (or arrays) as an argument to something 12:15
gfldex m: my @a; @a[0] = $('foo', 'bar'); @a[1] = $('baz', 'quux'); dd @a;
camelia rakudo-moar b3e9f5: OUTPUT«Array @a = [("foo", "bar"), ("baz", "quux")]␤»
psch m: sub f(+@args) { say @args.perl }; f [<foo bar>, <baz quux>]
camelia rakudo-moar b3e9f5: OUTPUT«[("foo", "bar"), ("baz", "quux")]␤»
12:16 vendethiel left
gfldex m: sub f(@args) { dd @args }; f [<foo bar>, <baz quux>] 12:16
camelia rakudo-moar b3e9f5: OUTPUT«[("foo", "bar"), ("baz", "quux")]␤»
psch m: my @a; @a[0] = ('foo', 'bar'); @a[1] = ('baz', 'quux'); dd @a;
camelia rakudo-moar b3e9f5: OUTPUT«Array @a = [("foo", "bar"), ("baz", "quux")]␤»
psch right, +@args is actually somewhat different... 12:17
m: sub f(+@args) { say @args.perl }; f <foo bar>, <baz quux>
camelia rakudo-moar b3e9f5: OUTPUT«[("foo", "bar"), ("baz", "quux")]␤»
literal ah, sorry, my problem is that the flattening occurs when any of the lists have 1 element
m: my @ar = [<foo>, <baz>]; say @ar
camelia rakudo-moar b3e9f5: OUTPUT«[foo baz]␤»
psch m: sub f(*@args) { say @args.perl }; f <foo bar>, <baz quux> # in contrast to this
camelia rakudo-moar b3e9f5: OUTPUT«["foo", "bar", "baz", "quux"]␤»
psch m: sub f(**@args) { say @args.perl }; f [<foo>, <bar>]
camelia rakudo-moar b3e9f5: OUTPUT«[["foo", "bar"],]␤»
psch m: sub f(**@args) { say @args.perl }; f [<foo>, <bar>], <baz> 12:18
camelia rakudo-moar b3e9f5: OUTPUT«[["foo", "bar"], "baz"]␤»
literal but maybe that's just <> which refuses to construct a List when presented with a single word
psch m: sub f(@args) { say @args.perl }; f [<foo>, <bar>]
camelia rakudo-moar b3e9f5: OUTPUT«["foo", "bar"]␤»
psch m: say <foo>.WHAT
camelia rakudo-moar b3e9f5: OUTPUT«(Str)␤»
psch yeah, seems like it :)
gfldex what does +@ in a signatuer do and how is it called?
psch m: sub f(@args) { say @args.perl }; f [('foo'), ('bar')]
camelia rakudo-moar b3e9f5: OUTPUT«["foo", "bar"]␤»
psch m: sub f(+@args) { say @args.perl }; f [('foo'), ('bar')] 12:19
camelia rakudo-moar b3e9f5: OUTPUT«["foo", "bar"]␤»
psch m: sub f(**@args) { say @args.perl }; f [('foo'), ('bar')]
camelia rakudo-moar b3e9f5: OUTPUT«[["foo", "bar"],]␤»
psch gfldex: i have no idea what it's called, but it's the user-facing way of declaring a sub that follows the single-arg rule wrt flattening
gfldex m: sub f(+@a){ dd @a }; f(1); f(1,2); f([1,2]); 12:21
camelia rakudo-moar b3e9f5: OUTPUT«[1]␤[1, 2]␤[1, 2]␤»
gfldex m: sub f(@a){ dd @a }; f(1); f(1,2); f([1,2]);
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5===␤Calling f(Int) will never work with declared signature (@a)␤at <tmp>:1␤------> 3sub f(@a){ dd @a }; 7⏏5f(1); f(1,2); f([1,2]);␤Calling f(Int, Int) will never work with declared signature (@a)␤at <tmp>:1␤------> 3sub f(…»
gfldex i shall doc
jnthn: do you got a good name for +@a in a sig? 12:22
12:24 vendethiel joined 12:25 raiph left 12:26 raiph joined
perlpilot gfldex: single-arg rule slurpy? (I don't think I've ever seen or heard a concise name for that) 12:28
12:28 _mg_ left
gfldex it does promote a single argument to a list, so "single argument promotion" is at least not wrong. 12:29
psch i like 'single-arg rule slurpy', although it depends on the single-arg rule being documented i suppose
m: my @a = <foo bar baz>; for (@a) { .say } 12:30
camelia rakudo-moar b3e9f5: OUTPUT«foo␤bar␤baz␤»
psch m: my @a = <foo bar baz>; for (@a,) { .say }
camelia rakudo-moar b3e9f5: OUTPUT«[foo bar baz]␤»
psch m: my @a = <foo bar baz>; for @a, { .say } # bit confused this doesn't work anymore..?
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5===␤Expression needs parens to avoid gobbling block␤at <tmp>:1␤------> 3my @a = <foo bar baz>; for @a, { .say }7⏏5 # bit confused this doesn't work anymor␤Missing block (apparently claimed by expression)␤at <tmp>:1␤------> …»
psch oh, that's probably somehow related to the STATEMENT_LIST fiddling lizmat++ did recently..? 12:31
12:31 rurban left
masak psch: I don't think I would expect that to work, now or earlier. 12:35
psch: what I see when I read this is an incomplete `for` loop over two elements: `@a` and a block
that's what a comma means :) 12:36
psch: do you have any evidence that `for @a, { .say }` used to work?
psch masak: yeah, i probably misremember the single-arg rule blog post
masak it's very different with and without the parens, is my point
psch yes, i know 12:37
and i am convinced already ;)
masak m: my @a; for @a, { .say } { say .^name }
camelia rakudo-moar b3e9f5: OUTPUT«Array␤Block␤»
masak there ya go :)
12:39 Zero_Dogg left, Zero_Dogg joined, Sgeo left
jnthn gfldex: "single argument rule slurpy" would maybe do it so long as that can be linked to a good description of the rule 12:41
Or maybe it's worth inlining that 12:42
And expressing it in terms of args
12:44 zakharyas joined 12:45 vendethiel left 12:49 rurban joined 12:54 Actualeyes left 12:58 vendethiel joined
dalek c: 9efd021 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Signature.pod:
doc +@ in signatures
 13:01
gfldex i think ...
[Coke] my brain wants "the same way then with a list" to be "the same way as with a list" 13:02
dalek c: 3d708d6 | (Wenzel P. P. Peppmeyer)++ | doc/Type/Signature.pod:
change listig wording
 13:03
gfldex your grammar is my command
13:04 jack_rabbit left
baest pqmD5or2pw 13:08
sorry
masak baest: what did you just call my dog? :P 13:09
gfldex don't be shy, tell us what this password is for
masak gfldex: it says right there, it's to unlock his pqmD5or2
13:10 vnbg joined, vytas left
melezhik Hi! If perl6 has analogous of perl5 our varibales? like "our $foo" 13:15
masak m: our $our-variable = "yes"; say $our-variable 13:16
camelia rakudo-moar b3e9f5: OUTPUT«yes␤»
melezhik probably here - doc.perl6.org/language/variables#Th...Declarator ...
?
masak melezhik: sorry, what's the question you have? 13:17
melezhik ah, nevermind, it seems you answered ...
;))
13:18 donaldh joined 13:19 vytas joined
masak note that Perl 6 `our` variables are not *exactly* like Perl 5 `our` variables. 13:20
13:20 vendethiel left
masak but chances are they are close enough for your needs. 13:20
I usually think of the `our` declarator as what allows `$MyModule::foo` access.
melezhik <masak> note that Perl 6 ... - sure! thank you 13:21
13:22 Actualeyes joined 13:23 vnbg left 13:24 sufrostico joined 13:25 ufobat joined 13:29 skids joined
stevieb m: my $x = "morning, " ~ <112 101 114 108 54>>>.chr.Str; $x~~s:g/\w+//.Str.say' 13:32
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 03108 54>>>.chr.Str; $x~~s:g/\w+//.Str.say7⏏5'␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modi…»
stevieb m: my $x = "morning, " ~ <112 101 114 108 54>>>.chr.Str; $x~~s:g/\w+//.Str.say
camelia rakudo-moar b3e9f5: OUTPUT«morning p e r l 6␤»
stevieb why is the whitespace in "p e r l" not being removed? 13:33
psch m: my $x = "morning, " ~ <112 101 114 108 54>>>.chr.Str; $x~~s:g/\w+//; $x.Str.say
camelia rakudo-moar b3e9f5: OUTPUT«, ␤»
psch stevieb: (1) \w means "word characters", no "whitespace characters", and (2) you're printing the unmodified $x 13:34
13:34 darutoko joined
stevieb psch: d'oh!!! wow. Not enough coffee this morning. particularly the '\w'. Think I'll go back to bed 13:35
jnthn m: my $x = "morning, " ~ <112 101 114 108 54>>>.chr.join
camelia ( no output )
jnthn m: say "morning, " ~ <112 101 114 108 54>>>.chr.join
camelia rakudo-moar b3e9f5: OUTPUT«morning, perl6␤»
stevieb m: my $x = "morning, " ~ <112 101 114 108 54>>>.chr.Str; $x~~s:g/\s+//.Str; say $x
camelia rakudo-moar b3e9f5: OUTPUT«morning,perl6␤»
stevieb jnthn: thanks!
