»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.
00:00 cpage_ joined
sena_kun AlexDaniel, are you here? 00:01
.tell AlexDaniel: if you have a time, can you update github.com/perl6/doc/issues/561 table? Since S26 is okay now, I assume many items were fixed and it'll be easier to search for issues with updated output. 00:06
yoleaux sena_kun: What kind of a name is "AlexDaniel:"?!
sena_kun .tell AlexDaniel if you have a time, can you update github.com/perl6/doc/issues/561 table? Since S26 is okay now, I assume many items were fixed and it'll be easier to search for issues with updated output.
yoleaux sena_kun: I'll pass your message to AlexDaniel.
00:06 sena_kun left 00:09 BrassLantern left 00:12 TEttinger left 00:13 dha left, skids joined, mcmillhj left
AlexDaniel . 00:15
yoleaux 00:06Z <sena_kun> AlexDaniel: if you have a time, can you update github.com/perl6/doc/issues/561 table? Since S26 is okay now, I assume many items were fixed and it'll be easier to search for issues with updated output.
AlexDaniel .tell sena_kun the output is very ugly because of this: github.com/perl6/doc/issues/710 00:16
yoleaux AlexDaniel: I'll pass your message to sena_kun.
00:17 TEttinger joined
AlexDaniel .tell sena_kun and the situation got even worse when another “Chat with us” link got added. The thing is, these links are on *every* page, so the tool is nice enough to warn about it every time… I will update it per your request but yeah… 00:17
yoleaux AlexDaniel: I'll pass your message to sena_kun.
00:20 BenGoldberg joined 00:22 grondilu left
AlexDaniel .tell sena_kun ah, just found that I can --suppress-temp-redirects. OK then, will update shortly 00:24
yoleaux AlexDaniel: I'll pass your message to sena_kun.
AlexDaniel .tell sena_kun ah no, I need -b option instead. Anyway, it will be OK. 00:28
yoleaux AlexDaniel: I'll pass your message to sena_kun.
jdv79 are there any decent xml libs yet? 00:32
timotimo we do have the one 00:35
one pure-perl6
AlexDaniel jdv79: what do you need exactly?
timotimo i think froggs or someone else is working on libxml binding or something?
AlexDaniel jdv79: I am asking because sometimes people attempt to parse html with XML libs. In such cases, it is better to use Gumbo 00:36
jdv79 i just remember XML was slow 00:37
and froggs wasn't done
just to look at some rss
wow. just fetching the rss feed with HTTP::UserAgent takes 8s. curl does it in .3 00:41
maybe i'll drop to I::P5
AlexDaniel jdv79: hmmm I am using HTTP::UserAgent too and it is not so slow here 00:45
or was it a very big page?
jdv79 ah, HTTP::Tinyish in .8. much better:) 00:47
00:47 mohae_ joined
jdv79 its just www.youtube.com/feeds/videos.xml?user=vsauce 00:47
only 21K. i don't know what HTTP::UserAgent is up to. 00:48
00:50 mohae left 01:09 kaare__ joined 01:16 kaare__ is now known as kaare_
dalek c: c8f8763 | (Aleks-Daniel Jakimenko-Aleksejev)++ | doc/Type/Any.pod6:
Fix “gather take” link

Newlines in links have never worked right.
01:33
01:46 ilbot3 left 01:47 kalkin- left, kalkin-_ joined, ilbot3 joined 01:56 itaipu joined 01:57 s34n_ joined
s34n_ it appears that Rakudo pre-installs a lot of default packages including DBIish and PSGI. It's interesting to me that those would be part of a "base" install 01:59
I am interested in the state of PSGI in perl6 02:01
MasterDuke Rakudo Star does, regular Rakudo doesn't 02:02
s34n_ ah 02:05
It looks like supernovus is very active in perl6 support for PSGI 02:06
Is anybody else here doing much with PSGI support? 02:07
jdv79 random warnings and segfault 02:14
lovely
02:18 kid511 joined 02:21 kid51 left 02:33 wamba joined, noganex joined 02:36 noganex_ left 02:42 holyghost left, holyghost joined
jdv79 File::Temp has a bug. can't nail it down though. 02:43
somehow $/.chars is Any at github.com/perlpilot/p6-File-Temp/...emp.pm#L38 02:45
gfldex jdv79: what is $tempfile in that case? 02:46
jdv79 even though $template is always "**********"
always default 02:47
could it be async related?..
gfldex m: my @filechars = flat('a'..'z', 'A'..'Z', 0..9, '_'); my $f = '**********'; $f ~~ s/ '*' ** 4..* /{ @filechars.roll($/.chars).join }/; dd $f; 02:48
camelia rakudo-moar f1313d: OUTPUT«Str $f = "8DsA9kS8ZO"␤»
jdv79 i'll try to golf it quick 02:49
gfldex @filechars may cause problems on case insensitive FS
if threads are involved you have to expect very odd errors 02:52
jdv79 yup. async related.
gist.github.com/anonymous/05dcaef3...47d536de70 02:53
if you comment out the start block it stops for me 02:54
same for you?
oh, i left it commented. oops.
gfldex PROCESS::<$SCHEDULER> = ThreadPoolScheduler.new(initial_threads => 0, max_threads => 1);
try this to have just one thread. If the problem goes away you will have to wait for the bug to be fixed. 02:55
jdv79 its a known bug are you saying?
gfldex yes
02:55 TEttinger left
jdv79 that seems to squelch it 02:56
where is this bug reported?
02:58 TEttinger joined
gfldex there is no bug report because golfing seams not to be fruitful but jnthn is informed 02:58
jdv79 my example is fairly small but ok 02:59
gfldex if it's quite small may be worth to rakudobug it 03:00
i have the suspicion that modules need to be involved to trigger it. It would be very helpful if you could confirm that. 03:01
jdv79 maybe the XML bug i also hit that was similar is related 03:03
03:03 bob777 joined 03:22 buggable joined 03:26 cdg joined 03:31 wamba left 03:40 kid511 left
jdv79 gfldex: doesn't require modules. rt.perl.org/Public/Bug/Display.html?id=128809 03:52
03:53 cpage_ left 04:03 pdcawley left 04:05 khw left, itaipu left 04:16 cdg left 04:17 cdg joined 04:21 cdg left 04:27 cpage_ joined 04:29 andrzejku joined
s34n_ zostay: ping 04:30
what pattern is?: identifer { ... } 04:35
that's not a sub. what is it?
is it some kind of Routine? 04:37
andrzejku hi 04:38
04:39 Cabanossi left
s34n_ andrzejku: hi 04:39
andrzejku I am looking for Perl friend 04:40
TimToady s34n_: for an ordinary identifier, that would be a sub call passing a block as the first argument, unless you're asking about regex
if it's regex, the two are not related 04:41
s34n_ ah. of course
TimToady identifier would match itself, and then the block would execute for its side effects, if any
04:41 Cabanossi joined
s34n_ is 'start' a reserved word? 04:42
gfldex m: dd &start 04:43
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared routine:␤ start used at line 1. Did you mean 'spurt', 'sort', 'sqrt'?␤␤»
gfldex looks like it
s34n_ I miss perldoc -f 04:44
gfldex m: sub start { say 'oi!' }; start;
camelia rakudo-moar f1313d: OUTPUT«oi!␤»
s34n_ is there a p6doc -f ?
gfldex but then, reserved isn't all that reserved in Perl 6
TimToady s34n_: it is if there's whitespace after it
m: say WHAT start { say "hi" } 04:45
camelia rakudo-moar f1313d: OUTPUT«(Promise)␤hi␤»
gfldex m: sub start(&c) { say 'oi!' }; start {};
camelia ( no output )
04:45 skids left
TimToady m: sub start(&c) { say 'oi!' }; start() 04:45
camelia rakudo-moar f1313d: OUTPUT«Too few positionals passed; expected 1 argument but got 0␤ in sub start at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
TimToady m: sub start(&c) { say 'oi!' }; start({})
camelia rakudo-moar f1313d: OUTPUT«Type check failed in binding to &c; expected Callable but got Hash (${})␤ in sub start at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
TimToady m: sub start(&c) { say 'oi!' }; start({;})
camelia rakudo-moar f1313d: OUTPUT«oi!␤»
TimToady you'd need the parens there to force the sub call 04:46
just as if you called if() or while()
so you can have functions that are the same name as reserved words, but they're kind of second-class citizens 04:47
you can't use listop form on them
s34n_ does start actually spawn a process? or a thread?
gfldex no
s34n_ it just creates a Promise? 04:48
TimToady it schedules something to do that might get assigned to a worker thread at some point
s34n_ "might"?
TimToady well, probably will eventually, depending on how many outstanding promises there are and how many worker threads, and whether you exit first :) 04:49
s34n_ k
TimToady m: say WHAT start { sleep 3; say "hi" }
camelia rakudo-moar f1313d: OUTPUT«(Promise)␤»
TimToady that one exits first
m: await start { sleep 3; say "hi" } 04:50
camelia rakudo-moar f1313d: OUTPUT«hi␤»
TimToady that one doesn't :)
s34n_ hm
holyghost s34n : futures for example in google 05:00
harmil I'm not sure how to read this error: Only identical operators may be list associative; since '..' and '..' differ, they are non-associative and you need to clarify with parentheses 05:01
05:01 quester joined
harmil m: say 1..3 R+ 2..5 05:02
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Only identical operators may be list associative; since '..' and '..' differ, they are non-associative and you need to clarify with parentheses␤at <tmp>:1␤------> 3say 1..3 R+ 2.7⏏5.5␤»
gfldex m: say 1..(3 R+ 2)..5 # that's what Perl 6 thinks you mean
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Only identical operators may be list associative; since '..' and '..' differ, they are non-associative and you need to clarify with parentheses␤at <tmp>:1␤------> 3say 1..(3 R+ 2).7⏏5.5 # that's wha…»
harmil gfldex: oh, it's nonsense, pure and simple. 05:03
But the error is borderline hilarious
gfldex it's very sensible actually
m: say 1..3*2;
camelia rakudo-moar f1313d: OUTPUT«1..6␤»
harmil gfldex: are you suggesting that the error "since '..' and '..' differ" is ... correct? 05:04
geekosaur different multis, I suspect
gfldex m: say 1..3 R+ 2x5
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Confused␤at <tmp>:1␤------> 3say 1..3 R+ 27⏏5x5␤»
gfldex m: say 1..3 R+ 2 x 5 05:05
camelia rakudo-moar f1313d: OUTPUT«1..55555␤»
geekosaur although if so then it's a bit LTA, but fixing that case would make for other LTA messages :/
05:05 dainis left
TimToady m: say 1 .. 2 .. 3 05:05
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Only identical operators may be list associative; since '..' and '..' differ, they are non-associative and you need to clarify with parentheses␤at <tmp>:1␤------> 3say 1 .. 27⏏5 .. 3␤ expectin…»
TimToady it's LTA for sure
.. should really be considered non-associative, probably 05:06
harmil TimToady: Great example. I'll rakudobug that before I turn in.
