»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.
samcv which is why it's 153 times slower than nqp::index 00:00
MasterDuke could try turning on tracing, or line coverage
samcv it does call nqp::index but idk what else it does...
MasterDuke or do a perf profile 00:01
samcv what i assumed though is that moarvm would compile the regex. and then when ran it would just have to do a few ops 00:02
00:02 mcmillhj joined
samcv but that was just my assumption 00:02
perf is so weird. i did perf record -a. then run `perf annotate perf.data` and it tells me no symbols 00:06
i'll try perl6-callgrind-m
00:06 mcmillhj left 00:08 pilne joined
MasterDuke did you build moarvm with --debug=3? 00:09
samcv yes
it seems to load if i just do `perf annotate` but shows a lot of cpu instructions
MasterDuke i do `perf record -g --call-graph dwarf`
and then usually `perf report --call-graph=none --no-children` 00:10
samcv ah ok now i have some info 00:11
though i do get: Failed to open /tmp/perf-16739.map, continuing without symbols
none of this stuff looks like stuff i can fix 00:12
MasterDuke what are the top things? 00:13
samcv gist.github.com/samcv/3b0c2c2dbce9...192ff00294
MVM_string_index is 0.01% of the cpu usage
00:15 samcv left, samcv joined 00:16 cdg joined, pierre_ joined
MasterDuke it's interesting to compare to putting that same string as a literal in the regex 00:19
samcv well it's a pretty long string but i could do that
MasterDuke no, $a
00:19 char_var[buffer] left
samcv oh yeah 00:20
00:20 cdg left
MasterDuke none of the frame and alloc and gc functions in the top 00:21
samcv well this seems to be a pretty big problem
and at least it makes me very happy that my work hasn't been in vain :P
just a problem with having it be a variable
6 00:22
00:22 wamba left 00:32 skids left 00:36 leont left, skids joined 00:45 Cabanossi left 00:46 newb2 joined
newb2 The documentation embedded in the perl6-readline file lib/Readline.pm contains 00:46
numerous instances of the construction 'C<something>' where "something" is a perl6 scalar,
function name or function parameter; plus a few instances of the construction 'L<something>' where
"something" is one or more words of ordinary English. What do these two things represent? 00:47
samcv they are Pod6
docs.perl6.org/language/pod
00:47 Cabanossi joined
samcv C<blah> is like `blah` in markdown. and L< > is for linking things 00:47
newb2 C<blah> executes the blah instruction? 00:48
samcv MasterDuke, can you help me with this? gist.github.com/ecacbea31a89e14d5d...d41dfe118e
m: gist.github.com/ecacbea31a89e14d5d...d41dfe118e
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3https:7⏏5//gist.github.com/ecacbea31a89e14d5d0b1d
expecting any of:
colon pair
samcv evalable6, gist.github.com/ecacbea31a89e14d5d...d41dfe118e
evalable6 samcv, It looks like a URL, but mime type is ‘text/html; charset=utf-8’ while I was expecting something with ‘text/plain’ or ‘perl’ in it. I can only understand raw links, sorry.
geekosaur newb2, markdown, not shell. displays as code 00:49
MasterDuke evalable6: gist.github.com/samcv/ecacbea31a89...g_bench.p6
evalable6 MasterDuke, Successfully fetched the code from the provided URL.
Cannot test 593fa5f87fcc97cf3cbafd62f34b3119c9b1fbe3 (Commit exists, but a perl6 executable could not be built for it)
geekosaur like github
samcv good job evalable6
MasterDuke committable6: HEAD^^ gist.github.com/samcv/ecacbea31a89...g_bench.p6 00:50
committable6 MasterDuke, Successfully fetched the code from the provided URL.
MasterDuke, ¦HEAD^^: «Test 1: 0.0059511 s␤Cannot resolve caller Real(Block: ); none of these signatures match:␤ (Mu:U \v: *%_)␤ in block <unit> at /tmp/Jx3rPKvSs6 line 4␤ «exit code = 1»»
samcv this error makes no sense to me
i'm not sure what is triggering this.
MasterDuke maybe try adding a `->` before the block you pass to test-it? 00:51
samcv ok the problem is $b2 00:53
my $b2 = { 0 ≤ $_ }; # if i pass $b1 to both tests then it works
it works fine if i do it manually though
m: my $b2 = { 0 ≤ $_ }; $b2(-1).say; $b2(10).say
camelia False
True
00:57 cdg joined 00:59 Sgeo left 01:01 cdg left
MasterDuke samcv: `$rtrn = &code;` need to add () to &code 01:01
newb2 samcv: OK now understand what these constructions are for in the Pod 'language' 01:05
samcv: But "p6doc Readline" doesn't render C<something> at all, just strips away the 'C<' and the '>'
samcv: Is there a way to render Pod documentation into something other than text, say ps or pdf?
samcv: Otherwise I see no purpose for these constructions
samcv you can render to markdown 01:06
01:06 raschipi joined
samcv i've contributed some of the code for the Markdown generator. but POD is meant so you can intersperse documentation and code 01:06
01:08 mst joined
newb2 Such as the markdown.deb 'markdown' ? 01:08
samcv newb2, install Pod::To::Markdown (github.com/softmoth/perl6-pod-to-markdown)
well there is a PDF and html render too
newb2 Very good.