13:37 vendethiel joined 13:44 sno left 13:46 Sgeo joined 13:51 rurban left 14:02 cdg joined 14:03 cdg left, cdg joined 14:06 zakharyas left, zakharyas joined 14:10 _mg_ joined 14:11 zakharyas left
dalek pan style="color: #395be5">perl6-examples: 00f090f | (Shlomi Fish)++ | categories/euler/prob463-shlomif.p6:
Start porting my #463 solution to p6.
 14:13
14:17 daxim joined
dalek c: d03ad1b | (Aleks-Daniel Jakimenko-Aleksejev)++ | / (2 files):
Slight changes to the “Not in Index” search result

Red background instead of purple; slightly bigger font; more noticeable link; parens outside of the link (otherwise it looks weird with underline); inverted font weight.		 14:21
AlexDaniel .tell Zoffix please review github.com/perl6/doc/commit/d03ad1b472 14:22
yoleaux AlexDaniel: I'll pass your message to Zoffix.
14:22 huggable joined 14:24 mohae joined 14:26 mst left, mst joined, brrt left
sexy-coder-girl It wasn't purple! It was "Can Can" on "Camile" :P 14:27
or "Carmile" rather :)
RabidGravy :)
14:29 cdg left
AlexDaniel
.oO( it wasn't colorful! It was “Camelia”… )
 14:30
.oO( we need colorful letters in unicode )
 14:32
14:34 abruanese left
timeless m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || ]!; 14:36
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || 7⏏5]!;␤ expecting any of:␤ infix stopper␤»
psch m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*)]!; 14:37
camelia rakudo-moar b3e9f5: OUTPUT«hi </a/><b><c>␤»
psch timeless: the alternation cannot be empty
14:38 acrussell joined
psch m: say "/a/" ~~ / '/' ~ '/' .+ / # also this might be an operator of interest 14:38
camelia rakudo-moar b3e9f5: OUTPUT«｢/a/｣␤»
timeless AlexDaniel: keeping a cursor position in a line doesn't seem hard...
timeless doesn't see liz
psch: I can't read that 14:39
in fact, p6 makes my head hurt
dalek pan style="color: #395be5">perl6-examples: 0838dc2 | (Shlomi Fish)++ | categories/euler/prob463-shlomif.p6:
Add a working solution of #463.
psch timeless: / $a ~ $b <$c> / translates to "match $a, then match $c (as pattern) and then match $b again" 14:40
m: say "[foo]" ~~ / '[' ~ ']' 'foo' /
camelia rakudo-moar b3e9f5: OUTPUT«｢[foo]｣␤»
timeless p5: $a = '/a/b#c'; $a =~ m{(/[^/]+/)([^#]+)(?:#(.*)|)};
14:40 mst left, mst joined
timeless is what I want/need 14:41
psch that's still an null pattern in the alternation - also 'p5: '
m: //
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3//7⏏5<EOL>␤»
psch timeless: ^^^ that is the problem you're introducing
timeless: alternative to that is an empty string literal, e.g. ''
14:41 mst left, mst joined
psch oh, that was an example in perl5 14:41
timeless shrugs
psch i see
timeless I just need something that works
psch m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m![(<-[#]>*)[<[#]>(.*)||'']!; 14:42
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse expression in metachar:sym<[ ]>; couldn't find final ']' ␤at <tmp>:1␤------> 032>" if $a ~~ m![(<-[#]>*)[<[#]>(.*)||'']7⏏5!;␤ expecting any of:␤ infix stopper␤ …»
timeless the p5 concept is really fairly easy to understand
`infix stopper`?
timeless goes to check the docs
Your search - site:docs.perl6.org infix stopper - did not match any documents. 14:43
psch m: my $a = '/a/b#c'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || '']!;
camelia rakudo-moar b3e9f5: OUTPUT«hi </a/><b><c>␤»
timeless my $a = '/a/b#'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || '']!;
m: my $a = '/a/b'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || '']!;
camelia rakudo-moar b3e9f5: OUTPUT«Use of Nil in string context in block <unit> at <tmp> line 1␤hi </a/><b><>␤»
timeless bonus points for not giving me a Nil... 14:44
psch m: my $a = '/a/b'; say "hi <$0><$1><$2>" if $a ~~ m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.* || '')]!;
camelia ( no output )
timeless fwiw, with this fixed, i'm pretty close to having decent filenames/urls for doc.perl6 14:45
psch good luck :)
timeless but I can only spend ~15mins on this, and then I should go back to my past
psch: so, are you giving up on the nil?
m!(\/<-[/]>*\/)(<-[#]>*)[<[#]>(.*) || '']! ---- seems okish 14:46
14:49 xinming left
psch timeless: well, if you're capturing something that might not match, of course you get a Nil 14:49
timeless: perl5 probably just doesn't tell you
14:49 jack_rabbit joined
psch m: "foo" ~~ /(bar)/; say $0 # same case 14:50
camelia rakudo-moar b3e9f5: OUTPUT«Nil␤»
timeless well, p5 basically treats it as '', which makes playing w/ strings much less annoying :)
14:50 xinming joined
psch m: constant Nil = Nil but role { method Str { '' } }; say "{Nil}" 14:51
camelia rakudo-moar b3e9f5: OUTPUT«␤»
psch shrugs
14:51 rurban joined
sexy-coder-girl Perl 5 complains that it's undef, actually :) 14:52
timeless oops :)
timeless clearly hasn't used p5 much in a while
sexy-coder-girl m: my $x = quietly Nil ~ Nil; say "[$x]" 14:53
camelia rakudo-moar b3e9f5: OUTPUT«[]␤»
sexy-coder-girl Same behaviour as in P5, AFAICS
timeless ... my bandwidth here seems to have been limited to low-speed dialup :( 14:55
sexy-coder-girl AlexDaniel++ bisectable is a damn fine bot! 14:57
diakopter all the sexiness distracts me :o
sexy-coder-girl so useful
14:57 araujo joined, araujo left, araujo joined 14:58 araujo left
AlexDaniel sexy-coder-girl: yeah, I enjoy seeing people use it 14:58
especially after my latest changes :) 14:59
rindolf seems like Travis CI is failing here - travis-ci.org/perl6/perl6-examples.../137810286
14:59 vendethiel left 15:01 _mg_ left
moritz it does't just seem like it. It fails. 15:01
dalek pan style="color: #395be5">perl6-examples: 4d79874 | (Zoffix Znet)++ | Makefile:
Term::ANSIColor has been renamed
 15:02
sexy-coder-girl ^ should be the fix 15:03
AlexDaniel diakopter: I agree. This nickname is very annoying
sexy-coder-girl hm 15:04
AlexDaniel sexy-coder-girl: consider changing it to “womble”
perlpilot heh
15:05 jack_rabbit left
timeless sighs 15:05
dalek pan style="color: #395be5">perl6-examples: e3dabb9 | (Zoffix Znet)++ | t/000-check-dependencies.t:
Term::ANSIColor is deprecated in favour of Terminal::ANSIColor
timeless apparently it's a good idea to escape quotes....
sexy-coder-girl AlexDaniel: feel free to open an Issue so we could all participate in a debate :) 15:06
AlexDaniel sexy-coder-girl: sure, where's the repo?
15:07 sno joined
sexy-coder-girl LHF to contribute: build all modules in ecosystem and report any "occurances of deprecated code" vis-a-vis using from-json/to-json from core 15:07
15:07 stevieb left
AlexDaniel by the way, what's up with smoke.perl6.org/report ? 15:09
15:09 user9 left
timeless hrm, I just unwatched 188 repositories, and I have no idea what they were 15:10
15:10 stevieb joined 15:11 vibha left
awwaiid that smoke site looks neat 15:11
15:12 zakharyas joined
AlexDaniel yeah, it's awesome 15:13
dalek pan style="color: #395be5">perl6-examples: 4b40008 | (Zoffix Znet)++ | .travis.yml:
Don't build ancient panda
AlexDaniel except that it seems like it's not live anymore
can we get it up again?
sexy-coder-girl stifles a laugh
huf sexy-coder-girl: oi! what happan to meatcpan.org? 15:14
awwaiid oh, I didn't notice the date
timeless SEC7115: :visited and :link styles can only differ by color. Some styles were not applied to :visited.