TimToady m: 1 but 2 but 3
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Only identical operators may be list associative; since 'but' and 'but' differ, they are non-associative and you need to clarify with parentheses␤at <tmp>:1␤------> 031 but 27⏏5 but 3␤ expecti…»
TimToady yeah, non-associatives are somehow getting treated as list associatives
m: say 1 <=> 2 <=> 3 05:07
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Only identical operators may be list associative; since '<=>' and '<=>' differ, they are non-associative and you need to clarify with parentheses␤at <tmp>:1␤------> 3say 1 <=> 27⏏5 <=> 3␤ expe…»
TimToady it's just the wrong error
m: say &infix:<cmp>.assoc
camelia rakudo-moar f1313d: OUTPUT«Method 'assoc' not found for invocant of class 'Sub+{<anon|48574256>}+{Precedence}'␤ in block <unit> at <tmp> line 1␤␤»
05:07 jcallen left 05:08 BenGoldberg left 05:10 TEttinger left
TimToady m: say &infix:<cmp>.prec 05:10
camelia rakudo-moar f1313d: OUTPUT«{assoc => non, prec => n=}␤»
TimToady m: say &infix:<..>.prec 05:11
camelia rakudo-moar f1313d: OUTPUT«{assoc => non, prec => n=}␤»
TimToady that seems right
05:12 jcallen joined
TimToady m: say 1 cmp 2 leg 3 05:12
camelia rakudo-moar f1313d: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Only identical operators may be list associative; since 'cmp' and 'leg' differ, they are non-associative and you need to clarify with parentheses␤at <tmp>:1␤------> 3say 1 cmp 27⏏5 leg 3␤ expe…»
05:13 cpage_ left 05:14 TEttinger joined
TimToady it's my own fault too; made the error better for list-infix collisions, but that made it much worse for non-list-infix collisions 05:18
s34n_ .seen zostay 05:20
yoleaux I saw zostay 1 Aug 2016 20:56Z in #perl6: <zostay> np, i haven't used bailador in a while myself, i just read the code on github, but something like what i wrote is possible
05:21 TEttinger left 05:22 TEttinger joined
AlexDaniel I want my own class that will fit everywhere a Str can fit. In other words, how can I subclass a Str? I mean, subclassing is the easy part, how do I make it work? What kind of a build method should I have or something? 05:23
05:23 CIAvash joined
s34n_ where does your subclass not work? 05:24
05:26 bob778 joined
AlexDaniel m: class MyStr is Str {}; say MyStr.new.chars # OK, but what about having an actual string in there? 05:26
camelia rakudo-moar f1313d: OUTPUT«0␤»
AlexDaniel m: class MyStr is Str {}; say MyStr.new(“test”).chars # I've tried a couple of things to make this work but nothing so far
camelia rakudo-moar f1313d: OUTPUT«Default constructor for 'MyStr' only takes named arguments␤ in block <unit> at <tmp> line 1␤␤»
05:28 bob777 left, roguelazer left, bob778 is now known as bob777
s34n_ m: Str.new("test").chars # You subclass works as well as Str works 05:28
camelia rakudo-moar f1313d: OUTPUT«Default constructor for 'Str' only takes named arguments␤ in block <unit> at <tmp> line 1␤␤»
llfourn looks like Str doesn't have its own .new 05:30
I thought it did.
CIAvash m: class MyStr is Str {}; say MyStr.new(value => “test”).chars
camelia rakudo-moar f1313d: OUTPUT«4␤»
ShimmerFairy llfourn: it does, it's just the default one :)
llfourn ShimmerFairy: that's what I meant :)
AlexDaniel CIAvash: oh wow
05:30 roguelazer joined
llfourn mmhmm. It has its own BUILD. github.com/rakudo/rakudo/blob/nom/...Str.pm#L27 05:32
(though it's a submethod so maybe that's not it)
m: "Life, the Universe, and Everything".WHY.say # -_- 05:34
camelia rakudo-moar f1313d: OUTPUT«42␤»
s34n_ that's cute. but worries me a bit 05:45
are there a lot of easter eggs like that in perl6?
05:45 cpage_ joined
llfourn I don't think there's a lot 05:45
we got rid of the Date.new == christmas by default before release I think 05:46
s34n_ what could ever go wrong, eh?
ShimmerFairy Which I still think was unnecessary :P
05:46 TEttinger left
llfourn ShimmerFairy: let's not go there :P 05:46
ShimmerFairy You started it :P :P 05:47
05:59 TEttinger joined 06:01 s34n_ left 06:02 waghso joined, NEveD left 06:03 TEttinger left 06:04 waghso left 06:07 harmil left 06:08 NEveD joined 06:12 NEveD left 06:17 ufobat joined, NEveD joined 06:18 Sgeo_ joined 06:20 Sgeo left 06:21 TEttinger joined 06:28 firstdayonthejob joined 06:30 andrzejku left 06:34 Sgeo__ joined 06:36 Sgeo_ left 06:39 firstdayonthejob left, abraxxa joined 06:40 CIAvash left 06:45 quester left 06:52 sno left 06:54 rmmm left 06:55 rmmm joined 06:59 rmmm left 07:00 rmmm joined 07:02 CIAvash joined 07:04 rmmm left 07:09 quester joined 07:11 darutoko joined
lizmat clickbaits p6weekly.wordpress.com/2016/08/01/...of-an-era/ 07:12
07:12 rmmm joined
ufobat good morning :) 07:12
clickbait++ 07:13
07:13 jargan is now known as jast
moritz lizmat++ # p6weekly 07:14
abraxxa can you subscribe via rss to p6weekly? 07:30
seems so ;)
07:36 espadrine joined, bjz joined 07:38 sno joined, Actualeyes left 07:39 zakharyas joined 07:42 pdcawley joined 07:45 rindolf joined 07:52 dakkar joined 07:53 Rawriful joined 07:54 Actualeyes joined 07:58 TEttinger left 08:05 jonas2 joined 08:14 japanoise joined 08:16 cuonglm joined, paraboli` joined, [Sno] joined 08:17 zostay left, zostay joined
cuonglm Hi, I opened a ticket on RT, it's marked as 'open' rt.perl.org/Public/Bug/Display.html?id=127370 08:17
08:17 parabolize left 08:18 Cabanossi left
cuonglm But it seems it wasn't notied much 08:18
08:18 jdv79 left 08:19 CIAvash left, sno left 08:20 jdv79 joined 08:21 tomboy64 joined, Fleurety left 08:22 Cabanossi joined, espadrine left 08:25 Fleurety joined
moritz such things happen when there are many more users than developers 08:26
cuonglm moritz: Yep, so I intend to make the patch myself 08:27
moritz cuonglm: that's great! 08:28
cuonglm moritz: Thanks. But I need someone to help me zoning the code 08:29
I have looked into IO::Handle lines() method github.com/rakudo/rakudo/blob/nom/...le.pm#L550 08:30
it sounds like I have to take a look into rakudo compiler code 08:31
Would you mind pointing me out :)
08:32 domm left, domm joined
lizmat could this be the issue that when chomping, MoarVM doesn't see the difference between an EOF and an empty line at the end? 08:33
08:38 bisectable left, committable left
cuonglm lizmat: Doing `perl6 -pe '' /proc/1/statm | od -t a` 08:38
you see perl6 produces a bulk of `nl`
moritz cuonglm: is the problem that /proc/1/statm is somehow special because it's a device, or is the problem that the contents might be binary? 08:39
08:39 movl left
moritz oh, doesn't seem binary 08:40
08:40 pierre_ joined, bisectable joined
moritz cuonglm: so, what you should do: we have some low-level primiitves for working with files, nqp::open, nqp::readlinefh and such 08:40
08:40 committable joined
moritz cuonglm: try reading the file with these primitives. If that works, you know that the problem is in rakudo. If not, it's in one of the levels below 08:41
cuonglm moritz: Thanks, I will take a look 08:42
For your previous question, /proc/1/statm isn't a device nor binary 08:43
moritz Perl 6 doesn't seem to be the only one having problems with it :( 08:44
$ file /proc/1/statm
/proc/1/statm: empty
cuonglm `/proc` is peusudo file system, /proc files appear as text file when stat()/fstat() is concerned, but they behave differently 08:45
08:45 trnh joined
cuonglm their content is dynamic, so you can't do two read() syscall on them, the second one will return 0 08:47
lxr.free-electrons.com/source/kerne...tl.c#L2138
moritz: in my experience, only perl6 goes into infinitive loop 08:48
perl, python, ruby ..., *nix shell don't 08:49
psch cuonglm: did you try node.js?
i don't know how much libuv deals with file IO, but that's what we share with node
libuv that is, not file IO :)
moritz cuonglm: so, did you try the thing with nqp::open etc.? 08:50
jnthn Also, does it work if you slurp the file, vs. reading it line by line? 08:51
cuonglm moritz: I didn't, will try it tonight (I'm in UTC +7)
08:51 tadzik joined
jnthn I'm a bit suspicious of the stat/seek stuff going on here: github.com/MoarVM/MoarVM/blob/mast...ile.c#L226 08:52
dalek sectbot/perl6-rewrite: 22079c6 | (Aleks-Daniel Jakimenko-Aleksejev)++ | / (2 files):
Perl 6 rewrite of committable

The base class was rewritten also, which means that it wont be too hard to make bisectable and benchable work.
Basically it is a 1-to-1 port, so further improvements are expected.
This bot is now fully functional. The only missing parts are:
   * commit b19c8e2
   * It does not trigger on “commit:” yet, which is not needed at this point
   given that we will be relying on perl 5 version of committable until perl 6
   version is proven stable.
08:54 quester left
cuonglm jnthn: using `$fh.slurp-rest` worked 08:54
dalek sectbot: 22079c6 | (Aleks-Daniel Jakimenko-Aleksejev)++ | / (2 files):
Perl 6 rewrite of committable

The base class was rewritten also, which means that it wont be too hard to make bisectable and benchable work.
Basically it is a 1-to-1 port, so further improvements are expected.
This bot is now fully functional. The only missing parts are:
   * commit b19c8e2
   * It does not trigger on “commit:” yet, which is not needed at this point
   given that we will be relying on perl 5 version of committable until perl 6
   version is proven stable.
08:55
sectbot: afdb8b8 | (Aleks-Daniel Jakimenko-Aleksejev)++ | / (2 files):
Merge branch 'perl6-rewrite'
AlexDaniel woah woah
AlexDaniel dalek: no need to repeat that, right? :)
cuonglm psch: nodejs works 08:56
08:56 CIAvash joined
cuonglm psch: The code I tried pastebin.com/cPx1i5f2 08:57
jnthn: After doing `slurp-rest`, I did $fh.lines, it returned 08:58
( ...)
jnthn cuonglm: OK, then perhaps try doing away with the lines I pointed at to see if that resolves it. 09:00
cuonglm jnthn: thanks 09:02
dalek sectbot: 1298585 | (Aleks-Daniel Jakimenko-Aleksejev)++ | Perl6IRCBotable.pm:
Make sure to chdir back in perl 5 base class
09:03
09:03 RabidGravy joined 09:10 rindolf left
dalek ateverable: 9fc484b | (Aleks-Daniel Jakimenko-Aleksejev)++ | / (2 files):
Repo moved

Previous name is definitely no longer relevant, but there are probably even better ways to call this repo. Suggestions are welcome!