samcv Pod::Render contains an executable that makes it easy to convert from Pod to html or markdown or pdf 01:09
and depends on the different modules which render as those formats and then you only have to learn one command and not have to work with the other modules manually since they have different semantics 01:10
newb2 samcv: Sounds better than Pod::to::Html {{:-/) 01:13
Am investigating now .. Thanks 10000x
01:13 newb2 left 01:19 cdg joined 01:24 cdg left 01:41 ijneb left 01:46 ilbot3 left 01:51 ilbot3 joined, ChanServ sets mode: +v ilbot3 01:52 pierre_ left 01:54 cdg joined 01:55 bstamour joined
AlexDaniel MasterDuke: huh? HEAD is broken? 01:56
e: say 42
evalable6 Cannot test 593fa5f87fcc97cf3cbafd62f34b3119c9b1fbe3 (Commit exists, but a perl6 executable could not be built for it)
AlexDaniel I deleted the build to give it another go, but both travis and appveyor say that it is broken on HEAD 01:57
geekosaur m: say 1
camelia 1
geekosaur not sure what's up there
AlexDaniel or, maybe not… travis is still yellow actually 01:58
c: HEAD^ say 42
committable6 AlexDaniel, ¦HEAD^: «42»
AlexDaniel ah, I see 02:01
NQP revision in rakudo was bumped before the right commit to nqp was pushed 02:02
so both appveyor and whateverable failed
e: say 42
evalable6 42
AlexDaniel there we go
(and camelia does not update so often so by the time it started building it everything was fine) 02:03
mystery solved!
mspo is 08 out, then? 02:05
02:06 bstamour left, pierre_ joined 02:07 ufobat_ joined
AlexDaniel mspo: yes, more than two days ago actually 02:07
02:10 ufobat left
AlexDaniel m: say $*ARGFILES.slurp 02:15
camelia »Wann treffen wir drei wieder zusamm?«
»Um die siebente Stund‘, am Brückendamm.«
»Am Mittelpfeiler.«
»Ich lösche die Flamm.«
»Ich mit«

»Ich komme vom Norden her.«
»Und ich vom Süden.«…
02:22 Prgall joined 02:27 ShalokShalom left 02:31 pierre_ left 02:35 pierre_ joined, benchable6 left, committable6 left, bloatable6 left, coverable6 left, evalable6 left, Prgall left, noganex joined 02:38 noganex_ left 02:43 Sgeo joined 02:44 MasterDuke left 02:45 Cabanossi left 02:46 MasterDuke joined 02:47 Cabanossi joined, maettu joined 03:00 pgal_pug joined, BenGoldberg left 03:08 pgal_pug left 03:13 pgal_pug joined 03:41 araujo left 03:43 raschipi left 03:44 Sgeo left 03:45 bloatable6 joined, ChanServ sets mode: +v bloatable6, benchable6 joined, bisectable6 joined, coverable6 joined, committable6 joined, evalable6 joined, unicodable6 joined, ChanServ sets mode: +v evalable6, ChanServ sets mode: +v unicodable6, releasable6 joined, statisfiable6 joined, ChanServ sets mode: +v statisfiable6 03:49 Sgeo joined 04:00 Cabanossi left 04:01 Cabanossi joined 04:02 espent left 04:10 cdg left 04:23 AnotherNick joined 04:43 xtreak joined 04:54 xtreak left 04:56 skids left 04:58 skids joined 05:09 mr-foobar joined 05:18 snarkyboojum joined 05:26 u-ou- joined
u-ou- hi 05:26
I had too much coffee
05:26 khw left
teatime I know that feel, bro. 05:27
u-ou- is there a tool like perlbrew which will let me install 6 in my home directory without sudo? 05:28
teatime yar, rakudobrew 05:30
rakudo.org/how-to-get-rakudo/#Insta...tar-Source 05:31
u-ou- thank you
05:35 xtreak joined, espadrine joined 05:36 skids left 05:38 pilne left 05:39 xtreak left 05:41 hhhh joined, u-ou- left, pgal_pug left 05:42 hhhh left 05:43 pierre_ left 05:45 Cabanossi left 05:46 Cabanossi joined 05:50 nadim joined, u-ou- joined
timotimo 7i believe perlbrew can also build rakudo for you 05:52
05:53 xtreak joined 05:56 u-ou- left 05:57 xtreak left 06:00 renormalist joined 06:01 u-ou- joined 06:05 nadim left 06:07 ufobat_ left 06:10 cdg joined 06:14 cdg left 06:15 wamba joined 06:26 AnotherNick left 06:27 darutoko joined 06:29 nadim joined 06:31 lowbro joined, lowbro left, lowbro joined 06:33 kshannon joined 06:38 dakkar joined, pierre_ joined
stmuk_ prepares to fly to .ch 06:39
06:39 u-ou- left, stmuk_ is now known as stmuk 06:41 u-ou- joined 06:44 Cabanossi left 06:46 Cabanossi joined 06:47 eroux joined, domidumont joined 06:52 domidumont left 06:53 domidumont joined, u-ou- left 06:55 snarkyboojum left 06:58 espadrine left 07:01 pierre_ left, lizmat left, astj left 07:02 astj joined 07:04 astj left, astj joined 07:22 ShalokShalom joined 07:23 AnotherNick joined 07:28 snarkyboojum joined 07:29 snarkyboojum left 07:37 ufobat_ joined 07:40 llfourn joined 07:41 abraxxa joined