AlexDaniel: can someone change the perl css not to upset modern security conscious web browsers?
sexy-coder-girl huf: I let it expire. The joke wore itself off :)
15:15 _mg_ joined
AlexDaniel timeless: sure! Is that an output of some tool? 15:15
timeless F12 (edge debugger)
15:15 user9 joined
sexy-coder-girl timeless: stop using stupid debuggers :) Problem solved 15:15
timeless (probably also from IE's F12, but I'm too lazy to open that...)
AlexDaniel I don't think that this warning is stupid
timeless: are there any other warnings? 15:16
timeless: and does it tell any line number or something?
timeless I don't see any particular hint that it does ... 15:17
my bet is pygments.css 15:18
sexy-coder-girl AlexDaniel: why is it not stupid?
timeless and it might be that you want to change e.g. the split A:link:hover / A:visited:hover to
A:hover +:link/:visited
or just um, get rid of the split 15:19
since they're actually the same color
sexy-coder-girl timeless: what's "perl css"? Which website are you talking about?
timeless doc.perl6.org
ah
got it
here www.irccloud.com/pastebin/XKNVl1oM/
15:20 domidumont left
timeless someone is futzing w/ text-decoration 15:20
well, maybe
15:20 _mg_ left
timeless still isn't absolutely certain 15:20
I can give you the url where msdn explains what they're thinking 15:21
AlexDaniel yes please
timeless and perhaps next week I could be on the team involved in that ui...
msdn.microsoft.com/query/dev12.quer...p.SEC7115)
AlexDaniel timeless: I'd love you to get that . / # ? thing done 15:22
timeless but yeah, basically text-decoration shouldn't be in the rule for :visited
AlexDaniel: yeah, i'm testing it
it's close, but i'm missing a cigar
AlexDaniel timeless: sure I'm not the one who should tell you what to do, but I'm just saying that I'd be happy to see that
15:22 vendethiel joined
timeless chuckles 15:23
HTML1409: Invalid attribute name character. Attribute names should not contain ("),('),(<), or (=).
timeless is impressed by f12
moritz is impressed we are now at HTML 1409; thought HTML 5 was state of the art 15:24
15:24 user9 left
timeless DOM7011: The code on this page disabled back and forward caching. For more information, see: go.microsoft.com/fwlink/?LinkID=291337 msdn.microsoft.com/query/dev12.que...p.DOM7011) 15:24
moritz: including an id + a number = google/bing searching can work
15:25 zakharyas left
moritz timeless: I know, I'm just trolling a bit 15:25
timeless and it doesn't limit you to your language (esp if that language isn't en-* where someone localized the error message)
moritz: I actually worked on exposing useful (crypto) error tokens to gecko a lifetime ago..
AlexDaniel timeless: this one feels IE-specific, although I'm not sure
timeless AlexDaniel: the visited thing isn't
i'm involved w/ the standards group in this area 15:26
AlexDaniel timeless: sure, yes, though I'm still not sure what's going on there…
timeless browser vendors are converging on more or less that rule
sexy-coder-girl timeless: doesn't the visited thing would still pose the same issue for colour changes?
Or does getComputedStyle() return trash for colour?
timeless sexy-coder-girl: color is not usually detectable by most scripts
and the UA can lie if it wants to about color
timeless can't remember what the UAs actually do for it 15:27
developer.mozilla.org/en-US/docs/W...d_selector
sexy-coder-girl Seems they went with a backwards solution. Why not just make getComputedStyle() return :link styles for :visited and be done with it :S
timeless Little white lies 15:28
The first change is that Gecko will lie to web applications under certain circumstances. In particular, getComputedStyle() and similar functions such as element.querySelector() always return values indicating that a user has never visited any of the links on a page.
-- the problem is that most other things beyond color can result in layout changes
and the UA isn't willing to retain a complete alternate layout just to lie to gCS 15:29
sexy-coder-girl Ahh, I see
timeless tries to figure out why routine is doing the wrong thing :( 15:30
right now, i need to get `&` doing what i want www.irccloud.com/pastebin/GMGZMtBv/ 15:31
15:31 ggoebel116 joined
AlexDaniel so, first of all, we're talking about pages like this 15:32
doc.perl6.org/language/objects#Using_Objects
you can see a <pre> with some links in it
rindolf sexy-coder-girl: thanks for fixing the perl6-examples repository.
dalek c: 6ee80bc | (Zoffix Znet)++ | html/css/style.css:
Do not style :visited
15:32 ufobat left
AlexDaniel now, the css says that visited and unvisited links are black and have an underline 15:32
what's the problem?
15:33 darutoko- joined
AlexDaniel the fix above is going to work, I think, but why was it complaining? 15:33
timeless that fix isn't enough
pygments whatever... 15:34
15:34 Xliff_ left
sexy-coder-girl AlexDaniel: it's a security warning. A malicious script can use the alternate styling to determine your identity, yada yada 15:34
timeless and boy do I wish the debugger linked me to the css in question
timeless wonders if FF or Chrome are nicer here
sexy-coder-girl timeless: well, at least you can read yours. Mine opened in like a 50px x 200px window and I can't resize it and it doesn't even go away when I press F12 again :S 15:35
AlexDaniel sexy-coder-girl: how? If the style is exactly the same for both visited and unvisited links
15:35 ssotka joined, user9 joined
timeless sexy-coder-girl: there's a maximize button on the far right 15:35
sexy-coder-girl timeless: don't got any buttons. 15:36
15:36 darutoko left
timeless AlexDaniel: see the snippet I pasted earlier where text decoration isn't the same 15:36
usercontent.irccloud-cdn.com/file/.../image.png
dalek c: 09bfa0e | (Zoffix Znet)++ | html/css/style.css:
Toss another differentiation between :visited and :link
 15:37
AlexDaniel ah, so by creating two links (one that is in pre and another one that is elsewhere) you can query the style to check if the user has visited that link or not
that's pretty clever, really
15:37 raiph left
timeless there's a whole area of security research there 15:37
(seriously)
AlexDaniel and the fix is to not touch text-decoration, right? 15:38
timeless it's been driving browser engineers nuts for over a decade
well, i'm speculating, because i'm trying to focus on the other thing :)
sexy-coder-girl Seems like this shit would apply only to few people who are living off-the-grid. No Facebooks or anything.
timeless but, my fix would be to *only* touch color
sexy-coder-girl: no
anyone who has visited something obscure
they're sorta-wateringhole attacks 15:39
maybe I check to see if you've been to `pulse`
15:39 raiph joined
timeless or if you've been to the admin page of GitHub/{something} 15:39
AlexDaniel so, do I get it right that if you want to style your links differently, then these styles should apply to all links on the site?
sexy-coder-girl timeless: how? You'd need to install some malicious script on my box won't you? 15:40
timeless you can style things in A:link. But your :visited css should *only* touch color
sexy-coder-girl: no
sexy-coder-girl The more I think of this issue the more stupid it seems to my mind.
AlexDaniel aha… okay
timeless <a href="pulse.com">a....
AlexDaniel sexy-coder-girl: no, that's not stupid
timeless isn't sure what pulse's web site is
timeless presumes everyone here recognizes pulse as of today 15:41
sexy-coder-girl never heard of it
timeless sexy-coder-girl: meatspace-meetspace
[Coke] meta meat meet space.