  .oO( “battlebots”? :P )
09:10
moritz AlexDaniel: excutioner! 09:11
it, like, executes code
AlexDaniel yeah, something like that 09:12
“*able” is a bit off because there's also huggable and buggable… 09:13
Rawriful Hey guys, I'm sorta interested in perl6 but I don't really know it's strengths in terms of *why* I should use it. Can anyone point me to a blog post or tell me perl6 strengths (e.g Rust is about memsafety, scala functional, python readability/batteries included). What makes perl6 standout? 09:14
or perl in general. 09:15
AlexDaniel Rawriful: maybe this docs.perl6.org/language/faq#Why_sh..._about_it?
moritz Rawriful: excellent Unicode support, built-in grammars for parsing, high-level concurrency constructs (no direct fiddling with locks and threads anymore)
Rawriful: but all in all it's simply a language that's written with the programmer in mind: very helpful error messges, shortcuts where you need them etc. 09:16
AlexDaniel moritz: you can still fiddle with locks and threads if you want to :P
Rawriful So it's a bit more of a pragmatic lang? When you want to get things done and do it quick.
09:17 xinming left
AlexDaniel yeah 09:17
09:17 M-Illandan joined, Matthew[m] joined 09:18 xinming joined
Rawriful Awesome and that link is super helpful 09:18
Thanks guys :)
AlexDaniel Rawriful: sure. If you have any other questions, feel free to ask!
09:54 mcmillhj joined 09:59 mcmillhj left 10:08 tbrowder joined 10:15 tx0h_ joined
cuonglm If I make change to moar, must I recompile it invidually? 10:17
10:17 zacts joined
moritz yes 10:17
on the plus side, you typically don't have to recompile rakudo after you changed moarvm 10:18
unless some APIs changed
cuonglm Doing `make` and `make install` from rakudo root repo make no changes
moritz a "make install" in MoarVM should be enough
cuonglm: also, #perl6-dev exists :-)
cuonglm moritz: greate, thanks
10:18 tx0h left 10:19 tx0h_ is now known as tx0h 10:21 cuonglm left, paraboli` is now known as parabolize 10:22 cgfbee joined 10:31 bpmedley left 10:35 mcmillhj joined 10:39 canopus left 10:40 mcmillhj left 10:43 bpmedley joined 10:47 canopus joined 10:49 mcmillhj joined 10:52 felher left 10:54 mcmillhj left 10:57 labster left 10:58 felher joined 10:59 ovibos joined 11:07 kurahaupo left 11:10 felher left 11:16 sena_kun joined 11:17 felher joined 11:19 pierre_ left 11:21 wamba joined 11:26 mcmillhj joined 11:30 rindolf joined 11:31 mcmillhj left 11:36 itaipu joined
sena_kun . 11:50
yoleaux 00:16Z <AlexDaniel> sena_kun: the output is very ugly because of this: github.com/perl6/doc/issues/710
00:17Z <AlexDaniel> sena_kun: and the situation got even worse when another “Chat with us” link got added. The thing is, these links are on *every* page, so the tool is nice enough to warn about it every time… I will update it per your request but yeah…
00:24Z <AlexDaniel> sena_kun: ah, just found that I can --suppress-temp-redirects. OK then, will update shortly
00:28Z <AlexDaniel> sena_kun: ah no, I need -b option instead. Anyway, it will be OK.
11:58 lambd0x joined 12:04 itaipu left 12:14 kid51 joined 12:15 pmurias joined 12:22 maybekoo2 joined, cyphase left
sena_kun AlexDaniel, thanks! 12:24
lambd0x people is there a simpler way to iterate through a multidimensional array? So far I know how to do it like this: 12:26
my @array[2;2;2]; for 0..1 -> $a { for 0..1 -> $b { for 0..1 -> $c { say @array[$a;$b;$c];}}}
m: my @array[2;2;2]; for 0..1 -> $a { for 0..1 -> $b { for 0..1 -> $c { say @array[$a;$b;$c];}}}
camelia rakudo-moar f568db: OUTPUT«(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤»
lambd0x ?
12:27 cyphase joined
masak lambd0x: ¿ 12:27
12:28 pierre_ joined
lambd0x masak: haha 12:28
masak lambd0x: well, you could write `for 0..1 X 0..1 X 0..1 -> ($a, $b, $c)`
12:28 wamba left
[Coke] ("rakudo star does, regular rakudo doesn't) - This means there's a module that you can install that is bundled with R*, or you can install it yourself. Not that the compiler can't do it. 12:29
masak m: my @array[2;2;2]; for @array { .say }
camelia rakudo-moar f568db: OUTPUT«(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤»
masak lambd0x: that seems pretty straightforward, too ;)
lambd0x seriously?
yes, indeed.
I'm just asking for there's always a more straighforward way of doing things 12:31
moritz m: my @array[2;2;2]; for @array.pairs { say .perl }
camelia rakudo-moar f568db: OUTPUT«((0, 0, 0)) => Any␤((0, 0, 1)) => Any␤((0, 1, 0)) => Any␤((0, 1, 1)) => Any␤((1, 0, 0)) => Any␤((1, 0, 1)) => Any␤((1, 1, 0)) => Any␤((1, 1, 1)) => Any␤»
moritz lambd0x: that's an iteration over they indexes and values at the same time
m: my @array[2;2;2]; for @array.kv { say .perl } 12:32
camelia rakudo-moar f568db: OUTPUT«(0, 0, 0)␤Any␤(0, 0, 1)␤Any␤(0, 1, 0)␤Any␤(0, 1, 1)␤Any␤(1, 0, 0)␤Any␤(1, 0, 1)␤Any␤(1, 1, 0)␤Any␤(1, 1, 1)␤Any␤»
lambd0x [Coke]: moritz .pais does what?
exatcly...
moritz lambd0x: return a list of Pairs :-)
lambd0x ah
moritz: nice :)
so index and value
moritz m: my @array[2;2;2]; for @array.pairs -> :key($x, $y, $z), $value { say $value }
camelia rakudo-moar f568db: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Unable to parse named parameter; couldn't find right parenthesis␤at <tmp>:1␤------> 3rray[2;2;2]; for @array.pairs -> :key($x7⏏5, $y, $z), $value { say $value }␤»
moritz m: my @array[2;2;2]; for @array.pairs -> :key ($x, $y, $z), $value { say $value } 12:33
camelia rakudo-moar f568db: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3my @array[2;2;2]; for @array.pairs -> :7⏏5key ($x, $y, $z), $value { say $value }␤ expecting any of:␤ formal parameter␤ named parameter␤»
12:33 kid51 left
moritz m: my @array[2;2;2]; for @array.pairs -> (:$key ($x, $y, $z), :$value) { say $value } 12:33
camelia rakudo-moar f568db: OUTPUT«(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤(Any)␤»
moritz m: my @array[2;2;2]; for @array.pairs -> (:$key ($x, $y, $z), :$value) { say "$x:$y:$z => ", $value }
camelia rakudo-moar f568db: OUTPUT«0:0:0 => (Any)␤0:0:1 => (Any)␤0:1:0 => (Any)␤0:1:1 => (Any)␤1:0:0 => (Any)␤1:0:1 => (Any)␤1:1:0 => (Any)␤1:1:1 => (Any)␤»
12:33 itaipu joined
moritz I don't know if I've ever unpacked signatures two levels deep before :-) 12:34
lambd0x moritz: It's rather uncommon unless you reallyhave somthing in mind. 12:35
Thanks guys.
12:36 sena_kun left, ShimmerFairy left, kid51 joined
lambd0x moritz: the ':' in front of some gives which special meaning to the setence? 12:36
*some vars
perlpilot lambd0x: they are named parameters 12:37
lambd0x perlpilot: and what do they mean in this context? 12:40
perlpilot m: my @array[2;2;2]; for @array.pairs -> (:$value, :$key ($x, $y, $z)) { say "$x:$y:$z => ", $value } # for example :) 12:41
camelia rakudo-moar f568db: OUTPUT«0:0:0 => (Any)␤0:0:1 => (Any)␤0:1:0 => (Any)␤0:1:1 => (Any)␤1:0:0 => (Any)␤1:0:1 => (Any)␤1:1:0 => (Any)␤1:1:1 => (Any)␤»
perlpilot lambd0x: a Pair has 2 attrs, a key and a value. moritz did a structural mapping from the Pairs attrs, to named vars. 12:42
DrForr 10 people now in the Perl 6 training class @YAPC.
lambd0x yes yes...
perlpilot: Am asking for so far I've just seem and used : for adverbs no variables 12:43
*not.
perlpilot DrForr: Tell them #perl6 says "hi" :-) 12:46
DrForr I imagine most of us on the channel will be at the conference :)
masak m: say "tell them camelia says 'hi'"
camelia rakudo-moar f568db: OUTPUT«tell them camelia says 'hi'␤»
lambd0x ahahha 12:50
masak m: say get
camelia rakudo-moar f568db: OUTPUT«Céad slán ag sléibhte maorga Chontae Dhún na nGall␤»
masak leaves the newbies to chew on that one :P 12:51
lambd0x moritz: loved really your solution for my code. 1 line for 8 lines. haahaha
masak m: my @array[2;2;2] = 1..8 12:52
camelia rakudo-moar f568db: OUTPUT«Assignment to array with shape 2 2 2 must provide structured data␤ in block <unit> at <tmp> line 1␤␤»
masak ...how can I "pour" the 1..8 into the structured array?
perlpilot masak++ I was wondering the same thing a few minutes ago 12:53
masak I have no reason for asking. I'm just curious.
moritz
.oO( @array.floodfill(1..8) )
psch m: $_ = $++ for my @a[2;2;2]; say @a.perl
camelia rakudo-moar f568db: OUTPUT«Array.new(:shape(2, 2, 2), [[0, 1], [2, 3]], [[4, 5], [6, 7]])␤»
jdv79 oh its a yapc now?
perlpilot psch: not very satisfying
masak psch: that's cool for this problem, but doesn't feel like a general solution
[Coke] I thought we were supposed to say "TPC" these days. :) 12:54
perlpilot jdv79: YAPC::EU
psch oh, i agree. it's also a misleading idiom in general
it's the only thing i could think of though vOv
masak though `$_ = @oldarray[$++] for @shaped-array` gets you pretty far
but it doesn't feel very clear or idiomatic
lambd0x indeed.
moritz finds TPC about as obscure as YAPC 12:55
psch (i'm saying "for my ..." is misleading, although maybe "nuanced" is the better word)
arnsholt One issue with @shaped-array = 1..8 is whether you want column-major or row-major semantics
masak arnsholt: didn't think of that, but my instinct says row-major unless you specify otherwise
because that's how the rest of Perl 6 works
perlpilot In P5 PDL, you can say sequence(2,2,2); it would be nice to have something like that in P6
arnsholt I guess people will mostly want row-major, but there's always going to be that one Fortran programmer who expects column-major =) 12:56
masak yes, definitely
arnsholt masak: I agree, row-major is going to be 90% (at least?) of what people want
masak there might even be other use cases, except for "I'm that one FORTRAN programmer" :P
arnsholt Presumably =D 12:57
jnthn m: my @array[2;2;2] Z= 1..8; say @array
camelia rakudo-moar f568db: OUTPUT«[[[1 2] [3 4]] [[5 6] [7 8]]]␤»
perlpilot arnsholt: providing one good way and some operators to get you the other things will take you quite far I think
jnthn++ 12:58
masak jnthn++
lambd0x jnthn+++
masak that's even... logical... I *think*... :P
lambd0x :)
12:58 mcmillhj joined
lambd0x and simple. 12:58
arnsholt Ooooooh
Trust the implementor to know how =D
perlpilot I mostly think of zipping rvalues, not lvalues.
arnsholt Yeah, I'd never think to zip an lvalue 12:59
lambd0x jnthn: How did you come to think about this really...
haha
masak lambd0x: he implemented it
perlpilot: many times when Perl 5 and Perl 6 surprise me, it's by using as an lvalue something that I'd only thought of as an rvalue
arnsholt (For values of "it" being more or less everything from the run-time library down to virtual machine)
lambd0x masak: convinient. 13:00
:P
masak lambd0x: yes, it's a good way to learn about something
lambd0x Like, I've learned a lot just chatting for these 20 minutes or so.. 13:01
13:01 hwcomcn joined
perlpilot lambd0x: that's what #perl6 is all about :) 13:01
masak arnsholt: the "it" I was thinking of here was "multidim array support" :)
13:01 hwcomcn left
masak lambd0x: every day is like that here. well, almost 13:01
lambd0x perlpilot: Have to agreee.
arnsholt masak: Yeah, I know. I just felt the rest of it had to be acknowledged as well =)
lambd0x masak: Yes, most of them tend to.