El_Che hello perl6! 07:48
07:57 rindolf joined 08:05 nattefrost joined, jonas1 joined 08:09 TEttinger left, cdg joined 08:10 AlexDaniel left, cdg left 08:11 emeric joined, cdg joined, xtreak joined 08:12 cdg_ joined 08:15 Cabanossi left, cdg left, xtreak left 08:16 Cabanossi joined, cdg_ left 08:18 zakharyas joined 08:21 mr-foobar left, darutoko left 08:22 cdg joined 08:23 mr-foobar joined 08:24 xtreak joined 08:25 mr-foobar left 08:26 cdg left 08:27 mr-foobar joined 08:28 xtreak left 08:29 piojo joined 08:35 domidumont left 08:37 leont joined, wamba left 08:38 wamba joined
piojo p6: my $indent=2; say so '__x' ~~ / ^ <{'_' x $indent}> x $ /; 08:40
camelia True
piojo p6: my $indent=2; say so '__x' ~~ / ^ {'_' x $indent} x $ /;
camelia False
piojo The need to use <{}> to include code in a regex is undocumented, I think--is this a bug? 08:41
I found it mentioned on a blog, saying that braces alone should work
moritz piojo: more documentation would help, yes 08:42
<{ ... }> is an assertion; if it returns a false value, this regex branch fails
{ ... } just executes code inside a regex, no side effects on the success of the regex (or regex branch) 08:43
piojo moritz: so code braces only work within assertions?
moritz piojo: no, both { } and <{ }> work; they do different things
as I tried to explain above :(
piojo moritz: I think I get it--{} will execute code, but not interpolate 08:44
and assertions do interpolate.
(with braces, I mean)
moritz they don't interpolate; the result is cast to Bool, and if it's False, the mach fails 08:45
but if the code for example returns a "a", then that does *not* mean there needs to be an "a" in the string to match
m: say so 'abc' ~~ /a { say "in code block"; True } / 08:46
camelia in code block
True
moritz m: say so 'abc' ~~ /a { say "in code block"; False } /
camelia in code block
True
piojo my $indent=2; say so '__x' ~~ / ^ <{'_' x $indent}> x $ /;
my $indent=2; say so '__x' ~~ / ^ <{'not_' x $indent}> x $ /;
err, wrong paste
moritz oh
piojo p6: my $indent=2; say so '__x' ~~ / ^ <{'not_' x $indent}> x $ /;
camelia False
moritz m: say 'abc' ~~ / <{ 'b' }> /
camelia 「b」
moritz piojo: I'm sorry, I was wrong; what I thought about was <?{ ... }>
piojo moritz: no worries 08:47
moritz <{ ... }> does interpolate
08:47 grumble left
piojo moritz: So I gather the rule is that assertions interpolate, but naked regex text does not? 08:47
No, it's more complicated than that 08:49
moritz you mean if you use a variable in a regex?
then it depends on the type of the object in the variable 08:50
piojo just {}
moritz {} just runs code
piojo oh, that's interesting
whoever writes your documentation is a saint, by the way
moritz thanks (on behalf of the doc writers) :-)
there's more than one
piojo yeah, I thought there would be
the task of putting everything together in an organized fashion is something I can't comprehend 08:51
(I can explain things, but to include everything without turning it into a thousand pages of bullet points? Unimaginable)
moritz the key realization is that when you start, you can't possibly catch everything 08:52
even my book on regexes won't capture everything that is to know about Perl 6 regexes, and it's going to be 150+ pages
08:54 grumble joined 09:00 Cabanossi left 09:01 Cabanossi joined 09:03 mr-foobar left 09:05 mr-foobar joined
nadim make it bigger! or in two tomes. I'll buy both. 09:07
pretty please
piojo moritz: wow. That's intriguing. 09:08
moritz nadim: I understand. Problem is that I want to finish eventually, too :-) 09:10
if I set as my goal to cover everything that can be covered, I run into analysis paralysis 09:11
09:12 lizmat joined
nadim moritz: I understand it's a lot of work :) I'm just being greedy ;) 09:13
piojo moritz: I hope you'll be covering [stacks of]? state, like what's required to parse an indentation-based language
though I'm gonna try to study and learn that tonight
09:13 cdg joined
moritz piojo: I do cover stacked symbol tables, and consider a python parsing example too 09:13
piojo I found one blog entry that demonstrates the technique, though its descriptions are very high level so it's a bit of a struggle to understand: strangelyconsistent.org/blog/parsin...ented-text 09:14
moritz ah, masak++'s blog
piojo oh, I didn't realize. I'm pretty new around here... 09:15
09:15 darutoko joined
