»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by masak on 28 November 2015.
AlexDaniel I wonder if there is anything similar to metacpan.org/pod/less in Perl 6? 00:00
nadim how do I tel if a variable is type object or it it is an instance of a type? 00:02
p6_nb lizmat: thanks 00:03
00:03 p6_nb left 00:04 p6_nb joined 00:06 p6_nb left 00:07 virtualsue left
cygx 'night o/ 00:07
geraud I seem to remember that @*INC was gone... Did I imagine that?
Skarsnik nop 00:08
00:08 cygx left
Skarsnik it's gone 00:08
geraud okay, so I'm sane. that's nice to know. 00:09
If I found code that used @*INC, is there a construct I should replace it with? 00:10
jdv79 geraud: i think not yet 00:11
Skarsnik $*REPO I think
but depend what you want to do
geraud Skarsnik: I just want to majke p6doc work. Right now 'p6doc Type::Str' croaks with : 00:13
Earlier failure:
Dynamic variable @*INC not found
in sub search-paths at /home/geraud/local/share/perl6/site/resources/1AF4CBA7DE6DE70558B631FA6737B5B5C96C7FFA:31
also the '1AF4CBA7DE6DE70558B631FA6737B5B5C96C7FFA' is not so fun...
the offending line is : @*INC.map({CompUnitRepo.new($_)})».IO».Str.map({ /\/$/ ?? $_ !! $_ ~ '/' }).list).flat.list; 00:14
Skarsnik what
dalek ecs: 0391e7c | (Juerd Waalboer)++ | S32-setting-library/Numeric.pod:
Provide tau constant

The Tau Manifesto <tauday.com/tau-manifesto>
See also
  <en.wikipedia.org/wiki/Turn_(geomet...oposal>
and
  <tauday.com/state-of-the-tau>
00:16
ecs: c6e70b6 | (Juerd Waalboer)++ | S32-setting-library/Numeric.pod:
Add constants τ and π (aliases for tau and pi)

π is already provided by Rakudo; a patch for τ is on its way.
ecs: 2ed4112 | (Aleks-Daniel Jakimenko)++ | S32-setting-library/Numeric.pod:
Merge pull request #103 from Juerd/master

Provide tau constant, include unicode aliases in spec
Since tau is now in rakudo, I see no problem including it in the spec.
zengargoyle has a feeling p6doc is going to be borked for a while.
geraud I just wiped every perl6-related file from my system, build rakudo, cloned panda, bootstrap'd it and panda install'd p6doc. That's all I did. 00:17
zengargoyle it relies on plain *.pod files that are installed to the old moar-nom/install/share/perl6/site/lib/ file space
can a plain *.pod file even be a CompUnit by itself? 00:18
geraud zengargoyle: is there an alternative to p6doc?
zengargoyle perl6 --doc ~/.rakudobrew/moar-nom/install/share/perl6/site/lib/Type/Str.pod | less 00:19
:P
geraud Oh, I see... :) 00:20
zengargoyle i have long held unsubstanciated worries over the common Perl 5 POD usage of things like: Module::Cookbook::Index or Module::Manual::Intro type documentations. 00:21
and it not going to be easy to just ls a directory and less a file and read the POD when things are broken. 00:22
AlexDaniel
.oO( How many people does it take to introduce τ to the language? One to prepare all pull requests, three to merge these pull requests… )
00:23
zengargoyle 'cause there's no Module.pm file anymore, its some locked up bytcode thing in a repo under a sha1 somewhere that has to be loaded up into the VM to be accessible.
autarch zengargoyle: if that stuff ships in a distro won't it just get installed like everything else? 00:25
Skarsnik AlexDaniel, 10 to agree on the spec 00:26
autarch and assuming it's installed there should be a way to write p6doc Module::Cookbook
AlexDaniel Skarsnik: well,
zengargoyle autarch: what stuff?
autarch zengargoyle: any file with pod in it
AlexDaniel Skarsnik: as far as I know the specs are not holy, but tests are :)
autarch given that the compiler actually parses pod now, pod is just more code, right?
zengargoyle right, what gets installed though is a .moarvm byte blob right.. you can't read the pod directly without perl6/p6doc working. 00:27
autarch right
it would be nice to keep a copy of the "text" version around too 00:28
I mean for everything
zengargoyle or when you get an error: died with foo at line 1 in blah... you can't actually look at blah.pm and see what's going on.
you get to hunt down the *source* of blah, install the repo or whatnot, find the right version matching the one you have installed, just to look at the source. 00:29
00:30 Psyche^ joined, Psyche^_ left
autarch yeah, sounds like a use case that needs to be thought through 00:30
zengargoyle it just breaks that niceness of dynamic stuff that the source is easily findable and look-at-able.
autarch yep
awwaiid In gist.github.com/awwaiid/7ae087830e...rse-p6-L21 if I leave in the { } I get a nice parse, and if I take it out I do not -- what is the empty { } doing? If this was a rule then I'd guess somewhitespace thing, but this is a token
(this is a whitespace-indent parser) 00:31
00:33 leont left
japhb .ask ShimmerFairy Wouldn't the strings that require escapes in your advent post require double quotes instead of single quotes? 00:33
yoleaux japhb: I'll pass your message to ShimmerFairy.
ShimmerFairy japhb: what strings? 00:34
yoleaux 6 Dec 2015 22:45Z <cygx> ShimmerFairy: I had to do some emergency surgery on your Day7 posting as something went wrong with HTML-encoding the code samples; cf gist.github.com/cygx/0ac6aa9922627cc2c2ea for what I did in case I messed up somewhere
00:33Z <japhb> ShimmerFairy: Wouldn't the strings that require escapes in your advent post require double quotes instead of single quotes?
geekosaur ewps... that first message might explain the second...
ShimmerFairy japhb: ah, yes, I see now. I guess &quot; means " then? 00:35
00:35 pierre-vigier joined
japhb ShimmerFairy: yes, IIRC you what &dquot; but it's been a while. 00:35
*want 00:36
Sheesh
ShimmerFairy so either my mangled " chars were mis-replaced, or I mistyped in the first place :P
jdv79 is panda not able to install non-eco modules anymore?
iirc that used to work, no? 00:37
japhb Hmmm, nope, I'm wrong, it's &quot ...
I was thinking of the typographical variant
Yeah, something is definitely goofy. I cut and pasted some of the code, and it is as incorrect as it appears (so it's not a goofy font issue, it's actually wrong) 00:38
Looks like raw single quotes in the HTML, not &apos; 00:39
00:39 bakedb left
japhb How are you inserting source into the WP post? 00:40
zengargoyle sorta seems like there needs to be a %?RESOURCE thing that's just a filesystem path for storage of thing the old way for stuff that if changed wouldn't break a compunit.. like docs/icons/etc.
zengargoyle will ponder more when the compunit stuff is more stable... :P 00:43
awwaiid that post is messed up yeah - like it has / &lt;:Alpha&gt;+ / 00:45
in the "Not Just Strings" section, as an example
(in the advent day 7 post) 00:46
00:46 Zoffix is now known as ZoffixW
ZoffixW .botsnack 00:46
yoleaux 6 Dec 2015 23:34Z <AlexDaniel> ZoffixW: basically I care about two things: 1) Can I do async stuff? and 2) If one of the plugins receives a message and interprets it in some way, is there any way for it to tell other plugins to ignore this message? Because when making a bot you can throw in a bunch of stuff, but you probably don't want your bot to answer twice for some message if both plugins trigger on it. That being said, sometimes you want
it…
6 Dec 2015 23:34Z <AlexDaniel> ZoffixW: implement these and I will be a happy man 00:47
00:47 ZoffixW is now known as Zoffix
yoleaux 6 Dec 2015 23:41Z <AlexDaniel> ZoffixW: maybe there has to be some sort of big dispatch thing with priorities and stuff, I don't know actually. But basically the reason I said that Net::IRC::Bot needs some reworking was because my plugins were conflicting and I was unable to find any non-ugly solution to this. One simple example: you want your bot to reply to “MyBot: !commands” message and at the same time you might want it to reply to any 00:47
other message that was
6 Dec 2015 23:42Z <AlexDaniel> ZoffixW: not handled, e.g. “MyBot: hi bot” should answer with “Please don't talk to me!”. That's just something to consider, maybe it will give you some ideas for better implementation.
:D
Zoffix AlexDaniel, thanks for the comments. I saved them and will be sure to consider them while I hammer down the API
m: sub term:<(╯°□°)╯︵┻━┻> export { say "Noes" }; (╯°□°)╯︵┻━┻ 00:50
camelia rakudo-moar f24d2e: OUTPUT«===SORRY!===␤Cannot find method 'has_compile_time_value'␤»
Zoffix m: sub term:<(╯°□°)╯︵┻━┻> { say "Noes" }; (╯°□°)╯︵┻━┻
camelia rakudo-moar f24d2e: OUTPUT«Noes␤»
Zoffix Oh
zengargoyle wow, is advent supposed to be light-grey text on white bg... terribly hard to read. 00:51
Zoffix zengargoyle, configure your monitor correctly :P
It's nearly black on my screen TBH
Titles are slightly lighter
And code examples are black (at least on this one perl6advent.wordpress.com/2015/12/...6-and-you/ ) 00:52
japhb Zoffix: It's quite possible his monitor just has a different gamma setting, which is a perfectly normal thing. Mac v. Windows gamma was a painful thing for a very long time, until image formats started specifying expected gamma (and the renderers actually honoring it) 00:54
The great thing about actual black and white (as opposed to greys) is they map to themselves no matter the gamma. :-)
Zoffix I'm not so sure about the "normal" bit. 00:55
japhb Zoffix: What OS and hardware are you using?
Zoffix If you can't see the difference between X shades of gray on your monitor, it's either a shitty monitor or it's misconfigured.
00:56 zengargo1le joined, zengargoyle left
Zoffix japhb, Bodhi Linux, GeForce GTX 970 card, Asus ROG PG278Q monitor... and the site looks fine on my $work box too, which is a Win7 on crappy hardware 00:57
japhb Zoffix: Who says zengargoyle can't differentiate many shades of gray? They may just not be the *same* shades of gray you see. Plus there's human vision; there are lots of people that have contrast issues, such that they essentially need to posterize to see it.
Zoffix Because if they were able to see that many shades of gray, there would be too many shades between the dark gray the text actually is and white :P 00:58
japhb Zoffix: Neither of those is a Mac, whose display hardware is configured differently (or it was; it's possible they adopted a different standard to match the industry, but somehow I have problems believing The Company That Steve Built doing that)
jdv79 what is up with having to scroll every code part in today's advent? 00:59
i gave up 1/2 way through:(
japhb Oh, and ... not everyone has a good monitor.
Sorry, I'm just a bit sensitive to people making assumptions about vision and graphics, given that I'm color blind, and I have friends with low acuity, contrast problems, you name it. 01:00
Zoffix I really doubt the boxes preferred by "hipster designers" render much differently than my $work box I use for design :)
I'm colourblind too.
awwaiid er... is someone going to fix the &gt; stuff? :) 01:01
Zoffix heh
zengargo1le imgur.com/59O4loH
Zoffix awwaiid, what is it supposed to be?
japhb Yes, but hipster designers often have perfect vision or even better-than-"normal" vision, and usually very good hardware, and they can have trouble realizing not everyone does.
Zoffix zengargo1le, weird. It's much darker on my screen
zengargo1le, i.imgur.com/khU5e99.png 01:03
zengargo1le i'm just going by comparison... majority of sites look OK to Fine... then advent is WTF? it's not like i complain about every single site. :P
japhb Both of those are actually *way* less contrasty than I have on my monitor. In fact, if someone hadn't mentioned it, I'd have assumed the grey was actually black. 01:04
awwaiid Zoffix: in the "Not Just Strings" section a lot of < and > are &lt; and &gt; instead 01:05
01:06 zengargo1le is now known as zengargoyle, cygx joined
japhb It appears the body text color is #666, which is supposedly a darker-than-medium gray. 01:07
cygx awwaiid: I fixed those already once (but messed up by replacing &quot; with ' instead of ", my bad)
then, ShimmerFairy fixed the quotes and re-broke < >
autarch I agree the colors on the advent calendar could be easier on the eyes 01:09
01:09 Zoffix left
autarch darkening the text would be a good step 01:09
ShimmerFairy cygx: goddammit, wordpress is the worst about technical writing :) (I was in HTML mode if it makes a difference) 01:10
jdv79 I vote for green or orange on black:) 01:11
cygx ShimmerFairy: will you take another shot, or should I?
lucs (high contrast)++
zengargoyle i read the post in my rss reader first and didn't notice any HTML entity weirdness...
autarch if we could make crayon syntax highlight perl 6 wordpress would be ok
jdv79 i haven't used crayons in years 01:12
zengargoyle and no contrast probs since the reader strips most of that away... only saw weirdness when visiting the actual page in a browser 01:13
ShimmerFairy cygx: if you can, great, otherwise I can get to it in a few minutes :)
(advent also needs a search bar, last I checked)
Or more generally, the advent site currently seems to be designed around "what's happening now?", and doesn't give any consideration to digging through its archives :) 01:14
jdv79 who does that? 01:15
i found it odd that some questions were only answered in the advent pieces 01:16
cygx ShimmerFairy: I'm on it, just a sec 01:18
awwaiid probably a lot should be lifted into doc.perl6.org 01:19
cygx: it's looking better already :)
Juerd m: tau.Rat.nude
camelia ( no output )
Juerd m: tau.Rat.nude.say 01:20
camelia rakudo-moar f24d2e: OUTPUT«(710 113)␤»
cygx ShimmerFairy: ok, that *should* now look as originally intended
'night, #perl6 - if there are still problems with the post, someone else needs to fix them ;) 01:26
01:27 cygx left
Juerd m: say (-1).Complex.sqrt 01:30
camelia rakudo-moar f24d2e: OUTPUT«6.12323399573677e-17+1i␤»
Juerd Shouldn't that be i?
Or, well, 0+1i 01:31
6.12e-17 is tiny, but not quite 0
01:32 yeahnoob joined
ShimmerFairy m: say (-1).Complex.sqrt ≅ i 01:32
camelia rakudo-moar f24d2e: OUTPUT«Type check failed for return value; expected Bool but got Slip␤ in block <unit> at /tmp/NYljak_YPz:1␤␤»
ShimmerFairy m: say )(-1).Complex.sqrt) ≅ i
camelia rakudo-moar f24d2e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/B0eQAVeg2_␤Unsupported use of bare "say"; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argument, or use &say to refer to the function as a noun␤at /tmp/B0eQAVeg2_:1␤------> 3…»
ShimmerFairy m: say ((-1).Complex.sqrt) ≅ i
camelia rakudo-moar f24d2e: OUTPUT«Type check failed for return value; expected Bool but got Slip␤ in block <unit> at /tmp/4hXctfHLnY:1␤␤»
ShimmerFairy huh, that's weird
looks like ≅ doesn't like 'i' or certain kind of Complex numbers for whatever reason 01:34
01:35 Skarsnik left
skids m: my $p = Supplier.new; my $s = $p.unsanitized-supply; $s.serial.say; $s.tap({ sleep 1; now.say}); $p.emit(42); $p.emit(43); sleep 2; # Looks like the "make supplies serial" stuff went a little overboard 01:41
camelia rakudo-moar f24d2e: OUTPUT«False␤Instant:1449452545.083135␤Instant:1449452546.087158␤»
japhb .tell cygx Looks to me like you fixed it, thanks! 01:42
yoleaux japhb: I'll pass your message to cygx.
01:50 TimToady_ is now known as TimToady, TimToady left, TimToady joined 01:51 mr_ron joined
mr_ron ShimmerFairy: say 0 ∈ «42 -5 1».map(&amp;log ∘ &amp;abs); - the &amp; entity shows through on two lines of your advent post 01:54
ShimmerFairy screw you, wordpress!
mr_ron thanks for the interesting post btw 01:55
ShimmerFairy you're welcome :) 02:00
02:03 zacts left 02:04 kaare_ joined 02:05 mr_ron left
dalek c/qualified-names: 86d7824 | (Brock Wilcox)++ | doc/Language/packages.pod:
Remove unnecessary details and misc cleanup
02:06
02:10 zacts joined 02:18 herby_ joined
herby_ Good evening, everyone! 02:18
flussence ShimmerFairy++ # that's a great article! (WP ate your footnote btw) 02:23
ShimmerFairy thanks :) (and for the love of...) 02:24
flussence cygx++ too # making &open saner
(that's one part of programming I usually avoid, because it's so *easy* to cause data loss by getting it wrong)
ShimmerFairy ponders advent using a perl6-powered blogging platform next year :P 02:25
dalek c/qualified-names: aab793f | (Brock Wilcox)++ | doc/Language/syntax.pod:
Actually provide a few random examples
02:26
02:26 gtodd left
ShimmerFairy flussence: I just realized I didn't finish the thought in the footnote, so it's good my attention had to be drawn to it :P 02:26
02:26 yqt left
SmokeMachine__ Hi! Shouldn't this work? www.irccloud.com/pastebin/TZTJDtyW/ 02:29
flussence the link? or whatever it's meant to link to? 02:30
argh, who hides a *pastebin* behind js?
SmokeMachine__ m: multi bla(--> Str){ 42.Str } multi bla(--> Int){ 43 } my Int $a = bla; my Str $b = bla; $a.say; $b.say;
camelia rakudo-moar f24d2e: OUTPUT«5===SORRY!5=== Error while compiling /tmp/c8LeIqObE1␤Strange text after block (missing semicolon or comma?)␤at /tmp/c8LeIqObE1:1␤------> 3multi bla(--> Str){ 42.Str }7⏏5 multi bla(--> Int){ 43 } my Int $a = bl␤ expecting any of:␤ …»
SmokeMachine__ flussence: ^^ that's the code 02:31
the pastern code
flussence I don't think multis are smart enough to multi on the return type like that yet.
SmokeMachine__ flussence: will they be?
m: multi bla(--> Str){ 42.Str }; multi bla(--> Int){ 43 }; my Int $a = bla; my Str $b = bla; $a.say; $b.say; 02:33
camelia rakudo-moar f24d2e: OUTPUT«Ambiguous call to 'bla'; these signatures all match:␤:( --> Str)␤:( --> Int)␤ in block <unit> at /tmp/prcwRvqDeb:1␤␤»
zacts b 1
TimToady SmokeMachine__: no, it is impossible without time travel 02:34
yoleaux 6 Dec 2015 21:30Z <[Coke]> TimToady: if it's late enough to update our HEAD revision to v6.c
SmokeMachine__ TimToady: So, let's start working on time travel? Just kidding... but why is it impossible? 02:35
TimToady the context is not known at the time of the return
flussence that'd basically be reintroducing perl5's wantarray with all the complexity of perl6 smartmatching
TimToady you could bind it to something half an hour later
the best you can do is return an object that behaves polymorphically depending on context 02:36
s/poly/allo/
and while (as Perl 5 demonstrates) it's possible for a compiler to pass some context information in, in Perl 6 this would basically make multiple dispatch to the next call impossible 02:38
SmokeMachine__ but, is it needed to know the context on time of the return? shouldn't the context to be needed on time to choose the function?
TimToady yes, it's basically just never gonna work 02:39
suppose you call your bla function, and then want to call something else that differentiates on Str vs Num, but you don't know what you got back if you returned both things :) 02:40
multiple dispatch can only work in one direction, and Perl 5 kinda chose the wrong direction, in my (deeply humble) estimation 02:41
SmokeMachine__ but if i know the type of the variable that I got the result...
skids .tell jnthn Questions that arose while updating S17 gist.github.com/skids/fabd0304f40e24db33d0 02:42
yoleaux skids: I'll pass your message to jnthn.
TimToady it would be possible to predict occasionally, but it's not worth it
SmokeMachine__ TimToady: thanks for your the explanation 02:43
ShimmerFairy SmokeMachine__: if you called bla in any context that didn't specify a constraint (e.g. my $a = bla , that is without the type), it wouldn't be able to know. C++ has the same issue, if you're curious (C++ doesn't differentiate on return type either)
02:45 ilbot3 left
SmokeMachine__ ShimmerFairy: IMHO, that would be an error... but ok, I understand... 02:46
02:46 ilbot3 joined
SmokeMachine__ ShimmerFairy: I mean: call in Any context should be an error (again IMHO) but it should (IMHO) work when getting the return on the "right" var type... 02:48
(and probably a compiler error) 02:49
but ok, thanks for listening...
ShimmerFairy that's of course a possible design choice, but the issue in general is that there's no constraint on where to use function calls like there is for arguments (that is, the language enforces constraints on arguments, but doesn't/can't enforce constraints on where to call functions), as far as I understand it :)
SmokeMachine__ ShimmerFairy: thanks! 02:50
ShimmerFairy The way I saw it worded when I looked this up in C++-land went something like "because nothing says you _have_ to do anything with the return type, so there's no way to reliably differentiate on it" 02:51
SmokeMachine__ ShimmerFairy: that makes sense... 02:52
ShimmerFairy s/return type/return value/, but mostly the same meaning :P
02:54 kaare_ left
ugexe m: use Test; subtest { if 1 { return 1; }; }; # curious what changed recently that results in this no longer working? 02:57
camelia rakudo-moar f24d2e: OUTPUT«Attempt to return outside of any Routine␤ in block <unit> at /tmp/Q_s6Z1jVm4:1␤␤»
TimToady what is a subtest? 02:58
I think jnthn just made return work correctly, is all 02:59
return has always been specced as purely lexotic, but rakudo gave up on did a dynamic return, which is against spec, so jnthn++ fixed it
*on that and did
ugexe yea im not saying it was correct before, just curious what the change was. subtest is from Test.pm, and it takes a &code 03:00
TimToady if there isn't an explicit Routine declaration, return can't see it
I mean, it'll keep looking outward for one 03:01
and if found, that is the only routine that return can return from
this is so that we make no distinction between built-in and user-defined control structures 03:02
a dynamic return could very easily find the wrong sub to return from
ugexe m: sub subtest(&code) { &code(); }; subtest { return 1; } # this is the golf 03:03
camelia rakudo-moar f24d2e: OUTPUT«Attempt to return outside of any Routine␤ in block <unit> at /tmp/FNDxkhPNtw:1␤␤»
03:04 noganex_ joined
TimToady well, { return 1 } is golf :) 03:05
03:06 noganex left
TimToady m: return 03:07
camelia rakudo-moar f24d2e: OUTPUT«Attempt to return outside of any Routine␤ in block <unit> at /tmp/OEh2HMoMFE:1␤␤»
TimToady that's golfer :)
so, bad test
flussence m: use Test; use nqp; is(nqp::sprintf('%#b', [0]), '0', 'simple %b with zero value'); # half a dozen of the tests surrounding this one in nqp's test suite currently fail 03:10
camelia rakudo-moar f24d2e: OUTPUT«This type does not support positional operations␤␤»
ShimmerFairy TimToady++ for making the language not treat itself as a special case (something I've been recently reminded C++ doesn't do, when looking into user-defined literals) 03:11
flussence nqp: say(nqp::sprintf('%#b', [0]));
03:11 camelia left
flussence huh. anyway, that line prints "0b" for me here 03:12
03:12 camelia joined, pierre-vigier left
ShimmerFairy flussence: don't use bare nqp:, it kills camelia on account of nqp-p 03:13
flussence oh, whoops sorry
nqp-m: say(nqp::sprintf('%#b', [0]));
looks like I really killed it :(
camelia nqp-moarvm: OUTPUT«0␤»
..nqp-jvm: OUTPUT«#␤# There is insufficient memory for the Java Runtime Environment to continue.␤# pthread_getattr_np␤# An error report file with more information is saved as:␤# /tmp/jvm-7308/hs_error.log␤»
..nqp-parrot: OUTPUT«Can't exec "./rakudo-inst/bin/nqp-p": No such file or directory at lib/EvalbotExecuter.pm line 194.␤exec (./rakudo-inst/bin/nqp-p /tmp/tmpfile) failed: No such file or directory␤Server error occurred! Closing Link: ns1.niner.name (Quit: camelia)␤Lost connect…» 03:14
03:14 ChanServ sets mode: +v camelia
camelia nqp-moarvm: OUTPUT«0␤» 03:14
03:14 pierre-vigier joined
flussence huh, works there... 03:14
I wonder if it's because I'm building with system libs and the bundled ones aren't vanilla 03:15
03:16 pierre-vigier left, kid51 left
ugexe i wanted to demonstrate where the return value was coming from before vs what happens now with it 03:17
03:19 pierre-vigier joined 03:21 pierre-vigier left
uruwi question: how to allocate a buffer of a certain size? 03:27
03:28 herby_ left 03:33 csd_ left 03:35 lustlife joined 03:40 aborazmeh joined, aborazmeh left, aborazmeh joined
dalek kudo/nom: 1e1556b | TimToady++ | src/ (3 files):
provide thunk info for possible metaop thunking
03:44
03:46 lolisa left 03:52 lolisa joined 04:02 yeahnoob left 04:07 yeahnoob joined 04:08 pierre-vigier joined 04:11 pierre-vigier left 04:18 MadcapJake left 04:26 WizJin joined 04:30 pierre-vigier joined, mohae_ joined 04:31 pierre-vigier left 04:32 mohae left 04:36 mephinet left 04:40 pierre-vigier joined 04:48 mephinet joined
lucs With panda, obtained "==> Successfully installed Debugger::UI::CommandLine". 04:50
But where is the module, the .pm file?
04:52 pierre-vigier left 04:55 M-eternaleye left, M-eternaleye joined, M-eternaleye is now known as eternaleye 04:56 MadcapJake joined
lucs goes to bed; tomorrow will be better. 04:57
robnox overall is the moar or jvm faster? 04:59
05:07 MadcapJake left, khw left
AlexDaniel robnox: try it and see 05:09
robnox: startup time on jvm is slower for sure, but everything else depends I think
05:15 dcmertens joined
robnox is there an easy way to toggle between the backends? 05:15
AlexDaniel perl6-j and perl6-m 05:16
robnox sweet, ty
05:16 census joined
AlexDaniel these commands should be avaialable if you've built both 05:16
census hi is anybody around?
skids o/ 05:17
AlexDaniel census: sure, but everyone is kinda asleep at the moment. Not too crowdy :)
robnox I just woke up, but im an odd one.. lol
census AlexDaniel I'm in urgent need of some mac help
i know it's the wrong room
AlexDaniel robnox: 7:17 AM, did not sleep yet
robnox I can help you.. step 1. uninstall OSX. step 2, install linux. ;)
AlexDaniel robnox: that's gonna take some time, but it's urgent 05:18
census robnox i can't do that because it's a school computer
AlexDaniel census: ouch… that sucks
05:18 znpy left
AlexDaniel census: anyway, what's your question? :) 05:18
dcmertens Hello everyone. In case you missed it, the NSF just released a program solicitation that I thought might be of interest on this channel
www.nsf.gov/pubs/2016/nsf16515/nsf1...=USNSF_179
census My mac has a flashing ? mark 05:19
the ? is in a flashing folder
i'm not good with macs at all. i usually use pc
is there a way to take the battery out and reboot ?
robnox sounds like fs corruption, I would try to repair the disk 05:21
dcmertens census, is this the bootup screen?'
AlexDaniel census: #mac perhaps?
census i don't have a repair disk
it is my school computer
dcmertens seconds AlexDaniel's suggestion
robnox fsck is your friend 05:22
census robnox what can i do with fsck ?
AlexDaniel “To force restart your Mac: On some Macs, press control-command and the power button or control-command-eject at the same time. On most Macs, hold the power button for five seconds to force shutdown, and then push the power button again to restart.” – or this? 05:23
robnox fsck = "filesystem check".. tool that can sometimes fix problems with fs corruption, it's ancient but i'm going to assume macs support it since they run on code stolen from BSD
dcmertens census, the only time I saw this on my mac was in bootup, and it meant my system wouldn't boot 05:24
if you're in the same boat, you'll need a recovery disk of some sort
05:25 rhr left
robnox now that I think about it, most of windows was stolen from BSD too.. lol 05:26
dcmertens robnox, I was just wondering about that
do you have any good links about the history? 05:27
adhoc robnox: there were folks from VMS that helped implement things in windows nt, IIRR
05:27 lolisa left
adhoc robnox: also a bunch of stuff was jopintly developed with IBM OS/2, and Windows program manager 05:28
moritz_ .botsnack 05:29
yoleaux 6 Dec 2015 22:31Z <autarch> moritz_: I need access to the advent calendar blog to post my article at some point
:D
6 Dec 2015 22:34Z <ZoffixW> moritz_: never mind, I sent autarch an invite
robnox OS/2 wow, that's something I haven't heard of in a while!
05:29 moritz_ is now known as moritz
robnox dcmertens, don't have any links but if you look in the windows.h you should see some BSD licenses still there 05:29
05:29 skids left
robnox mostly in the network stack 05:29
probably some basic kernel structures as well like linked lists etc 05:30
05:31 rhr joined
census is there a way to take a battery out of a mac? 05:31
AlexDaniel Like this: s-media-cache-ak0.pinimg.com/736x/...f0c269.jpg 05:35
adhoc robnox: yes, thats cuase they outsourced that bit. 05:37
census AlexDaniel you are telling me that the battery is inside the computer?
adhoc well, for NT 3.5 and onwards, IIRR, there are articles explaining about why their implementation of ipconfig was quite similar to bsd 4.x 05:38
AlexDaniel census: you didn't tell us which model it is, so we don't know :)
05:39 aborazmeh left
census MacBook Air 05:39
i don't know more because the computer isn't working
AlexDaniel well, then… where else would it be? :)
robnox adhoc: years ago the windows source code was leaked, and everyone that got a copy was able to confirm that the network stack was indeed directly copied. I may or may not have seen this source, but am not at liberty to discus. lol 05:41
s/discus/discuss/
census i just drained the battery 05:43
and now it is not going to a white apple screen
robnox census, what happens when you hold shift at startup? 05:48
census hold one shift robnox? how long? 05:51
robnox while it boots.. 05:52
that should get you into safe mode, which supposedly will automatically run fsck. if that doesn't work you will probably need to reinstall from media. that's all the pointers i'll give you, if you still need help try #mac 05:53
05:55 lolisa joined
adhoc robnox: if thats the case, then the license allows that. whats the problem ? 05:55
census robnox, do i hit shift once or many times? 05:58
robnox it didn't work with shift 05:59
i tried hitting it many times 06:00
robnox the problem is I can't say whether I've seen the windows source or not. and I've worked as a contractor for Microsoft in the past so I'd rather not burn bridges, but if you are curious google it and make your own conclusion 06:01
06:02 adhoc left
robnox census: #mac 06:02
06:03 WizJin left 06:04 adhoc joined
census robnox it does not seem to go into safe mode 06:06
06:08 xfix joined, dcmertens left
xfix github.com/perl6/specs/commit/2ed4...63d8a26435 06:09
robnox What is the current status of DBI with perl6?
xfix Hm, I wonder what should I do about Math::Tau module I made 2 year sago.
06:11 WizJin joined 06:18 sno left 06:20 pierre-vigier joined
dalek osystem: b5270f6 | (Konrad Borowski)++ | META.list:
Remove Math::Tau obsoleted by perl6/specs@2ed4112
06:23
06:25 [Tux] left 06:27 Khisanth left 06:29 CIAvash joined 06:30 csd_ joined
census thanks robnox for your help. i'm not good with macs 06:31
robnox your welcome. I'm not really a mac guy either, but I've been forced to work with them at school. That and I live less than a mile from Apple:P 06:33
06:34 csd_ left 06:36 [Tux] joined 06:44 Khisanth joined, geraud left 06:53 bjz joined 06:55 zengargoyle left 06:57 bjz left 06:58 bjz joined 06:59 abaugher_ left
robnox cool, got some basic stuff working with DBIish and perl6. Is that really the only current option for perl6 idioms and database interface? 07:00
07:01 abaugher_ joined
AlexDaniel robnox: interesting… DBIish is currently broken 07:01
dalek kudo/nom: 0a2303c | TimToady++ | src/ (5 files):
implement transitive thunkiness for ! && R metaops