13:02 hwcomcn joined
arnsholt Mostly like that. Except when it's wall-to-wall puns 13:02
perlpilot arnsholt: even that is educational ;)
arnsholt FSVO of educational, definitely
Not to mention hilarious
13:03 sena_kun joined
masak arnsholt: "wall-to-*wall* puns"? really? :P 13:03
that only happens when both TimToady and quietfanatic are here at the same time
perlpilot heh, masak++ beat me to it
masak it was a low-hanging wall 13:04
arnsholt *groan*
lambd0x perlpilot: I was kinda lost here in the beggining because all I saw was m: ... rakudo output and so on, so on... hahah 13:05
masak arnsholt: I could go on like this wall day long
masak .oO( but I walln't )
lambd0x It's possible to get used. 13:07
masak lambd0x: anyway, if you're new here... welcome to #perl6! 13:13
make yourself at home
lambd0x masak: It has been a month since I started to learn Perl6 using the rakudo compiler. 13:16
But was using mostly the docs in perl6.org. Found out here to be a better source for learning and finding good people to talk about it :) 13:17
masak aye, that's my experience too
not because of the docs -- they're nice -- but because interaction and feedback are worth a lot
lambd0x masak: :D
masak: Exatly. But considering that Perl6 is so antagonised by it's performance aspect, it's been growing nice though.. 13:19
13:19 AlexDaniel left
lambd0x *exactly. 13:19
13:19 skids joined
masak lambd0x: it's also significantly faster now than earlier 13:20
not yet blazingly fast, mind. but a lot faster.
arnsholt So much faster... 13:21
lambd0x masak: Yes, I've noticed this I started programming with rakudo perl6 of june, the july version is better.
arnsholt I remember when the JVM backend was the fastest to compile Rakudo
lambd0x arnsholt: How long ago was that? 13:22
13:22 NEveD left
[Coke] pre Moar, no doubt. 13:22
masak what's the state of JVM nowadays? 13:23
psch j: say "kinda worky"
masak I remember it ran pretty fast, once you got over startup and once everything JITted
camelia rakudo-jvm cd19db: OUTPUT«kinda worky␤»
psch oh, performance
arnsholt lambd0x: Two, maybe three years ago? Not sure, to be honest =D
psch i never check that :)
masak I guess I should use rakudo-jvm some more...
psch still too many semantic problems
13:24 smls joined
tadzik I remember nom being new :D 13:24
arnsholt Yeah, I think it's sorta functional, but suffers from lack of care and feeding since Moar took over as the primary backend
tadzik and alpha being sometimes useful
arnsholt tadzik: Me too!
And taking some time to acclimatizing to the primary branch not being master! =D
13:24 pmurias left
tadzik :D 13:25
13:25 grondilu joined, itaipu left
lambd0x masak: Moar is more complete looking at that page of thngs that are done through all the different compilers that implements perl6 13:28
forgot the link...
ahaha
*things. 13:29
perl6.org/compilers/features here. Although there's not all of the implementations I think, is there guys? 13:31
[Coke] lambd0x: there's 1 implementation, 2 backends, that's it. 13:32
13:32 cdg joined
[Coke] niecza is unmaintained and predates some major rewrites at this point. 13:32
ilmari [Coke]: 2.5 backends, no? or doesn't js count? 13:33
arnsholt JS is only NQP at the moment, I think
[Coke] ilmari: there is no js backend for rakudo yet
gfldex and it counts rather slowly :)
ilmari ah
[Coke] ilmari: there's a js backend for -nqp-
smls Is there a nicer way than { [*] |$_ } to write a lambda that takes an array of numbers and returns the product? 13:36
For the sum one can use *.sum (whatever code), but there is no method .product
lambd0x Oh, ok. Question: Why is that Parrot backend was dropped? I read some docs over the Internet saying it is very good from 6 years ago I think... 13:37
So I never undestood really the reason.. for that and the construction over nothing of MoarVM... 13:38
DrForr See the 'from 6 years ago' bit there :)
grondilu Parrot was great because it existed and was better than Pugs and Niecza 13:39
[Coke] lambd0x: parrot had different design goals than "be the best backend for perl6"
so there was some painful work at the boundaries between the systems that we were able to entirely eliminate with Moar. 13:40
tadzik its selling point was to provide a cross-language compatibility, and it abandoned that idea a some point too
[Coke] tadzik: I would say never fulfilled, but not abandoned.
tadzik it did work at some point 13:41
I don't think it was when I was around, but there was a blogpost with working examples :)
I don't remember who it was who kept it working, but he said that the rest of the team kept breaking it all the time, so he gave up on maintaining it
stmuk_ also I think there was expectation parrot would be used outside the perl community
lambd0x I see. So the purpose of the development used in Parrot was not sufficient for the goals of Rakudo Perl. 13:43
13:43 buggable left, buggable joined
lambd0x And to develop MoarVM, someone already had something in done or was completely from scratch? How long did it take until the first official release? 13:44
smls lambd0x: MoarVM was written and optimized specifically for Perl 6. Even after it had only being worked on by one person for a year or so, it was already faster (for Perl 6) than Parrot after all its years of development.
13:44 sergot joined
dalek c: a56f0e6 | (Tom Browder)++ | doc/Language/pragmas.pod6:
fill in such details as are known
13:45
c: bd42216 | (Tom Browder)++ | doc/Language/pragmas.pod6:
Merge pull request #786 from tbrowder/pragmas

Add a page for pragmas
lambd0x smls: So initially just was person was working with this backend... He did a awesome job. Who was him?
smls jnthn
lambd0x ahhh..
had to be... jnthn
hahaah
nine I thought diakopter was pretty involved, too? 13:48
smls Also, I think performance was not the only issue. As I recall, there were some unsolved problems getting Parrot to support all features that Perl 6 needs.
Maybe those problems could have been solved in Parrot, but MoarVM got there faster, and seemingly more effortlessly. 13:49
jnthn diakopter contributed a lot early on to MoarVM too, yes 13:53
I worked on it alone to get the key data structures in place, the initial GC, interpreter, object system, call frames, etc. diakopter did much of the early work on Unicode support and, once we got threads, GC orchestration, plus many other pieces, though those are the ones I remember most. 13:55
Spesh (SSA-based dynamic optimization, together with type specialization, inlining, deoptimization etc.) was planned from the start, so much of the early design thought about how to allow that also. 13:56
But it didn't actaully get implemented until rather later on, when Moar already was being used quite actively. 13:57
13:57 cdg left 13:58 cdg joined
tbrowder ref docs: on the main docs page <perl6.org/documentation/> I believe the top title (Perl 6 Documentation) ought to be a link to the canonical (official) docs page <docs.perl6.org/>. In addition, where it is referenced in the bottom doc group, left column, it should have following its name (Perl 6 Documentation) some text or highlight or 13:58
footnote indicating it is the canonical (official) doc reference fpr Perl 6.
jnthn When I started MoarVM I'd actually no idea if I could pull it off. :-) Could I really in a short amount of time make something that would compete with and go on to do better than something many folks had contributed to over the years? So, it stayed a fairly private project until I was fairly comfortable it would do OK. :-) 14:00
I had the huge advantages of knowing what Parrot's pain points had been so far, knowing what Perl 6 largely looked like (much less known when Parrot started and for some years), and the various other dynlang VMs that had popped up over the years. 14:01
So it's worth keeping in mind that I started with something of an unfair advantage. 14:02
14:02 cdg left
perlpilot jnthn: even with the "unfair advantage", I think the primary thing you had that Parrot didn't was focus on providing a backend for Perl 6. That alone made a huge difference IMHO. 14:08
jnthn perlpilot: That's also true. 14:10
At the same time, I knew I wanted to keep coupling between Rakudo/NQP and MoarVM relatively loose, so there was still an amount of API design that wanted doing :)
14:12 margeas joined
dalek c: fa83f13 | (Wenzel P. P. Peppmeyer)++ | doc/ (3 files):
doc use fatal
14:17
14:18 itaipu joined 14:19 wamba joined, pmurias joined 14:21 Sgeo__ left
lambd0x jnthn: Still, you and everyone else that afterwards touched the project...did a very good job independently and that is still something worth a lot. 14:22
for rakudo and everyone else really.. I wouldn't start reading about perl6 if it wasn't because of MoarVM. Because I got intrigued with it's name and what the hack was that basically and here am I :P 14:24
gfldex and they say we are bad at marketing ... 14:25
14:25 nowan_ left
lambd0x gfldex: hahahaah 14:25
14:28 nowan joined 14:29 trnh left, trnh joined, trnh left, trnh joined
sena_kun gfldex, hi. Maybe this was already discussed, but still. I noticed that our routines-pages automatically form links like "From /type/$type#$method". Since all our method-anchors looks like a "#method_$name", it's obviously a tons of 404. If we form links with "method_" part appended, output of checklink will be almost empty. What can be broken by such change? Here is a diff: pastie.org/private/im3aep80wjcvh8cera1vaa 14:32
stmuk_ jnthn++ # not only great tech work but good explanation
sena_kun s/looks/look/ 14:33
timotimo also, it's only because of jnthn that i could contribute anything worthwhile :3
well, okay, i could probably have contributed stuff to the JIT just from brrt's guidance, too ... 14:34
14:34 BrassLantern joined
gfldex sena_kun: looks ok to me but we need to try to be sure. We can always revert the commit if it doesn't work out. That's what the link checker in debug mode is for. 14:34
sena_kun gfldex, I'm currently running this version locally and recursively walking over it with the checklink. I will play with debug-mode a bit also, and if everything will be okay, I'll commit it then. 14:36
lambd0x Later everyone o/ 14:40
14:40 lambd0x left
arnsholt is more than usually amused by the shell script idiom `yes "$(fluctuating-command)" | head -n 2` 14:49
14:51 hwcomcn left 14:52 kurahaupo joined
timotimo uh ... does that even do what i think it does? 14:53
wouldn't it just execute the fluctuating command once and spam its output, then grab the first two lines?
arnsholt Except it doesn't! =D 14:54
14:54 anser joined, pierre_ left
arnsholt head closes the fd (in this case the pipe attached to yes's stdout) when it's read its fill, which causes the write to fail and yes to terminate 14:55
timotimo ... so?
arnsholt So it prints an infinite amount (in the Haskell sense) of the output of a run of the command, of which we print the first two =D
ilmari but why do you need to copies of the output of the command? 14:56
arnsholt So I get two identical copies of the output of fluctuating-command, rather than two almost identical ones
timotimo oh, is that what that is for
arnsholt Because it gets shovelled into a script that generates a (TeX) table for me, which needs two entries
But for this particular row, the cells are identical in pairs (because this particular experimental setup doesn't have an additional bell and whistle the rest do) 14:57
So it's actually: yes "$(make -s majority-vote | tail -n 1)" | head -n 2
Which is even uglier
timotimo ... 14:58
i sometimes use "echo" to turn multi-line text into one-line text 14:59
dalek c: f40f638 | Altai-man++ | htmlify.p6:
Now all our routines links that point to original type page have correct anchor that starts with `method_`
15:00
timotimo way cool.
arnsholt Neato!