piojo moritz: that sounds like a fun example. being able to parse *real* data without needing to write a traditional parser is... well, it'll be liberating 09:15
nadim tools are good too, like this one in P5 search.cpan.org/dist/Regexp-Debugger/ there are entries in the issues that are wishes, some of them good I hope (I wrote them) 09:16
09:17 cdg left 09:21 dakkar left 09:22 dakkar joined 09:32 xtreak joined 09:36 parv joined, xtreak left 09:50 wamba left 09:54 wamba joined 09:58 AlexDaniel joined 10:10 gregf_ joined 10:14 Cabanossi left 10:16 Cabanossi joined 10:17 wamba left 10:20 wamba joined 10:24 wamba left 10:25 wamba joined, ShalokShalom_ joined 10:28 ShalokShalom left 10:31 Zoffix joined
Zoffix Jotted down some from-sauce instructions that avoid rakudobrew and all its issues and just create a bash alias instead. 10:35
huggable: sauce
huggable Zoffix, Install untested latest development version of Rakudo from source: github.com/zoffixznet/r#table-of-contents
Zoffix Though—just like rakudobrew by default—that still builds a dev commit, rather than latest tested release. 10:36
We could use rakudobrew-light. Something that's only for installing a single version (and updating it once in a while to latest release), so it avoids rakudobrew's issues, while keeping the convenience. 10:38
timotimo: I don't see an option to build rakudo in perlbrew. Is that some new feature? 10:41
it got cperl in it tho
10:47 zakharyas left 10:49 sena_kun joined
sena_kun . 10:49
10:50 lizmat left
Zoffix .hugu 10:52
huggable hugs you
10:52 dakkar left
sena_kun m: my $buf = Buf.new(1); say $buf; say "{$buf}"; 10:57
camelia Buf:0x<01>
Cannot use a Buf as a string, but you called the Stringy method on it
in block <unit> at <tmp> line 1
sena_kun shouldn't the result be the same? 10:58
10:59 Cabanossi left 11:01 Cabanossi joined
Zoffix sena_kun: no, say() calls .gist, while interpolation calls .Str 11:05
m: Buf.new(1).Str
camelia Cannot use a Buf as a string, but you called the Str method on it
in block <unit> at <tmp> line 1
Zoffix m: Buf.new(1).gist.put
camelia Buf:0x<01>
sena_kun m: my $buf = Buf.new(1); say $buf; say "{$buf.gist}";
camelia Buf:0x<01>
Buf:0x<01>
sena_kun great. 11:06
thanks!
Zoffix Any time.
11:09 leont left 11:16 ShalokShalom_ is now known as ShalokShalom 11:22 atroxaper joined
atroxaper Hello! 11:26
p6: so (%(1, 2), %(1, 2), %(3, 4)).unique.elems ~~ 2
camelia WARNINGS for <tmp>:
Useless use of "so " in expression "so (%(1, 2), %(1, 2), %(3, 4)).unique.elems ~~" in sink context (line 1)
atroxaper p6: so (%(1, 2), %(1, 2), %(3, 4)).unique.elems == 3 11:27
camelia WARNINGS for <tmp>:
Useless use of "so " in expression "so (%(1, 2), %(1, 2), %(3, 4)).unique.elems ==" in sink context (line 1)
moritz m: say (%(1, 2), %(1, 2), %(3, 4)).unique.elems == 3
camelia True
atroxaper Yeah.
How can I change that to get 2 instead of 3 ? 11:28
11:28 domidumont joined
moritz m: say (%(1, 2), %(1, 2), %(3, 4)).unique( as => *.perl).elems 11:30
camelia 2
moritz docs.perl6.org/type/List#routine_unique 11:31
atroxaper Hm... Thanks. I had to go in documentation. Sorry :)
Zoffix No need to be sorry :)
11:33 parisba_ is now known as DrParis
nadim %(1, 2)i === %(1, 2) 11:33
Zoffix There's also a `with` argument you could give a different op to use to compare with (e.g. &[eqv]), but that gets dog slow quickly if you have a lot of items
nadim m:(%(1, 2)i === %(1, 2)).say
evalable6 False
nadim m:(%(1, 2) eqv %(1, 2)).say 11:34
evalable6 True
Zoffix atroxaper: note that not everything can have .perl called on it. Lazy lists, for example, will throw
nadim there is an example with @list.unique(:with(&[eqv])) 11:35
atroxaper Zoffix: good point. I see. 11:37
Thanks you all :)
Zoffix Any time! 11:38
atroxaper One more question. Is there documentaion about "Perl6 Memomy Model"? Like in Java. I see docs.perl6.org/language/concurrency but it is more like ConcurrencyFramework but not memory model. I what to know what I have to hide on lock eth. And where is read/write lock?! xD 11:40
nadim Not an answer to you question but have you seen this? www.jnthn.net/papers/2015-yapcasia-...rrency.pdf there's also a video for it if you google 11:45
atroxaper I will watch then. 11:50
11:50 zakharyas joined 11:53 parv left 12:00 Cabanossi left 12:01 atroxaper left, Cabanossi joined 12:16 Skarsnik joined
Skarsnik Hello 12:17
Zoffix \o 12:20
Is it me you looking for?