... team
timeless and not to beat a dead horse, but um... dead meat
:(
AlexDaniel timeless: ok, that's an interesting observation. Thanks for reporting
timeless sexy-coder-girl: is that enough for you to recognize the ref? 15:42
sexy-coder-girl timeless: well, what is that link? What does it mean? Where would it be used? I can just stuff that on any page I want and tell you "Here, take a look!" and gather all the URLs you visited.
timeless sexy-coder-girl: there were attacks precisely like that, yes
sexy-coder-girl I don't see how the styling of visited links on doc.perl6.org is in any way exploitable by some dude that never installed any malicious scripts in my browser. 15:43
timeless lemme see if I can find a sample
sexy-coder-girl: :visited isn't limited to local links
it in theory applies to links to other domains
so wateringhole.attack.example.com
15:43 vendethiel left
timeless could color w/ :visited a link to admin.perl6.org 15:43
sexy-coder-girl timeless: sure, but the styling will be applied only when I'm on doc.perl6.org and there are ain't any malicious people gathering stats 15:44
awwaiid that would only matter if perl6.org loaded evil content, right?
timeless no
:visited would apply if your browser knew you had visited admin.perl6.org
and it would apply to the link on wateringhole.attack.example.com
sexy-coder-girl timeless: right, but it would apply only when I'm browing docs.perl6.org
timeless no
if your history says you went to admin.perl6.org 15:45
then any page you visit anywhere on the web w/ a link to admin.perl6.org can style :visited
sexy-coder-girl timeless: exactly, but what does it matter WHAT that style is on docs.perl6.org?
timeless instead of thinking of `doc.perl6.org` as an attackable site
awwaiid well what's the worst case here -- someone at some point can get a whitelist of my browsing history, and htereby narrow an attack for me, right?
timeless awwaiid: so...
imagine you were a frequenter of Pulse 15:46
awwaiid k
timeless but you hadn't told people
awwaiid right
timeless if they can determine that you visited Pulse, they can out you from the closet
awwaiid right
privacy leak
timeless maybe you're in a place where they pehead people
15:46 darutoko- left
timeless s/peh/beh/ 15:46
awwaiid and worse it can also allow them to target an attack at me -- like now that they know I go to pulse they can craft a targeted fishing link
timeless now, it could be that they choose to attack you because it turns out you have useful credentials for admin.perl6.org
sexy-coder-girl timeless: but they would just give you a URL to visit. I still don't see how docs.perl6.org styles help with anything whatsoever 15:47
timeless sexy-coder-girl: lemme see if I can find a sample page that talks about this
sexy-coder-girl Sure.
awwaiid sexy-coder-girl: I think browsers have worked-around this for the color attribute
timeless the point is that they could have a page with 1000 links
and it's not about asking the user to visit them all
it's about asking the browser has the user already visited them 15:48
awwaiid right -- they just introspect the page to see if the browser knows they've been :visited
sexy-coder-girl timeless: yes, but this still has nothing to do with styles on docs.perl6.org
timeless sexy-coder-girl: the way browsers protect against this is where it crosses docs
sexy-coder-girl Hm?
timeless docs is using some styles which browsers have decided they consider risky and don't want to honor
sexy-coder-girl So it's not a security issue. It's browsers thinking we're a malicious website. 15:49
timeless because browsers want to protect their users
browsers have to treat *all* sites as malicious
dbaron.org/mozilla/visited-privacy
AlexDaniel but the point is that unless we start loading weird shit on docs.perl6.org there's no problem
sexy-coder-girl Exactly. And thus the circle is complete with my original statement: that warning is stupid. 15:50
timeless well
sexy-coder-girl leaves for lunch
AlexDaniel there are no ads or anything, the site is also non-dynamic
timeless the point is that your UX designer needs to know that the style he's using won't work
AlexDaniel so XSS is out as well
timeless and he should stop using it
blog.mozilla.org/security/2010/03/...tory-leak/
AlexDaniel timeless: check it out again, did the warning disappear? 15:51
awwaiid ah, this makes sense now. Not protecting users directly, just complying with browsers trying to protect users.
timeless sighs 15:52
awwaiid I heard about this info-leak when it came out, but didn't know the latest in how browsers try to fix
timeless I can't figure out if (NotSAFE!) www.haveyourfriendsbeenthere.com/ is a different attack or something else 15:53
AlexDaniel: still there
you can get a VM from ms
15:53 ggoebel116 left
timeless modern.ie 15:54
> Test Microsoft Edge and also IE6 up to IE11 using virtual machines that you download and manage locally for free.
cssfingerprint.com/results
is probably the right thing
but it looks like it's gone :( 15:55
timeless asks wayback for help
awwaiid aww that's too bad. there's the eff one for fingerprinting that works similarly...
but is a little different 15:56
timeless web.archive.org/web/20100411060604/...print.com/
15:57 rurban_ left 15:59 sno left, sno joined 16:01 rindolf left
AlexDaniel timeless: how can it be there if :visited links are not touched. Are you sure that you're not using a cached version of css or something? 16:01
16:01 khw joined
sexy-coder-girl timeless: I'm not getting the warning on Edge 13 16:02
16:02 pmurias joined
sexy-coder-girl Oh, nm.. Had to move my mouse over a link to trigger it 16:02
timeless AlexDaniel: i'll leave you and sexy-coder-girl to work this through 16:03
since it seems like sexy-coder-girl hit it :)
timeless goes back to &
16:03 raiph left
timeless sighs 16:04
16:04 raiph joined, rindolf joined 16:05 wamba joined
pmurias sexy-coder-girl: re warning is stupid, why should the browsers give docs.perl6.org access to the users browser history? 16:06
sexy-coder-girl Nah, I just wanted to see if there was a way to find a line number.
timeless (we should ask google/bing if there's a way) 16:07
sexy-coder-girl pmurias: it's up to them to decide. If they see no reason, they can ignore any styles but `color` applied to :visited. It's pretty dumb expecting the devs to comply with some sort of warning
timeless the problem is that enabling that adds cost to the engine
awwaiid so does detecting it to give a warning :)
16:08 rindolf left
sexy-coder-girl I mean really, the bad guys won't comply. I see it more as "in the future, your non-color styles might stop working" warning. 16:08
timeless awwaiid: detecting is much cheaper than maintaining the info necessary for a verbose message
AlexDaniel yeah, so we should fix it
awwaiid yeah, I suppose
timeless sexy-coder-girl: precisely
awwaiid sexy-coder-girl: I agree
16:09 mr-foobar joined
sexy-coder-girl timeless: I was just trying to see if I could ease out the line number from it, but I don't see much return on investment in attempting to fix this. It says some styles not applied to "visited." So that'll involve digging to jquery UI CSS, pygments and other shite on the site to find it 16:09
16:09 ggoebel116 joined
timeless so, back to perl6 16:10
I want to find out where some data came from :/
(I really want a custom tainter) 16:11
16:11 araujo_ joined
moritz the second most annoying question when debugging: "where did this data come from?" 16:11
the most annyoing being "why doesn't this code run?"
timeless #2 is much harder than #1 16:12
16:12 setty1 joined
AlexDaniel timeless: by the way, is it a good idea to separate those with -? That is, why amp-amp instead of ampamp? :) 16:12
timeless #2 can be basically solved by abusing #1
geekosaur wonders if "but" can be (ab?)used for data tracking/tainting
timeless AlexDaniel: some other words get confusing
awwaiid I always teach new devs to put a gigantic syntax error in their code to prove it CAN break and that they're running it at all :)
pmurias sexy-coder-girl: you mean it's stupid that there is a warning but the styles are applied despite it? 16:13
timeless awwaiid: indeed
AlexDaniel: i'm not 100% certain that the -'s are needed 16:14
it's a lot easier to understand that dash-dash-x is two dashes and not some special dashdash
AlexDaniel timeless: dunno, hm… what about using spaces? :)
timeless AlexDaniel: consider `/ /` 16:15
AlexDaniel uhh
timeless vs. `//`
i'm tempted to use `_` instead of `-` -- there already was code trying to use `_` fwiw... 16:16
16:16 vendethiel joined
timeless AlexDaniel: so, space is a bad idea, agreed? :) 16:17
AlexDaniel huggable: dunno
huggable AlexDaniel, ¯\_(ツ)_/¯
timeless `/ /` -> `slash slash`; `//` -> `slash slash` ==> 16:18
huggable: frown
huggable timeless, nothing found
AlexDaniel should be ‘slash slash’ vs ‘slash slash’ but yeah, that's not cool
16:19 zakharyas joined
AlexDaniel let's just hope that we're not going get into the same situation with dashes 16:19
sexy-coder-girl pmurias: I guess it's stupid that Microsoft shoves this under "Security errors" section, when it really isn't about security. I dunno, maybe it's not the warning but the way this discussion proceeded to imply that we somehow endager users of docs.perl6.org by styling :visited links (or at least that's how I perceived it).
timeless it escapes dash
but yeah, it's not wonderful, `_` is probably better, but it's harder to type
and it isn't the English token for the use case
in principle, `-` is what English uses for this
so, if `-` works w/o *too* much pain, i'd rather 16:20
if it doesn't, i'm not wed to it
AlexDaniel yeah, it's probably ok
16:22 g4 left, cdg joined
timeless cuts doc down to 5 pod files 16:23
16:28 abraxxa left 16:31 rurban left, andreoss left 16:34 donaldh left
timeless AlexDaniel: my biggest concern is `#infix &&` 16:34
because sometimes `#` should be hash, and sometimes it's `#` (anchor ref) 16:35
16:37 rurban joined 16:38 stevieb left 16:40 FROGGS[mobile] joined
FROGGS[mobile] o/ 16:40
16:41 dakkar left
timeless m: ' ' ~~ /&/; 16:42
camelia rakudo-moar b3e9f5: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Null regex not allowed␤at <tmp>:1␤------> 3' ' ~~ /&7⏏5/;␤ expecting any of:␤ term␤»
timeless ok, so, can someone please remind me why `perl6` doesn't give a line number when I do that? 16:43
AlexDaniel timeless: there's a line number
tony-o at <tmp>:1
AlexDaniel timeless: filename <tmp>, line number :1
timeless perl6 --ll-exception htmlify.p6 --no-highlight
Initializing ...