15:00 harmil_wk joined
arnsholt timotimo: Anyways, I've one week left of dedicated time (modulo baby) for completing this last chapter of my thesis before I start my new job. I threw pretty code out the window a while ago, in favour of fully embracing the worse-is-betterness of Unix 15:01
Whatever gets the job done is my new motto!
timotimo OK :) 15:02
as long as you're not sharing that code, it's guaranteed to be not-a-problem
when you start sharing code on that basis, that's when it gets a little tricky :)
arnsholt Yeah. Welcome to academia, where getting a Makefile showing how tings are run is good documentation =) 15:04
harmil_wk So... about bug work. What's a good place to start if I have a trivial number of weekend tuits to spend?
15:06 mcmillhj left
timotimo hm, good question 15:08
sena_kun .tell AlexDaniel - I'm sorry to bother you again, but since github.com/perl6/doc/commit/f40f638965 we need checklink once again. If I'm right, we eliminated almost all output. Can you give me a right command for a checklink? I tried `checklink localhost:3000/ -X '.*.(com|org)' -r`, but it seems it has timeout for every request, because I get only one page per second. That's toooo slow. Thanks for your efforts.
yoleaux sena_kun: I'll pass your message to AlexDaniel.
arnsholt harmil_wk: The perl6 RT queue is where you want to go 15:10
I've no useful hints on how to find manageable bugs though, sadly =( 15:11
perlpilot look for ones tagged with LHF
harmil_wk Oh nice!
perlpilot: thanks
smls Unfortunately, the LHF tag doesn't seem to be used: rt.perl.org/Search/Results.html?Fo...%2C%0A%27%
3Cb%3E__LastUpdatedRelative__%3C%2Fb%3E%27%2C%0A%27__NEWLINE__%27%2C%0A%27%27%2C%0A%27%3Csmall%3E__QueueName__%3C%2Fsmall%3E%27%2C%0A%27__%3Csmall%3E%3C%2Fsmall%3E__%27%2C%0A%27%3Csmall%3E__CreatedBy__%3C%2Fsmall%3E%27%2C%0A%27%3Csmall%3E__LastUpdatedBy__%3C%2Fsmall%3E%27&Order=DESC%7CASC%7CASC%7CASC&OrderBy=LastUpdated%7C%7C%7C&Query=Queue%20%3D%20%27perl6%27%20AND%20Subject%20LIKE%20%27%5BLHF%5D%27%20AND%20%28%20%20Status%20%3D%20%27new%27%20OR%
20Status%20%3D%20%27open%27%20OR%20Status%20%3D%20%27stalled%27%20%29&RowsPerPage=50
timotimo oh lord :D 15:12
smls oop, sorry for that
timotimo buggable: rt
buggable timotimo, TOTAL: 1355, UNTAGGED: 611, BUG: 414, LTA: 90, JVM: 62, NYI: 32, SEGV: 28, UNI: 25, RFC: 24, PERF: 19, POD: 14, CONC: 11, @LARRY: 10, TODO: 9, PRECOMP: 8, GLR: 6, BUILD: 5, STAR: 4, WEIRD: 3, LTA ERROR: 3, MOARVM: 2, OSX: 2, FEATURE REQUEST: 1, CPP: 1, SPEC: 1, LIST: 1, LHF: 1, SPESH: 1, DOCS: 1, NATIVECALL: 1 Details: bug.perl6.party/1470150741.html
timotimo apparently we have both LTA and LTA ERROR?
harmil_wk I guess I'll take the @LARRY ones ;-)
smls here, click the LHF link: gist.github.com/smls/eed26639e60f1...114856f41c
arnsholt harmil_wk: rt.perl.org/Public/Bug/Display.html?id=127395 might be doable. We already have edit distance help for variable names, so should be doable to make that work for method lookup failures
timotimo and a single LHF tagged one, it seems?
oh, that's a slightly tricky one 15:13
because the object itself is allowed to handle method lookup; you'll probably want to first check if the metaobject is just ClassHOW without any roles mixed in or anything
and not replaced by the user and such 15:14
and that there's no FALLBACK, or any other such stuff
perlpilot harmil_wk: maybe start with the LTA ERROR then; they're often uncomplicated 15:15
15:15 mcmillhj joined
timotimo also, it'd be beneficial if the work done to find similar-named methods is only done if the error message actually gets output, not when the error is thrown 15:15
i don't know if we do that for others, though
15:18 anser left 15:20 itaipu left, itaipu joined, mcmillhj left 15:22 zacts left 15:24 acrussell joined
harmil_wk Do we send docbugs to the same place as rakudobugs, or is that another address? 15:25
15:25 mcmillhj joined
llfourn harmil_wk: doc bugs live on GH 15:26
github.com/perl6/doc/issues 15:27
15:27 BrokenRobot joined 15:28 s34n joined
s34n zostay: ping 15:28
zostay sup?
s34n I'm interested in your P6W efforts
BrokenRobot harmil_wk: I have a private list of ticket tags and I have these marked as "Easy"... they may or may not be easy, but take a look: gist.github.com/zoffixznet/b7e1e11...768e12c874
15:29 BrokenRobot is now known as unmatched}
harmil_wk BrassLantern: thansk 15:30
zostay cool, sadly, i can't work on it this month, but i can answer questions and try to help you
harmil_wk *thanks
llfourn: thanks
15:32 maybekoo2 left
unmatched} Or semi-private, I guess. The tag db is at temp.perl6.party/release.db and can be loaded with github.com/zoffixznet/perl6-Ticket-Trakr 15:35
15:37 cuonglm joined 15:41 khw joined 15:43 abraxxa left 15:46 trnh left
harmil_wk Aw... I'm made sad. It seems you can't pass around nqp ops as Perl6 functions ;-) 15:47
m: use nqp; my &unboxer = &nqp::p6box_n
camelia rakudo-moar 3c2849: OUTPUT«Type check failed in assignment to &unboxer; expected Callable but got Any (Any)␤ in block <unit> at <tmp> line 1␤␤»
harmil_wk That should have been "boxer" but you get the idea
jnthn no, you can't :) 15:48
harmil_wk I admit, I only have vague ideas of what I was expecting to have happen there.
jnthn They're not Perl 6 functions :)
Or anything functions :)
harmil_wk They should get autoboxed, it's the Java way! ;-)
unmatched} m: use nqp; my &unboxer = -> |c { nqp::p6box_n(|c) }; unboxer 15:49
camelia rakudo-moar 3c2849: OUTPUT«This representation (VMHash) cannot unbox to a native num (for type BOOTHash)␤ in block <unit> at <tmp> line 1␤␤»
zostay s34n, feel free to ask your questions or pm me if you need an introduction 15:51
15:55 mcmillhj left 15:57 holyghost left 16:01 djbkd left
timotimo what are "the P6W efforts"? 16:02
16:02 mcmillhj joined
zostay the software spec formerly known as P6SGI 16:03
which i may refer to as P6WAPI to avoid confusion with P6 Weekly
timotimo oh 16:04
if it was actually me who did the weekly, i'd totally rename it TPS Report
unmatched} P6WASP would sound cooler... P6 Web Application Something-something Programming
hoelzro I can see the logo already... 16:05
timotimo Perl Six Web App Supercharging Protocol
nine Web Application Server Programming?
jnthn Server Protocol, surely :P 16:06
16:07 mcmillhj left
dalek c: 473fd59 | Altai-man++ | doc/Type/ (52 files):
All header-like class definitions are written as {} instead of { ... } now. It fixes many docscompilation

I've found all definitions using grep and checked them manually, but feel free to revert if needed.
16:10
sena_kun Ah, nano never makes newlines great. 16:11
16:13 acrussell left 16:14 smls left, mcmillhj joined
jnthn
.oO( Make newlines great again! )
16:14
huf MANGA?
sena_kun jnthn, I just need to LEARN THIS MAGIT ALREADY, but laziness and command-line git with nano are so tempting. 16:16
timotimo what?
16:16 NEveD joined
jnthn Sounds a bit magital... 16:17
16:19 mcmillhj left
mst thought magit was a type of pokemon tbh 16:23
ilmari magit is very handy
huf i've heard about magotes, but never magit 16:24
16:24 lichtkind joined
sena_kun It seems there are not so many emacs-users here. (: 16:25
huf i only use it when i'm pointing out a cat in a rather poetic manner 16:26
16:30 mcmillhj joined
harmil_wk From the journal of a background character: 16:30
m: unit class Today; first rule was {do not ask "Doctor who?"}
camelia ( no output )
harmil_wk Wow, I camel-cased "Today" without even thinking. :) 16:31
cuonglm jnthn: How should I write a test only run when /proc available 16:32
jnthn cuonglm: Do you have a good way to know when it's available? 16:33
nine cuonglm: why would /proc ever not be available?
jnthn Windows? :P
nine jnthn: if that's the reason, there's also a very easy way to detect that ;) 16:34
16:34 mcmillhj left
jnthn Indeed. But anyway, if it's-avaialable { ok ... } else { skip 'No /proc on this platform' } or something like that 16:34
cuonglm nine: /proc is not available anywhere, many *nix don't have it 16:35
jnthn: thanks. And for our discussion earlier, the line you pointed out is the culprit
cuonglm working on the patch 16:36
geekosaur nine, what of available but possibly not compatible? freebsd /proc is not linux /proc
moritz write a virus that wipes out all FreeBSD installations? :-) 16:38
nine geekosaur, cuonglm: so wouldn't the better question be "is a specific feature in /proc available"? 16:39
16:40 lizmat joined, dakkar left
jonadab Didn't OpenBSD remove procfs entirely? 16:41
16:42 [Sno] left
cuonglm nine: right, so in this case, I think check the OS is Linux is best 16:42
jonadab Then what do you do when systemd subsumes the functionality formerly provided by procfs and two thirds of all Linux distros don't have /proc any more in favor of systemd? 16:43
;-)
hoelzro jj
gah
timotimo they'd probably have a good reason for that if they did that
ZZ
16:45 mcmillhj joined
harmil_wk I think the number of Linux scripts that rely on /proc would make its deprecation impratical 16:45
timotimo probably
geekosaur systemd is not going to replace /proc. systemd undoubtedly *relies* on /proc. it's the primary channel for getting information that otherwise lives only in internal kernel structures 16:49
16:49 mcmillhj left
timotimo *shrug* 16:53
16:53 mcmillhj joined
jonadab geekosaur: It was a joke. 16:55
geekosaur not especially happy about /proc's scope creep on linux, but it's better than the old freebsd solution (libkvm)
people believe all sorts of stupid things about systemd :/
timotimo that's right
moritz geekosaur: systemd eats small children for Breakfast, no? :-) 16:56
harmil_wk Someone on reddit/r/perl getting upset about Perl 6's design goals changing over time and why it hasn't been renamed...