For the moment, we disable it for X and R because it breaks things.
AlexDaniel xfix: you could have kept the module for those who are against tau addition :D 07:02
07:03 Khisanth left
xfix AlexDaniel: It was already added to Perl 6 core. 07:03
07:03 ribasushi left 07:04 domidumont joined
AlexDaniel well, yeah, I know… 07:04
07:06 adu joined 07:08 bjz left
robnox AlexDaniel, thanks for the heads up. I'll now expect to encounter some brokenness soon, lol 07:08
07:08 domidumont left
AlexDaniel robnox: well, there's nothing wrong with DBIish itself, as far as I can see 07:09
robnox: something somewhere else is broken…
robnox ahh
AlexDaniel robnox: what's your rakudo version?
robnox: pre curli or not?
robnox should be the latest 07:10
AlexDaniel robnox: rakudobrew?
robnox rakudostar
AlexDaniel 2015.11 then
robnox aye
07:10 Khisanth joined
robnox about a week old 07:10
AlexDaniel robnox: okay, DBIish should break when you upgrade, but there is a workaround 07:12
07:12 pierre-vigier left
robnox what's the workaround? 07:12
AlexDaniel robnox: the problem is that this line fails github.com/perl6/DBIish/blob/maste...sh.pm6#L14 07:13
robnox: here is the error: travis-ci.org/perl6/DBIish#L962
TimToady m: say 3 Rxx rand 07:14
camelia rakudo-moar 0a2303: OUTPUT«(0.0592491323031192 0.00578079871473254 0.290929631967117)␤»
AlexDaniel robnox: the ugly workaround is to change that line to 「use DBDish::Pg;」 or whatever database you are using
dalek osystem: 72bb079 | ugexe++ | META.list:
Sort::Naturally: s/META.info/META6.json/
07:15
robnox that doesn't seem too bad. why ugly? 07:16
07:16 WizJin left
AlexDaniel robnox: well, it shouldn't use DBish::pg if you are not using Pg… 07:17
robnox oh.. im using pg lol
AlexDaniel this problem with 「require」 is beyond me, I hope somebody has better brains to dive into it 07:18
robnox yeah, going to have to shelve this code for the moment.. can't believe the weekend is already over, have to study integrals for finals week 07:19
I got a 95 on the last exam. never really considered myself a math person, but most of this is just memorizing the integral formulas and recognizing when they occur/when to apply 07:20
07:23 FROGGS joined 07:24 bjz joined
dalek ast: 1a09681 | usev6++ | S32-str/samemark.t:
Skip tests for samemark on JVM
07:24
ast: aac0489 | usev6++ | S32-num/power.t:
Fudge tests for vulgar fractions and big exponents on JVM
ast: 877fa65 | usev6++ | S15-literals/numbers.t:
Fudge tests for unival literals on JVM
ugexe ive "fixed" 2 modules that were failing due to precomp by putting "no precompilation" in every single .pm and .t file of the project
07:24 WizJin joined 07:25 lolisa left 07:26 hankache joined
hankache ShimmerFairy++ #Advent post 07:28
07:30 tarski joined 07:32 hankache_ joined 07:33 hankache_ left, hankache_ joined 07:34 hankache_ left 07:35 hankache left
nadim Good morning, Can I tel if a variable is type object or it is an instance of a type? 07:35
07:35 tarski left 07:36 tarski joined
nadim arf, I hate it when I ask a question, after l looked for the answer, and then find the answer a second later! 07:37
TimToady m: say Int.DEFINITE; say Int.new.DEFINITE
camelia rakudo-moar 0a2303: OUTPUT«False␤True␤»
TimToady though usually you can just use .defined
nadim TimToady: thank you 07:38
robnox nadim, that's how it usually goes!
nadim robnox: Yup! make one look a bit silly (/me takes pointy hat out) 07:40
07:40 xfix left 07:42 pierre-vigier joined 07:45 darutoko joined
masak goooood morning :) 07:45
masak .oO( perl6etnam )
07:48 census left
moritz \o 07:52
07:56 ribasushi joined 07:59 Khisanth left
dalek kudo/nom: 67202d6 | TimToady++ | src/Perl6/Actions.nqp:
make sure nested metaops see inner meta's thunky
08:01
ast: 44804e7 | TimToady++ | S03-metaops/reverse.t:
test that RRxx and RRRxx flip thunky transitively
AlexDaniel robnox: yeah, that's why I've never had high scores on this stuff
robnox: my brain refuses to memorize when I know that I will forget it after the exam 08:02
robnox: I love math though, just not the part when you have to memorize a bunch of stuff just to pass the test 08:03
08:03 km3 joined
moritz math school teaching is notoriously bad at motivating 08:06
for me that usually wasn't a problem, but a lot of my peers struggled 08:07
(for me, bad motivation was a problem in other subjects, such as history; I liked to learn how things lead to each other, but I was very bad at memorizing date, and so had quite some trouble with tests) 08:08
El_Che moritz: that a classic. History at university level is all about relations and not about memorizing data. In my first year only 33% passed. 08:12
moritz: a lot of people thought it was about learning phonebooks by heart :)
08:13 Khisanth joined
AlexDaniel well, my solution was to ignore scores completely and just focus on learning 08:13
08:13 Khisanth is now known as Guest98452
AlexDaniel that is, I was setting my own goals. I think that worked 08:13
by worked I mean I got highest stuff learned/time spent ratio, though some people around me were not so happy about some of my marks 08:15
moritz I hope that when my daughters are in school, I'll be sufficiently relaxed as long as they get passing grades
s/are/will be/ 08:16
AlexDaniel moritz: what about homeschooling? 08:17
moritz AlexDaniel: not practical 08:18
AlexDaniel moritz: why?
moritz AlexDaniel: because I work, and my wife works part time
AlexDaniel: and I don't think the two of us could cover all topics well either 08:19
08:19 CIAvash left
moritz AlexDaniel: and then there's the social aspect to consider 08:19
and I don't consider myself to be a good teacher either
08:19 sno joined
moritz (maybe for very motivated pupils in subjects I love, but certainly not for many of the topics covered in school) 08:19
AlexDaniel moritz: yeah, well. I wonder if I will be able to pull it off 08:20
moritz even though some of my school education was shitty, I don't think I could offer my kids a better experience if I did it myself 08:21
08:23 RabidGravy joined 08:25 bjz_ joined 08:26 zakharyas joined, bjz left 08:31 pierre-vigier left
dalek c: e2054c9 | (Brock Wilcox)++ | doc/Language/ (2 files):
Add basic info on qualified names, taken from S02
08:33
c: ea23fcd | (Brock Wilcox)++ | doc/Language/ (3 files):
Move into a dedicated Packages doc
c: 86d7824 | (Brock Wilcox)++ | doc/Language/packages.pod:
Remove unnecessary details and misc cleanup
c: aab793f | (Brock Wilcox)++ | doc/Language/syntax.pod:
Actually provide a few random examples
c: 353e898 | RabidGravy++ | doc/Language/ (2 files):
Merge pull request #224 from perl6/qualified-names