Null regex not allowed
at gen/moar/m-CORE.setting:21553 (/home/timeless/hg/perl6/rakudobrew/moar-nom/install/share/pe
AlexDaniel: perl6 ^ doesn't; camelia does.
tony-o that :\d+ is the line number
m-CORE.setting line 21553 16:44
AlexDaniel timeless: what if you use no --ll-exception? :)
timeless at gen/moar/m-CORE.setting:21553 (/home/timeless/hg/perl6/rakudobrew/moar-nom/install/share/perl6/runtime/CORE.setting.moarvm:throw)
tony-o ll-exception shows a the longer version of the error, so you get more line numbers 16:45
timeless AlexDaniel: that works
AlexDaniel hm 16:46
tony-o i think timeless has me on ignore
16:46 Xliff joined, stevieb joined
timeless so, w/o --ll-exception it gives something useful www.irccloud.com/pastebin/1npGnJl6/ 16:46
Xliff \o 16:47
timeless tony-o: i'm not ignoring you
well, I am, but i'm not
AlexDaniel timeless: can you paste full backtrace with --ll-exception?
timeless AlexDaniel: sure
16:48 TimToady left
Xliff When retrieving pointers returned from Nativecall backed lib, what is the best way to test for equivalence? =:= seems to be giving me mixed results. 16:48
timeless full www.irccloud.com/pastebin/0ZhwDW9i/
16:48 zakharyas left
timeless tony-o: the problem is that while i'm getting line numbers, they aren't related to my code 16:48
AlexDaniel yeah, it does not look right 16:49
nine ^H^H^H^H^H^H^H/win 14
AlexDaniel nine: yeah
16:50 TimToady joined
AlexDaniel timeless: I think that it needs a shorter snippet to reproduce the issue, eh. But you are right, the problem is there 16:50
I don't think that people are supposed to use --ll-exception since normal error messages are meant to give good enough error messages
if they don't, that's a bug :) 16:51
but it's the first time I see somebody complain about --ll-exception, so I don't know
timeless here you go www.irccloud.com/pastebin/IOKyc4hD/
the help doesn't say "and eat the normal stuff" :-( 16:52
I think the problem is that it really isn't what I want 16:53
i want a normal backtrace, not a low level one
tony-o timeless: ahh, that is a problem - i've had that problem too and resorted to doing 'say' debugging to figure out where it's happening. that isn't ideal if you have a lot of code
timeless every sane language supports this
dalek c: f312715 | (Zoffix Znet)++ | / (4 files):
Add 404 page with menu bar on it

Fixes #129
timeless tony-o: yep
the debugger is pretty much unusable, i tried yesterday 16:54
tony-o timeless: agreed, it isn't great and i'm sure it will improve
timeless so, i'm basically w/ say debugging
but i was using --ll-exception, naïvely assuming it'd at least help a little
at this point, it seems more like a footgun
tony-o timeless: modules.zef.pm/modules/github:Altai...vel::Trace 16:55
that may be interesting
another: modules.zef.pm/modules/not%20in%20m...l6::Tracer
timeless depends on how slow things are
16:55 domidumont joined
tony-o i need to implement short url generation for the modules.zef 16:55
nine timeless: it's a compile time error. By definition at that time, the only stack in existence is the compiler's very own and that's what you get with --ll-exception
timeless i tried perl6-debug-m and it slowed things down to a crawl
nine: um... 16:56
timeless polishes off a fancy "end user" hat
timeless points to it proudly
... wearing my end user hat... i expect to not lose information when i ask for more information
w/o that flag, i get something vaguely useful
when i add that flag, i lose it
proudly wearing my end user hat, i'm unhappy 16:57
timeless slowly removes that hat and places it back on the shelf
nine timeless: but you did _not_ ask for more information but for "a low level backtrace" and that's what you got.
timeless nine: this is not a useful argument 16:58
and i hope you can understand that
i carefully put on an end user hat to explain context
i'm trying to improve the doc site, so i'm mostly wearing a UX polisher hat 16:59
i'm also very skittish
geekosaur so, the message for end users is --ll-exceptionis not for you
it's for debugging rakudo
timeless geekosaur: footguns should be well hidden
mst so it probably needs to make it more obvious
jnthn Well, where did you find out about it?
16:59 rurban left
timeless perl6 --help 16:59
(where every normal user starts when they have a problem) 17:00
17:00 pmurias left
timeless perl5 --help does not to my knowledge include footguns 17:01
i mean, sure `-U allow unsafe operations` may be a footgun
but it's well labeled
i'm happy to suggest alternate text for your footgun if you are interested 17:02
17:02 vendethiel left
timeless although, personally, i'd much rather it just include the normal output in addition to what it's doing 17:02
geekosaur would move internal options to a --help-all or something, really
timeless --- and to finish my point about being skittish. i have dozens of other things i can do 17:03
jnthn The point of it is to help us debug situations where producing the nroaml output *is* the problem.
I'd sooner just remove it from --help if it's going to cause confusion, than try and find other wording.
17:03 vendethiel joined
timeless either works. 17:04
thanks for taking my feedback. i appreciate it
17:04 FROGGS joined
timeless tony-o: ok, so... i can try -DDevel::Trace 17:04
jnthn Could also add a --help-dev to list such things. --stagestats is arguably also better under there, and we could mention the RAKUDO_MODULE_DEBUG envvar or whatever it's called :) 17:06
timeless hrm, that might be vaguely usable
offhand --target and --optimize probably don't belong in --help either 17:07
nine hopes that we can one day get rid of RAKUDO_MODULE_DEBUG
Xliff Isn't there a way to have rakudo print out all of the operations it is performing?
nine All the debug output does not exactly make the code more readable
timeless i'm also not sure whether --profile* belongs in --help -- i think it depends on whom your target audience is 17:08
Xliff Or did I really imagine that capability was in there. (maybe confused with RAKUDO_MODULE_DEBUG)
timeless if it's more for debugging moar than for debugging my program, then i'd suggest you move it too
oh!
perl6 --doc ~ perldoc ? 17:09
would it hurt much to add `per6-doc` which was a wrapper for perl6 --doc ?
*perl6-doc 17:11
17:12 spider-mario joined
nine Why don't I have permission to view ticket #63956? 17:13
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=63956
17:15 ingy joined
jnthn --profile is certainly intended for normal users (or at least intermediate-advanced level users) to understand where their program is taking time, and it's pretty normal for compilers to take a --optimize flag also. So I think those two are fine under --help. 17:16
dinner & 17:17
AlexDaniel well, I don't think that we can toss out --ll-exception right now
because error messages are not that good yet
timeless AlexDaniel: this isn't about removing it, so much as moving it :)
jnthn AlexDaniel: The proposal isn't to remove it, just to not include it in --help.