Response, "I imagine Perl6's name will change right about the time that C++'s name changes to --D. At which point, I expect we'll become Swine 5 :-)"
unmatched} timotimo: are you dalek's bot master? Would you consider adding github.com/perl6/marketing/ to the list of repos announced on #perl6? There's not much there and it probably won't be updated often, but when it is, it'd be handy for people to see those updates to encourage them to use whatever marketing materials we store there. 16:57
timotimo i'm not the botmaster, but i do have root on the machine it's run on 16:58
unmatched} \o/
harmil_wk: which post? 16:59
harmil_wk www.reddit.com/r/perl/comments/4up...rl/d60zmrv
moritz unmatched}, timotimo: it's easier to add a web hook in the github repo 17:00
unmatched} doesn't know how :( 17:02
moritz unmatched}: I've done it for now; for the record, github.com/perl6/marketing/settings/hooks add a hook with URL hack.p6c.org:8088/dalek?t=freenode,perl6 17:03
content-type application/x-www-form-urlencoded
github.com/perl6/mu/blob/master/mi...k-push.txt has some docs 17:04
unmatched} moritz++ thanks
moritz if you have any idea about where to link this stuff so that it's easier to discover, please add it :-)
unmatched} dalek: source 17:09
:)
or dalek: help
psch .help
yoleaux psch: I'm yoleaux. Type .commands to see what I can do, or see dpk.io/yoleaux for a quick guide.
psch right, that was the other one 17:10
17:10 firstdayonthejob joined
TimToady m: my &prod = &prefix:<[*]>; say prod 1..5 17:10
camelia rakudo-moar e1c772: OUTPUT«120␤»
TimToady .tell smls you can autogen the function directly: my &prod = &prefix:<[*]>; say prod 1..5 17:11
yoleaux TimToady: I'll pass your message to smls.
TimToady (one of the last things I hacked in before christmas, as it happens)
dalek c: 423a055 | coke++ | doc/Language/pragmas.pod6:
use nbsp
17:15
c: 14a6114 | coke++ | doc/ (2 files):
remove trailing whitespace
c: 6304b24 | Altai-man++ | doc/ (10 files):
Now files that start with 'a' letter can be compiled.

just skipped this pieces of code, but it will be great if someone can find more awesome solution. Rest of the changes is trivial: empty bodies for a method signatures, catching of exceptions with formatted output and some skipping for console-line command, etc.
17:18
c: e17bf57 | Altai-man++ | doc/Type/X/TypeCheck/Assignment.pod6:
Exception output message was fixed
17:21
c: 9810f0d | coke++ | util/extract-examples.p6:
remove trailing whitespace
17:23
sena_kun Documentation repository was attacked this evening. 17:24
unmatched} sena_kun: what do you mean attacked? 17:25
17:25 Actualeyes left
TimToady actually, meta autogen was early november, so not as late as I thought... 17:26
unmatched}
.oO( *this* evening? ZOMG! We're getting attacked in the future! )
It's pretty crazy. 17:28
m: my &lulz = &prefix:<[RRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRRZ]>; say lulz <trololo>
camelia rakudo-moar e1c772: OUTPUT«((trololo))␤»
TimToady .tell smls and the reason we don't have a .product to go with .sum is that .sum can be heavily optimized for huge ranges like sum 1 .. 10¹⁰⁰⁰ whereas heavy use of factorials would probably just cache the sequence from [\*] 17:31
yoleaux TimToady: I'll pass your message to smls.
unmatched} m: say sum 1 .. 10¹⁰⁰⁰ 17:32
camelia rakudo-moar e1c772: OUTPUT«5000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000…»
unmatched} :D
17:32 jonas2 left
unmatched} m: say (sum 1 .. 10¹⁰⁰⁰).Num 17:32
TimToady m: say (sum 1 .. 10¹⁰⁰⁰).chars
camelia rakudo-moar e1c772: OUTPUT«Inf␤»
rakudo-moar e1c772: OUTPUT«2000␤»
unmatched} m: say (sum 1 .. 10¹⁰⁰⁰⁰⁰⁰⁰).chars 17:33
camelia rakudo-moar e1c772: OUTPUT«(timeout)»
TimToady it was trying :) 17:34
rindolf Hi all. 17:35
unmatched} \o
rindolf Which queries were the replies for exactly? 17:36
unmatched} ?
rindolf Ah, I think I see.
unmatched} .Num is Inf and .chars is 2000
rindolf Regarding the "m:" lies
unmatched}: ah. 17:37
TimToady it must be using a relatively stupid exponentiator to calculate 10¹⁰⁰⁰⁰⁰⁰⁰ since it's been running 5 minutes and not growing memory much 17:39
I'm beginning to wonder whether it's actually in an infinite loop 17:40
17:40 Tonik joined 17:41 margeas left
TimToady virtual memory seems to wobble around 200m without long-term growth 17:42
unmatched} m: my $x = 10¹⁰⁰⁰⁰⁰; say now - BEGIN now
camelia rakudo-moar e1c772: OUTPUT«0.04616005␤»
unmatched} m: my $x = 10¹⁰⁰⁰⁰⁰⁰; say now - BEGIN now
camelia rakudo-moar e1c772: OUTPUT«1.9618868␤»
unmatched} m: say 1.9618/0.046 17:43
camelia rakudo-moar e1c772: OUTPUT«42.647826␤»
unmatched} m: my $x = 10¹⁰⁰⁰⁰⁰⁰⁰; say now - BEGIN now
camelia rakudo-moar e1c772: OUTPUT«(timeout)»
TimToady so maybe I just need to wait another 10 minutes... 17:45
gfldex m: say $?TABSTOP 17:49
camelia rakudo-moar e1c772: OUTPUT«8␤»
unmatched} m: use MONKEY-SEE-NO-EVAL; sub p { my $t = now; EVAL "10¹" ~ "⁰" x $^a; now - $t }; say (p(0), p(1), p(2), p(3) … ∞)[0..5] 17:50
camelia rakudo-moar e1c772: OUTPUT«===SORRY!===␤Unable to deduce arithmetic or geometric sequence from 0.0056914,0.0050271,0.00517571 (or did you really mean '..'?)␤»
unmatched} aaww
17:52 acrussell joined 17:54 [Sno] joined
TimToady actually, it's probably going to take closer to an hour 17:58
18:00 BrassLantern left 18:02 JYOTI joined, JYOTI left
unmatched} And it'll give 42 as the answer? :P 18:04
18:04 holyghost joined
unmatched} m: say WHY "Life, the Universe, and Everything": 18:05
camelia rakudo-moar e1c772: OUTPUT«42␤»
sena_kun unmatched}, I was joking about big amount of commits to docs repo(which is good). My jokes today are pretty lame, so don't mind it. 18:06
harmil_wk The intersection between ops defined in Grammar.nqp, ops listed in the "Operators" doc page and ops listed in the "Routines" doc page... is a sort of fractal interface. 18:07
unmatched} harmil_wk: well, Perl 6 has an infinite number of ops :)
m: say [RRRR~] <foo bar> 18:08
camelia rakudo-moar e1c772: OUTPUT«foobar␤»
unmatched} m: say [RRRRRRRRRRRR~] <foo bar>
camelia rakudo-moar e1c772: OUTPUT«foobar␤»
unmatched} etc
Unless I'm misunderstanding what you mean :)
18:08 andrzejku joined
unmatched}
.oO( huh? fractal interface? I better nod and pretend to know what it is so I look smart... )
18:09
TimToady m: say [RRR~] <foo bar> 18:10
camelia rakudo-moar e1c772: OUTPUT«barfoo␤»
TimToady just checking... :)
harmil_wk unmatched}: yeah, but it's ops like ∩ that show up in two but not the other, notandthen that only shows up in the Grammar, etc.
TimToady m: say [RXZXZRXXXXXZZZZR~] <foo bar>
camelia rakudo-moar e1c772: OUTPUT«(((((((((((((barfoo)))))))))))))␤»
TimToady notandthen is really just there for the 'without' modifier 18:11
18:11 Merovingian joined
timotimo that's a very helpful metaop 18:12
andrzejku hi
I'm looking for perl friend 18:13
TimToady timotimo: yes, if you want more parens around something
andrzejku: we try to be friendly 18:14
unmatched} m: say [RXZXZR[&say]XXXXXZZZZR~] <foo bar>
camelia rakudo-moar e1c772: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing required term after infix␤at <tmp>:1␤------> 3say [RXZXZR[&say]XXXXXZZZZR~7⏏5] <foo bar>␤ expecting any of:␤ prefix␤ term␤»
unmatched} andrzejku: have you tried a local Perl mongers group?
andrzejku: www.pm.org/
[Coke] TimToady: NO AND THEN!? 18:15
unmatched} andrzejku: www.pm.org/groups/237.html
[Coke] suspects that movie ref may be lost on TimToady.
unmatched} got it :)
18:15 Merovingian left
andrzejku there no group in my city 18:15
:/
TimToady Dude! Sweet!
unmatched} :D 18:16
andrzejku: that's an awesome reason to start one!