Skarsnik me? no lol 12:21
me? no lol 12:22
Zoffix Skarsnik: www.youtube.com/watch?v=PPQNbTPb-F0
Oh wait, that's the wrong song ^_^ 12:25
www.youtube.com/watch?v=62XB9IbMnxQ there :) 12:26
12:26 AlexDaniel left
Skarsnik I wanted to work on a nativecallable bot but I forget to commit one file in gptrixie. I have to wait to get back at home now x) 12:26
I wanted to work on a nativecallable bot but I forget to commit one file in gptrixie. I have to wait to get back at home now x) 12:27
Zoffix :)
Skarsnik: for some reason all of your lines get duplicated in the chat: irclog.perlgeek.de/perl6/2017-08-24#i_15065068
Skarsnik Hm blame the web client of freenode on a very unreliable connectivité 12:29
I could maybe try an irssii on a term but the connectivity is bad
12:32 aborazmeh joined, aborazmeh left, aborazmeh joined 12:37 andrzejku left 12:39 Skarsnik left 12:43 skarsnik joined 12:44 skarsnik is now known as Skarsnik 12:49 mcmillhj_ joined 12:56 epony left, olinkl left, olinkl joined 12:59 Cabanossi left
nadim Skarsnik: nativecallable? something new? 13:00
13:01 Cabanossi joined 13:02 doc_ joined
Skarsnik Well, just a bot that call GPTrixie to produce the perl 6 code definition corresponding to C definition 13:02
13:02 domidumont1 joined
doc_ Executing Perl 6 within Rmarkdown so that the resulting html output will inculde perl 6 output with the input code. See question here: stackoverflow.com/questions/458579...-rmarkdown 13:03
Skarsnik like nativecallable, char *piko(int a, char* b); and it give you the NC code to call it
13:03 cdg joined
Zoffix wow cool! :D 13:04
13:04 domidumont left 13:07 Skarsnik left, Skarsnik joined
perlpilot A module that does that along with a small utility that can be run on C code would be awesome. :) 13:08
13:10 kerframil joined
Skarsnik gah I get disconnected and irssi lost me the backlog 13:10
13:12 cdg_ joined
Skarsnik well if you want to generate defintion from C headers you can use github.com/Skarsnik/gptrixie 13:12
(does not work currently since since I forget to commit a file x)
but generating perl6 code from C code (not declaration) I don't know if there is an easy way 13:13
13:15 cdg left 13:16 aborazmeh left 13:18 aborazmeh joined, aborazmeh left, aborazmeh joined 13:19 aborazmeh left, doc_ left
perlpilot yeah, I dream a little bit sometimes. gptrixie is awesome though. Skarsnik++ 13:21
Skarsnik I can try relooking at stuff that work with llvm/clang 13:24
13:26 leont joined, mr-foobar left 13:28 mcmillhj_ left 13:29 mcmillhj_ joined 13:30 mr-foobar joined 13:32 domidumont1 left 13:33 domidumont joined 13:42 MilkmanDan left 13:44 MilkmanDan joined
Zoffix Seems Spreadsheet::XLSX doesn't wanna get used via Inline::Perl5. Dies with "Type Inline::Perl5::Object does not support associative indexing" when trying to compile the example from synopsis 13:45
13:56 mr-foobar left 14:06 leont left
Zoffix Managed to workaround it by making a wrapper Perl package that extracts stuff I want and returns an arrayref 14:10
14:11 mr-foobar joined
Skarsnik a perl5 class could not be binded? 14:12
14:12 geekosaur left
Zoffix doesn't understand the question 14:13
14:14 DerAlex joined
Zoffix it doesn't look like the Inline::Perl5::Object exposes the blessed Perl type, so you can't poke at a blessed hash with .AT-KEY for example 14:14
14:14 leont joined
Skarsnik I know nothing about I::Perl5 ^^ I was just trying a wild guess 14:15
are all perl5 objects just a weid blessed hash? 14:16
14:17 bdmatatu joined
perlpilot Skarsnik: no, not all. But many (most?) are. 14:18
tadzik some are a weird blessed array %) 14:19
Skarsnik I stopped doing perl5 stuff like 10 years ago. was like perl 5.10?
Zoffix Skarsnik: you can make an object by blessing any ref, but most objects are hashes, just due to convenience of mapping attribute <-> hash key 14:20
14:21 geekosaur joined 14:25 skids joined
nadim Skarsnik: why not patch www.swig.org/? I used it a lot a few years ago for perl5, I think it still has a lot of users. 14:27
14:30 Kyo91_ joined 14:31 Levex joined
Levex ⚛️ ⚛️ ⚛️ 14:31
14:31 eroux left
Zoffix u: ⚛️ ⚛️ ⚛️ 14:32
unicodable6 Zoffix, U+0020 SPACE [Zs] ( )
Zoffix, U+269B ATOM SYMBOL [So] (⚛)
Zoffix, 11 characters in total: gist.github.com/20f73c9a23c34c5e96...9cad96ce16
Zoffix Nothing but the atom renders for me :)
14:33 Skarsnik left
Levex :-) 14:34
14:35 wamba[m] joined 14:37 jonas1 left 14:38 flatwhite joined 14:39 Skarsnik joined
Skarsnik re? 14:40
<join to #perl6 was synced in 74 secs> fun time 14:41
14:42 kerframil left 14:44 khw joined 14:45 Cabanossi left 14:46 Cabanossi joined 14:47 Skarsnik left, nadim left 14:51 mst left 14:52 mst joined, mst left, mst joined 14:54 kannan joined 14:55 raschipi joined
Zoffix . 14:55
Well, it took only 2 hours of downloading stuff and learning R language, but I managed to answer doc_'s question above. #Success :) 14:56
japhb Nice. In the notebook space, do we have integration with Jupyter working? ISTR someone was working on that .... 15:10
sena_kun japhb, working - dunno, but you can look at github.com/bduggan/p6-jupyter-kernel 15:15
maintainer says it's working fine.