Add package (namespace) documentation, taken from S02
08:35 pierre-vigier joined 08:37 sarya joined
sarya hii 08:37
masak sarya: hi
sarya: you are here for the Perl 6, I take it?
masak grins
sarya ok 08:38
what to do
masak install it!
write a blog engine! :D
take over the world
MWHAHAHA
sarya how
masak oh, hm
sarya yes
masak go to perl6.org/
press the big blue friendly "Download Rakudo Perl 6" button 08:39
sarya i did
masak didn't work? :/
sarya nopes
08:39 rurban joined
sarya anyone? 08:40
masak could you describe your problem in more detail, please? 08:41
in what way did it not work?
08:41 csd_ joined
sarya i have already a perl 5.22 08:42
masak that should be fine
so do I 08:43
are you on Windows?
sarya i want to do something in 6
yes
masak great!
RabidGravy wonders what perl 5 version there is on this machine
08:43 ollej left
masak m: say "try this :)" 08:43
camelia rakudo-moar 67202d: OUTPUT«try this :)␤»
masak sarya: ^
08:44 abraxxa joined
sarya yes 08:44
masak did you try one of the .msi files on rakudo.org/downloads/star/ ?
sarya masak
wait a minute
masak takes on an air of infinite patience 08:45
AlexDaniel RabidGravy: try this: perl6 -e 'run «perl --version»' # :D
masak :P
08:45 csd_ left
TimToady m: say 43 Rand 0 08:46
camelia ( no output )
TimToady m: say 43 Rand 1 08:47
camelia rakudo-moar 67202d: OUTPUT«43␤»
masak Rand! o.O
masak .oO( TimTatlas Shrugged )
AlexDaniel phaha :D
sarya masak wat is this file msi? 08:48
i didn't get it
masak sarya: better now? :)
sarya y
masak sarya: those are install files. should be able to just run them.
well, one is enough
take the latest! :D
AlexDaniel RabidGravy: perl6 -e '.flip.say for run(«perl --version», :out).out.lines.list.reverse' 08:49
sarya ok thanks
08:49 Ven joined 08:50 g4 joined, g4 left, g4 joined
masak sarya: best of luck »ö« 08:51
sarya :)
AlexDaniel actually, I wish there was Rfor 08:52
08:53 rindolf joined
AlexDaniel otherwise, how can you make this left-to-right without sacrificing anything? .flip.say for run(«perl --version», :out).out.lines.list.reverse 08:53
oh right you can use { }! 08:54
「for run(«perl --version», :out).out.lines.list.reverse { .flip.say }」 # stupid me
moritz do you even need the .list?
AlexDaniel moritz: it is Seq otherwise
hm, it works, indeed 08:55
moritz m: say <a b c>.Seq.reverse
camelia rakudo-moar 67202d: OUTPUT«(c b a)␤»
AlexDaniel The reason I wrote .list is because 「reverse」 is not listed here: doc.perl6.org/type/Seq
moritz ah, it's an Any method 08:56
08:57 csd_ joined 09:00 sarya left, [ptc] left 09:01 aigan joined, csd_ left 09:02 [ptc] joined
timotimo the doc page for "Stringy" mentions "Cat" 09:05
09:05 Actualeyes left
sjn \o 09:07
sjn notes it's tadzik's birthday today. Happy birthday! o/
tadzik: ^^
masak tadzik.age++ \o/
timotimo oooh!
tadzik, happy birthday :) 09:08
masak and the "Happy Birthday to You" song is in the public domain now, so we can actually...
AlexDaniel tadzik.age++ \o/
masak 􏿽xF0􏿽x9D􏿽x85􏿽x9F happy birthday to youuuuuu ♫
􏿽xF0􏿽x9D􏿽x85􏿽x9F happy birthday to you! ♫ 09:09
masak apparently doesn't know the lyrics so well
gfldex m: with Seq.^find_method("reverse") { say .file, .line }
camelia rakudo-moar 67202d: OUTPUT«gen/moar/m-CORE.setting1878␤»
masak actually, since the third line is meant to be parametric on the name of the recipient...
...could you actually *patent* "Happy Birthday"? :P
"Technique for creating a person-adaptive birthday celebration song" 09:10
gfldex if you find a patent office that takes your money, you can patent anything
timotimo this is critical to my business and my competitors must not benefit from my money and time spent in research without me getting compensated.
Ven mh, mh, my `make install` fails with "No writeable path found"... did I miss some directions? 09:12
tadzik \o/ thanks everyone :) 09:14
RabidGravy HB tadzik ! 09:15
Ven happy birthday tadzik :)
tadzik thank you :)
Woodi hi #perl6 :)
timotimo: moust of the world just waits few years until patents are outdataed :) 09:16
RabidGravy Anyone in a position to put nine's %?RESOURCES stuff back? (If you weren't following at home, it got partially reverted by accident last night)
09:16 dakkar joined
gfldex m: with Seq.^find_method("reverse") { say .file, .line } 09:17
camelia rakudo-moar 67202d: OUTPUT«gen/moar/m-CORE.setting1878␤»
gfldex ^^^ where is this bugger declared?
i can't doc what i can't see :(
FROGGS Happy Birthday tadzik <3 09:20
09:20 cdc joined
Woodi as for education: it is broken seence thousands of years... Socrates was sentenced to death becouse his students dropped democratic ideals... I hear Edison was banned from his school (and it was quite common practice in that times), there was famous german discipline later we had do not press your childrens! 09:21
timotimo Woodi: how did the world go for "a few years" without double-clicking? 09:23
tadzik FROGGS: thank you :}
timotimo Woodi: "drücke deine kinder nicht"?
Woodi many years ago I hear north Europe is perfect place for childrens and now there is epidemy of suiciding teenagers... (but I would like to hear "insiders" voice on this)
timotimo: translating, pleas wait ;) 09:24
gfldex those insiders are all dead, no?
pdcawley m: my $foo′ = 10; say $foo′
camelia rakudo-moar 67202d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/7Ub2nHdDIe␤Bogus postfix␤at /tmp/7Ub2nHdDIe:1␤------> 3my $foo7⏏5′ = 10; say $foo′␤ expecting any of:␤ constraint␤ infix␤ infix stopper␤ postfix␤ …»
pdcawley Dang... kind of hoped that, what with the unicodeness I might be able to get away with a prime at the end of a variable name. 09:25
Woodi timotimo: yes, keybouard and rulers are prohibited :)
gfldex i found .reverse, it's declared in Any. I shall doc! 09:26
timotimo pdcawley: the rule isn't "if its unicodeness is sufficient, it's allowed in an identifier", you know? ;) 09:27
pdcawley Sure, but prime doesn't really make sense as an operator...
timotimo why not? 09:28
pdcawley Unless, I suppose, I made it an operator on a variable name :)
timotimo: generally used in notation as a 'like the variable named 'foo', but modified in some way'
s/variable/thing/
timotimo the thing is, we leave the decision to the unicode consortium ... mostly 09:30
09:30 hankache joined
timotimo we just implement a decision based on "general category" 09:30
however, you can make it a slang
just like Piersing
hankache o/
who owns the twitter handle @perl6 ? 09:32
shouldn't we have an *official* twitter account?
I think we should get ready to trend 6.c when it's out 09:33
09:33 vytas joined
nadim two terms in a row across lines (missing semicolon or comma) is a fun error message for something that starts with an elsif (IE no if block before it). I like the new message but they get time to get used to 09:34
hankache twitter/facebook/HN/reddit
09:34 yeahnoob left, Jonis joined
timotimo we also should get some content onto twitter.com/perlhex, because we do control that 09:34
09:34 TEttinger left
hankache perlhex? 09:35
timotimo yes 09:36
you know, like a hexagon has 6 corners?
hankache aha 09:37
09:37 lolisa joined, Skarsnik joined
hankache but wouldn't it be better if we take control of @perl6? 09:38
timotimo sure
we probably do, but whoever has it just didn't care about twitter enough? ;)
i mean, last tweet's back in 2012, when twitter wasn't that big yet
Ulti ooc there was talk of splitting Rakudo dev chat into another channel did that happen? 09:39
timotimo i can privmsg you some bug reports, then we'll have #perl6 and our privmsg; does that count? :) 09:40
sorry, read to much SMBC 09:41
RabidGravy If it became necessary I'd go with making a #perl6-users 09:42
09:43 bakedb joined
timotimo you think so? then people would stumble in here to discuss using perl6 and we'd be sending them off to #perl6-users 09:43
if we have #perl6 and #perl6-dev, we'd have to send people over who are interested in dev talk
so if we expect more users than core devs, we should go the one way
RabidGravy this is also true
timotimo otherwise, the other
RabidGravy it won't make a difference to the noise either way 09:44
timotimo that depends on whether you decide to do your backlogging and main reading in #perl6 or "the other channel" ;)
RabidGravy as people, even if they know there is a split, will want to jump the queue and go where the "big kids are" 09:45
timotimo so we'll instead make a second channel that is populated only by markov bots to keep noisy people away from us? 09:47
RabidGravy :)
hankache if #perl6 is still manageable why make another?
timotimo we're approaching a point where suddenly #perl6 may become "unmanagable" 09:48
RabidGravy I'm actually surprised it isn't already unusable but I'm a cynic when it comes to people's behaviour on the internet
El_Che RabidGravy: you're just like Hitles <--- see what I dead there? :) 09:49
r
timotimo ideally, we'd be infecting every new person coming in here with our ... what do you call that?
El_Che saliva?
Ven am I the only one with a failing `make install`? 09:50
I should backlog more
hankache how does perl irc function?
one channel?
timotimo the perl irc channel is a terrible place, if i'm not mistaken
Skarsnik Hello, I am for a #perl6-ecosystem channel too x) 09:51
RabidGravy haven't been in the #perl for years
hankache we can try mimicking what they did. if it is good
El_Che isn't this an academic discussion? Something for when perl6 is big? #perl6 is pretty cozy as it is
RabidGravy it's a different thing, most of p5 development is done on the mailing list 09:52
nine #perl has improved a lot in recent years
yoleaux 6 Dec 2015 22:05Z <lizmat> nine: I've given up on fixing the merge/revert mess that killed %?RESOURCES
6 Dec 2015 22:48Z <masak> nine: this may help a bit, too: gist.github.com/masak/32f236bca8f2857d82fc -- I spent some time building various combinations of parts of the commit that fails.
09:52 AlexDaniel left, Actualeyes joined 09:53 adu left
timotimo El_Che: either you have an "academic discussion" up front and know at least 5 ways that don't work when reality catches up to your fears & expectations ... 09:53
RabidGravy nine, just the person - you may want to check whether parts of what made %?RESOURCES work didn't get reverted by accident last night
timotimo or you don't discuss it up front, things suddenly happen, and you end up going with one of these 5 ways that don't work that you could have figured out in advance :)
hankache my fear is that #perl6 will not be attractive anymore 09:54
RabidGravy because it stopped working last night
hankache if a split happens
people want to hang out with the *big boys*
timotimo put some make-up on it. by which i mean, turn it into a "slack" chatroom
RabidGravy frankly I'll just return to my default position and ignore IRC altogether at some point 09:55
El_Che How do you make the split? what about non core-dev people that just want to program and do the ocassional commit (code, doc) for things they encouter after checking on the channel?
nine RabidGravy: I'm on it. May take a while though. Haven't had a coffee yet and was up late celebrating my birthday :) 09:56
El_Che you'll loose the paper cuts fixes
timotimo hasn't a clue and is just being obnoxious
RabidGravy hay, happy birthday for yesterday nine :)
09:56 azawawi joined
azawawi hi 09:56
nine Thank you :)
hankache hi azawawi 09:57
Skarsnik hm, where is stdout? the spec (old spec?) says $STDOUT but it's not that x)
RabidGravy $*OUT?
azawawi when I see an error message mentioning ./.rakudobrew/moar-nom/install/share/perl6/site/sources/BF0504AF8F51E3C7D2286B7FA5CDF41EB923632D (i.e. Panda::Common on my machine), does that look sane to you? :) 09:58
Skarsnik Oh thx :)
timotimo i didn't know!
happy birthday nine! 09:59
nine :)
hankache happy birthday nine :)
RabidGravy azawawi, it could probably be improved :)
timotimo Skarsnik: where is that comment about $STDOUT?
nine blushes a little
azawawi well it was improved lol... that's a regression :)
09:59 pierre-vigier left
azawawi RabidGravy: hi 10:00
Skarsnik www.perl6.org/archive/rfc/30.html x)
hankache haaaaaaappppyyy birthdaaaaaaaay
timotimo lol.
Skarsnik: why are you looking there?! :)
hankache start singing
Skarsnik well google perl6 stdout
timotimo right. well, it says "This file is part of the Perl 6 Archive" at the top 10:01
in big bold letters :)
but that probably makes it less visible :D
dalek kudo/nom: 6f85793 | ronaldxs++ | tools/build/Makefile-common-macros.in:
Update Makefile-common.in to work with mingw

Fix "/usr/bin/sh: C:Strawberryperlbinperl.exe: command not found" type error the same way we did for nqp about a week ago by effectively setting the SHELL to cmd under windows and 'sh' for posix environments.
kudo/nom: fd8e7a9 | lizmat++ | tools/build/Makefile-common-macros.in:
Merge pull request #616 from ronaldxs/nom

Update Makefile-common.in to work with mingw
10:02 pierre-vigier joined
Skarsnik timotimo, well searching stdout on docs.perl6.org did not help me so... x) 10:02
hankache can we not force old documents not to be indexed by google?
.htaccess 10:03
timotimo right, it should probably mention the word "stdout" in there somewhere
"Error due to tapping a Proc::Async stream before spawning its process" - this is the description of X::Proc::Async::TapBeforeSpawn
however
hankache or robots.txt
10:03 lustlife left
timotimo You must call stdout before you call method start. Otherwise an exception of class X::Proc::Async::TapBeforeSpawn is thrown. 10:03
Skarsnik hm not even mentionned in doc.perl6.org/language/io 10:04
timotimo i don't actually even have a X::Proc::Async::TapBeforeSpawn in my rakudo ... not anywhere to be found :o 10:05
um, there it is
ah, searched from the wrong base directory
Ven ShimmerFairy: irclog.perlgeek.de/perl6/2015-12-07#i_11667977 technically wrong, because of SFINAE.
hankache see you later everyone 10:06
10:06 hankache left
Ven but then, SFINAE is probably one of the worst thing ever :P 10:06
dalek c: bc26214 | timotimo++ | doc/Type/X/Proc/Async/TapBeforeSpawn.pod:
TapBeforeSpawn was described the wrong way around
10:07
10:07 pierre-vigier left
lizmat m: (1..5).Supply.throttle( 4, { say "$*THREAD.id(): start $_"; $_ } ).wait; 10:08
camelia rakudo-moar 67202d: OUTPUT«(timeout)3: start 1␤4: start 2␤3: start 3␤4: start 4␤5: start 5␤»
lizmat m: (1..5).Supply.throttle( 2, { say "$*THREAD.id(): start $_"; $_ } ).wait; # does not hang 10:09
camelia rakudo-moar 67202d: OUTPUT«3: start 1␤4: start 2␤5: start 3␤6: start 4␤7: start 5␤»
lizmat m: (1..5).Supply.throttle( 4, { say "$*THREAD.id(): start $_"; sleep 0.01; $_ } ).wait; # doesn't hang
camelia rakudo-moar 67202d: OUTPUT«3: start 1␤4: start 2␤5: start 3␤6: start 4␤7: start 5␤»
10:09 brrt joined
lizmat .tell jnthn I think there's a race condition in supply syntax exposed by "(1..5).Supply.throttle( 4, { say "$*THREAD.id(): start $_"; $_ } ).wait;" 10:09
yoleaux lizmat: I'll pass your message to jnthn.
lizmat .tell jnthn fwiw, I've seen S17-supply/syntax.t hang occassionally: feels like the throttle example is exposing this more reliably 10:10
yoleaux lizmat: I'll pass your message to jnthn.
Skarsnik duh, http::message, why sometime your $.content is Buf, sometime Str x) 10:11
dalek c: fdb6c21 | timotimo++ | doc/Language/variables.pod:
make $*OUT & friends more discoverable for google

by mentioning their more common names "stdout" and such.
timotimo Skarsnik: do you think this will help? ^ 10:12
in general, doc.perl6.org ought to get a better pagerank than learningperl6.com and the archives.
and by "ought to" i mean "we should make it so", not "it's likely the case"
nine RabidGravy: %?RESOURCES looks fine here?
.tell tadzik Happy birthday to you :) 10:13
yoleaux nine: I'll pass your message to tadzik.
lizmat nine: I can't install Inline::Perl5 because it cannot find the helper
RabidGravy nine, that's odd. let me prod this a bit harder
lizmat nine: Unhandled exception: Cannot locate native library '/Users/liz/Github/panda/.panda-work/1449483191_1/resources/p5helper.dylib': dlopen(/Users/liz/Github/panda/.panda-work/1449483191_1/resources/p5helper.dylib, 1): image not found
RabidGravy yeah, I'm seeing the same behaviour (albeit with Sys::Utmp) which worked earlier yesterday 10:15
Skarsnik timotimo, add it in language/io maybe? it's a likely place where it interesting or a way to find it?
brrt good * #perl6 10:16
timotimo right. i can do tthat later. will be AFK for a bit first
nine lizmat: oh, that's to be expected! 10:22
Skarsnik Can I remove @*INC from the doc? 10:23
nine lizmat: I'm sorry! We do have %?RESOURCES now, but we still don't know how to handle the platform dependentness of the shared object file. Currently the META.info has p5helper.so hard coded which won't work on OS X or Windows
Skarsnik nine, you should have typed ressource declaration in the meta file. like Ressource : [ dynlib : "filewithoutext" ]. there is so weird way to find the extension (NC does it) 10:25
nine Yes, I was thinking about "resources": [ { "dynlib": "p5helper" } ] 10:26
S22 says on this: "If a leaf of the "resource" hash is a hash, it indicates files with special properties (such as being provided by the builder, so not part of the distribution). This will allow a good MANIFEST to be created from the META6.json information." 10:27
But in this case, it would be nice, to tell the builder, too, that this is platform dependent.
10:27 geekosaur left
nine Except if we want panda to guess that from the file extension. 10:28
Skarsnik I think everything that help distrib packager is better 10:29
nine Or you specifiy just the base name, panda notices that there's no file with that name but one with the dynlib extension added, so it automatically adds that to the name.
Skarsnik like rt.perl.org/Ticket/Display.html?id=126744
tadzik nine: thank you :)
yoleaux 10:13Z <nine> tadzik: Happy birthday to you :)
nine That would work in the cases we know but I'm nervous about such automatisms while we have so little experience with actual use cases.
tadzik yay, double happy birthday :)
nine Skarsnik: external dependencies are something the Perl 5 people also still struggle with. I strongly recommend talking to them about such issues. 10:31
10:31 espadrine joined, aborazmeh joined, aborazmeh left, aborazmeh joined
Ven Interesting that irclog.perlgeek.de/perl6/search/?ni...path+found doesn't have any result with the sentence I searched :( 10:32
Skarsnik I did not do a perl5 module in a while, but I remember that you have nothing to even be informative about this kind of stuff x) 10:33
RabidGravy nine, you can probably ignore me too the "Cannot locate native library '/home/jonathan/.panda-work/1449484225_2/resources/lib/utmphelper.so': /home/jonathan/.panda-work/1449484225_2/resources/lib/utmphelper.so: cannot open shared object file: No such file or directory" seems to suggest that I'm not quite doing the right thing with the paths
10:34 eviltwin_b joined
RabidGravy that is, when one does "panda install ." (which works) it is subtley different to doing "panda install Sys::Utmp" 10:34
10:35 Actualeyes left
gfldex .seen ZoffixW 10:36
yoleaux I saw ZoffixW 00:46Z in #perl6: <ZoffixW> .botsnack
Skarsnik so what is the new @*INC? (so I can remove it from the doc ~~)
it's @*REPO?
lizmat $*REPO
but it's a different beast altogether 10:37
Skarsnik m: say $*REPO.perl
camelia rakudo-moar fd8e7a: OUTPUT«CompUnit::Repository::Installation.new('/home/camelia/.perl6/2015.11-404-gfd8e7a9')␤»
10:38 pierre-vigier joined, eviltwin_b is now known as geekosaur, rindolf left
gfldex .tell ZoffixW does htmlify.p6 work for you right now? I'm asking because after updating rakudo/panda (precomp fallout) it got a type check error in Perl6::Documentable::new . 10:38
yoleaux gfldex: I'll pass your message to ZoffixW.
10:40 simcop2387 left
dalek c: 172351c | (Wenzel P. P. Peppmeyer)++ | doc/Language/classtut.pod:
link type/Method in Language/classtut
10:41
c: 3e69312 | (Wenzel P. P. Peppmeyer)++ | doc/Language/functions.pod:
Merge remote-tracking branch 'upstream/master'
c: ed94543 | (Wenzel P. P. Peppmeyer)++ | / (6 files):
Merge remote-tracking branch 'upstream/master'
c: 5031508 | (Wenzel P. P. Peppmeyer)++ | / (21 files):
Merge remote-tracking branch 'upstream/master'
c: bc7876b | (Wenzel P. P. Peppmeyer)++ | doc/Type/Any.pod:
doc Any::reverse
c: c774629 | (Wenzel P. P. Peppmeyer)++ | doc/ (2 files):
Merge pull request #226 from gfldex/master

doc Any::reverse and fix a link problem
Skarsnik $*REPO A variable holding information about modules installed/loaded 10:42
- is that enought? (yes it's rather none descriptive)
lizmat Skarsnik: I guess that will do for now 10:43
10:43 pierre-vigier left 10:44 geekosaur left
|Tux| test 50000 22.123 22.016 10:45
test-t 50000 20.297 20.190
csv-parser 50000 24.296 24.189
dalek c: d2ddc1a | (Sylvain Colinet)++ | doc/Language/variables.pod:
Remove @*INC from language/variables.pod. Replaced with a small description about REPO
10:46
c: 69799f8 | (Sylvain Colinet)++ | doc/ (2 files):
Merge branch 'master' of github.com/perl6/doc
Skarsnik dafuq this dumb additionnal commit
git is really confusing
gfldex .oO( in soviet git bad tools blame the craftsman ) 10:48
Skarsnik: does htmlify.p6 work for you? 10:49
Skarsnik probably? my rakudo is still from october
but I don't get it. I try to push, it tell me I am behind master. I pull to update, it create me a commit 10:50
gfldex i merged a commit 9 min ago 10:51
Skarsnik Yes, but why it created me another commit just by merging my copy with master?
brrt Skarsnik; git pull is git fetch + git merge
that's what git pull does 10:52
if you want the other behaviour, i.e., pull-and-apply-my-commit-on-top, you want to do git pull --rebase
it's not like svn in that way
Skarsnik git update was to complicated to have x)
dalek kudo/nom: 29ba648 | lizmat++ | src/core/Supply.pm:
Fix race condition in Supply.throttle