18:16 zakharyas left
[Coke] TimToady++ 18:16
18:16 Tonik left
andrzejku I am not good at perl 18:16
I am even not work in Perl
:/
unmatched} andrzejku: I don't see how that precludes you from starting a Perl Mongers group. 18:17
andrzejku: seems there are even instructions for you: www.pm.org/start/
pochi m: say [&infix:<*>] 1,2,3
camelia rakudo-moar e1c772: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 3say [&infix:<*>]7⏏5 1,2,3␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement modifi…»
perlpilot
.oO( <andrzejku> I am not even passion for Perl )
TimToady m: say [[&infix:<*>]] 1,2,3 18:18
camelia rakudo-moar e1c772: OUTPUT«6␤»
pochi aha
TimToady you can't make the [] do two different things
note also that you can't extend this to any arbitrary lambda because reductions don't allow internal spaces 18:19
pochi so [[ ... ]] is a special thing, or is it that [ ... ] interprets its content special if it's in []?
andrzejku I am Perl passionate but right now I learn more
TimToady (to help distinguish them from [1, 2, 3] and such 18:20
andrzejku C++
unmatched} TimToady: so that's by design? I saw a ticket last night saying disallowed whitespace is a bug
TimToady nope
andrzejku however I want to have a good Perl friend
unmatched} andrzejku: well, you have ~300 people in here. 18:21
TimToady maybe you could find a friend and then teach them Perl?
andrzejku :)
Perl is also not popular in my country
perlpilot TimToady: what about that earlier [RXZXZR[&say]XXXXXZZZZR~] ? It looks like [&whatever] messes up the metas. Should that be?
unmatched} andrzejku: well, sounds like if you start a Perl mongers group you can both learn Perl and make it more popular! 18:22
TimToady [&say] is not a meta
TimToady channels MJD
andrzejku ye, but all of I want it is to have virtual perl friend
:/
unmatched} m: say [RXZXZRXXXXXZZZZR[&say]] <foo bar>
camelia rakudo-moar e1c772: OUTPUT«barfoo␤(((((((((((((True)))))))))))))␤»
unmatched} andrzejku: I can tell you an easy way to get on, wanna know the secret? 18:23
TimToady yes, the real operator can only come when you're done with the metas
unmatched} *one
mst what the flying fornication was that m:
unmatched} mst: just common Perl 6 :P 18:24
timotimo what kind of emoticon is m: ?
unmatched} I can't live without that operator :)
andrzejku unmatched}, yeah sure
TimToady mst: if you'd bother to learn the language, it would be obvious :P
unmatched} timotimo: "that m:" == "that eval"
timotimo ah :)
andrzejku unmatched}, tell me please 18:25
unmatched} andrzejku: help us out. Perl is a volunteer effort. You can help with the docs github.com/perl6/doc/issues you can help with tickets on rt.perl.org/ (here are some that may be easy: gist.github.com/zoffixznet/b7e1e11...768e12c874 or you can just process the ticket queue perl6.party/post/A-Date-With-The-Bu...e-Help-You ) or you can join 18:26
#perl and ask how you can help the Perl 5 effort too.
andrzejku: and I think you'll make many Perl friends in the process. 18:27
perlpilot "friends" should be in scare-quotes for #perl ;)
andrzejku yeah 18:28
mst perlpilot: wuss
unmatched} mst: it's a say() sub used as an operator that's then reversed, then zipped, then zipped, then zipped, then zipped, then crossed, then crossed, then crossed, then crossed, then crossed, then reversed, then zipped, then crossed, then zipped, then crossed, then reversed, and then used as a metaop on the list containing 'foo' and 'bar'
andrzejku right now I learn C++ because I am tired with my actually job
;s
and want to get hired somewhere what I want to do
mst unmatched}: yeah, when you and TimToady both decided to be completely unhelpful I went and hunted down the documentation :D 18:29
andrzejku but I know Perl basics since I used it from time to time
unmatched} :)
andrzejku however I have no friends and any Perl support in my job
when I write something in Perl ppl laugh at me or told me to do what others do
mst TimToady: I'm starting to learn. I was able to clearly articulate why your BUILD semantics are bad and should feel bad the other day :P 18:30
andrzejku this mean write Python and Bash
TimToady mst: yes, we've been thinking for a couple years now that a function that you could call inside BUILD that says "set defaults now" might be a useful addition
but round tuits and all that 18:31
unmatched} andrzejku: well... lead, follow, or get out of the way... I think the saying is. You can either do what you like or you can always change to what people tell you to be. 18:32
mst TimToady: I do feel like being able to opt out of the standard slot population code on a per attribute rather than per class basis (ala init_arg =>) would be even better
TimToady there were also proposals to tie defaults to the initial FETCH so it would appear as if they were already set
andrzejku unmatched}, yes I know
TimToady mst: under the function approach, you just do things before or after the function call 18:33
the reason we didn't go with the FETCH approach is because it's not clear it would work right with natives
mst yeah, it's definitely fixable, once one decides what 'fixed' means in specific, and the current system is obviously a superset of what M* permits functionality wise even if not huffman coded the way I expected
most importantly from my POV, mentally regarding perl6's BUILD as being more POPULATE made my mental model consistent again 18:34
unmatched} I rather see improvement with the parameter default handling, even if in module space. I think it's a bigger "bad and should feel bad" than BUILD 18:35
m: sub foo (:$what = 42) { dd $what }; sub bar (:$what) { foo :$what }; bar :42what; bar;
camelia rakudo-moar cc3932: OUTPUT«Int $what = 42␤Any $what = Any␤»
mst so while I'm not 100% happy with the current functionality sugar wise, I'm no longer nigglingly troubled when trying to think about it at all
unmatched} I tried to invent a trait to avoid this, but failed.
mst so, yeah, getting there, slowwwwwly 18:36
andrzejku hey how can I create website
in pm.org domain?
psch unmatched}: maybe allowing METAOP_ASSIGN works there? e.g. :(:$what //= 42) 18:37
unmatched}: might be worth a try, although you'd have to Slang it for module space
unmatched} k, I'll try to learn slangs and try the slang approach 18:38
18:38 TimToady left
psch well, i'd try it in Perl6::Grammar first, to see how badly it breaks :) 18:38
unmatched} :)
psch and you'd need to know the spots to tweak anyway with what currently passes for our Slang API 18:39
...although that probably wouldn't go away vOv
18:40 TimToady joined
harmil_wk ⊍ vs ⊎ ... I anticipate some code readability issues, there... 18:40
unmatched} psch++ I really like the //= syntax for this
harmil_wk: I don't see a problem: i.imgur.com/O2YhyfM.png :P 18:42
Ah, found the white-space-in-meta-op ticket: rt.perl.org/Ticket/Display.html?id...et-history 18:43
unmatched} rejects
harmil_wk unmatched}: well, you technically have the same problem I do, just for a different reason :-)
It's www.fileformat.info/info/unicode/ch.../index.htm vs www.fileformat.info/info/unicode/ch.../index.htm
In most fonts in most programming-friendly-scales they look the same. 18:44
psch "programming-friendly-scales"? 18:45
unmatched} Yeah, they actually do display on my IRC and I can sorta see two extra pixels on the right one :P
18:45 freeside joined
harmil_wk psch: By that I mean "you can fit a lot of text on the screen" 18:45
psch i suppose that means i'm weird with programming in 16pt? :) 18:46
unmatched} This is how I see them on "programming-friendly-scales": i.imgur.com/DRkqdVY.png 18:47
psch well, on 1920x1080 on a 21" widescreen
unmatched} codes in 1440p res... with three 80-char panels displayed at the same time.
psch putty say 146x36 terminal resolution for fullscreen 18:48
irssi usually is at 80x24, but we have lots of >80 chars lines in rakudo
dalek c: 8481062 | (Jan-Olof Hendig)++ | doc/Type/Setty.pod6:
Added docs for Setty.Bool
18:49
harmil_wk unmatched}: yeah, that's more or less what I see
psch most of the time i don't feel like "more lines at the same time" would help me having better insight into the code, fwiw
unmatched} I mostly care for "files at the same time" 18:50
JS+CSS+Markup+Backend
psch heh, yeah, my screen usually has around 15 or so windows
18:50 freeside left
psch or, well, screens :P 18:50
18:52 tmtowtdi joined
TimToady yeah, well, it's pretty hard to see the difference between 礻 and 衤 too 18:53
harmil_wk TimToady: I'd suggest that's why we don't use pictograms for programming, but someone's going to pull out an emoji operator library if I do that... 18:54
18:55 spider-mario joined 19:01 ufobat left
dalek c: 4255fa6 | (Wenzel P. P. Peppmeyer)++ | doc/Language/variables.pod6:
make variables.pod6 compile
19:02
unmatched} m: sub 礻 { $^礻 + $^衤 }; say 礻 2, 2, 19:06
camelia rakudo-moar cc3932: OUTPUT«4␤»
pochi you should name it ネ for full confusion 19:09
19:09 Woodi left
unmatched} m: sub 礻 { $^礻 + $^衤 }; my \衤 = { say 礻 2, 2 }; 衤 19:09
camelia rakudo-moar cc3932: OUTPUT«WARNINGS for <tmp>:␤Useless use of 衤 symbol in sink context (line 1)␤»
unmatched} m: sub 礻 { $^礻 + $^衤 }; my \衤 = { say 礻 2, 2 }; 衤()
camelia rakudo-moar cc3932: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Variable '&衤' is not declared␤at <tmp>:1␤------> 3 { $^礻 + $^衤 }; my \衤 = { say 礻 2, 2 }; 7⏏5衤()␤»
unmatched} m: sub 礻 { $^礻 + $^衤 }; my \衤 = { say 礻 2, 2 }; 衤.() 19:10
camelia rakudo-moar cc3932: OUTPUT«4␤»
cuonglm jnthn: I pushed the patch and creating PR for MoarVM github.com/MoarVM/MoarVM/pull/388 and roast github.com/perl6/roast/pull/137 19:13
Please take a look, thanks.
19:16 cuonglm left 19:17 zakharyas joined 19:24 jdoege joined 19:25 jdoege left, jdoege joined 19:26 rindolf left 19:30 wamba left 19:31 rindolf joined 19:33 darutoko left 19:35 zacts joined
harmil_wk unmatched}: Actually, I'd use 礻 and 衤 more as a prefix op with texas equivalents, "iscult" and "isdressed" 19:36
unmatched} Jebus... This person has 904 modules! metacpan.org/author/PERLANCAR 19:37
I guess my title of releasing copious amounts of shite has been thoroughly stolen.
harmil_wk There was a time when people were publishing to CPAN faster than new Star Trek novels were coming out! ;-) 19:39
unmatched} doesn't know enough about Star Trek to know whether the number is high or low... 19:42
mst unmatched}: he doesn't like the fact that people don't always apply patches immediately, so he just releases his own version instead 19:45
unmatched} heh 19:46
mst on the upside, he does tend to document what it's a version -of-, so idea theft is still totally viable
dalek c: e7d7d27 | (Wenzel P. P. Peppmeyer)++ | doc/Language/variables.pod6:
push slipped index entires back to where they belong
mst it's a little bit aggravating given pulling in any one of his modules can easily pull in half his author directory
unmatched} :D
mst but, like, it beats him keeping them all on a personal darkpan so 19:47
unmatched} I still remember drowning in bit rot when I had just over 200 modules. Can't imagine what it's like to maintain 900...
mst this is why I try and give mine a team as fast as possible 19:49
19:50 brrt joined 19:55 prammer joined
harmil_wk Huh, when sorting the list of Perl 6 ops, I found that Linux "sort -u" will drop any purely non-ascii lines unless you set LC_COLLATE to something (anything, even something invalid) 19:57
dalek c: c8d65cc | coke++ | doc/Language/variables.pod6:
remove trailing whitespace
19:58
jdoege I'm working my way through the documentation, trying to understand how introspection works and I have developed a couple of questions. 1) inside a grammar, if I want to reference the regex method itself in a block inside the named regex, how can I get a pointer to the regex so it can be the subject of introspection. For instance, I may want some executable code to discover the name of the regex of which it is a part. 20:03
20:03 setty2 joined 20:04 travis-ci joined
travis-ci Doc build errored. Will "Coke" Coleda 'remove trailing whitespace' 20:04
travis-ci.org/perl6/doc/builds/149289130 github.com/perl6/doc/compare/e7d7d...d65cc8dd11
20:04 travis-ci left
jdoege 2) when I tried to iterate over the methods in a grammar, it threw a warning when I tried to say the iterator but it did say the name of the regex. If I asked for the name, I got "Regex" which makes sense since the object is a Regex. But that raises the question of how do I legitimately determine the name of a named Regex through introspection without throwing a warning? 20:06
for $parser.^methods -> $i { say $i; } # does print the name of the regex but throws a warning. $i.name returns "Regex" 20:08
20:10 brrt left
gfldex jdoege: you may have luck reading the source of Grammar::Tracer 20:11
jdoege Thanks for the pointer. I'll check it out.