15:19 zakharyas left 15:21 ShalokShalom left, lowbro left 15:26 epony joined 15:28 margeas joined
japhb sena_kun: Thanks! 15:42
15:42 kerframil joined, koto joined 15:44 Cabanossi left, sena_kun left 15:45 flatwhite left 15:46 Cabanossi joined 15:48 nadim joined 15:55 cdg_ left 15:57 abraxxa left
timotimo japhb, it might be broken with very latest net-zmq 15:59
koto It's already fixed.
timotimo cool
i was on a car ride for a long time
now i'm feeling sick
15:59 Skarsnik joined
Skarsnik re hello :) 16:00
@Home
raschipi Skarsnik: please name your bot trixie. 16:04
Skarsnik lol
16:05 AlexDaniel joined, ShalokShalom joined 16:07 sena_kun joined
Xliff Got a segfault for t/03-service.t in p6-jupyter-kernel 16:08
sena_kun: You guys did a nice job! ;)
sena_kun Xliff, zmq version?
16:09 koto left
Xliff The latest via panda. 16:09
sena_kun nope, I mean, zeromq library version.
whatever_package_manager_you_use --key-to-see-package-info zeromq
Xliff 4.2.1 16:10
sena_kun Bang!
Xliff ?
Need to downgrade to 4.1, then?
sena_kun I wrote that we support 4.1 and 4.2 doesn't work. For me either.
Yes.
Or send me a patch.
Xliff I'll try to track it down... *sigh*
sena_kun Because it seems like a ATMOSPHERIC BLACK SUICIDE METAL to fix that, tbh. 16:11
Xliff I honestly think that some of the explicitly managed references are at fault.
sena_kun It can be, but it not very obviously what exactly... And it works like a charm with 4.1. :/ 16:12
Xliff Yeah. Most likely a memory management issue.
Or perhaps 4.1 had a memory leak that was fixed in 4.2 and rakudo ends up doing a double free? 16:13
sena_kun github.com/zeromq/libzmq/releases - here is the changelog.
Xliff I dunno... I'd have to look. And gods, I hope the problem is not in libzmq
"This release introduces new APIs, but it is ABI compatible with 16:15
libzmq 4.1.2 and up."
Suuuure.
16:15 geekosaur left
Xliff Oh well, time for a nap. 16:15
sena_kun o/ 16:16
16:20 robertle joined, cdg joined 16:22 geekosaur joined, domidumont left 16:29 st_elmo joined
nadim Hi, I started 8 threads on my 4 core machiineand timed it, it took appx 1/4 of the time of running them separately. then I put everything in a ^50 loop and I am surprised to see that at the 50th iteration it is half as fast. is that a known behavior? 16:32
ugexe how do you know it was 8 threads 16:35
nadim good question, I use start 8 times. the first results come 4 times faster, so it is at least 4 threads, if P6 started the other threads or not, I don't know. 16:38
I actually wnated to see how that can be controlled but there was nothing in the "start" documentation", I gett the number of threads in the pool is controlled somehwere else 16:39
ugexe m: say (await (^5).map({ start { $*THREAD.id } })).unique.elems
camelia 3
ugexe yes you would use $*THREAD to create threads directory, which is generally discouraged (but useful for poking around) 16:40
s/directory/directly/
nadim I get 3 too, I need to check if I really had it run 4 times faster 16:41
yes 4 times faster 16:42
ugexe m: say (await (^50).map({ start { $*THREAD.id } })).unique.elems
camelia 15
ugexe now do you see?
nadim that it starts threads as necessary? yes 16:43
ugexe you don't notice the difference in threads used vs loop count?
16:44 Cabanossi left
ugexe of course a loop of 5 (which can essentially do 5 threads and parralelize the entire thing) is faster than trying to spawn more threads than the limit of 16 16:44
nadim I always do 8 starts, but i loop that 50 times 16:45
ugexe unless you are awaiting those 8 starts before iterating again, same thing
nadim I await of course 16:46
nopaste.linux-dev.org/?1160980
16:46 Cabanossi joined
nadim the loop time varies between 0.25 to 0,40, gradually going up 16:47
nopaste.linux-dev.org/?1160981
ugexe i see. does it get slower if you loop more, or does it eventually top out? 16:48
nadim 2 secs
after 100 iterations it stabilizes around 0.5s halp the speed of the first iteration 16:51
I try 500 iterations
16:51 bstamour joined
nadim It looks quite stable after a few hundred iterations, maybe a very slight degradation. I guess the garbage collector may also take some cpu time but 50% less effect between the 1st and the 100th iteration is a lot. 16:56
jnthn: any ideas or explanations? 16:57
16:58 bstamour left, mr_ron joined 16:59 domidumont joined, nattefrost left 17:02 bstamour joined 17:03 lizmat joined
mr_ron m: grammar g { token DIGIT { <[0..9]> }; our token DIGIT is export { <.g::DIGIT> } }; say so "1" ~~ /<g::DIGIT>/; import g; say so "2" ~~ /<DIGIT>/ # working way to export grammar token but don't see in roast - IS THERE BETTER / MORE STANDARD WAY? 17:08
camelia True
True
[Coke] docs: I think I am going to start adding labels for docs, so we end up with doc-missing, doc-clarify (probably as two tags, actually, so we can still easily find all doc tags, but then I can skip "clarify") 17:09
(and hopefully someone with more fortitude than I can deal with the "clarify" ones, the majority of which seem to be asking for very specific things but not actually providing the text for them.) 17:10
sorry, specific is the wrong word. particular is better. 17:11
17:14 Cabanossi left
mr_ron github.com/perl6/roast/blob/master...hods.t#L38 suggests it may be valid. Any objection to doing it that way? 17:15
17:16 setty1 joined, Cabanossi joined 17:18 grumble left, grumble joined 17:22 Sgeo left
Zoffix .ask AlexDaniel is quotable's database available somewhere? Does it include timestamps? I'm curious to extract data about what first timers—nicks who never been on the channel before—ask about. 17:32
yoleaux Zoffix: I'll pass your message to AlexDaniel.