Although I'm not sure why this fixes it, it appears that the order of the whenevers is important.
10:53
brrt yes, it was :-)
exactly because of the distinction i mentioned just now 10:54
10:54 Actualeyes joined
brrt your branch on your local computer, it's really only symbolically connected to the same-named branch on another repo 10:54
10:55 pecastro left
brrt there may be the same tree objects and the same commit objects, and that redundancy may be used to optimize the traffic between to git repos 10:55
10:55 simcop2387 joined
brrt but there is no such thing as update because that would assume there to be a single source of commit truth, and there is none 10:56
arnsholt Cf. distributed systems =) 10:57
yoleaux 6 Dec 2015 22:05Z <lizmat> arnsholt: I've given up on fixing the merge/revert mess that killed %?RESOURCES
10:57 Ven left
lizmat .tell autarch could you test Supply.throttle again and tell me if it fixed your problem? 10:58
yoleaux lizmat: I'll pass your message to autarch.
nadim The documentation of can, and candidates, is a bit sparse. Can I ask if a class does a method with specific signature, including return type?
of course they are multis
arnsholt lizmat: Does that mean something's still broken post-reverting the precomp-singleprocess merge, or is that that notification that you reverted the merge? 10:59
lizmat arnsholt: I reverted the merge, but after that certain things still appeared to be broken
arnsholt Weird 11:00
lizmat today, it turns out, they may have been broken already and I spent several hours on fixing something that was broken to begin with :-(
that Inline::Perl5 doesn't install on OS X, was apparently a known problem :-( 11:01
arnsholt Ah, right
lizmat brunch& 11:03
11:03 pecastro joined, geekosaur joined
RabidGravy lizmat, I'm pretty certain now that my "%?RESOURCES not working" is actually my fault 11:03
gfldex nadim: you can get hold of all candidates of a method and their signatures, then create your own signature object and ~~ them. If that makes sense for a dynamic language is a different matter.
nadim: if you want to i can dig that out for you 11:04
brrt can you .^can with a signature?
arnsholt Usually when people call .can they just want to ensure it exists before calling it. Shouldn't $obj.?method(...) work?
brrt anyway, lecture 11:05
^
&
11:05 brrt left
Skarsnik actually I have a similar question. how to get the signature of a sub/method? 11:05
gfldex if you .?method your don't know if there is no such method (name) or if the signature didn't match
nadim gfldex: that would be great
gfldex: I can't .ĉan with a signature but it is exactly what I want 11:06
11:06 SmokeMachine__ left 11:07 SmokeMachine__ joined, aborazmeh left, km3 left, pierre-vigier joined
gfldex m: class C { multi method m(Int){}; multi method m(Str){}; }; say .signature for C.^methods[0].candidates; 11:08
camelia rakudo-moar fd8e7a: OUTPUT«(C $: Int, *%_)␤(C $: Str, *%_)␤»
gfldex ^^^ step one
nadim testing ... 11:09
Skarsnik m: class C { multi method m(Int){}; multi method m(Str) returns Bool{True}; }; say .signature for C.^methods[0].candidates; 11:10
camelia rakudo-moar fd8e7a: OUTPUT«(C $: Int, *%_)␤(C $: Str, *%_ --> Bool)␤»
Skarsnik and Discord, y u change your webapi (╯°□°)╯︵ ┻━┻ 11:11
DrForr Skarsnik: Because it's true to its name?
Skarsnik yes that's too 11:12
11:14 _nadim joined, nadim left
lucs With panda, obtained "==> Successfully installed Debugger::UI::CommandLine". 11:16
But where is the module, the .pm file?
Skarsnik It should not exist if I understand correctly? There is a SHAencoded filename that is the prcompiled version of the file 11:17
lucs :/
_nadim can one get the methods by name rather than going through an indexed list? 11:18
11:18 nadim joined
Skarsnik m: class C { multi method m(Int){}; multi method m(Str) returns Bool{True}; }; say .signature for C.^methods<m>.candidates; 11:18
camelia rakudo-moar 29ba64: OUTPUT«Type List does not support associative indexing.␤ in block <unit> at /tmp/lr6rMj7rsO:1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp:2883␤ in block <unit> at /tmp/lr6rMj7rsO:1␤␤»
11:19 zengargoyle joined
gfldex m: class C { multi method m(Int){}; multi method m(Str){}; }; my $sig = (my method (C: Int) {}).signature; say $sig; say .signature, ' ', so .signature ~~ $sig for C.^methods[0].candidates; 11:20
camelia rakudo-moar 29ba64: OUTPUT«(C $: Int, *%_)␤(C $: Int, *%_) False␤(C $: Str, *%_) False␤»
gfldex that should work but doesn't
may be a bug or may be a disagreement between me and Perl 6 11:21
nadim: you may have to as jnthn for details
11:22 _nadim left
lucs Skarsnik: As far as you know, how can I look at the module's code? 11:22
(re. Debugger::UI::CommandLine for example) 11:23
gfldex m: class C{}; my $sig1 = (my method (C: Int) {}).signature; my $sig2 = (my method (C: Int) {}).signature; say so $sig1 ~~ $sig2;
camelia rakudo-moar 29ba64: OUTPUT«False␤»
11:23 pierre-vigier left
gfldex m: class C{}; my $sig1 = (my method (C: Int) {}).signature; my $sig2 := $sig1; say so $sig1 ~~ $sig2; 11:24
camelia rakudo-moar 29ba64: OUTPUT«False␤»
gfldex i'm quite sure that the very same signature should smartmatch :)
Skarsnik lucs, no idea, sorry :(
lucs Skarsnik: Ok, thanks. 11:25
So, I installed Debugger::UI::CommandLine, and when using it, I get "Circular module loading detected involving module 'Debugger::UI::CommandLine'"
How do I find out what the heck is going on?
Skarsnik blame nine 11:27
lucs :)
By the way, this procompiling, is it done for performance reasons? 11:28
Skarsnik *like his patch on http::ua* my $a = HTTP::UserAgent.new(:debug) # debug on stdout; my $a = HTTP::UserAgent.new(:debug($*ERR)) # debug on stderr; my $a = HTTP::UserAgent.new(:debug("file")) # debug on file;
nine lucs: yes 11:29
Skarsnik hm, how could I say to append to file.txt actually x)
lucs nine: Okay, but is there no way be able to see the source somehow? 11:30
nine Skarsnik: the :append flag to .open
lucs: what do you mean by "to see the source"? Rakudo does need the source to be able to precompile, so the source is available anyway
Skarsnik I mean interface wise? like having a way to tell to append or not. I could not bother since I can give you a IO::Handle but
gfldex rakudobugged as #126835
lucs nine: I guess rakudo saw it, but how can _I_ see it? 11:31
RabidGravy Skarsnik, I'm pondering changing that :debug on H::UA so that it can also be an object that has a .debug() method
gfldex Skarsnik: see: perl6advent.wordpress.com/2015/12/...-projects/
nadim gfldex: thanks, I'll play a bit and get back if I find out, or if I havent ;) 11:32
nine lucs: Debugger::UI::ComandLine does "use nqp;" twice. This may trigger a bug in Rakudo complaining about the circular module loading
gfldex nadim: for the time being you could .Str and compare those. I have the feeling that you will bump into problems in any case because of where-clauses.
lucs nine: Ah, could be. How do you (or I) find out that it's doing that?
11:33 kid51 joined
DrForr github.com/jnthn/rakudo-debugger/b...andLine.pm # lines 3 and 5 :) 11:33
gfldex nadim: what do you actually want to do in the end anyways? I'm asking because the answer you be "Have a role noob!"
Skarsnik RabidGravy, if you are willing to do that, add a .Debug or something to http::Message x) for now I use .Str but it's ugly
nine lucs: I cloned the git repo ;) But rakudo also prints file names in backtraces.
lucs: pm files should usually end up in your rakudo/install/share/perl6/site/sources 11:34
DrForr ^^ for lucs - Looks like a tyop in the file but it could be an obscure requirement of Terminal::ANSIColor.
RabidGravy me-- # totally didn't test his %?RESOURCES amendments properly 11:36
lucs nine: Aha, I see. Hmm... Does each file there correspond to a single original perl file, or are they mangled somehow?
nine lucs: they are 1:1 - only the names change 11:37
nadim gfldex: I amm writing a dumper, either the class to be dumped implements a method for the dump or the dumper itself implements it. I need to know if the dumper implements it.
lucs nine: Okay. So currently, if I want to look at some of that source, I need to go throught those files and see what I can find? 11:38
*through
nadim gfldex: I agree that were clauses will be a problem if I just compare strings, it's already frustrating in the error message one get from P6 to not see the clauses. it will work for what I want to do rigt now though :)
nine lucs: unless the error message already told you the file name, which it should
nadim a super newbie question, why does map take a semi colon but for doesn't? I absolutely hated python semicolon, now I almost want it to be added to P6 for consistency. 11:39
gfldex nadim: if the dumper are all witten by you, you should have them do a role with that method and match against the role. Otherwise you may end up with some object that implements a method that looks like it might be the right one. 11:40
nadim gfldex: that's how it works
gfldex for is syntax, map is a method or sub
nadim gfldex: but the end user can also add a role, thus I need to know if the dumper has had the role added 11:41
lucs nine: In this case, it turns out that the error only mentions 'Debugger::UI::CommandLine' (which, ack'ing trough that directory you mentioned, is found in my "3982B8FE2[...]" file)
gfldex roles have identity that goes beyond their name, you can even hide the name with some anonymous role or by not exporting it 11:42
lucs nine: Would it be a good idea to have .../Debugger/UI/CommandLine somewhere be a symlink to "3982B8FE2..." 11:43
?
(...CommandLine.pm even)
lucs can't be the only one who wants to be able to easily look at installed modules code, can he? 11:44
nine lucs: we're using those SHAs because different distributions may contain files with similar names. Also it helps with file systems that don't support non-ASCII file names and it makes lookup of sources of precompiled files trivial. 11:47
RabidGravy I think I have detected a new variant of Godwins law in the wild: "... because the right wing media made you think that" 11:48
nine lucs: and not all systems support symlinks. That said, we may utilize them for such convenience features on systems that do.
lucs nine: I understand, and it all makes sense. 11:49
RabidGravy nine, lizmat everyone you are all entirely exhonerated regarding Sys::Utmp not installing, entirely my stupidity
Skarsnik maybe have a perl6 --display ModuleName ?
nine RabidGravy: thanks 11:50
11:51 vividsnow joined
nine Skarsnik: or just a separate tool. I use perldoc -l to find the sources of installed Perl 5 modules. And I have a pmversion alias to find the version of an installed module. We could ship similar helpers with rakudo. 11:51
Maybe a p6mod tool that queries the repositories 11:52
nadim gfldex: I will look more into the role's identity. my problem is a bit simplified by the fact that the signature of my multis has a single element but even if I get all the signatures I need to know if the object being dumped is derived from on of those signature types. 11:56
DrForr There's a perl6 --doc flag, maybe 'perl6 --doc -MFoo' to look at Foo's documentation? Probably just meant to work on individual .pm6 files, haven't looked closely yet. 11:57
nadim gfldex: I honnestly think that it is better to eval the whole thing, P6 can handle it already, I am not re-inventing the wheel, and an exceptions can handle the whole multi with the right signature does not exist
are eval and exceptions terribly expensive? 11:58
11:58 RabidGravy left
gfldex eval is, exceptions can be 11:58
nadim thx 11:59
zengargoyle good * #perl6 12:06
Skarsnik nine, do you think is there a way to exploit the compile unit with stuff like Mojo::Template6? (that allow for perl6 code in the template file) 12:12
nine Skarsnik: I'm not sure I can follow you? 12:15
Skarsnik Mojo::Template6 file look like www.nyo.fr/~skarsnik/story.tm . There is clearly perl6 code. I was wondering is there was a way to precompile that? 12:17
nine If you turn the whole template into perl6 code, you surely can precompile it. And without having looked, I guess that's what Mojo::Template6 does anyway 12:19
12:19 zakharyas left
nadim if someone is interested in code that segfaults nopaste.linux-dev.org/?877888 the redeclaration warning point at the problem which is more than trivial to fix but maybe this should be handled somewhere. 12:19
dalek kudo/nom: b668da5 | (Stefan Seifert)++ | src/core/CompUnitRepo.pm:
Fix "Circular module loading detected" failure with use Foo:from<Perl5>

When encountering a use Foo:from<Perl5> the Perl5ModuleLoaderStub first loads Inline::Perl5 and then calls CompUnitRepo.load_module again to try give the real Perl5ModuleLoader a chance. At this point "Foo" is already in the list of modules to load and would get added again, triggering the "Circular module loading detected" error. Fixed by tidying up the load stack before trying the second time.
nine Now there's only one of Inline::Perl5's test files still failing :) 12:20
12:21 ZoffixMobile joined
ZoffixMobile TIL proto.perl6.org is a thing... m.slashdot.org/story/303623 12:21
so much hate in the comments of that article :p 12:23
lizmat m: my @a; @a[1] = Any; say @a.sum
camelia rakudo-moar 29ba64: OUTPUT«X::Multi::NoMatch exception produced no message␤ in block <unit> at /tmp/K_G4saOp0i:1␤␤»
12:26 azawawi left
nadim ZoffixMobile: it's always easier to write a hateful comment than write code! 12:26
ZoffixMobile Eh, they're referencing perl 4 of all things. :) 12:27
12:27 ZoffixMobile is now known as ZoffixW
ZoffixW . 12:27
yoleaux 10:38Z <gfldex> ZoffixW: does htmlify.p6 work for you right now? I'm asking because after updating rakudo/panda (precomp fallout) it got a type check error in Perl6::Documentable::new .
ZoffixW glfdex, no idea. has been a million years since I've ran htmlify.p6
12:30 rindolf joined 12:31 [ptc] left 12:33 [ptc] joined
gfldex ZoffixW: i opened an issue on github with versions and stacktrace 12:33
12:33 pierre-vigier joined
nadim ZoffixW: Perl3, the good old days of perl 3 ;) 12:35
Skarsnik I only knew perl 5.6 I think x) 12:36
gfldex, want me to run htmlfy on the doc?
dalek kudo/nom: 9b336d6 | lizmat++ | src/core/List.pm:
Make List.sum handle sparse arrays correctly
12:37
12:38 rurban left
psch m: say :($: $) 12:41
camelia rakudo-moar b668da: OUTPUT«5===SORRY!5=== Error while compiling /tmp/JsrfOEyPkh␤Can only use the : invocant marker in the signature for a method␤at /tmp/JsrfOEyPkh:1␤------> 3say :($: $7⏏5)␤ expecting any of:␤ constraint␤»
psch not sure i like that :/
i mean, i get that a sub or multi signature shouldn't allow the invocant marker, but a bare signature..? 12:42
12:42 ZoffixW left
psch nadim: fwiw, you don't need eval to catch MMD failures 12:43
m: class A { method f() { } }; { A.new.f(1); CATCH { default { .perl.say } } }
camelia rakudo-moar b668da: OUTPUT«X::AdHoc.new(payload => "Too many positionals passed; expected 1 argument but got 2")␤»
gfldex psch: you may want to rakudobug that signature thing 12:44
psch gfldex: well, i know why and how it was changed, and i'm not sure it's actually wrong 12:45
nadim psch: MMD? multi ??? dispatch?
psch gfldex: it is admittedly inconvenient when trying to match a made-up sig against a method, but maybe that's not really something one should (have to) do because MMD is pretty great, actually
nadim: multi method dispatch
nadim: ...not quite right in this context, 'cause method f in my example wasn't a multi... :) 12:46
12:46 ChoHag joined
nadim inconvenient till one need it then whateverway if convenient ;) 12:46
gfldex psch: i can create some anon mathod and steal it's signature to get hold of the signature object, not being able to create one with :() is silly
12:47 Ven joined
psch gfldex: yeah, that still hides some magic from you though, as seen in your attempts to compare earlier 12:49
gfldex: as in, a non-class bound anon method doesn't get the implicit *%_ for interface consistency
nadim I have bee trying to create local method with its signature created from an object, doesn't seem to do what I want. my SomeClass $o; (my method($o)).signature.say.
psch ...or does it? o.O 12:50
m: (my method f($: $) { }).signature.perl.say # 'cause this *does*
camelia rakudo-moar b668da: OUTPUT«:($: $, *%_)␤»
psch oh duh 12:51
Signature.ACCEPTS is "can we dispatch the LHS against the Signature?" not "are these Signatures equivalent"... :P
m: class A { method f(A:D $a: $b) { } }; my $sig = (my method f(A:D $a: $b) { }).signature; say $sig.perl; say A.can('f')[0].signature.perl; say $sig eqv A.can('f')[0].signature
camelia rakudo-moar 9b336d: OUTPUT«:(A:D $a: $b, *%_)␤:(A:D $a: $b, *%_)␤True␤»
12:52 kid51 left, [ptc] left
psch m: class A { method f(A:D $a: $b) { } }; \(A.new(), 2, 2) ~~ A.can('f')[0].signature 12:52
camelia ( no output )
psch m: class A { method f(A:D $a: $b) { } }; say \(A.new(), 2, 2) ~~ A.can('f')[0].signature
camelia rakudo-moar 9b336d: OUTPUT«False␤»
12:54 kaare_ joined
gfldex and even if i cant just ~~ signatures, it would be nice to have a proper error message 12:54
12:55 [ptc] joined
psch gfldex: well, infix:<~~> does "smart" things (FSVO). i understand being surprised by some of them, but i don't think always warning with "this might not do what you assume" is useful 12:56
gfldex: if you want equivalence, use infix:<eqv> vOv
nadim psch: as of nopaste.linux-dev.org/?877908 I am surely doing it right what I want is the new method to accept and object of the type of the argument I used
psch nadim: i don't understand your code example. you want to create a method at compile time with a typed parameter that has the type derived from a value you fetch at runtime..? 12:58
gfldex m: class C { multi method m(Int){}; multi method m(Str){}; }; my $sig = (my method (C: Int) {}).signature; say $sig; say .signature, ' ', so .signature eqv $sig for C.^methods[0].candidates;
camelia rakudo-moar 9b336d: OUTPUT«(C $: Int, *%_)␤(C $: Int, *%_) True␤(C $: Str, *%_) False␤»
nadim psch: I want to do it at run time of course, but do I need to eval it? and can I do it altogether? 12:59
12:59 zakharyas joined
gfldex nadim: see my last m: (s/~~/eqv/) 12:59
nadim gfldex: trying it 13:00
gfldex psch: every time i see some oddity in Perl 6 i ask myself: "How would I document that?". Sometimes the answer smells. 13:01
13:01 brrt joined 13:02 pierre-vigier left
dalek ast: 614d76a | lizmat++ | S02-types/array.t:
Add basic Array.sum tests
13:06
psch gfldex: this is about smartmatch in particular right now? 13:07
gfldex: 'cause ACCEPTS (and thus smartmatch) behavior should be documented with the class that implements it
13:07 n0tjack joined, pierre-vigier joined, csd_ joined
psch gfldex: as in, the Signature should probably say something like "using a Signature as the RHS to infix:<~~> checks if the LHS can be dispatched against the given Signature" or so 13:08
gfldex indeed, and in a dynamic language you may not now what the LHS might be
psch s:1st/Signature/Signature docs/
gfldex how do you look that up?
psch ...why does that matter? you know the signature, you know what dispatching against a signature means
so if it fails you know it's something that doesn't fit the signature 13:09
and if it matches it's something that can be passed to a Routine with that signature and dispatch successfully
gfldex it doesn't fail, it return False
13:10 pierre-vigier left
psch yes, to indicate that passing the LHS to a Routine with the given signature would fail the dispatch 13:10
13:10 lucasb joined
gfldex psch: do i understand the 2nd halve of the following right? doc.perl6.org/type/Signature#method_ACCEPTS 13:13
Skarsnik the search bar does not find this btw x) 13:14
psch m: say :($, $) ~~ :($, $, $?)
camelia rakudo-moar 9b336d: OUTPUT«True␤»
dalek ast: d557444 | lizmat++ | S02-types/list.t:
Add some basic List.sum tests
psch gfldex: that gets a bit more interesting now...
gfldex "may not work with methods" smells :)
psch m: class A { method f($) { } }; say A.can('f')[0].signature ~~ A.can('f')[0].signature 13:15
camelia rakudo-moar 9b336d: OUTPUT«False␤»
13:15 pierre-vigier joined
psch gfldex: right, thanks for clarifying 13:15
m: class A { method f($) { } }; say A.can('f')[0].signature.ACCEPTS(A.can('f')[0].signature)
camelia rakudo-moar 9b336d: OUTPUT«False␤»
psch hm, i wonder if we just mistreat invocant markers there 13:16
gfldex infix:<eqv> is defined for Signature but not documented 13:17
psch m: class A { method f($, *%) { } }; say A.can('f')[0].signature.ACCEPTS(A.can('f')[0].signature) 13:18
camelia rakudo-moar 9b336d: OUTPUT«False␤»
psch m: class A { method f($, %) { } }; say A.can('f')[0].signature.ACCEPTS(A.can('f')[0].signature)
camelia rakudo-moar 9b336d: OUTPUT«False␤»
psch m: sub f(*%) { }; say &f.signature ~~ &f.signature
camelia rakudo-moar 9b336d: OUTPUT«False␤»
gfldex looking at how eqv is implemented i can understand why :)
psch slurpy hashes are wrong in Signature.ACCEPTS 13:19
m: sub f(*@) { }; say &f.signature ~~ &f.signature
camelia rakudo-moar 9b336d: OUTPUT«True␤»
13:19 pierre-vigier left, zakharyas left
lucasb m: say :(*@a) ~~ :(*@a) 13:20
camelia rakudo-moar 9b336d: OUTPUT«True␤»
lucasb m: say :(*%h) ~~ :(*%h)
camelia rakudo-moar 9b336d: OUTPUT«False␤»
Skarsnik m: say int32.perl; 13:22
camelia rakudo-moar 9b336d: OUTPUT«int32␤»
psch m: :(*%).signature.params.classify({.named}).perl.say
camelia rakudo-moar 9b336d: OUTPUT«Method 'signature' not found for invocant of class 'Signature'␤ in block <unit> at /tmp/5hZCxSF_tX:1␤␤»
psch m: :(*%).params.classify({.named}).perl.say
camelia rakudo-moar 9b336d: OUTPUT«(my Any %{Any} = Bool::True => $[*%])␤»
Skarsnik m: say int32.REPL.perl;
camelia rakudo-moar 9b336d: OUTPUT«Method 'REPL' not found for invocant of class 'int32'␤ in block <unit> at /tmp/3HxjNJ5dPH:1␤␤»
psch m: :(*%).params.classify({.named}).grep({.slurpy}).say 13:23
camelia rakudo-moar 9b336d: OUTPUT«Method 'slurpy' not found for invocant of class 'Pair'␤ in block <unit> at /tmp/50ik98fYI5:1␤␤»
psch hm, i'll look closer at this later, got an appointment in a bit o/
Skarsnik does the int32 and co type have a specific trait to distinct them from like Int?
13:24 zakharyas joined
dalek ast: b327867 | lizmat++ | S17-supply/ (43 files):
add some perl6 vim markers
13:24
lucasb m: class C { method m($,$,$,$,$) {} }; say C.can('m')[0].signature 13:25
camelia rakudo-moar 9b336d: OUTPUT«(C $:,,,,, *%_)␤»
lucasb I was thinking that the parameters was getting ignored, but they just don't show up in the output (psch: ^^) :) 13:26
13:27 znpy joined
Skarsnik m: say Int.^roles 13:28
camelia rakudo-moar 9b336d: OUTPUT«((Real))␤»
[Coke] m: say ‘g'day’
camelia rakudo-moar 9b336d: OUTPUT«g'day␤»
Skarsnik m: say int32.^roles
camelia rakudo-moar 9b336d: OUTPUT«Method 'roles' not found for invocant of class 'Perl6::Metamodel::NativeHOW'␤ in block <unit> at /tmp/nufBOHJjk_:1␤␤»
Skarsnik m: say int32.^can('Native')
camelia rakudo-moar 9b336d: OUTPUT«()␤»
Skarsnik m: say int32.^can('NativeHOW') 13:29
camelia rakudo-moar 9b336d: OUTPUT«()␤»
13:30 zakharyas1 joined, vividsnow left 13:32 zakharyas left, rurban joined
dalek c: 22c9d07 | [email@hidden.address] | doc/Language/functions.pod:
Remove typo
13:34
c: b388df6 | nxadm++ | doc/Language/functions.pod:
Merge pull request #228 from nxadm/master