really dinner :) &
AlexDaniel yeah, I mean keep it in --help
timeless should lunch
AlexDaniel in fact, I'd vote to make it more visible… 17:18
:/
17:21 sjn joined 17:22 donaldh joined
timeless nine: out of curiosity, how'd you find that number? 17:23
nine timeless: it's referenced in t/spec/S11-modules/nested.t
timeless frowns 17:24
17:25 kurahaupo joined, vendethiel left 17:27 sexy-coder-girl is now known as BrokenRobot 17:29 vendethiel joined
timeless [timeless@gcc2-power8 doc]$ perl6 --doc Perl6::Documentable::Registry 17:33
Could not open Perl6::Documentable::Registry. Failed to stat file: no such file or directory
17:33 user3 joined
timeless ok, i give up, how does `perl6 --doc` work/ 17:34
user3 im also interested in this answer
17:36 rindolf joined 17:40 donaldh left 17:41 domidumont left
timeless tony-o: the second one seems promising 17:43
gfldex see design.perl6.org/S26.html#How_Pod_i..._processed 17:44
timeless [timeless@gcc2-power8 doc]$ perl6 --doc perlrun 17:45
Could not open perlrun. Failed to stat file: no such file or directory
gfldex: that's based on the example in your url
17:46 stux|RC-only left, stux|RC-only joined 17:47 wamba left
gfldex you didn't tell it what module to use to render pod 17:47
17:47 user3 left
gfldex perl --doc $foo # will look for POD::To:$foo 17:47
Pod::To::$foo even 17:48
17:49 FROGGS[mobile] left
timeless so, we should talk about help 17:50
--profile=kind write profile information to a file (MoarVM)
means kind is mandatory
--target=[stage] specify compilation stage to emit
should mean stage is optional
?? www.irccloud.com/pastebin/cVjJ8I8H/ 17:52
17:53 vendethiel left
stevieb I absolutely LOVE being able to easily look at what methods a class has while I'm learning. I just dedicate one window for things like this, so I don't have to look up in the docs all the time: "Str>>.^methods>>.say" 17:53
...what I *really* meant was "Str.^methods>>.say" 17:54
17:54 user9 left
timeless m: Str.^methods>>.say 17:54
camelia rakudo-moar 66ebd8: OUTPUT«BUILD␤Int␤Num␤chomp␤chop␤pred␤succ␤simplematch␤match␤ords␤samecase␤samemark␤samespace␤word-by-word␤trim-leading␤trim-trailing␤trim␤encode␤NFC␤NFD␤NFKC␤NFKD␤wordcase␤trans␤indent␤codes␤chars␤uc␤lc␤tc…»
17:54 vendethiel joined
stevieb timeless: I was going to do that here, but was afraid it would list one item per line ;) 17:56
17:58 user9 joined
timeless gfldex: ok, so, how does one find a list of Pod::To::{things}} ? 17:59
gfldex modules.perl6.org/#q=Pod%3A%3ATo 18:00
skids (and there are a couple shipped with rakudo core)
timeless assume i only have core 18:01
(the rrakudo world build system is a bit of a mess -- i couldn't get panda to build, i'm happy i have zef) 18:02
skids OK it looks like just "Text.pm6"
18:02 Sqirrel joined
timeless [timeless@gcc2-power8 doc]$ perl6 --doc Text perlrun 18:02
Could not open Text. Failed to stat file: no such file or directory
skids --doc=Text works for me 18:03
timeless [timeless@gcc2-power8 doc]$ perl6 --doc=Text.pm6 perlrun
Could not open perlrun. Failed to stat file: no such file or directory
[timeless@gcc2-power8 doc]$ perl6 --doc=Text perlrun
Could not open perlrun. Failed to stat file: no such file or directory
p6: say 3 18:04
camelia rakudo-moar 66ebd8: OUTPUT«3␤»
timeless m: say 4
camelia rakudo-moar 66ebd8: OUTPUT«4␤»
stevieb panda list | grep Pod
timeless doesn't understand the difference 18:05
stevieb: as i noted above, i don't have panda
stevieb oh crap, sorry :)
skids $ perl6 --doc=Text -e '=begin pod
> foo
> =end pod'
foo
timeless www.irccloud.com/pastebin/5gjMOqaL/ 18:06
also, we already know that `tohtml` works, since i'm able to run the htmlify.p6 script 18:07
fwiw, your -e test also works 18:08
so, the problem is not the --doc= notation
it's the `perlrun` side of the equation
the doc= bits are indeed optional www.irccloud.com/pastebin/MQ0DJFr4/ 18:09
BrokenRobot timeless: what exactly `perlrun` supposed to be? 18:16
timeless design.perl6.org/S26.html#How_Pod_i..._processed 18:17
18:17 vendethiel left
timeless Hence, to read Pod documentation you would type things like: 18:17
perl --doc perlrun
BrokenRobot timeless: that's just an example (Perl 5 has perlrun pod) 18:18
timeless BrokenRobot: well. if that isn't supposed to work, then remove the example or replace it. 18:19
AlexDaniel is happy to see somebody who is unhappy with the way things work :)
BrokenRobot timeless: you're reading an archived, historical speculation.
timeless BrokenRobot: i'm reading it because someone pointed me to it
don't point people to things if you don't want people to read them.
or remove them.
you left a landmine in a playground and are surprised someone stepped on it. 18:20
BrokenRobot wonders wtf happened to the red message at the top of the page
timeless also, please note that my time and patience are limited
if i hit a certain number of failures / certain number of negative pushbacks, i'll leave 18:21
heck, i shouldn't be here now. i should be reviewing something else
18:22 huggable left
BrokenRobot timeless: as you've probably noticed, there are many people working on the Perl 6 project, constantly improving it. Complaining that something isn't perfect and threatening to leave (for the second time) isn't going to win you any favours. 18:22
18:22 huggable joined
BrokenRobot timeless: if you're frustrated, take a break. 18:23
timeless also, please don't rely on a red box at the top of the page.. i was given an #anchor, so unless css brings that box into view, the box isn't worth anything.
AlexDaniel this red box is fixed :)
BrokenRobot timeless: the box were visible even with the anchor.
AlexDaniel: for some reason it's gone
18:23 Actualeyes left
AlexDaniel huh? Indeed… 18:24
it's visible in other files, just not in S26
timeless ...
BrokenRobot Oh, indeed. Maybe something wasn't rebuilt properly 18:25
gfldex IIRC S26 is causing troubles when turned into HTML and is processed by hand
TimToady timeless++ for finding a bug; timeless-- for playing emotional blackmail games, so I guess it's a wash :)
timeless TimToady: i'm more of a 3 steps forward 2 steps back kinda guy
TimToady we value being stubborn :) 18:26
AlexDaniel I think that it is totally fine that timeless is complaining about things. It's not like he does nothing and is only complaining, he is working on stuff and is constantly reporting problems. It may sound a bit negative, but hey, I think that it pushes things forward.
BrokenRobot There's a note "(HTML rendering of S26 is known to be incomplete)" I guess whatever is causing it to be incomplete is causing it to not get refreshed with proper CSS <link> 18:27
18:27 vendethiel joined
AlexDaniel timeless: by the way, going to close #597 in favor of what parabolize has suggested 18:28
timeless which is that? NaN/Inf or the doc link?
i'm about to push the doc link
AlexDaniel github.com/perl6/doc/pull/597/files
ah
18:28 ggoebel116 left
timeless is waiting for a silly process to finish 18:28
18:29 donaldh joined
timeless and yeah, i'm using his text... in some other world, i'd be tempted to credit him w/ it 18:29
note to self: use cp -r instead of locally cloning repos... :(
BrokenRobot S15 also ain't got a message 18:30
Fixed it manually. No idea if it'll get rewritten back on next cron run 18:32
timeless BrokenRobot: thanks
AlexDaniel hm, let me try this: guys, let's move design.perl6.org to speculations.perl6.org?
BrokenRobot +1 18:33
18:34 ggoebel116 joined
moritz -1 18:34
I still find the name "design docs" pretty fitting 18:35
anybody who has a bit of experience with real-world software development knows they aren't always 100% accurate
timeless last i checked there were still links into it from doc.perl6 -- although perhaps we've removed them
AlexDaniel sure, it's ok if it works for you. Except that people still keep thinking that these are up to date 18:36
moritz also, last time the URL changed, I went everywhere I could and changed links (like Wikipedia, other wikis, my pages etc.)
AlexDaniel we can have a redirect 18:37
moritz AlexDaniel: then maybe we should either update them, or mark specific areas as outdated
I think that would have a much better effect than changing the domain name
timeless AlexDaniel: if i know it works, should i wait for travis or just push it? :) 18:38
dalek c: b7e4c71 | (Josh Soref)++ | doc/Language/5to6-perlsyn.pod:
use pod link instead of url for 5to6-nutshell
[Coke] (mark areas as outdated) - every page is marked as outdated. 18:40
except, probably the ones generated from POD6 instead of POD5
AlexDaniel outdated.perl6.org is not too bad too, I agree ;) 18:41
timeless: if it's obvious then just push it
timeless nods
AlexDaniel timeless: sometimes it turns out to be non-obvious and the whole thing blows up, but that's not the end of the world ;) 18:42
18:45 TEttinger joined 18:47 itaipu joined 18:50 vendethiel left, labster joined
BrokenRobot is a bit baffled seeing Pod::To::HTML actually references design.perl6.org's CSS file :/ 18:50
I guess this is an excuse enough: " 18:51
# FIXME: this code's a horrible mess."
moritz is utterly fascinated by the gravitational wave detections from binary bloack hole mergers 18:52
18:53 travis-ci joined
travis-ci Doc build passed. Josh Soref 'use pod link instead of url for 5to6-nutshell' 18:53
travis-ci.org/perl6/doc/builds/137882850 github.com/perl6/doc/compare/pod-nutshell
18:53 travis-ci left
AlexDaniel haha 18:54
parabolize is design.perl6.org/S26.html#Pod what we are using for design docs or is there some Pod on it?