[Coke] is there a way to provide travis with a pre-built rakudo with teh modules needed?
(for doc)
20:12 brrt joined
hoelzro [Coke]: that would be nice - you could pre-build it and throw a tarball up on hack, and then download it in the travis job 20:13
zacts does rakudo support 32bit platforms? (and if so, will this likely be the case for a while?)
hoelzro that would be nice for module testing, tbh
20:17 lizmat left
dalek c: c2818df | (Jan-Olof Hendig)++ | doc/Type/Setty.pod6:
Spruced up keys, values and kv a bit and fixed a couple of minor errors
20:20
20:21 zacts left 20:22 zacts joined
hoelzro actually, having a nightly build available would be handy for that kind of stuff 20:23
[Coke] say, a nightly build that if it passed make test got swapped out for the last daily installed one? 20:24
20:25 zakharyas left 20:26 prammer left 20:27 prammer joined
hoelzro yeah, that'll do 20:30
20:30 labster joined
hoelzro we could even keep a ring buffer of sorts of one week of nightlies should we desire 20:30
jdoege gfldex: Thank you. That provided what I needed. 20:31
20:32 ZoffixMobile joined
ZoffixMobile zacts, yes, it does, although I don't think there's JIT on 32bit version, so it's slower. There's also no Rakudo Star version for recent releases, so you'd need to build from sauce 20:33
well, no windows 32 binaries for R* is what ai meant 20:34
zacts ah ok cool 20:35
will 32bits likely be dropped within the next 4 years?
20:41 andrzejku left, mcmillhj left 20:42 mcmillhj joined, AlexDaniel joined, brrt left 20:45 margeas joined
[Coke] we're not going to go out of our way to keep purchasing old hardware to run it on. 20:46
(if some company does, that's another thign entirely)
AlexDaniel purchasing old hardware? 20:49
yoleaux 15:08Z <sena_kun> AlexDaniel: - I'm sorry to bother you again, but since github.com/perl6/doc/commit/f40f638965 we need checklink once again. If I'm right, we eliminated almost all output. Can you give me a right command for a checklink? I tried `checklink localhost:3000/ -X '.*.(com|org)' -r`, but it seems it has timeout for every request, because I get only one page per second. That's toooo slow. Thanks for your efforts.
AlexDaniel sena_kun: I use this: checklink -b -D 25 -q doc.perl6.org | tee "$(date '+%F')" 20:51
sena_kun AlexDaniel, thanks, I will now run it with local server...
[Coke] AlexDaniel: or keeping old hardware running. same thing 20:52
AlexDaniel [Coke]: are you talking about 32-bit builds? If so, can't you install a 32-bit system on any hardware? At least that's what I thought… 20:56
20:56 brrt joined 20:57 brrt left 20:58 kurahaupo left, ZoffixMobile left 21:00 skids left
dalek c: b49dc5a | coke++ | doc/Language/5to6-perlfunc.pod6:
Close #656; document use VERSION
21:01
[Coke] AlexDaniel: can you include a timestamp at the top of gist.github.com/AlexDaniel/fc6cb8d...43ecc47c33 so we can tell when it was last updated? 21:03
AlexDaniel: let me rephrase: we're not going to go out of our way to make stuff work on old hardware or OSes. 21:04
21:04 AndyBotwin left
[Coke] we are still understaffed and are lucky we're not breaking stuff we do support. :) 21:05
AlexDaniel [Coke]: how precise you want this timestamp to be? I update the gist whenever I update this page: github.com/perl6/doc/issues/561
also, the gist itself has a “Revisions” tab with all the timestamps 21:07
dalek c: 6f34839 | coke++ | doc/Language/nativecall.pod6:
Remove invalid L<> - Pointer isn't a type
21:09
[Coke] AlexDaniel: ah, duh, revisions is fine. Thanks!
sena_kun Currently, if we resolve webchat/examples links, then almost all output will be gone. It would be actually useful to search for dead links, not just staring at huge output. 21:10
AlexDaniel sena_kun: webchat/examples warnings go away if you use -b
21:10 mcmillhj left
[Coke] AlexDaniel: line 68 of your gist; complains the the following fragment needs to be fixed (put on Line:121) 21:11
I don't see a #put anywhere in that page. 21:12
What's it complaining about, or has that one already been fixed?
sena_kun AlexDaniel, hm, it's strange, since I'm currently running `checklink -b -D 25 -q localhost:3000/ -X '.*.(com|org)' | tee "$(date '+%F')" > output_check` and I still get 5 warnings about them.
[Coke] not sure the local host is 100% the same as the version deployed to the web. 21:13
21:13 jdoege left
sena_kun Yes, it may be. 21:13
Current revision is outdated, since it was created 13 hours ago and github.com/perl6/doc/commit/f40f63...d7c4002b45 was pushed 6 hours ago. 21:14
[Coke] AlexDaniel: it's also complaining about the fragment "method_put", but that seems to be a valid id on that page. 21:15
If you can explain what those 2 are complaining about, I can probably fix a bunch of them. (but they look spurious to em.) 21:16
*me
21:16 kaare_ left, obfusk left 21:17 obfusk joined
AlexDaniel [Coke]: these are some very good questions 21:17
[Coke]: previously the output contained other stuff that looked wrong, but then the number of such warnings decreased as we fixed the HTML that we prouce
[Coke]: in this case, perhaps it is complaining about links inside the links? 21:18
<a class="u" href="#___top" title="go to top of document"><a href="/routine/put#class_Mu">method put</a></a>
sena_kun Ah, we actually need a fix.
AlexDaniel so these warnings tend to indicate that there's a problem, but in case of broken html they are actually complaining about broken html… or so it seems 21:19
however, it talks about line 121 which is this: <p>From <a href="/type/Mu#method_put">Mu</a></p> 21:21
right above the line with broken html… so I don't know
21:21 zacts left
sena_kun I'll fix up "From $type" for routines now. 21:22
Xliff Does anyone know where nqp::nativecallrefresh() is documented 21:23
21:23 prammer left 21:24 NEveD left, prammer joined
psch Xliff: i wouldn't assume it's documented. it's implemented in MoarVM/src/core/nativecall.c:676 21:25
Xliff: well, the C function that the nqp op maps to is :)
21:29 rindolf left
Xliff OK. Thanks, psch++ 21:29
psch Xliff: if it's transparent enough it'd be nice to have it doc'd in nqp/doc/ops.markdown 21:30
Xliff: we have a #nativecall header there, but that op seems to be missing
Xliff psch: I would be happy to write it since I'm making use of it, but I'm not precisely sure when it should be used. 21:32
psch Xliff: well, i have no idea what it even does. if you can describe that it'd be already notably more than what we have
Xliff All I know is that sometimes, if changes are made to a repr('CStruct'), those changes aren't reflected in the P6 object unless that op is used on it.
psch that sounds like we should be calling it internally more often 21:33
Xliff Yeah. :/
Just got done with a weird test in p6-XML-LibXML where I could make a change via nqp::bindattr and then check that change and see it, but then a few scopes up, that change would not appear. 21:34
arnsholt Structs passed as arguments to a function are checked after the function returns
psch i don't think you should ever need to bindattr in user-facing code 21:35
if so we're missing a layer of abstraction
Xliff arnsholt: Right, but say a call "$obj.somethingChangesHere" is made. Changes in $obj may not be seen in that scope unless "nqp::nativecallrefresh($obj)" is called first.
21:35 prammer left
arnsholt Depends on what the method calls and such 21:36
Xliff psch: Yeah, well I've been running into that necessity a lot, but only in complex NativeCall projects.
psch (FSDO "user-facing", obviously...)
arnsholt But the refresh only refreshes the struct in question, *not* recursively
But that holds for the op as well, so that shouldn't matter 21:37
Xliff arnsholt, yeah. Recusively. And thar be dragons... thar....
I don't know the performance hit of nqp::nativecallrefesh() so I am using it as sparingly as possible.
psch Xliff: what's a simple example of where you need bindattr? 21:39
dalek ateverable: 4811e59 | (Daniel Green)++ | Whateverable.pm6:
Minor simplification of to-full-commit()
21:40
21:41 harmil_wk left
Xliff github.com/perl6/nqp/pull/302 21:43
psch: LOL! "simple" he says... 21:44
gist.github.com/Xliff/96e48eaad469...80a73a918a
^^ psch
21:45 webstrand left
dalek ateverable: e52a215 | (Daniel Green)++ | Whateverable.pm6:
Whoops, missing semi
21:45
psch Xliff: ops.markdown is mostly compiler developer documentation. references to users calling ops feel kind of off there
21:45 webstrand joined
psch s/mostly // 21:46
okay, so it's a recursively nested struct..?
Xliff psch: Seems that way. 21:47
psch that's definitely beyond my C and NC knowledge :|
Xliff: well, the typedef references itself
i mean, i can accept that it's possible, but that's probably because C is crazy 21:48
Xliff psch++ .... no arguments there.
psch Xliff: oh, also there's an 's' too many in the op name in your PR 21:49
21:50 itaipu left
Xliff psch: Updated. 21:50
Yeah. And that that extra s is now gone. 21:51
This is what I get for using the web interface.... oh... and typing tired.
psch Xliff++
Xliff Thankee 21:52
21:52 acrussell left
dalek c: a7aab6d | Altai-man++ | htmlify.p6:
Now 'routine_'-like links are fine, just like 'method_'-like links
21:55
21:56 RabidGravy left 21:57 espadrine joined 22:07 CIAvash left 22:13 TEttinger joined 22:16 grondilu left 22:21 erl6 joined 22:25 pmurias left 22:26 spider-mario left, Sgeo__ joined, erl6 left 22:29 aries_liuxueyang left 22:30 firstdayonthejob left 22:32 setty2 left 22:33 skids joined, cpage_ left 22:40 canopus left 22:48 canopus joined, profan joined 22:49 lizmat joined 22:58 mcmillhj joined 23:04 rpburkholder joined 23:05 cpage_ joined 23:09 cpage_ left 23:11 mcmillhj left, edehont joined 23:21 margeas left 23:23 sena_kun left 23:29 cyphase left 23:31 cpage_ joined 23:39 cpage_ left 23:40 cyphase joined 23:41 ruoso left 23:50 edehont left 23:53 rpburkholder left 23:58 kid51 left