Remove typo
El_Che crap, I need to change my git email
brrt yes, that happens 13:39
masak maybe add a user.name too :) 13:44
El_Che on the other hand, f*ck the nsa. Try to catch me now :) 13:56
13:56 ChristopherBotto joined
El_Che It has been fun to write p6. Although it takes a lot longer than p5. Not because of the language missing batteries, but because wanting to try out al the new goodies :) 13:58
13:59 ChristopherBotto left
FROGGS El_Che: they won't try to catch you directly... they'll organise weapons for .lu :o) 14:00
El_Che :)
14:00 n0tjack left 14:02 cdg joined
El_Che what the overhead of functions in perl6, btw? compared to perl6? better due to types? 14:03
llfourn m: say sub foo:bar {}; 14:04
camelia rakudo-moar 9b336d: OUTPUT«sub foo:bar () { #`(Sub|58055904) ... }␤»
lizmat El_Che: well, small functions already get inlined on heavy use, so YMMV
llfourn m: say sub foo:bar(){};
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/FwEwsl9CXd␤Null operator is not allowed␤at /tmp/FwEwsl9CXd:1␤------> 3say sub foo:bar()7⏏5{};␤»
llfourn bug or LTA error?
lizmat El_Che: but in general, there's an overhead for every scope, just as in P5
El_Che thx
lizmat but with a lot more introspectability, so after 6.c, we can optimize a lot of stuff to be *much* faster 14:05
El_Che \o/
llfourn m: say sub foo:bar () {}; # seems like spaces needed for : routine names 14:07
camelia rakudo-moar 9b336d: OUTPUT«sub foo:bar () { #`(Sub|65318704) ... }␤»
El_Che in an unrelated note, in my use case IO seemed fine (certainly to the easy use of channels). 14:08
14:09 pmurias joined
llfourn std: say sub foo:bar () {}; # seems like spaces needed for : routine names 14:15
camelia std : OUTPUT«No such file or directory»
llfourn wonders if he did it wrong
camelia std : OUTPUT«(timeout)cat: /home/camelia/std/snap/revision: No such file or directory␤»
llfourn thinks std is broken
14:17 MilkmanDan left
moritz looks like it 14:18
14:20 Some-body_ joined 14:22 csd_ left 14:23 skids joined, burnersk left, eiro left, DarthGandalf left, matt_ left, burnersk joined, Some-body_ is now known as DarthGandalf
[Coke] ... You should get that STD checked out. 14:24
14:24 xjrK joined
[Coke] seriously, though, we can probably drop std from camelia at this point. 14:24
timotimo aye, especially since it's not being kept up to date any more 14:25
moritz if the fix is trivial, I'll try to fix it. 14:26
or it seems somebody is already on it?
timotimo i'm not 14:27
[Coke] gah. I have RSS feeds for the various projects, and try to review commits. I'm currently 445 commits behind on nom.
14:27 corbyhaas left 14:28 eiro joined 14:29 matt_ joined, matt_ is now known as Guest50923, aborazmeh joined, aborazmeh left, aborazmeh joined 14:30 corbyhaas joined 14:35 lucasb left, hankache joined
xjrK p6 emacs syntax highlighting? 14:37
github.com/hinrik/perl6-mode works? anyone confirm?
moritz xjrK: I'm not an emacs user, but that's the thing that we recommend when somebody asks 14:38
hankache ohai moritz
xjrK moritz: awesome, figured that. cheers (-:
hankache hi xjrK
xjrK hi hankache, how are you 14:39
hankache xjrK all good mate, how are you?
moritz \o hankache, *
xjrK hankache: not too bad, updating emacs plugins currently... waiting for class to start
hankache xjrK good luck :) 14:40
14:40 nowan left
xjrK hankache: thanks! (-: 14:40
Now starts the commit from p5 to p6....
hankache any volunteers to translate perl6intro.com to your native language? 14:43
moritz hates writing technical content in German
14:43 nowan joined
dalek ast: 19bbd5c | lizmat++ | S17-supply/throttle.t:
Add some more Supply.throttle tests
14:44
moritz I know all the words and phrases and English, and have to search and translate endlessly to express something in German
lizmat
.oO( Datenfernübertragungsanlage )
hankache moritz indeed, I am translating to French and I can tell you it's a pain in the ass
El_Che hankache: it's a difficult group you're looking for. People that know enough english to attempt the translation, imho, read their technical documentation in english, so the native terms would sound strange for them 14:45
hankache i have to search each technical term
14:45 Ven left
hankache indeed El_Che 14:46
14:46 ZoffixW joined
lucs hankache: Fabrique un glossaire et envoie-le moi! :) 14:46
hankache salut lucs
lucs Allo
lucs also sometimes struggles with finding appropriate French technical computer terms. 14:47
hankache merci lucs 14:48
El_Che start with megaoctets and numeration :)
ZoffixW I don't get it, how do I catch "fail" exceptions?
m: sub foo { fail "kaboom!"; my %h = 1 => 2, foo => 'bar'; return %h }; try my %h = foo; CATCH { exit }; say %h<foo>
camelia rakudo-moar 9b336d: OUTPUT«(Any)␤»
Skarsnik maybe use some know dictionary like kde? (but there are probably more user oriented)
CATCH { fail {} } ?
or X::Something? 14:49
hankache afk for 15 min
brb
lucs hankache: You're welcome, but I'm not sure for what :)
ZoffixW m: sub foo { fail "kaboom!"; my %h = 1 => 2, foo => 'bar'; return %h }; my %h = try { foo(); CATCH { default { exit } } }; say %h<foo>
camelia ( no output )
ZoffixW This is sooo messy. Weren't there a cleaner syntax without a ton of curlies?
m: sub foo { fail "kaboom!"; my %h = 1 => 2, foo => 'bar'; return %h }; my %h = try { foo() } or exit; say %h<foo> 14:50
camelia rakudo-moar 9b336d: OUTPUT«Odd number of elements found where hash initializer expected␤ in block <unit> at /tmp/fr8naZnpXm:1␤␤»
ZoffixW :(
14:50 aigan left
Skarsnik yes execptin syntax is weird 14:51
ZoffixW It's way too verbose
14:51 virtualsue joined
ZoffixW m: sub foo { fail "kaboom!"; my %h = 1 => 2, foo => 'bar'; return %h }; my %h = foo() or exit; say %h<foo> 14:52
camelia rakudo-moar 9b336d: OUTPUT«kaboom!␤ in sub foo at /tmp/nmfp9H9UNS:1␤ in block <unit> at /tmp/nmfp9H9UNS:1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp:2883␤ in block <unit> at /tmp/nmfp9H9UNS:1␤␤»
ZoffixW The docs suggest that should work, since I'm checking the returned failure for truth: docs.perl6.org/type/Failure
m: sub foo { fail "kaboom!"; my %h = 1 => 2, foo => 'bar'; return %h }; my %h = foo() || exit; say %h<foo>
camelia ( no output )
ZoffixW Yey \o/
14:53 tyrus joined
moritz yes, precedence 14:53
lucs ZoffixW: By the way, after unsuccesfully trying to build a greasemonkey script to bypass the doc.perl6.org search box, I decided to clone the repo and build the HTML at home.
14:53 hankache left
ZoffixW Now, if I could get the error message from failure, then I'll cream myself 14:53
moritz ZoffixW: "use fatal;"
14:53 xjrK left
[Coke] <steve>Language.</steve> 14:54
lucs ZoffixW: But: make html ... 15/32: doc/Language/grammars.pod => language/grammars -- Type check failed in assignment to @!categories; expected Str but got Array
ZoffixW lucs, that seems like an overkill
m: use fatal; sub foo { fail "kaboom!"; my %h = 1 => 2, foo => 'bar'; return %h }; my %h = foo() || exit; say %h<foo>
camelia rakudo-moar 9b336d: OUTPUT«kaboom!␤ in sub foo at /tmp/BR7k9x4XKO:1␤ in block <unit> at /tmp/BR7k9x4XKO:1␤␤»
lucs ZoffixW: Perhaps, but there it is.
14:55 aborazmeh left
Skarsnik hm 14:55
How do I test if a sub has a specific trait?
ZoffixW moritz, well, I have two types of fails "Network error" and "Location not found"... I'm really hoping for something nice and clean like: my %h = foo() || $! ~~ /Network/ ?? die "Network error" !! die "Location not found" 14:56
moritz m: sub f { fail 'kaboom' }; my %h = f() orelse die "Oh noez: $_.message()\n"; say 42
camelia rakudo-moar 9b336d: OUTPUT«kaboom␤ in sub f at /tmp/xv7usNGlBY:1␤ in block <unit> at /tmp/xv7usNGlBY:1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp:2883␤ in block <unit> at /tmp/xv7usNGlBY:1␤␤»
cdc Skarsnik: &foo.does(MyTrait) I suppose.
ZoffixW
.oO( orelse?? )
moritz std: 42
camelia std : OUTPUT«No such file or directory»
vytas m: use Readline;
camelia rakudo-moar 9b336d: OUTPUT«===SORRY!===␤Could not find Readline:ver<True>:auth<True>:api<True> in:␤ /home/camelia/.perl6/2015.11-407-g9b336d6␤ /home/camelia/rakudo-m-inst-1/share/perl6/site␤ /home/camelia/rakudo-m-inst-1/share/perl6/vendor␤ /home/camelia/rakudo…»
ZoffixW Learn something new every day
camelia std : OUTPUT«(timeout)cat: /home/camelia/std/snap/revision: No such file or directory␤#perl6 <vytas> rakudo-moar: use Readline;␤»
ZoffixW m: sub f { fail 'kaboom' }; my %h = f() orelse say "Oh noez: $_.message()\n"; say 42
camelia rakudo-moar 9b336d: OUTPUT«kaboom␤ in sub f at /tmp/PFLcbYDTmW:1␤ in block <unit> at /tmp/PFLcbYDTmW:1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp:2883␤ in block <unit> at /tmp/PFLcbYDTmW:1␤␤»
cdc m: &foo.does(Callable) 14:57
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8_3iErr_4Y␤Undeclared routine:␤ foo used at line 1␤␤»
14:57 csd_ joined
cdc m: sub foo() {}; &foo.does(Callable) 14:57
camelia ( no output )
cdc m: sub foo() {}; &foo.does(Callable).say
camelia rakudo-moar 9b336d: OUTPUT«True␤»
14:57 uruwi left
Skarsnik Oh thx 14:57
cdc you're welcome :)
moritz m: sub f { fail 'kaboom' }; my %h = f() orelse die "Oh noez: $_\n"; say 42 14:59
camelia rakudo-moar 9b336d: OUTPUT«kaboom␤ in sub f at /tmp/9qde_ZEzkP:1␤ in block <unit> at /tmp/9qde_ZEzkP:1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp:2883␤ in block <unit> at /tmp/9qde_ZEzkP:1␤␤»
14:59 uruwi joined
moritz m: sub f { fail 'kaboom' }; my %h = (f() orelse die "Oh noez: $_\n"); say 42 14:59
camelia rakudo-moar 9b336d: OUTPUT«Oh noez: (HANDLED) kaboom␤ in sub f at /tmp/KKu_wuyyo9:1␤ in block <unit> at /tmp/KKu_wuyyo9:1␤␤␤ in block <unit> at /tmp/KKu_wuyyo9:1␤␤»
14:59 khw joined
moritz precedence strikes again. 14:59
vytas how does Perl6 know what i have installed, I just rebuild my perl, panda. and perl -e 'use Readline' gives me "[..]/dist/B6EB284194164A4B20F99B6D70B0E6CC6CCC5E18: no such file or directory" 15:00
lizmat cycling& 15:01
ZoffixW vytas, it's because its META file is broken. See this PR: github.com/drforr/perl6-readline/pull/2/files
vytas, pull it and install with panda install .
DrForr Woopsie. 15:02
ZoffixW Tun-tun-tuuuun
vytas ZoffixW, i appreciate this issue github.com/drforr/perl6-readline/issues/3
DrForr Merged.
vytas but why the hell Rakudo thinks I got it installed?
DrForr I've been focusing on perl6-ANTLR4 because I'm mentioning it in talks. 15:03
moritz vytas: because you installed it before?
vytas i did not, cause I couldn't because of that issue, so I nuked rakudo and built it again
ZoffixW vytas, yes, that issue can be closed now. The "Failed to copy '/tmp/.panda-work/1449234898_1/lib/Readline.pm6' to '/home/" is due to the wrong META, which just got fixed 15:04
tyrus DrForr Long time!
ZoffixW m: sub f { fail 'kaboom' }; my %h = (f() orelse die "Oh noez: $_.message\n"); say 42
camelia rakudo-moar 9b336d: OUTPUT«Oh noez: (HANDLED) kaboom␤ in sub f at /tmp/DVnk9HddGV:1␤ in block <unit> at /tmp/DVnk9HddGV:1␤.message␤␤ in block <unit> at /tmp/DVnk9HddGV:1␤␤»
ZoffixW m: sub f { fail 'kaboom' }; my %h = (f() orelse die "Oh noez: $_.message()\n"); say 42 15:05
camelia rakudo-moar 9b336d: OUTPUT«kaboom␤ in sub f at /tmp/lpvGY4hNZB:1␤ in block <unit> at /tmp/lpvGY4hNZB:1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp:2883␤ in block <unit> at /tmp/lpvGY4hNZB:1␤␤»
ZoffixW shrugs
DrForr rubs a few neurons together... uh... good *god*. Yes.
vytas all i care is should i worry and report a bug - if so I need to know what to include in bug report, or should I just carry on and install Readline ?
moritz ZoffixW: it's $_ or $_.Str(), not .message
ZoffixW: my mistake 15:06
DrForr vytas: Please report the bug.
ZoffixW m: sub f { fail 'kaboom' }; my %h = (f() orelse die "Oh noez: $_.Str()\n"); say 42
camelia rakudo-moar 9b336d: OUTPUT«Oh noez: (HANDLED) kaboom␤ in sub f at /tmp/XbrUf72LgM:1␤ in block <unit> at /tmp/XbrUf72LgM:1␤␤␤ in block <unit> at /tmp/XbrUf72LgM:1␤␤»
ZoffixW I kinda hate the (HANDLED) is in there
vytas, what's the bug? The Issue you opened? Try again, it's just been fixed
Skarsnik hm, is that a bug? gist.github.com/Skarsnik/bae6082aa16fb6c20b4b (same erreor with is(native) 15:07
timotimo can't you get the original exceptino object out of the failure and get its message?
vytas ZoffixW, nope, issue is that rakudo falsely thinks i got readline installed
moritz timotimo: yes, .exception.message should also work 15:08
ZoffixW m: sub f { fail 'kaboom' }; my %h = (f() orelse die "Oh noez: {$_.exception.message}\n"); say 42
camelia rakudo-moar 9b336d: OUTPUT«Oh noez: kaboom␤␤ in block <unit> at /tmp/O4NtxGOdiB:1␤␤»
ZoffixW timotimo++ moritz++ thanks.
15:09 cdg left
Skarsnik vytas, try reinstalling readline? 15:09
vytas, it probably get lost because the filename was wrong
vytas Skarsnik, that way i would loose environment where it's broken
DrForr I'd better reinstall it myself to see what broke during GLR :( 15:10
15:10 virtualsue left
ZoffixW vytas, I suspect your issue is there because panda failed to copy the files *when the META was broken*. So it marked it as installed, but failed to copy actual files. Just reinstall it with --force 15:10
15:11 cdg joined
Skarsnik fill an issue for panda 15:11
it should not have did that x)
15:11 xjrK joined
vytas thanks all :) 15:11
Skarsnik probably
15:12 cdg left
Woodi hankache: from perl6intro, jargon: rakudobrew isn't *Perl6* installer 15:12
*manager
15:13 tyrus left, znpy left 15:15 MilkmanDan joined, cdg joined
dalek c: 8dc82a5 | (Zoffix Znet)++ | doc/Language/testing.pod:
Add actual code examples for like/unlike
15:21
15:23 RabidGravy joined 15:24 lolisa left 15:25 jkva joined, nadim left
Skarsnik hm, how to work around to test if a sub has the native trait ~~ 15:25
15:25 nadim joined 15:27 Ven joined
timotimo "work around"? 15:27
15:27 cdg left
timotimo m: use NativeCall; sub printf() is native(*); say &printf ~~ does Native 15:27
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tWmT_DfpN_␤A unit-scoped sub definition is not allowed except on a MAIN sub;␤Please use the block form.␤at /tmp/tWmT_DfpN_:1␤------> 3e NativeCall; sub printf() is native(*);7⏏5 say &printf ~~ does…»
Skarsnik gist.github.com/Skarsnik/bae6082aa16fb6c20b4b 15:28
timotimo ah, it's private to NativeCall
m: use NativeCall; sub printf() is native(*); say &printf.^roles.perl
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/cdApVt978A␤A unit-scoped sub definition is not allowed except on a MAIN sub;␤Please use the block form.␤at /tmp/cdApVt978A:1␤------> 3e NativeCall; sub printf() is native(*);7⏏5 say &printf.^roles.…»
timotimo m: use NativeCall; sub printf() is native(*) { }; say &printf.^roles.perl
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/fK4_QK6FN5␤No appropriate parametric role variant available for 'NativeCall::Native'␤at /tmp/fK4_QK6FN5:1␤»
timotimo m: use NativeCall; sub printf() is native(*) { }; say &printf.^roles>>.name.perl
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/W48jmJ8lzZ␤No appropriate parametric role variant available for 'NativeCall::Native'␤at /tmp/W48jmJ8lzZ:1␤»
timotimo darn, it tries to pun that role
however
m: use NativeCall; sub printf() is native(*) { }; say &printf.native_symbol() 15:29
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/5hqYxaTgEh␤No appropriate parametric role variant available for 'NativeCall::Native'␤at /tmp/5hqYxaTgEh:1␤»
timotimo ... huh?
oh, it's the whatever star
m: use NativeCall; sub printf() is native(Str) { }; say &printf.^roles.perl
camelia rakudo-moar 9b336d: OUTPUT«(NativeCall::Native[Sub,Str], Callable)␤»
timotimo m: use NativeCall; sub printf() is native(Str) { }; say &printf.native_symbol()
camelia rakudo-moar 9b336d: OUTPUT«Method 'native_symbol' not found for invocant of class 'Sub+{NativeCall::Native[Sub,Str]}'␤ in block <unit> at /tmp/aZuy9PE7rt:1␤␤»
timotimo ah, that's only for when you "is symbol(...)"
Skarsnik: the thing after "is" isn't what gets mixed in; it's what trait_mod:<is> gets called with 15:30
Skarsnik: that's why it complains about "native" as a term
ZoffixW m: use Test; cmp-ok 42, '<=', 100, 'ayup'
camelia rakudo-moar 9b336d: OUTPUT«not ok 1 - ayup␤␤# Failed test 'ayup'␤# at /tmp/sv26hSqRuk line 1␤# Could not use '<=' as a comparator␤»
ZoffixW :(
Why not? 15:31
m: use Test; cmp-ok 42, '==', 100, 'ayup'
camelia rakudo-moar 9b336d: OUTPUT«not ok 1 - ayup␤␤# Failed test 'ayup'␤# at /tmp/si1MVL3pcB line 1␤# expected: '100'␤Sub+{<anon|73581040>}+{Precedence} object coerced to string (please use .gist or .perl to do that) in block at /home/camelia/rakudo-m-inst-1/share/perl6/sources…»
Skarsnik should does(NativeCall::Native) work?
moritz m: use Test; cmp-ok 42, &infix:[ '<=' ], 100, 'ayup'
camelia rakudo-moar 9b336d: OUTPUT«ok 1 - ayup␤»
moritz ZoffixW: because it's not callable
ZoffixW m: use Test; cmp-ok('my spelling is apperling', '~~', /perl/, "bad speller"); 15:32
camelia rakudo-moar 9b336d: OUTPUT«ok 1 - bad speller␤»
Skarsnik m: use NativeCall; sub printf() is native(Str) { }; say &printf does NativeCall::Native;
camelia rakudo-moar 9b336d: OUTPUT«Could not find symbol '&Native'␤ in block <unit> at /tmp/saVdKhHOOA:1␤␤Actually thrown at:␤ in block <unit> at /tmp/saVdKhHOOA:1␤␤»
ZoffixW Bottom of these docs suggest that at least '==' should work: docs.perl6.org/language/testing#Com...ual_output
m: use Test; cmp-ok 42, &infix:[ '==' ], 42, 'ayup'
camelia rakudo-moar 9b336d: OUTPUT«ok 1 - ayup␤»
Skarsnik m: use NativeCall; sub printf() is native(Str) { }; say &printf ~~ does NativeCall::Native; 15:34
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/dSHdpscUgA␤Undeclared routine:␤ does used at line 1␤␤»
timotimo Skarsnik: it should not. that role is private to the NativeCall module
moritz ZoffixW: ah yes, it's a bug in cmp-ok 15:35
if $op ~~ Callable ?? $op !! try EVAL "&infix:<$op>" -> $matcher {
of course &infix:<$op> won't work if < is part of the operator name
DrForr &infix:<< ... >> for those, no?
Skarsnik m: use NativeCall; sub printf() is native(Str) { }; say &printf.^roles.Str ~~ /NativeCall/; 15:36
timotimo my first thought was it'd have to be because "&infix:<foo>" already tries to interpolate
camelia rakudo-moar 9b336d: OUTPUT«Use of uninitialized value of type NativeCall::Native[Sub,Str] in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed. in any at gen/moar/m-Metamodel.nqp:1693␤Use of uninitialized value of type Callable in st…»
moritz DrForr: that won't work for << in the operator name
Skarsnik m: use NativeCall; sub printf() is native(Str) { }; say &printf.^roles.Str ~~ rw/NativeCall/;
camelia rakudo-moar 9b336d: OUTPUT«5===SORRY!5=== Error while compiling /tmp/tloixOWK5w␤Missing required term after infix␤at /tmp/tloixOWK5w:1␤------> 3say &printf.^roles.Str ~~ rw/NativeCall/7⏏5;␤ expecting any of:␤ prefix␤ term␤»
moritz probably easier to use use &CLIENT::something or so
m: my $name = '<='; say &infix:[ $name ] 15:37
camelia rakudo-moar 9b336d: OUTPUT«Use of uninitialized value <element> of type Any in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed. in any canonicalize_pair at src/Perl6/World.nqp:4012␤Use of uninitialized value <element> of type Any in…»
moritz m: my $name = '<='; say &infix:[ $name ].^name 15:38
camelia rakudo-moar 9b336d: OUTPUT«Use of uninitialized value <element> of type Any in string context␤Any of .^name, .perl, .gist, or .say can stringify undefined things, if needed. in any canonicalize_pair at src/Perl6/World.nqp:4012␤Use of uninitialized value <element> of type Any in…»
timotimo Skarsnik: you really want to use >>.name rather than .Str
Skarsnik m: use NativeCall; sub printf() is native(Str) { }; say &printf.^roles>>.name ~~ /NativeCall/;
camelia rakudo-moar 9b336d: OUTPUT«Method 'name' not found for invocant of class 'NativeCall::Native[Sub,Str]'␤ in block <unit> at /tmp/96vcDaLYnE:1␤␤»
Skarsnik m: use NativeCall; sub printf() is native(Str) { }; say &printf.^roles.perl ~~ /NativeCall/; 15:39
camelia rakudo-moar 9b336d: OUTPUT«「NativeCall」␤»
timotimo BBL
15:39 lucasb joined 15:40 FROGGS left 15:41 g4 left
moritz ZoffixW: I'm currently testing a fix that escapes < in operator names 15:41
ZoffixW cool 15:42
But == also doesn't work
m: use Test; cmp-ok 100, '==', 100
camelia rakudo-moar 9b336d: OUTPUT«ok 1 - ␤»
ZoffixW Oh :S
moritz m: say &infix:<==>.perl
camelia rakudo-moar 9b336d: OUTPUT«sub infix:<==> (Mu $?, Mu $?) { #`(Sub+{<anon|73581040>}+{Precedence}|64656048) ... }␤»
ZoffixW Ah, I was comparing 42 == 100 and that's why it failed, not because of the issue
autarch lizmat: looks like Supply is fixed now
yoleaux 10:58Z <lizmat> autarch: could you test Supply.throttle again and tell me if it fixed your problem?
ilmari m: use Test; cmp-ok 100, &infix:<==>, 100
camelia rakudo-moar 9b336d: OUTPUT«ok 1 - ␤»
ZoffixW The error is LTA tho
m: use Test; cmp-ok 42, '==', 100 15:43
camelia rakudo-moar 9b336d: OUTPUT«not ok 1 - ␤␤# Failed test at /tmp/qShrj579lG line 1␤# expected: '100'␤Sub+{<anon|73581040>}+{Precedence} object coerced to string (please use .gist or .perl to do that) in block at /home/camelia/rakudo-m-inst-1/share/perl6/sources/E44B9B5D5CB949…»
ZoffixW The anon stuff.
ilmari moritz: say ~&infix:<==>
m: say ~&infix:<==>
camelia rakudo-moar 9b336d: OUTPUT«Sub+{<anon|73581040>}+{Precedence} object coerced to string (please use .gist or .perl to do that) in block <unit> at /tmp/S5xMSbAL0e:1␤infix:<==>␤»
ilmari moritz: sorry, tab-complete habit, please ignore 15:44
[Coke] is rt.perl.org/Ticket/Display.html?id=126805 rejectable since a fresh checkout worked? 15:45
Skarsnik doc.perl6.org/routine/signature the link on Signature (grey text) is not there btw x) 15:46
15:48 Guest98452 left 15:49 xpen joined
lucasb If I have PERL6LIB=/path/to/lib set at installation time, it seems precomp ends in /path/to/lib/.precomp instead of ~/.perl6/* . Is this intended/expected? 15:49
15:51 oka^ left, zakharyas1 left 15:52 Khisanth joined
Skarsnik not bad x) 15:55
Mu gumbo_destroy_output - You should not return a non NC type (like Int inplace of int32), truncating errors can appear with different architectures
gumbo_normalized_tagname - Not a valid parameter type for parameter [0] : Int
-->For Numerical type, use the appropriate int32/int64/num64...
root@testperl6:~/piko/nativecall-typediag#
15:57 xjrK left
moritz m: say &::("infix:<\\<=>") 15:57
camelia rakudo-moar 9b336d: OUTPUT«No such symbol '&::infix:<\<=>'␤ in block <unit> at /tmp/UgpON2O8WH:1␤␤Actually thrown at:␤ in block <unit> at /tmp/UgpON2O8WH:1␤␤»
moritz I believe this is a bug :(
15:58 abraxxa left
dalek osystem: 6b5251c | (Zoffix Znet)++ | META.list:
Add WWW::Google::Time to eco