moritz parabolize: much of the design docs is still in Perl 5 Pod 18:55
BrokenRobot Except for S15/S26 apparently
18:55 donaldh left
timeless moritz: i'm unfamiliar w/ bloack holes 18:57
18:58 vendethiel joined 18:59 CIAvash left
BrokenRobot Weird. Getting illegal sub declaration from trying to use the module, but if I run the same thing manually it's fine: gist.github.com/zoffixznet/58fc667...46a1515087 19:00
That's with github.com/perl6/Pod-To-HTML/blob/...To/HTML.pm
moritz timeless: physics.aps.org/articles/v9/68 if you want something to read :-) 19:02
basically, black holes are super dense (just about a hundred kilometers, but containing multiple sun masses), and when two of them get close to each other, they orbit around each otehr 19:03
and emit gravitational waves, which are small distortions in space time
timeless moritz: oh, i know a little about black holes. i was wondering about bloack holes :)
moritz oh
timeless: then en.wikipedia.org/wiki/Typographical_error is the proper URL for you 19:04
:-)
timeless :-)
huf that article should be under Tyopgraphical_error
BrokenRobot Fuck it. Was gonna attempt to fix the S15/S26 missing the message, but I'm stymied with Pod::To::HTML 19:05
geekosaur BrokenRobot, I don't think perl 6 wants you to define names in other people's namespaces like that? (although maybe it should... I cannot say)
huf in fact, the correct spelling of typo should be changed to tyop quietly
so people still keep spelling it typo
geekosaur correct way is probably package main { sub ... {} }
BrokenRobot geekosaur: but isn't it in Pod::To::HTML namespace? I've no idea where main:: is comign from
geekosaur eyes file... 19:06
just foir grins and giggles, try each of these separately: 19:07
(1) remove the leading space on line 30
(2) comment out line 30 (which will cause a different error later, of course)
19:07 rindolf left
geekosaur being nasty suspicious at parser >.> 19:08
BrokenRobot Nothing changed with neither (1) nor (2). Same error about main::escape_html 19:09
geekosaur odd
BrokenRobot :S 19:10
I think I may be hitting a precomp bug. Just typed utter trash into the file, but same error
Gonna try updating perl6 19:11
geekosaur updates so he can play...
yeh
but, that's just weird. it's defined in a unit module, it's not a qualified name, it should have that module's package as qualification... 19:12
and it's happening at compile time so hard to call it precomp
oh wait
geekosaur just reexamined gist
why are you running perl 6 source in perl (aka perl 5)? 19:13
BrokenRobot EGAWDS
geekosaur ...because that is a parl 5 error
*perl
BrokenRobot Damn, yeah, I was using `perl` instead of `perl6` ~_~
geekosaur++
moritz that's a good reason to always "use v6;" at the top 19:14
geekosaur ^
BrokenRobot heresy! :)
moritz heresay!
geekosaur I even noticed that when looking at it initially but didn't twig that you were actually running it in perl5
BrokenRobot I think I can handle being burnt by that one more time before I'll start using v6 :P
moritz erm, hearsay
19:15 girafe joined
geekosaur aaand, it kinda tells you something that p5 didn't see anything wrong until it hit that sub definition, and even then gave a less than sensible error (oh yes, I know why, but damn perl5 accepts so much trash...) 19:16
BrokenRobot :) 19:17
19:17 yqt joined
moritz it still surprises me that perl 5 accepts whitespace between sigils and variable names 19:19
geekosaur that iirc is a specific hack 19:20
moritz as is basically everything in toke.c :-) 19:21
geekosaur becaue perl language modes treated $ as an escape (to get names like $! right) and then blew up on ${expr}
so $ {expr} was made to work
and that means any other $name can also have whitespace
19:22 domidumont joined
timeless heh 19:25
19:25 itaipu left
geekosaur (to be clear I'm talking about editor modes there, not perl itself, wrt $ as escape. admittedly hacking the language parser to work around editor deficiencies is itself weird, but at the same time very perl5) 19:26
dalek : 22dfad9 | (Zoffix Znet)++ | util/update-design.perl6.org.sh:
Make Pod6 pages also have historic message
 19:28
19:29 hankache joined
BrokenRobot Someone with sudo access on hack needs to log in as design.perl6.org and update Pod::To::HTML module to include my css fix: github.com/perl6/Pod-To-HTML 19:30
19:31 jjido joined, sivoais joined
moritz doing it now 19:32
(ftr the HTML is built on hack
19:37 acrussell left 19:38 huggable left 19:41 hankache left
timeless is jnthn ~ jonathanstowe? 19:42
konobi nope
jonathanstowe == RabidGravy 19:43
RabidGravy yes
19:43 huggable joined
timeless pines for a way to run htmlify faster 19:46
BrokenRobot Weird. docs.perl6.org/fasdfdsfsdf shows proper 404 page, but docs.perl6.org/routine/%2F%2F doesn't :/ 19:47
timeless BrokenRobot: the app server version doesn't have that problem ... 19:48
BrokenRobot timeless: it's an entirely different beast, so... :) 19:49
timeless argues that this is a bug
RabidGravy: also, i don't suppose you can explain how documents get their names... 19:50
BrokenRobot timeless: the purpose of the app server version is for people who don't want to install Apache for dev. If you got it installed, feel free to experience the same version as on the live site.
AlexDaniel umm 19:51
docs.perl6.org/routine.html
trait… which trait?
(first item in the list)
timeless AlexDaniel: neat
i hope i didn't break that..
timeless finally figures out why the NF* classes are so broken 19:52
timeless will commit fix once the build finishes for NaN 19:53
RabidGravy timeless, I have no special knowledge in this regard 19:54
19:55 skrshn joined
timeless RabidGravy: so, i tried adding Constants to Num 19:55
and it doesn't appear
afaict, there's some black magic elsewhere blocking it www.irccloud.com/pastebin/GOIjVcxB/ 19:56
dalek c: 9765cb3 | (Josh Soref)++ | doc/Type/NF (3 files):
fix articles for NF* unicode classes
 19:57
timeless i don't understand what's special about `Num` 19:59
RabidGravy well, it appears exactly like that in Slip so a bit unsure 20:00
[Coke] (htmlify faster) I will hopefully have some time this week to hack on doc's makefile branch, which might help repeat builds go faster, anyway.
timeless i tried getting htmlify to fork
but i really don't know what it was doing
RabidGravy incidentally Num could do with having its methods documented 20:02
[Coke] the makefile branch is trying to split it out into chunks so that we can || build that way. 20:03
RabidGravy the only difference I can see in Slip is that it already has a =head1 Methods as well as =head1 Constants
but I've never looked at the way the HTML is produced 20:04
20:05 donaldh joined 20:07 donaldh left 20:11 colomon left
dalek c: 4fba789 | (Jan-Olof Hendig)++ | doc/Type/Baggy.pod:
Added docs for Baggy.pickpairs
 20:17
20:19 ufobat joined 20:22 jjido left
parabolize When building html for the docs is it normal to need to remove precompiled/ before running `make html`? `make clean` doesn't remove the files in that directory. If I don't remove them the web pages don't seem to change. 20:23
20:26 jjido joined 20:31 cdg left, aries_liuxueyang left 20:32 kaare_ left, aries_liuxueyang joined
moritz I've never used the makefile, I just run htmlify.p6 20:33
[Coke] ...all the makefile does is run htmlify
parabolize right
[Coke] someone added precomp support, but I'm not sure they added its cleanup to clean
20:34 zakharyas joined, domidumont left
[Coke] nope. no precomp in any clean* targets. 20:34
though ideally precomp would figure out if it was valid or not, and work properly. 20:35
but we're doing it manually instead of relying on `perl6`
so, "normal"'s not the right word, but it's not unexpected, no. 20:36
20:38 domidumont joined
timeless RabidGravy: this doesn't work either www.irccloud.com/pastebin/7QQPo3fe/ 20:38
here's the generated content (with most other files deleted to speed generation time) www.irccloud.com/pastebin/vzurfPbq...2FNum.html 20:40
20:42 jjido left 20:43 domidumont left, vike left 20:44 jjido joined, fhorck joined 20:45 vike joined 20:46 FROGGS left
RabidGravy well, I'm so unfamiliar with the mechanics of the doc generation I would be tempted to just commit what you had in Num.pod, then if it doesn't show up on docs call it a bug 20:51
it works fine with the standard Pod::To::HTML - I guess the other stuff is post processed in 20:52
parabolize timeless: could you upload those files to something I could copy / paste 20:53
nevermind / I found the raw button 20:54
timeless parabolize + RabidGravy : thanks for working w/ me :) 20:59
21:02 skids left, avuserow_ joined
timeless so doesn't understand this 21:04
is it possible for Devel::Tracer to spit out the values of variables? :) 21:05
ok, so, what exactly is the story on `perl6 --doc`?