AlexDaniel .
yoleaux 17:32Z <Zoffix> AlexDaniel: is quotable's database available somewhere? Does it include timestamps? I'm curious to extract data about what first timers—nicks who never been on the channel before—ask about.
AlexDaniel one second…
Zoffix: alright, what about this? gist.github.com/AlexDaniel/183ed52...0db27143e2 17:34
Zoffix: it should create a couple of folders with .json files in them 17:35
I think should be trivial to process them
Zoffix AlexDaniel: thanks. 17:36
AlexDaniel Quotable takes these json files and puts all of this stuff into one file with fields separated by nulls
(so that there's no json decoding or other unnecessary stuff, this way it's relative fast to process millions of messages) 17:37
Zoffix: does it work?
I haven't tried running it in a while…
Zoffix AlexDaniel: I'm gonna try it a bit later :) I didn't expect you to respond right awy. 17:38
Is anyone interested in adding Rakudo support to knitr? From what I understand it's a code-executer for R lang's markdown. A person here suggesting a PR to be made to to knitr's repo to add support: stackoverflow.com/questions/458579...5_45864801 17:41
AlexDaniel Zoffix: yeah, does work. It will take a while the first time you run it because there's a lot of stuff to fetch and it is not parallelized, but it works 17:48
17:48 Sgeo joined, AlexDaniel left
Zoffix cool 17:48
17:50 nhywyll joined, Sgeo_ joined 17:51 geekosaur left, domidumont left 17:52 Sgeo left 17:54 mr_ron left, AlexDaniel joined, itaipu joined 17:57 geekosaur joined 18:00 Cabanossi left 18:01 Cabanossi joined
Geth doc: 18f5adcedd | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Type/Cool.pod6
Improve Cool's description/summary

Fixes #1095
18:12
18:15 aindilis joined 18:20 geekosaur left 18:22 geekosaur joined 18:23 geekosaur left 18:24 st_elmo left 18:33 geekosaur joined 18:37 cdg_ joined 18:41 cdg left, cdg_ left
Geth doc: 52fd65b266 | (Zoffix Znet)++ (committed using GitHub Web editor) | doc/Type/Cool.pod6
Use more precise language in Cool summary

Addresses #1095
18:44
18:50 storer joined
Geth doc: 6843510b5c | (Zoffix Znet)++ | 2 files
Clearify our timezone offsets are in seconds

Fixes #1446
18:57
18:58 devmikey joined 19:00 espadrine joined 19:09 samb1 left 19:15 darutoko left 19:17 AnotherNick left 19:20 samb1 joined 19:25 aindilis left 19:31 MilkmanDan left 19:33 ryu0 left 19:44 mr-fooba_ joined 19:45 mr-foobar left 19:59 mr-foobar joined, mr-fooba_ left 20:00 DerAlex left 20:08 ShalokShalom left 20:18 dozn joined, dozn left 20:19 kannan left 20:23 kannan joined 20:30 lizmat_ joined 20:32 lizmat left 20:39 bdmatatu left 20:41 MilkmanDan joined 20:42 cdg joined 20:44 Cabanossi left 20:46 Cabanossi joined 20:50 MilkmanDan left
kannan hello friends! yesterday was a humbling deep dive on day 1 (after reading the perlintro resources) lol. i am working thru the free edition of thinkperl6 now. Glad to be here. whether perl6 is the best 1st language or not, this is the friendliest community one can see. 20:53
20:54 MilkmanDan joined 20:55 Celelibi left 20:57 ufobat_ left 21:02 Celelibi joined 21:04 AlexDaniel left, Skarsnik left, raschipi left, TEttinger joined 21:08 skids left
b2gills kannan: github.com/perl6/specs/blob/master...draft.pod6 21:11
21:14 lizmat_ is now known as lizmat
kannan b2gills : very good, thanks! 21:17
21:20 mcmillhj_ left
Xliff b2gills: Minor typo fix for S27. PR submitted. 21:21
b2gills I would like to see it moved out of “draft” status at some point 21:22
Xliff: PR's for that repository get posted to #perl6-dev, no need to notify 21:26
Xliff kk 21:29
Xliff is now playing: Dug Pinnick - If You Fuk Up
\o\ /o/ \o\
21:29 Kyo91_ left
Xliff www.youtube.com/watch?v=XNYaySOx_2...dUgPinnick 21:30
21:31 setty1 left 21:40 rindolf left 21:57 imcsk8 left 22:01 imcsk8 joined 22:13 leont left 22:15 |oLa| left 22:17 |oLa| joined 22:18 ShalokShalom joined 22:20 |oLa| left 22:23 espadrine left 22:25 leont joined 22:26 ShalokShalom left 22:27 kannan left 22:30 emeric left, nadim left 22:32 devmikey left
Zoffix We're also one of the few communities with free hugs! 22:50
22:50 skids joined
Zoffix .hug 22:50
huggable hugs everyone