Get time for various locations using Google: github.com/zoffixznet/perl6-WWW-Google-Time
15:58
moritz TimToady: what's the proper name to look up an infix operator by name if it happens to contain a < or «?
Skarsnik Ho interesting 15:59
ZoffixW, should it be webservice::Google:Time?
ZoffixW Skarsnik, why? 16:00
Skarsnik Well there is already some WebService:: module
ZoffixW ¯\_(ツ)_/¯ I really don't see any additional clarity offered by the additional verbosity. 16:01
[Coke] does rt.perl.org/Ticket/Display.html?id=126718 still matter post-precomp? 16:02
ZoffixW I was even this ->| |<- close to naming it Google::Time
Skarsnik Well WWW for me imply you scrap some html x)
but I am weird
ZoffixW Skarsnik, well... I am
lucasb m: say &infix:['<']
camelia rakudo-moar 9b336d: OUTPUT«sub infix:«<» (Mu $?, Mu $?) { #`(Sub+{<anon|73581040>}|65095944) ... }␤»
ZoffixW [Coke], IIRC there's still an issue where use lib conflicts with -I... lemme test real quick 16:03
16:03 xpen left
hoelzro o/ #perl6 16:08
ZoffixW [Coke], couldn't reproduce what I thought the issue was. I closed that ticket. 16:09
hoelzro .seen MadcapJake 16:11
yoleaux I saw MadcapJake 5 Dec 2015 03:07Z in #perl6: * MadcapJake is off to bed!
ZoffixW Skarsnik, but I think I already talked about the whole namespace thing. In most cases, these generic namespaces add nothing but extra characters into your source code... Like this: backpan.perl.org/authors/id/Z/ZO/ZO...001.tar.gz 16:12
Skarsnik great win7 that open me tar.gz with snes9x xD 16:13
ZoffixW m: say 'use POE::Component::IRC::Plugin::CPAN::LinksToDocs::No404s::Remember 0.001;'.chars
camelia rakudo-moar 9b336d: OUTPUT«75␤»
ZoffixW 75 chars! That's almost at the line limit :)
hoelzro .tell MadcapJake do you need more help on your changes to Perl6::Compiler? I would be happy to help you out with it!
yoleaux hoelzro: I'll pass your message to MadcapJake.
16:13 uruwi left
moritz use shorter::identifiers; 16:14
ZoffixW Skarsnik, and there's plenty more examples in backpan.perl.org/authors/id/Z/ZO/ZOFFIX/
Skarsnik use PCIPCLNR
ZoffixW moritz, way ahead of you: backpan.perl.org/authors/id/Z/ZO/ZO...001.tar.gz
Skarsnik, which is why I'm now more reluctant to use extra-long-winded module names :D
16:14 Actualeyes left 16:15 adu joined
Skarsnik Where is Acme::Alltheunicodecodeinthename ? 16:16
RabidGravy right finally fix0rificated Sys::Utmp and Sys::Lastlog
nine++ # resources stuff 16:17
ZoffixW Skarsnik, won't work, the indexer has like 32-char limit
Oh wait never mind, that's for the version
ZoffixW really wish to see more modules per day for P6 on www.modulecounts.com/ 16:20
Skarsnik zoffix: say Module-count;
ZoffixW Go is pretty insane 252 modules per day. Talk about wheel reinvention factory.
Skarsnik lol
16:20 MilkmanDan left 16:21 Ven left
moritz ZoffixW: well, they have no generics, so they probably need an Int32BTree, an Int64BTree, a StrBTree, ... :-) 16:22
arnsholt m: EVAL("say Parameter.WHERE"); say Parameter.WHERE 16:23
camelia rakudo-moar 9b336d: OUTPUT«47722176␤47722176␤»
ZoffixW :D
16:24 MilkmanDan joined 16:25 abaugher_ left 16:26 abaugher_ joined
RabidGravy I think the modules *will* pick up when people start making applications in it 16:26
grondilu are those modules written by humans though? 16:27
16:27 adu left
RabidGravy depends whether you consider go programmers human or not ;-) 16:27
Skarsnik *wrote 3 modules, patched 6-8.. to do stuff with ponies*
pmurias the future according to modulecounts doesn't seem especially good for Perl5 :/ 16:28
yoleaux 4 Dec 2015 21:34Z <[Coke]> pmurias: github.com/coke/perl6-roast-data/b...ummary.out Cannot find module 'tap', so everything dies.
ZoffixW pmurias, how do you figure? If you look at the CPAN chart its climb barely tapers off, which is understandable since there's a module for everything so there's not much new to make... 16:30
16:30 molaf joined 16:31 uruwi joined, pmurias left
ZoffixW okay then :P 16:31
RabidGravy I'm actually thinking of *removing* some modules
ZoffixW RabidGravy, from P6 or P5? 16:32
RabidGravy from P5
Linux::Svgalib will never work again for instance
ZoffixW Oh yeah. I think I've made ~250 modules in my lifetime, but currently there are only ~100 listed.
16:33 pmurias joined
Juerd RabidGravy: Because it's the CP**A**N, I'm reluctant to remove modules. 16:33
16:33 brrt left
Juerd RabidGravy: Someone, some day, may still need your modules or like to be inspired by it. 16:33
RabidGravy: This feeling was made stronger when I visited a Retro Gaming space (like a hackerspace), and learned how hard it is to get drivers and libraries from 10 or 20 years ago. 16:34
RabidGravy: Storage and bandwidth are incredibly cheap, so why remove anything? 16:35
RabidGravy oh there's some twenty year old modules in there that aren't going anywhere
Juerd awesomeretro.org/ by the way
ZoffixW Because there's not much use for bit-rotten code that you don't wish to maintain anymore. There's always BackPan
16:36 tyrus joined
pmurias ZoffixW: it's just the impression from the numbers without actually looking what's inside the repos 16:36
Juerd And awesomespace.nl/
ZoffixW: I hate the whole "there's always backpan" thing that keeps popping up.
16:36 micxjo joined
Juerd ZoffixW: Backpan isn't mirrored as much, is hardly indexed, hard to search through, etc. 16:36
pmurias ZoffixW: the other "dynamic" languages seem to have more modules and faster growth 16:37
RabidGravy yeah, it's useful for getting the older versions of things
Juerd It's a solution to a problem that didn't have to exist, but now because there's a solution, people seem to be very motivated to make the problem bigger :)
ZoffixW Juerd, I've just spent too much time digging through CPAN search results before realizing the module is broken in some way and the author is long AWOL
Juerd ZoffixW: A simple boolean would have done that job so much better.
ZoffixW Boolean? 16:38
Juerd Yeah.. Let's call it is_maintained
It could even be a datetime.
ZoffixW I think mst has given this issue more thought with the MSTPan thing
Juerd I don't know mstpan
RabidGravy could even put an "I want to abandon this" button on PAUSE
Juerd RabidGravy: Yep, that would be good. 16:39
ZoffixW Datetime? Some modules' authors are well and active and the module iself works well without updated
16:39 tyrus left
Juerd ZoffixW: Still, though, I'd like to indicate for DBIx::Simple that it's still relevant. 16:39
ZoffixW RabidGravy, there is. Just give perms to ADOPTME
Juerd, shadow.cat/blog/matt-s-trout/mstpan-1/
Juerd I'm not putting my modules up for adoption because I fear that the future maintainer will remove them and depend on backpan. 16:40
RabidGravy :)
ZoffixW hahaha
Juerd ZoffixW: tl;will read another time, but thanks for the link )
:)
16:44 jkva left 16:50 ZoffixW left 16:52 telex left 16:57 telex joined 16:59 Hor|zon left 17:01 nowan left 17:03 nowan joined, nowan left 17:06 nowan joined
dalek kudo/nom: 7fe2481 | TimToady++ | docs/announce/2015.12.md:
remove a couple of dups
17:08
kudo/nom: f650749 | TimToady++ | docs/announce/2015.12.md:
nick format preference
zostay so... is perl6 going to continue using an oldish version of libuv for the foreseeable future?
i ask because i suspect i've run into a bug in libuv that is causing aborts in my work with supplies when it tries to cleanup one of the threads it's using 17:09
lucasb I've been saying this for months now :)
zostay and by perl6 i actually mean MVM
17:09 TEttinger joined
lucasb I wish it updated to the latest libuv... but now, before the release, it seems risk 17:10
zostay there's always the point release ;)
nine Well we do have the spectest suite for the risk 17:13
TimToady which regularly blows up on threading things... :/
dj_goku is there a reason, you couldn't, git pull the most recent, build, run a roast against it?
nine TimToady: which could be a reason to upgrade if it helps with those 17:14
TimToady troo
dj_goku oh btw libuv has moved from github.com/joyent/libuv to github.com/libuv/libuv
nine Today's advent post made it into the lwn.net news! 17:15
lucasb nine: Can you give me a hand with precomp? If I install rakudo with 'PERL6LIB=/some/path make install', it installs the precomp modules into /some/path/.precomp . Is this intended/expected? 17:17
nine lucasb: yes, that's intended
lucasb nine: oh, ok. I wasn't sure.
nine lucasb: we always store the precompiled modules into the precomp store of $*REPO, i.e. the head repository of the chain. 17:18
lucasb nine: thanks 17:19
nine lucasb: modules are precompiled for a certain chain of repositories which satisfy the module's dependencies. If the chain changes, other versions of those dependencies may be used, invalidating the precompiled modules.
lucasb I'm still figuring it out this whole precomp thing. It just seemed strange 17:20
One example is: 'PERL6LIB=/no/permission make install' fails to install because the user doesn't has permission to mkdir that dir 17:21
nine lucasb: well we assume that PERL6LIB or -I or use lib (they're all the same) are for development purposes where you usually have write permissions. System wide installations of modules go into CompUnit::Repository::Installation repos 17:22
17:23 cognominal_ joined, xjrK joined, AlexDaniel joined
nine However we should probably make those precompilation failures non-fatal. That would probably be an easy patch for anyone annoyed enough to give it a try. 17:23
lucasb Another unrelated thing is that there are 3 repos: share/perl6, share/perl6/site, share/perl6/vendor. What is your opinion about making the first one be share/perl6/core? 17:24
17:24 cdg joined
zostay if you install a new version of rakudo, should it fail on old .precomps? 17:24
or even just rebuild the same version
17:25 cognominal left
nine lucasb: I would very much like that. Or even remove it entirely. What we have right now is nothing but a mirror of Perl 5's directory layout. 17:25
lucasb nine: ah, ok. thanks again 17:26
nine zostay: the SHA hashes in the names of precompiled files include the compiler-id which changes on every rakudo build. If you rebuild rakudo, it won't use those files anymore.
ugexe lucasb: i had the same problem... a read only network share where the modules can't use each other via -I unless they have actually been installed. it makes testing the validitity of multiple packages before installing a huge pain 17:30
zengargoyle can one say: PERL6LIB=/path1 make install; then move the repo: mv /path1 /path2/elsewhere; and then PERL6LIB=/path2/elsewhere perl6 ...; ?? 17:31
ugexe paths are hard coded in the byte code i believe
geekosaur hrm 17:33
jnthn You can't relocate it. 17:34
yoleaux 02:42Z <skids> jnthn: Questions that arose while updating S17 gist.github.com/skids/fabd0304f40e24db33d0
10:09Z <lizmat> jnthn: I think there's a race condition in supply syntax exposed by "(1..5).Supply.throttle( 4, { say "$*THREAD.id(): start $_"; $_ } ).wait;"
10:10Z <lizmat> jnthn: fwiw, I've seen S17-supply/syntax.t hang occassionally: feels like the throttle example is exposing this more reliably
zengargoyle that's sad. :(
geekosaur not so good for staged deployment (binary OS packages, AFS/staged NFS, etc.) 17:35
jnthn Then...submit patches!
I've nothing against it being relocatable provided it's done in a sane way
zengargoyle strangely enough, my one Perl 5 contribution was related to relocation. 17:36
jnthn ;-)
TimToady do you count putting each path into a padded 1k buffer and providing a binary patcher as "sane"?
jnthn TimToady: I tried to binary patch Parrot like that but it didn't go too well :P 17:37
I'd prefer if we could find a better way, but I can probably live with that option :)
geekosaur (-headerpad-max-install-names) 17:38
vytas perl6 doesn't have goto, does it ?
[Coke] m: HERE: goto HERE;
camelia rakudo-moar f65074: OUTPUT«Label.goto() not yet implemented. Sorry. ␤ in block <unit> at /tmp/cjf9tmH_ul:1␤␤»
zengargoyle is there a reason paths can't be relative?
17:39 silug joined 17:40 pistacchio joined
TimToady m: say 3 RRRxx rand 17:41
camelia rakudo-moar f65074: OUTPUT«(0.33914404302382 0.618942086312358 0.188375246650337)␤»
TimToady jnthn: ^^
! works too, but still thinking about Z and X...
and [&&]
I can naïvely make rand Xxx 1,2,3 work, but that sort of approach breaks <a b c> Xxx 1,2,3 17:42
nadim hi, is cached is not implemented yet on methods, wouldn't it be better t have a warning (that one can hide somehow) rather than a compile error? This will mean modifying the code when is cached is available rather than just use it when available. 17:43
jnthn TimToady: \o/ 17:44
nadim: I'm going to make "is cached" experimental in the next week or so, and we'll get it fixed up for 6.d or something.
[Coke] nadim: going from an error to providing an implementation is ok. going from being quiet to dying is bad.
jnthn++ 17:45
jnthn We didn't figure out what the cache keying mechanism should be, there's a difficult to fix issue with its interaction with multis, and the method semantics need a decent bit of thought. I'd rather just postpone it and get it right than rush it now and live with it being broken forever.
nadim Thank you for the explanation :) 17:46
jnthn Also, this may encourage exploration of the issue in module space, and then we may get to pick a better solution from there :)
17:46 domm left
TimToady "use experimental" should probably be "use MONKEY-BUSINESS" or so 17:48
nadim You know better, I'll implemet a little, specific, cache in the mean time
17:51 virtualsue joined, dakkar left
[Coke] monkey-wrench? monkey-bread? monkey-island? monkey's-uncle/ 17:56
dj_goku running through a promise example I got this error: fpaste.scsys.co.uk/502519
17:59 vendethiel joined
TimToady monkey-hear-no-evil? 18:00
dalek c: 2dcd172 | (David H. Adler)++ | doc/Language/5to6-perlfunc.pod:
Adjusted entry on <fc> in 5to6-perlfunc to reflect that it is now
18:00 dha joined
dj_goku m: my $p = Promise.new; $p2.break('bad'); say 'result ' ~ $p2.result; 18:03
camelia rakudo-moar f65074: OUTPUT«5===SORRY!5=== Error while compiling /tmp/gVifQpnoNB␤Variable '$p2' is not declared␤at /tmp/gVifQpnoNB:1␤------> 3my $p = Promise.new; 7⏏5$p2.break('bad'); say 'result ' ~ $p2.re␤»
flussence zostay: you can build moarvm with --has-libuv, it runs fine with 1.7.5 18:04
dj_goku ug
m: my $p = Promise.new; $p.break('bad'); say 'result ' ~ $p.result;
camelia rakudo-moar f65074: OUTPUT«bad␤ in block <unit> at /tmp/875uSn1GQD:1␤␤»
18:05 hankache joined
dj_goku m: my $p = Promise.new; $p.break('bad'); say $p.status; say $p.result; 18:05
camelia rakudo-moar f65074: OUTPUT«Broken␤bad␤ in block <unit> at /tmp/pNZ2FII_vF:1␤␤»
18:05 llfourn left
dj_goku am I doing something wrong to get that error? 18:07
[Coke] you're trying to get the result of a promise that died, and when you try to access it, it dies for you right then. I -think- that's desired behavior. 18:08
18:08 pmurias left
flussence m: my $p = Promise.new; $p.break('bad'); say $p.result.WHAT; 18:09
camelia rakudo-moar f65074: OUTPUT«bad␤ in block <unit> at /tmp/U2FBSklZAX:1␤␤»
[Coke] m: my $p = Promise.new; $p.keep('yay!'); say $p.result;
camelia rakudo-moar f65074: OUTPUT«yay!␤»
[Coke] m: my $p = Promise.new; $.break('bah!'); { say $p.result; CATCH { say "oh, it died" } } 18:10
camelia rakudo-moar f65074: OUTPUT«5===SORRY!5=== Error while compiling /tmp/wDJgKW1LiR␤Variable $.break used where no 'self' is available␤at /tmp/wDJgKW1LiR:1␤------> 3my $p = Promise.new; $.break('bah!')7⏏5; { say $p.result; CATCH { say "oh, it d␤»
[Coke] m: my $p = Promise.new; $p.break('bah!'); { say $p.result; CATCH { say "oh, it died" } }
camelia rakudo-moar f65074: OUTPUT«oh, it died␤bah!␤ in block <unit> at /tmp/3t4AEusqKs:1␤␤»
[Coke] ^^
18:10 pistacchio left
dj_goku [Coke]: so then this example should be updated accoridingly? doc.perl6.org/language/concurrency#Promises 18:12
m: my $p = Promise.new; $p.break('bad'); say $p.status; {TRY say $p.result; CATCH { say "oh, it died" } } 18:13
camelia rakudo-moar f65074: OUTPUT«5===SORRY!5=== Error while compiling /tmp/YS1t9KbAIZ␤Undeclared name:␤ TRY used at line 1␤␤»
RabidGravy no TRY
don't need it
dj_goku m: my $p = Promise.new; $p.break('bad'); say $p.status; {say $p.cause; CATCH { say "oh, it died" } } 18:14
camelia rakudo-moar f65074: OUTPUT«Broken␤bad␤»
dj_goku m: my $p = Promise.new; $p.break('bad'); say $p.status; {say $p.result; CATCH { say "oh, it died" } } 18:15
camelia rakudo-moar f65074: OUTPUT«Broken␤oh, it died␤bad␤ in block <unit> at /tmp/hc0zOttF6A:1␤␤»
dj_goku RabidGravy: interesting. :D
RabidGravy that CATCH still dies though
dj_goku why is that? 18:16
RabidGravy m: my $p = Promise.new; $p.break('bad'); say $p.status; {say $p.result; CATCH { default { say "oh, it died - $_ " } } }
camelia rakudo-moar f65074: OUTPUT«Broken␤oh, it died - bad ␤»
RabidGravy CATCH is like a magic "given" for exceptions 18:17
you need to match the exception to stop it being thrown
timotimo well, anything that sets a $_ acts as a given, no? 18:18
psch m: with "foo" { when "foo" { say "dunno" } }
camelia rakudo-moar f65074: OUTPUT«dunno␤»
psch well, maybe :P
RabidGravy so without at least a default it will throw
18:19 spider-mario joined
timotimo the thing is, the failure stores a "handled" flag, which gets set to True when it gets matched against and returns true 18:22
so if a "when" with the right type or a default (not sure how exactly that part works, tbh) checks against it, it'll be handled
when a check for the wrong exception type comes up, the flag won't be set
ah 18:23
default calls "succeed" on the exception payload 18:24
psch m: 5 + "foo"; CATCH { say $_.defined }
camelia rakudo-moar f65074: OUTPUT«WARNINGS:␤Useless use of "+" in expression "5 + \"foo\"" in sink context (line 1)␤True␤Cannot convert string to number: base-10 number must begin with valid digits or '.' in '3⏏5foo' (indicated by ⏏)␤ in block <unit> at /tmp/MWTTrwPKqS:1␤…»
18:24 uruwi left
flussence ooh, green eject 18:25
Skarsnik I should install my modules: root@testperl6:~/piko/Ferrel# perl6 -I ../perl6-gumbo/lib -I . -I /home/skarsnik/devel/http-useragent/lib -I ../Discord/lib/ -I ../config-simple/lib/ Ferrel.p6 18:27
x)
timotimo hehe 18:29
flussence just put them in a big META6.json somewhere and -I that :)
timotimo at some point we have to have something like FatPacker 18:30
dj_goku alright just to make sure I got everything. I need a default given option else it'll still throw an exception. 18:31
ahh: Waits for the promise to be kept or broken. If it is kept, returns the result; otherwise throws the result as an exception. from doc.perl6.org/type/Promise#method_result 18:32
I guess that is why $p.cause is there. 18:34
18:36 hankache left
timotimo hum. 18:42
the perl6advent.wordpress.com/ theme is super pretty, but now it's impossible to find the yearly summary/linklist posts without scrolling all the way down
dj_goku timotimo: maybe tag by year? add the ability to filter by year? 18:43
timotimo perl6advent.wordpress.com/?s=Table...20Contents <- btw. 18:45
18:45 ZoffixW joined
dj_goku ahh 18:45
ZoffixW m: 0/0 or "kabloom"; say "No crisis"
camelia rakudo-moar f65074: OUTPUT«No crisis␤»
ZoffixW :)
timotimo but nothing links to that - except for the perl6weekly i'm writing right now
ZoffixW: did you select the wordpress theme for the advent calendar?
grondilu m: say 0/0 18:46
camelia rakudo-moar f65074: OUTPUT«Attempt to divide by zero using div␤ in block <unit> at /tmp/0iij_1c0Iv:1␤␤Actually thrown at:␤ in block <unit> at /tmp/0iij_1c0Iv:1␤␤»
ZoffixW timotimo, yeah
m: 0/0; say "No crisis"
camelia rakudo-moar f65074: OUTPUT«No crisis␤»
grondilu it ignores it in sink context or something?
flussence it returns a failure
ZoffixW Yeah 18:47
timotimo can we get either a link to the search for "Table of Contents" somewhere near the top or side, or ... something else?
18:47 yqt joined
timotimo actually 18:47
0/0 gives you a Rat object
m: say (0/0).WHAT
camelia rakudo-moar f65074: OUTPUT«(Rat)␤»
timotimo m: say (0/0).nude
camelia rakudo-moar f65074: OUTPUT«(0 0)␤»
timotimo it only explodes when you try to do something else with it
flussence oh, I guess .Str gives you the failure then
timotimo m: 0e0/0e0; say "no problem"
camelia rakudo-moar f65074: OUTPUT«Attempt to divide by zero using /␤ in block <unit> at /tmp/2OAzWr5mXS:1␤␤Actually thrown at:␤ in block <unit> at /tmp/2OAzWr5mXS:1␤␤»
flussence m: 0/0 .Str
camelia ( no output )
flussence m: 0/0 .Str.WHAT
camelia rakudo-moar f65074: OUTPUT«===SORRY!===␤Method call must either supply a name or have a child node that evaluates to the name␤»
timotimo ^- Num does it immediately
flussence m: say (0/0).Str.WHAT 18:48
camelia rakudo-moar f65074: OUTPUT«Attempt to divide by zero using div␤ in block <unit> at /tmp/hIPcOGf5Of:1␤␤Actually thrown at:␤ in block <unit> at /tmp/hIPcOGf5Of:1␤␤»
timotimo wow, how did ... %)
grondilu m: (0/0)*0
camelia rakudo-moar f65074: OUTPUT«WARNINGS:␤Useless use of "*" in expression "(0/0)*0" in sink context (line 1)␤»
grondilu m: say (0/0)*0
camelia rakudo-moar f65074: OUTPUT«Attempt to divide by zero using div␤ in block <unit> at /tmp/6vxwzlXXWg:1␤␤Actually thrown at:␤ in block <unit> at /tmp/6vxwzlXXWg:1␤␤»
timotimo ZoffixW: also, for some reason, the wordpress extra bar that logged-in people get at the top ... it lags behind when scrolling
flussence yay presentational javascript! 18:49
timotimo very yay indeed
flussence there are many reasons to use js for positioning, emulating position:fixed is not one 18:50
ZoffixW It is if you're an ancient code base that powers 30% of the web :)
timotimo, TBH I don't see any way to mangle anything in there. There's an option to use a static home page (so no recent posts) or to choose another theme. 18:51
18:51 koo8 joined
timotimo ugh. 18:51
ZoffixW Unless I simply don't know how to use Wordpress :) 18:52
timotimo wordpress has things like "aside" or something that show up differently in the blog. maybe something like that could work? pinned to the top, linking to all table of contents posts?
ZoffixW I'll take a look later today, when I'm on my fast box at home. 18:53
timotimo cool. TYVM 18:54
18:54 zengargoylew joined
dalek kudo/nom: fb957f8 | grondilu++ | src/core/native_array.pm:
improve permutations