[timeless@gcc2-power8 doc0]$ perl6 --doc Devel::Trace 21:06
Could not open Devel::Trace. Failed to stat file: no such file or directory
avuserow_ so last night, I used perl6 to do parsing of a multi-gigabyte file (mostly line-by-line but with rather large lines) and it worked great. I haven't implemented the interesting logic yet but it's still awesome that it's feasible with perl6 :) 21:07
21:10 setty1 left 21:12 araujo_ left, araujo_ joined
parabolize timeless: Everything in www.irccloud.com/pastebin/7QQPo3fe/ seems to work for me. I had to add a `=begin pod` line but otherwise all the formatting seems to work. 21:17
timeless: I used perl6 --doc=HTML NaN.pod > NaN.html 21:18
21:19 cdg joined 21:21 cdg left, cdg joined
parabolize timeless: I think `p6doc Devel::Trace` does what you want. 21:24
21:24 brosenberg joined 21:25 ufobat left 21:27 firstdayonthejob joined, jonadab left 21:30 Emeric left 21:32 raiph left 21:38 vendethiel left 21:39 jonadab joined 21:45 vendethiel joined 21:47 zakharyas left
tbrowder avuserow_: did you compare times between your Perl 6 file reader and a Perl 5 version? My readers show a 10+ times greater elapsed time for the Perl 6 vs Perl 5 on the same file. 21:47
21:50 rurban joined 21:51 itaipu joined 21:52 jjido left, trnh joined
avuserow_ tbrowder: nope. certainly a Perl 5 version would be faster to execute, but then I'd need to rewrite a grammar into something that P5 could handle, and I'm not terribly interested in doing that. Doing a line-by-line version with some cheats is fast enough for my purposes. 21:53
22:00 jjido joined 22:07 cpage_ left, jonadab left 22:08 sjn left 22:11 mr-foobar left 22:30 ilogger2 joined, ChanServ sets mode: +v ilogger2
dalek c: cbaf749 | coke++ | doc/404.pod:
Update language on 404.
 22:32
Heuristic branch merge: pushed 160 commits to doc/makefile by coke 22:33
AlexDaniel m: ‘🙂’.uniname.say 22:35
camelia rakudo-moar c7cd00: OUTPUT«SLIGHTLY SMILING FACE␤»
AlexDaniel m: ‘🙼’.uniname.say
camelia rakudo-moar c7cd00: OUTPUT«VERY HEAVY SOLIDUS␤»
AlexDaniel wow, how heavy
Xliff Is there a unicode symbol for "Heavy Water" ?? 22:40
🙂
VERY HEAVY LIQUIDUS 22:41
22:42 stevieb9 joined
Xliff So here's a trick question. I have a C-lib that returns pointers to CStructs. However, most of this is done via CPointers. 22:42
If created using P6. We want to use classes based on the CStructs. However, how can we equate that to returned CPointers from the C-Lib? 22:43
I wish I had targetted code to show, but you can look at the tests here: github.com/Xliff/p6-XML-LibXML/blo...nts-port.t 22:44
Based on the classes
github.com/Xliff/p6-XML-LibXML/blo...ML/Attr.pm 22:45
github.com/Xliff/p6-XML-LibXML/blo...ML/Node.pm
Which are based on the structs found here: github.com/Xliff/p6-XML-LibXML/blo...Structs.pm
I have a feeling the simplistic approach I have gone with for the additions to Node.pm and Attr.pm may require nqp. 22:46
22:54 vendethiel joined 23:00 jack_rabbit joined 23:01 tardisx joined 23:04 maddingu1 joined, Gruber joined, jargan joined
stevieb9 does anyone here that I see regularly committing to perl6 use it at $work in production? If so, in what capacity? If not, do you know anyone who does? again... for what? 23:04
23:05 mithaldu__ joined, Lucas_One__ joined 23:07 obfusk_ joined, a3r0_ joined, klapperl joined, masak_ joined, atta_ joined
b2gills stevieb9: You could look through the irclog if you are really interested, I remember at least one person mentioning that they did 23:07
23:07 moritz_ joined, ponbiki_ joined, hcit_ joined, Jonis_ joined, bhm_ joined, gypsydav15 joined, saaki_ joined
ugexe i've had contracts where i've put perl6 into production. all parsing related 23:07
23:07 bonsaikitten joined, obarb joined, rudi_s_ joined, sunnavy joined, b^_^d joined 23:08 [Coke]_ joined, lucs_ joined, jnthn_ joined, Ulti_ joined, lnx_ joined, ambs_ joined, mre_ joined, jstimpfle joined, smash_ joined, ab5tract_ joined, kshannon_ joined, perlpilo1 joined, jdv79_ joined
stevieb9 b2gills: thanks, I'll do that 23:08
23:08 f3ew_ joined, leedo_ joined, Woodi_ joined, breinbaa1 joined
stevieb9 ugexe: I'm just a newb, but I can already see the benefits of perl6 parsing. I'm wading and playing with the basics, but regex looks extremely potent in perl6 vs perl5, and perl5 is no slouch 23:09
23:10 yeltzooo9 joined, emdashcomma_ joined 23:11 sjohnson` joined, remmie2 joined
stevieb9 what I mean is the way you can build regexes... not particularly regex capabilities themselves 23:11
23:11 exodist_ joined, gensym_ joined, remmie2 left 23:12 m0ltar_ joined, Zoffix joined, musca` joined, remmie2 joined, Bucciarati_ joined 23:14 ambs left, mithaldu_ left, tailgate left, exodist left, kshannon left, Ulti left, atta left, gensym left, klapperl_ left, sjohnsen left, f3ew left, saaki left, emdashcomma left, Jonis left, gypsydave5 left, mre left, jsimonet left, jstimpfl1 left, rudi_s left, bhm left, renormalist left, lnx left, yeltzooo left, smash left, moritz left, ab5tract left, mithaldu__ is now known as mithaldu_
Zoffix weeeeeeeeeeeeeeeeeeeee 23:14
yoleaux 14:22Z <AlexDaniel> Zoffix: please review github.com/perl6/doc/commit/d03ad1b472
23:14 gensym_ is now known as gensym, sjohnsen joined, matt joined, remmie2 left
Zoffix stevieb, I use it + a couple of Perl 5 modules (via Inline::Perl5) at work for scraping webpages. I'd use it much more if it had a web framework on the quality comparable to Perl 5's Mojolicious (most of my programming at $work is web-related) 23:14
23:14 matt is now known as Guest31458, remmie joined
stevieb9 at first, the amount of what I'd call 'symbolism' in perl6 appeared intimidating at first, but as I slowly get time to play around, much of it so far works logically. for instance the hyper operator 23:14
23:14 Guest31458 is now known as matt_
stevieb9 Zoffix: that's awesome to know. You've mentioned a bit before about what you use at work. I did not know you used perl6 at all though. I assumed that was your own time like me :) 23:15
23:16 as joined
stevieb9 my goal is, by the end of July, to have my first perl6 distribution (do we call them dists here?) included, and as I work on a second, keep patching the first with more perl6-isms as I learn 23:16
Zoffix yeah, dists
stevieb9 cool
I've already built a dist in perl6, but clearly not ready to be included yet, however I'm getting the process. 23:17
23:19 drforr1 joined
stevieb9 I've been going through docs one-by-one (Types, at this time and bouncing around), and making notes in a scripts repo of the concepts I wouldn't conceive of to force it into memory. That also allows me to spot any issues in the docs, which, of course, I will patch 23:19
23:19 mtj_ joined, risou joined, tailgate joined, jsimonet joined 23:21 tomboy64 joined
stevieb9 today, I printed a couple of copies of the Types mapping image. One is pinned to the wall in my office, the other at home. It really helps. I think it should be at the top of the doc, but that's something I'd ask more experienced folks about before I changed. That's not like a typo or something ;) 23:22
23:22 jargan is now known as jast 23:27 eam joined
stevieb9 Zoffix: if you're still located where your github profile says you are, I'd really like to take you out for dinner or something downtown next time I fly home. You've been helpful across the board for years 23:29
Zoffix Thanks, but I don't like interacting with people in meatspace :) 23:32
stevieb9 lol, fair enough 23:33
I'm trying to convince #p5p to do the next conference in Calgary, as I've never been to one, and would really like to meet some perl people in person. 23:35
my sales pitch: AB is in recession, so there's likely deals on hotels to be had
Zoffix The Weekly says "TimToady removed a warning when creating an Enum from a Range."... What exactly does Enum from a Range look like? 23:36
stevieb9 15 years into perl, figure I could get involved physically
23:43 kerframil joined, huggable joined 23:49 cpage_ joined 23:50 musca` is now known as musca 23:51 cdg joined 23:56 cdg left, pochi joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!