Zoffix \o/
22:51 cdg left, storer42 joined, storer42 left
Zoffix Some more writing on friendliness and culture by audreyt++ medium.com/@audrey.tang/lessons-i-...f5d8107e34 pugs.blogs.com/audrey/2009/08/my-ho...gging.html 22:53
22:53 storer42 joined, storer left
Zoffix Xliff: I sent you an invite to perl6 GitHub org (surprised you didn't have one already). You can accept it on github.com/perl6/ and that'll give you direct commit access to most of our repos. 22:55
23:00 raiph joined, leont left 23:01 robertle left
Zoffix is playing Babymetal - Headbanger 23:05
www.youtube.com/watch?v=2IzR_ClTE8Y :)
sena_kun www.youtube.com/watch?v=jdobHDjcPLw 23:08
Xliff Zoffix++ 23:10
.hug Zoffix
huggable hugs Zoffix
Zoffix \o/
23:10 Xal left 23:12 Xal joined
Xliff Nice links for music BTW. 23:13
A bit more screaming than I like. That's why I listen to Pinnick (and King's X) and other bands he's in.
samcv argh more interpolate issues
Xliff There are actual vocals :)
samcv m: my $var = 'st'; say 'st' ~~ /:i $var/
camelia Nil
Xliff Although Babymetal was kyute!
samcv i need to find someone who really knows the internals of perl 6 regexes... 23:14
Xliff m: my $var = 'st'; say 'st' ~~ /:i {$var}/
camelia 「」
samcv literals are treated differently than variables
Xliff Well, that's the one workaround I've come up with.
m: my $var = 'st'; say 'st' ~~ /:i <$var>/ 23:15
camelia Nil
Xliff *shrug*
Besides, MoarVM internals give me the heebies.
I'm still trying to grok nqp.
samcv this is nqp/perl6 which has the regex engine
not moarvm stuff
Xliff See whut I mean! I didn't even know that. 23:16
There are a whole bunch of things I wanna implement in nqp.
Xliff thinks about working on a MangledHash class. 23:17
23:18 nhywyll left
Zoffix Note that there's no user-facing interface for nqp. As in: it can be changed at any time without any notice. If you use an nqp op, your code might break in the future if that op is changed, renamed, or deleted. 23:18
samcv and i need to figure out what takes /$var/ 150 times more time to do stuff than /literal/ 23:19
but the regex engine is really confusing and i have only worked on the literal part and not sure what happens for variables. and if the code for that is in rakudo or nqp... 23:20
since i see no nqp tests that have variables in the regex needle like /$this/
skids www.youtube.com/watch?v=cjB23s_mZGM <-- quite more mellow screaming, at risk of dating myself. 23:21
Zoffix nqp: my $x := "foo"; say("foobar" ~~ /$x/)
camelia foo
23:21 gregf_ left
Zoffix My wild guess would be it's in nqp. 23:21
More mellow vocals from Laura Nichol :) www.youtube.com/watch?v=0JxYsNTKG5k 23:24
Xliff skids: Ooh! Nice! 23:26
Zoffix: Eeh... Laura isn't bad. I can deal with some screaming. Hell, I listen to Sevendust 23:27
Still. I prefer my metal bass-heavy with harmonics. Either in the vocals or the guitars. Preferably both.
Zoffix :) 23:28
skids That and Clutch's Robot Hive were pretty much my last metal albums. Well 10000 days of course.
23:29 kerframil left 23:32 margeas left
skids Some guy wrote a book making the case in earnest that the very best year for music was the year I was born. 23:32
So to *really* date myself: www.youtube.com/watch?v=7gGPv9hngk4 23:33
samcv argh there's so much broken i just found 23:35
this works: my $var = 'A'; 'Ea' ~~ /:i $var/ but if i do ignoremark+ignorecase it doesn't
i'm seeing it looks like some things in the AST are not passing along the regex subtypes (i think) 23:36
Xliff \o/ HAWKWIND!
Xliff just dated himself, too.
skids++
skids samcv: Past a certain point, you don't need RT anymore to find bugs :-) 23:38
samcv i should probably add RT's though
sena_kun night coding session is over... www.youtube.com/watch?v=m7Knc3l8l9o and sleep. o/ 23:40
23:41 hunter_ joined
Zoffix \o 23:42
23:42 sena_kun left 23:43 Cabanossi left 23:46 Cabanossi joined 23:49 raschipi joined 23:54 hunter_ left 23:55 hunter_ joined, hunter_ left, hunter_ joined 23:56 ShalokShalom joined
samcv going to focus on what i can do though. like making nqp::index 6.79579631 faster when the haystack is made up of strands and the needle is only one grapheme 23:57
23:59 snarkyboojum joined