avoid using C<first> for better performance
18:56
kudo/nom: d7dfa68 | grondilu++ | src/core/native_array.pm:
Update native_array.pm

use native integers in permutations
kudo/nom: 6af4098 | lizmat++ | src/core/native_array.pm:
Merge pull request #621 from grondilu/patch-3

improve permutations
ZoffixW .tell MadcapJack Found a really annoying breakage of the syntax highlighter: apostrophes in regexes: say so "You're wrong" ~~ /'You\'re wrong'/; That breaks all the following code after that line :( 18:58
yoleaux ZoffixW: I'll pass your message to MadcapJack.
timotimo grondilu: i think you might be able to get a tiny bit more performance out of "permutations" by replacing "if !@!a { return (@!a = ^$!n).List }" with "return (@!a = ^$!n).List if !@!a" (because we avoid an inner scope that would likely get a closure taken)
18:58 patrickz joined
timotimo grondilu: also i have no clue if our .= code-gen is any good 19:00
19:01 virtualsue left, virtualsue joined 19:02 Peter_R joined
patrickz Hey, rakudo master on moarvm fails to install Inline::Perl5 on linux here. Tests fail with p5helper.sh not found. Is that a known error? 19:02
dalek osystem: cf45f31 | (Zoffix Znet)++ | META.list:
Add WWW::You'reDoingItWrong to eco

Why say someone is doing it wrong when you can SHOW it?: github.com/zoffixznet/perl6-WWW-Yo...ingItWrong
ZoffixW :)
ZoffixW is pleased with an obvious troll of apostrophes in identifiers for the naysayers :P 19:03
19:03 rurban left
Skarsnik what 19:03
19:03 llfourn joined
timotimo hm. what name does cygx want to appear with in the weekly i wonder? 19:03
ZoffixW Skarsnik, there was a discussion a while back that apostrophes should not be allowed in identifiers and there's no use for them :) 19:04
dalek kudo/nom: 2d3bb09 | lizmat++ | docs/announce/2015.12.md:
dolmen === Olivier Mengué
Skarsnik Oh yeah lol
AlexDaniel m: my @a = [ [1800, 20], [2800, 40] ]; for @a -> [ $foo, $bar ] { say “$foo ↔ $bar” } # didn't expect [ ] thingy to work, nice!
camelia rakudo-moar f65074: OUTPUT«1800 ↔ 20␤2800 ↔ 40␤»
19:05 domm joined
dalek kudo/nom: 7e0a4fc | lizmat++ | docs/announce/2015.12.md:
Simplify daxim
19:05
kudo/nom: ce64674 | lizmat++ | docs/announce/2015.12.md:
Add nicks for lizmat and woolfy
19:06
19:07 llfourn left
timotimo ShimmerFairy: do you want me to put your full name into the weekly where i link to your advent post? 19:11
psch m: say :(*%).params.classify(*.named && *.slurpy){True}[0].type # this doesn't fit the expectation of src/core/Signature.pm:66, i think... 19:12
camelia rakudo-moar ce6467: OUTPUT«(Associative)␤»
psch m: say so :(*%).params[0] ~~ { !(.type =:= Mu) } 19:13
camelia rakudo-moar ce6467: OUTPUT«True␤»
psch m: :(*@).params[0].type.say
camelia rakudo-moar ce6467: OUTPUT«(Positional)␤»
psch hm, a named slurpy has to be (a sub class of) Associative, doesn't it? 19:14
psch .oO( or is that "hasn't it"..? ) 19:15
m: say Associative.HOW.^name
camelia rakudo-moar ce6467: OUTPUT«Perl6::Metamodel::ParametricRoleGroupHOW␤»
ZoffixW timotimo, BTW, my complaint with the last Weekly was that real names without nicks were used. So all the bunch of seemingly random people were named, even though I see them all the time here under nicks 19:16
awwaiid why does panda on nom hate me?
(retorical)
ZoffixW awwaiid, try feeding it eucalyptus
awwaiid bamboo
lizmat ZoffixW: I thought I *did* add the nicks in the last P6W ? 19:17
awwaiid I don't want to kill my panda!
timotimo ZoffixW: lizmat prepared a bit for the weekly and it had both full names and nicknames in parens
dha awwaiid - you are insufficiently covered in bamboo
timotimo oh there liz is :) 19:18
Skarsnik moar hate everything and make panda angry x)
lizmat ah, yes, for this week's...
hehe
lizmat was a bit timewarped
ZoffixW :P
dha If Perl 6 doesn't give you a TARDIS, what good is it? 19:19
timotimo github.com/masak/tardis
ZoffixW :o
dha :-p
DrForr o/ Drinking / Those moments when o/ 19:20
dha I am slightly annoyed that I'm going to be in London in a period where Doctor Who is not airing. 19:21
El_Che dha: use your tardis to go to the right time
awwaiid I've implemented some fork() based timetravel debuggers as well, which can serve as an alternate theory of time travel to build upon
timotimo gfldex, ZoffixW, are you all right with appearing with your realnames in the weekly?
awwaiid (haven't ported to perl6 yet)
DrForr El_Che: But that trick *never* works!
psch it might work if he releases the hand brake..? 19:22
ZoffixW timotimo, I'm fine :)
dha I don't think time travel debuggers actually have traveling in time and space implemented as yet.
We have names??
Skarsnik timotimo, you did not mention skarsnik-stuff.blogspot.fr/2015/11/...nt-to.html (if it's interesting enought that it)
timotimo move all variables half a meter to the right
i did not? but i didn't rite last weeks' post! 19:23
i'll include it this time, though!
19:23 espadrine left
DrForr timotimo: You forgot. Variables don't, constants aren't and functions won't. 19:23
19:24 uruwi joined
dha "scrapping"? 19:24
19:24 lolisa joined
DrForr I'd like to do that ... pretty much every working day. 19:24
dha No offense, but shouldn't that be scraping?
Skarsnik maybe? 19:25
El_Che I had to secretly propose something small so that the kids could buy for xmas. I suggested a miniature weeping angel and dalek :)
Skarsnik fr.wikipedia.org/wiki/Web_scraping ah yes, thx
ZoffixW Skarsnik, "scraping" is like scraping a barrel. "scrapping" is like... recycling :) 19:26
ZoffixW is reminded of Fallout 4
flussence
.oO( the dalek activates in the middle of the night and yells out commit ids )
Skarsnik the url will stay the same probably :(
timotimo i found the post extra amusing when it still said "scrapping" everywhere
19:27 micxjo left
ZoffixW :P 19:27
Skarsnik I just edit it x)
dha El_Che - I'm pretty sure I've seen christmas tree ornament versions of both. 19:29
El_Che dha: I want them for the office. So I can filter people dropping by :)
ZoffixW "weeping angel"? That's a helluva ornament.. very festive :S
El_Che it's not for a tree :) 19:30
ZoffixW Ah
Hah :D
DrForr They have those in the new LEGO release, I think.
El_Che my kids got me a bunch of the simpsons one :)
DrForr Which I need to see if they have down at Forbidden Planet when I get back. 19:31
El_Che s
dha Yeah, I'm sure you can find figures for your desk. Certainly Daleks. Those come in all kinds of sizes. Some even talk.
19:31 tipdbmp joined
El_Che does perl6 has a similiar shortcut like perl5's moose for an attribute: init_arg => undef? (it made de argument no settable by new, but it generates a getter) 19:32
dha: no, I alread send the amazon links to my wife :)
flussence El_Che: «has $.attr is readonly»? 19:33
timotimo flussence: that'll still let you initialize it, though
El_Che (is readonly the default?)
Skarsnik Yes
for $.foo 19:34
19:34 lucasb left
El_Che I can of course use a $! var and create a getter myself 19:34
just wondering is there was an not-yet-documented-shortcut :)
weeping angel: www.amazon.de/Doctor-Who-Light-Up-W...ping+angel
flussence
.oO( maybe we should copy IO.open's mode flags wholesale for attrs )
timotimo should be possible to write such a thing as a module
El_Che dalek: www.amazon.de/gp/product/0762449314...ailpages00 19:35
timotimo: the getter will not be a lot longer than the import of the module
timotimo yeah, but if you have more than one attribute of that sort ... :) 19:36
El_Che I liked that feature about Moose, hence the question
dj_goku timotimo: RabidGravy so my confusion about Promises earlier. Would updating the doc/Language/concurrency.pod be worth stating to use $p.cause to get the cause of the failure vs $p.result which will throw an exception?
Skarsnik m: class A { has $.foo; method foo() {$!a = 42}} class B { has A $.a }; my B $f = B.new; $f.a.foo();
camelia rakudo-moar ce6467: OUTPUT«5===SORRY!5=== Error while compiling /tmp/DlScwSX5KK␤Attribute $!a not declared in class A␤at /tmp/DlScwSX5KK:1␤------> 3 A { has $.foo; method foo() {$!a = 42}}7⏏5 class B { has A $.a }; my B $f = B.new;␤ expecting any of:␤ h…»
Skarsnik m: class A { has $.foo; method foo() {$!foo = 42}} class B { has A $.a }; my B $f = B.new; $f.a.foo();
camelia rakudo-moar ce6467: OUTPUT«5===SORRY!5=== Error while compiling /tmp/50SwVrqOIm␤Strange text after block (missing semicolon or comma?)␤at /tmp/50SwVrqOIm:1␤------> 3 { has $.foo; method foo() {$!foo = 42}}7⏏5 class B { has A $.a }; my B $f = B.new;␤ expecting an…»
timotimo dj_goku: um, i don't remember talking to you about that?
Skarsnik m: class A { has $.foo; method foo() {$!foo = 42}}; class B { has A $.a }; my B $f = B.new; $f.a.foo();
camelia rakudo-moar ce6467: OUTPUT«Cannot look up attributes in a type object␤ in method foo at /tmp/SfpnP3RMfx:1␤ in block <unit> at /tmp/SfpnP3RMfx:1␤␤»
Skarsnik oups 19:37
m: class A { has $.f; method foo() {$!f = 42}}; class B { has A $.a }; my B $f = B.new; $f.a.foo();
camelia rakudo-moar ce6467: OUTPUT«Cannot look up attributes in a type object␤ in method foo at /tmp/RxtYAw578T:1␤ in block <unit> at /tmp/RxtYAw578T:1␤␤»
dj_goku timotimo: oh I guess I asked [Coke]
timotimo Skarsnik: the "scrapping" typo is still alive in the URL <3
Skarsnik timotimo, I don't know if blogger will let me change it
flussence Cool URLs Don't Change™ 19:38
timotimo dj_goku: but that text already tells you about try $promise.result and say $promise.cause
Skarsnik Can't change it :(
ZoffixW Just leave it :)
timotimo though ... huh 19:39
why is the comment mentioning the output where say $promise.cause is?
19:39 daniel90 joined 19:41 jkva joined
daniel90 Why does "my int @grid[10,10];" work but "my Int @grid[10,10];" does not? The error message is "Method 'gist' not found for invocant of class 'NQPMu'" 19:44
Skarsnik what? 19:45
jnthn m: my Int @grid[10,10];
camelia ( no output )
jnthn m: my Int @grid[10,10]; say @grid
camelia rakudo-moar ce6467: OUTPUT«Method 'gist' not found for invocant of class 'NQPMu'␤ in block <unit> at /tmp/XttDCgyQNM:1␤␤»
jnthn m: my Int @grid[10,10]; say @grid.WHAT
camelia rakudo-moar ce6467: OUTPUT«(Array)␤»
jnthn m: my Int @grid[10;10]; say @grid.WHAT
camelia rakudo-moar ce6467: OUTPUT«(Array)␤»
jnthn m: my Int @grid[10;10]; say @grid
camelia rakudo-moar ce6467: OUTPUT«Method 'gist' not found for invocant of class 'NQPMu'␤ in block <unit> at /tmp/Fv61lgOwqF:1␤␤»
jnthn wtf
I have no ide
*idea 19:46
El_Che (I wanted to yell: netbeans!) :)
19:46 lolisa left 19:47 sno left
timotimo m: my Int @grid[10;10]; say @grid.shape 19:48
camelia rakudo-moar ce6467: OUTPUT«(10 10)␤»
timotimo m: my Int @grid[10;10]; say @grid.Str
camelia rakudo-moar ce6467: OUTPUT«Cannot unbox a type object␤ in block <unit> at /tmp/CY_LU1RSBr:1␤␤»
timotimo m: my Int @grid[2;2] = (1, 2), (5, 6); say @grid.Str 19:49
camelia rakudo-moar ce6467: OUTPUT«1 2 5 6␤»
timotimo m: my Int @grid[2;2] = (1, 2), (5, 6); say @grid
camelia rakudo-moar ce6467: OUTPUT«[[1 2] [5 6]]␤»
timotimo so if nothing gets assigned, things go boom?
ZoffixW m: use MONKEY-TYPING; augment class Mu { method WTF { self.gist } }; say Mu.WTF
camelia rakudo-moar ce6467: OUTPUT«(Mu)␤»
ZoffixW Skarsnik, ^
Skarsnik, though that doesn't make it available on everything :(
timotimo yeah
ZoffixW Skarsnik, how did you manage to have Test::META in prereqs and yet have a failing META? :D 19:50
m: say "6" before v6.c 19:51
camelia rakudo-moar ce6467: OUTPUT«True␤»
ZoffixW m: say v6 before v6.c
camelia rakudo-moar ce6467: OUTPUT«False␤»
ZoffixW hm. I wonder what values are acceptable for "perl" key in META. Seem a lot of folks are putting "v6" or just "6"
flussence m: say Version.new("6") eqv Version.new("v6") 19:52
camelia rakudo-moar ce6467: OUTPUT«False␤»
ZoffixW m: say Version.new("6") before Version.new("v6") 19:53
camelia rakudo-moar ce6467: OUTPUT«False␤»
ZoffixW Skarsnik, PR sent: github.com/Skarsnik/acme-wtf/pull/1 19:54
timotimo m: say Version.new("6"); say Version.new("v6")
camelia rakudo-moar ce6467: OUTPUT«v6␤vv.6␤»
timotimo ^-
hoelzro is there an advantage to using the SHA-1 of a name for the compunit files? it kind of makes it hard to figure out which files belong to which distribution 20:00
dj_goku timotimo: I see a couple examples down yes it does try $promise.results and then say $promise.cause. I was referring to the first example though. I should have read the comment that it was going to die/throw an exception.
ZoffixW This is funny: ecosystem's META.list has 474 lines. modules.perl6.org build log says it found 473 dists. Two are errored out and the final site says we have 470 dists in the ecosystem :(
"There are two hard things in computer science: cache invalidation, naming things, and off-by-one errors."
flussence modules.perl6.org/#q=foo 20:01
ZoffixW Well, 474 <-> 473 is just my Wrong module not built yet, but where is that 1 missing dist going off to *squints*
Juerd ZoffixW: The hardest thing for me as a programmer continues to be estimating work time.
dalek kudo/nom: 328e95e | TimToady++ | docs/announce/2015.12.md:
unicodify name
Juerd ZoffixW: Something that feels like it should take a few days is sometimes completely done in an hour, and some things that really shouldn't take more than 15 minutes can easily take a week. 20:02
ZoffixW Yeah, I hear ya
timotimo definitely 20:03
flussence hoelzro: it's a mixture of not having to deal with IO::Spec limitations of various OSes and needing a unique ID keyed on VM+source so we don't get those old "wrong version of serialization blah" errors
[Coke] Juerd: I've been getting estimates wrong for 20 years!
Skarsnik thx ZoffixW++
psch uhm, help 20:04
Juerd [Coke]: I've stopped estimating most projects. Clients hate that, but I feel so much better. And everything gets done faster because I'm not wasting time on estimations.
psch m: say :(*%).params.classify({.named}){True}.SetHash.keys[0] ~~ { !( .type.does(Associative) ) } # junction-less
camelia rakudo-moar ce6467: OUTPUT«False␤»
psch m: say any(:(*%).params.classify({.named}){True}.SetHash.keys[0]) ~~ { !( .type.does(Associative) )} # lhs junctioned
camelia rakudo-moar ce6467: OUTPUT«True␤»
psch there clearly is only one element, right..?
timotimo i thought i'd take quite a long time to get the basic line-coverage reporting thing in moar done and then a bit of extra time to make it work well .. turned out the first working version was done very quickly
ZoffixW flussence, thanks, I only now realized what your link meant. Still dists should be indexed by name + auth + url
ZoffixW suspects some part just makes it all about name + auth 20:05
hoelzro flussence: that makes sense; I was just wondering if something two-way (so I could tell what dist I'm looking at) would be better
but maybe it's just the lack of tooling for dealing with compunit repos
ZoffixW Oh wait.
timotimo psch: i wonder if you get the junction on the RHS of ~~ as the $_
m: say any(1, 2) ~~ { $_.WHAT.say }
camelia rakudo-moar ce6467: OUTPUT«(Junction)␤True␤»
timotimo psch: there you go. that's your problem
ZoffixW That should—correctly—be one dist, right? 'cause author is the same and the name are the same
psch m: my $x = 0; say any($x) ~~ { .defined }; say $x ~~ { .defined }
camelia rakudo-moar ce6467: OUTPUT«True␤True␤»
ZoffixW k, mystery solved
Skarsnik yep definitly need a tool for compiled stuff x) 20:06
psch timotimo: well, not "mine" particularly, but yes, seems junction dispatch to ACCEPTS in that case is wrong...
m: my $x = 0; say any($x) ~~ { .isa(Junction) }; say $x ~~ { .isa(Junction }
camelia rakudo-moar ce6467: OUTPUT«5===SORRY!5=== Error while compiling /tmp/05graXxRJu␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/05graXxRJu:1␤------> 3(Junction) }; say $x ~~ { .isa(Junction 7⏏5}␤»
psch m: my $x = 0; say any($x) ~~ { .isa(Junction) }; say $x ~~ { .isa(Junction) }
camelia rakudo-moar ce6467: OUTPUT«True␤False␤»
timotimo yeah, we explicitly allow ~~ to check for junctions on the LHS
so ~~ dispatches on Mu, not on Any 20:07
psch timotimo: as in every ACCEPTS has to deal with Junction itself?
timotimo aye
psch that seems brittle :/
Hotkeys how terrible of an idea would it be to have ≈ as a pretty unicode alternative to ~~
psch Hotkeys: i don't think the not-Perl-6 pragmatics for ≈ match what infix:<~~> does particularly well
ZoffixW .u ≈ 20:08
yoleaux U+2248 ALMOST EQUAL TO [Sm] (≈)
Hotkeys fair enough
timotimo psch: wait ... this could actually be about { } having $_ as Mu
m: ({ $_ }).signature.perl.say
camelia rakudo-moar ce6467: OUTPUT«:(;; $_? is raw)␤»
timotimo m: ({ $_.perl.say })(any(1, 2)) 20:09
camelia rakudo-moar ce6467: OUTPUT«any(1, 2)␤»
timotimo m: (-> Any $_ { $_.perl.say })(any(1, 2))
camelia rakudo-moar ce6467: OUTPUT«1␤2␤»
timotimo m: any(1, 2, 3) ~~ -> Any $_ { .perl.say }
nadim can someone help me re-write this unsigtly code nopaste.linux-dev.org/?878167, specially with those nifty '>>' that i haven't well understood yet and that I haven't found documentation for either.
camelia rakudo-moar ce6467: OUTPUT«1␤2␤3␤»
timotimo psch: there you go. this is your *actual* problem :)
gfldex timotimo: my realname is very weekly friendly
20:09 davercc joined
psch timotimo: right, that seems cleaner. as in, we still have something fundamentally WAT-y with parametricity of an implicit $_ in any block, but at least ACCEPTS is still sensible 20:10
Hotkeys speaking of hyperoperators, does any documentation of them exist on the docs.perl6.org site exist?
nadim what's the name of the >> operator?
psch m: say $_.VAR
camelia rakudo-moar ce6467: OUTPUT«Any␤»
Hotkeys >> and << are hyperoperators
psch m: say { $_.VAR }(0)
camelia rakudo-moar ce6467: OUTPUT«0␤»
moritz nadim: it's "hyper", and it's a meta operator 20:11
timotimo RabidGravy is Jonathan Stowe, right?
20:11 Peter_R left
RabidGravy yes 20:11
20:11 Sqirrel joined
Hotkeys nadim: perl6advent.wordpress.com/2009/12/...yperspace/ 20:11
pretty old but I think the information is still accurate 20:12
20:12 darutoko left
dha A troll through the doc directory shows almost no information relating to the world "hyper operator" 20:12
Hotkeys at least wrt syntax and usage
dha bah, autocorrect put in that space.
nadim Hotkeys: thanks, don't know where it is on doc.P6
Hotkeys dha: they are technically metaoperators, according to moritz (see above) they're just referred to as 'hyper' 20:13
psch timotimo: { } implicitly getting a signature of Mu $_ is worth an RT, isn't it?
timotimo: 'cause i remember one about $_ in the mainline being Mu and that being wrong...
(well, unless it's already bugged...)
dha And the two instances that *are* there are not terribly informative.
timotimo psch: it's designed that way
except if you mean {} without any code in it 20:14
well, has to be {;} in that case
otherwise it'd be a hash literal
psch timotimo: no, i'm talking about a block with one or more statements in it
dha Hotkeys - looking just for "hyper" doesn't look like it gives much more of use.
timotimo you're saying it shouldn't be Mu?
Hotkeys /o\
brb 20:15
getting on the bus
psch timotimo: i think it should be Any, yes
timotimo: to allows Junctions ~~ Block to DWIM
El_Che related to my question about unsettable attributes (like "init_args => undef" in Moose):
psch s/Junctions/Junction/
El_Che m: class Foo { has $!priv = 42 }; Foo.^add_method('priv', my method priv(Foo:) { Foo!priv }); Foo.^comp; my $foo = Foo.new(); $foo.priv
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/wyDHTcuDPP␤Private method call to priv must be fully qualified with the package containing the method␤at /tmp/wyDHTcuDPP:1␤------> 3('priv', my method priv(Foo:) { Foo!priv7⏏5 }); Foo.^comp; my $foo …»
El_Che so, what would be the MOP syntax for Foo!priv ?
(was checking of I would change the behaviour by trait_mod) 20:16
(not a lot of doc on the subject)
dha There are a couple of places where it says hyper operators can be used, with an example, but at a glance, I'm not seeing an actual section that explains hyper operators in any detail.
El_Che m: class Foo { has $!priv = 42 }; Foo.^add_method('priv', my method priv(Foo:) { 'stringthing' }); Foo.^comp; my $foo = Foo.new(); $foo.priv 20:17
camelia rakudo-moar 328e95: OUTPUT«Method 'comp' not found for invocant of class 'Perl6::Metamodel::ClassHOW'␤ in block <unit> at /tmp/M_2kBWWGO5:1␤␤»
timotimo psch: i can't find where in the design it says that { $_ } should have Mu, but i think it's explicitly wanted
jnthn El_Che: Think it's add_private_method 20:18
El_Che jnthn: same 20:19
jnthn: the idea was to add a public method as a getter for a private attribute
psch timotimo: similarly, i remember an RT ticket that complains about $_ being Mu when it should be Any, which i also can't find right now... :) 20:20
timotimo here's an idea: just fix it in the code, find which spec tests break. that'll show you what exact synopsis requires that :)
ZoffixW m: my @a = 4, 8, * + 4 ... *; say @a
camelia rakudo-moar 328e95: OUTPUT«[...]␤»
ZoffixW m: my @a = 4, 8, * ... *; say @a
camelia rakudo-moar 328e95: OUTPUT«[...]␤»
psch i don't think i want to dig into that. i'm gonna work-around in Signature.pm for now
ZoffixW :/ 20:21
psch as in, provide the -> Any $_ signature for the block on the smartmatch RHS
jnthn El_Che: Oh...then you don't want add_private_method at all, you need to get the Attribute meta-object and use get-value on it
Hotkeys Hello
Made it to the bus
El_Che jnthn: thx, I'll give it a try
timotimo congrats, Hotkeys
Hotkeys Thanks
I got to the stop just as it was arriving 20:22
ZoffixW m: my @a = 4, 8, 12 ... *; say @a
camelia rakudo-moar 328e95: OUTPUT«[...]␤»
timotimo m: my @a = 4, 8 ... *; say @a[^10]
camelia rakudo-moar 328e95: OUTPUT«(4 8 12 16 20 24 28 32 36 40)␤»
timotimo m: my @a = 4, 8 ... *; say @a.perl
camelia rakudo-moar 328e95: OUTPUT«Cannot .elems a lazy list␤ in block <unit> at /tmp/jqsVeX_OA5:1␤␤Actually thrown at:␤ in block <unit> at /tmp/jqsVeX_OA5:1␤␤»
timotimo m: my @a = 4, 8 ... *; say @a.gist
camelia rakudo-moar 328e95: OUTPUT«[...]␤»
timotimo ..right
ZoffixW weird
k, got it 20:23
m: my @a = 2, 4, 6 ... 12; @a >>/=>> 2; say @a
camelia rakudo-moar 328e95: OUTPUT«[1 2 3 4 5 6]␤»
ZoffixW Pretty neat linenoise :)
20:24 labster joined
Hotkeys Looks nicer when you use the pretty version 20:25
psch ...wait
20:25 espadrine joined
psch actually, i don't get what that line is supposed to do in the first place? 20:25
20:25 sno joined
ZoffixW m: my @a = 2, 4, 6 ... 12; @a »/=» 2; say @a 20:26
camelia rakudo-moar 328e95: OUTPUT«[1 2 3 4 5 6]␤»
ZoffixW psch, how about now?
timotimo psch: look better than ^6, i suppose?
Hotkeys Does it need to be /=?
ZoffixW m: my @a = 2, 4, 6 ... 12; @a »/» 2; say @a
camelia rakudo-moar 328e95: OUTPUT«[2 4 6 8 10 12]␤»
ZoffixW m: my @a = 2, 4, 6 ... 12; say @a »/» 2;
camelia rakudo-moar 328e95: OUTPUT«[1 2 3 4 5 6]␤»
psch ZoffixW, timotimo: uhm, sorry. i'm talking about src/core/Signature.pm:66 :)
ZoffixW lol
Hotkeys Ah
ZoffixW m: my @a = 2, 4, 6 ... 12; say @a «/» 2;
camelia rakudo-moar 328e95: OUTPUT«[1 2 3 4 5 6]␤»
Hotkeys I would've tried it but my phones irc client is bad at copy pastint 20:27
timotimo psch: oh! that's definitely a bug, then
Hotkeys And yet andchat is still the best option for irc on Android iirc
ZoffixW points Hotkeys to IRC::Client
psch timotimo: huh? yes, the bug caused by that line is :(*%) ~~ :(*%) giving False
ZoffixW May the force be with you
Hotkeys Do you want me to write a perl 6 app for android 20:28
Is that even possible
ZoffixW Hotkeys, YES!
psch timotimo: further it also showed me the Blocks-without-Sig-get-Mu-typed-topic
ZoffixW Yes, I want you to. Possible? Probably
timotimo yeah
psch timotimo: but i think that's not even needed to fix the bug in Signature smartmatch, and line 66 can just be removed
'cause it doesn't actually do anything useful
Hotkeys Also I've been making a bot with janky irc stuff for a little while now why did you have to make me aware of this module
psch as in, we know the RHS has a slurpy named, we're in the conditional for a slurpy named for the LHS, so we can just match 20:29
well, except for the case that we don't know the order of nameds..?
Hotkeys Does IRC::Client support irc v3 stuff?
psch m: say :(:$a, *%) ~~ (:$b, *%) # probably False, and should be?
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/BXAB8mUrBu␤Variable '$b' is not declared␤at /tmp/BXAB8mUrBu:1␤------> 3say :(:$a, *%) ~~ (:7⏏5$b, *%) # probably False, and should be?␤»
psch m: say :(:$a, *%) ~~ :(:$b, *%) # probably False, and should be? 20:30
camelia rakudo-moar 328e95: OUTPUT«False␤»
Hotkeys If not should I make an IRC::Client::v3 module
dalek ast: 6757126 | usev6++ | S05-substitution/subst.t:
Fudge for JVM (samemark, ordbaseat NYI)
Hotkeys :p
psch m: say :(:$a, *%) ~~ :(:$a, *%) # probably False, and shouldn't be!
camelia rakudo-moar 328e95: OUTPUT«False␤»
20:30 _nadim joined
moritz iirc we had a bug report about that today 20:30
20:30 nadim left
psch moritz: yeah, i was around then. i'm just not quite sure how to solve 20:31
it'll probably come to me over the next day or so, i don't feel quite thinky enough anymore anyway... :)
ZoffixW m: sub infix:<❤> ($a, $b) { "$a ❤ $b" }; say ^6 »❤» 'Perl 6' 20:32
camelia rakudo-moar 328e95: OUTPUT«(0 ❤ Perl 6 1 ❤ Perl 6 2 ❤ Perl 6 3 ❤ Perl 6 4 ❤ Perl 6 5 ❤ Perl 6)␤»
ZoffixW \o/
20:32 hankache joined
ZoffixW Hotkeys, ATM it's a barebones socket really :P Stay tuned for January 2016 20:33
timotimo ZoffixW: travis-ci.org/zoffixznet/perl6-WWW-Google-Time
Hotkeys Should I stick with my janky bot irc functions for now then 20:34
timotimo ZoffixW: also, shouldn't that go into the WebService:: namespace?
Skarsnik funny you say that x)
Hotkeys My irc bot is currently one file and is in my opinion a mess of spaghetti
ZoffixW timotimo, but why? irclog.perlgeek.de/perl6/2015-12-07#i_11671157 20:35
Hotkeys But that's okay
For my eyes only right now
One day I will dejankify it
ZoffixW timotimo, and travis hasn't yet indexed it (same goes for WWW::You'reDoingItWrong)
timotimo OK
i didn't realize it's that new
ZoffixW m: say 'WebService'.chars - 'WWW'.chars 20:36
camelia rakudo-moar 328e95: OUTPUT«7␤»
20:36 geraud joined
awwaiid How do you pass a 3rd param to an infix operator, like =~= tolerance? 20:37
timotimo with =~= we have a dynamic variable that it looks up
otherwise you can append an adverb
awwaiid adverb was my suspect, tell me more 20:38
(how to do an adverb on =~=) 20:39
timotimo m: sub infix:<^-^>($a, $b, :$carefully) { $a + $b + ($carefully ?? 10 !! 0) }; say 10 ^-^ 5 :carefully; say 10 ^-^ 5
camelia rakudo-moar 328e95: OUTPUT«25␤15␤»
ugexe i vote for InternetSuperHighway::Google::Time
Hotkeys I'm disappointed my keyboard on Android doesn't have the Unicode =~=
psch infix:<=~=> doesn't take an adverb, it uses $*SIGNIFICANCE (iirc)
timotimo m: say &infix:<=~=>.candidates>>.signature 20:40
camelia rakudo-moar 328e95: OUTPUT«((|c is raw))␤»
timotimo can't tell from this, but yeah, i think it only accepts the dynamic variable
ZoffixW I was kinda eyeing InternetSuperHighwayOfInformation::GoogleTheInformationGiant::TimeLookupInVariousLocations
awwaiid m: say 0.1e0 + 0.2e0 =~= 0.3e0 :tolerance(1e-16)
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/kTyzwPGDoQ␤You can't adverb &infix:<=~=>␤at /tmp/kTyzwPGDoQ:1␤------> 3.1e0 + 0.2e0 =~= 0.3e0 :tolerance(1e-16)7⏏5<EOL>␤»
Hotkeys I support this idea ZoffixW
timotimo not bad
awwaiid well looking at the code ≅ does 20:41
Hotkeys But ZoffixW keep in mind the internet is less a highway and more a series of tubes
20:41 abaugher_ left
ZoffixW ehehe 20:41
awwaiid m: say &infix:<≅>.candidates>>.signature
camelia rakudo-moar 328e95: OUTPUT«(($?) (\a, \b, :$tolerance = { ... }) (Complex:D \a, Complex:D \b --> Bool) (Complex:D \a, Real \b --> Bool) (Real \a, Complex:D \b --> Bool))␤»
awwaiid m: say 0.1e0 + 0.2e0 ≅ 0.3e0 :tolerance(1e-16)
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/z92dB0tqcd␤You can't adverb &infix:<≅>␤at /tmp/z92dB0tqcd:1␤------> 3 0.1e0 + 0.2e0 ≅ 0.3e0 :tolerance(1e-16)7⏏5<EOL>␤»
psch ah, TOLERANCE it is
well, $*TOLERANCE
m: say 1e-10 =~= 1e-9; my $*TOLERANCE = 5; say 1e-10 =~= 1e-9 # somewhat..? vOv 20:42
camelia rakudo-moar 328e95: OUTPUT«Use of uninitialized value of type Any in numeric context in block <unit> at /tmp/lzYHBJbxmU:1␤False␤True␤»
20:42 abaugher_ joined
psch not sure what the uninit warning does there... 20:42
awwaiid I mean ... ≅ signature has :$tolerance... 20:43
psch m: say infix:<=~=>(1e-10, 1e-9, :tolerance(1))
camelia rakudo-moar 328e95: OUTPUT«True␤»
psch m: sub infix:<foo>($, $, :$a) { $a ?? "yup" !! "nup" }; say 1 foo 2 :a 20:44
camelia rakudo-moar 328e95: OUTPUT«yup␤»
psch well, i'm off to bed o/
Hotkeys Nighty night
Don't let the perl bugs bite 20:45
(I assume the major contributors here sleep in the codebase)
ZoffixW \o
Hotkeys I need to contribute more
My only contribution is a single line removal in nqp 20:46
ZoffixW Hotkeys, go wild :) rt.perl.org/Search/Results.html?Qu...lled%27%29
Hotkeys Oh nooooo
Juerd Hotkeys: That's still more than > 99,9999% of all people ever did for Perl 6 :)
timotimo why doesn't Acme::WTF appear in the ecosystem site? 20:47
Hotkeys ZoffixW: are you trying to xss me because the site thinks you are
timotimo hahaha
Skarsnik Meta was wrong?
timotimo yeah, rt.perl.org thinks supplying search parameters for you is a very dangerous thing 20:48
Skarsnik not sure why
timotimo if you don't fix it soon, it won't appear in the weekly!
ZoffixW Stupid RT
dalek ast: bc4b127 | usev6++ | / (3 files):
Some more fudging for JVM
Skarsnik It's fixed x)
ZoffixW timotimo, it was already fixed! :) It will appear on next modules.perl6.org rebuild
timotimo oh, ok
when is that?
ZoffixW looks
timotimo there it is 20:49
ZoffixW Aha! :)
timotimo Skarsnik, ZoffixW, you'll appear as "Skarsnik" and "Zoffix" in the post, unless you tell me you want your realname there, too 20:50
Skarsnik I don't mind if you use my real name
ZoffixW timotimo, "Zoffix Znet"
El_Che jnthn: no get-value or similiar (with ^methods(:all) or something from here: doc.perl6.org/type/Metamodel::ClassHOW). Adding methods works, but I am not every to retrive the value from a private attribute in the MOP: Private method call to priv must be fully qualified with the package containing the method
I'll try tomorrow. Will do the dishes now :) 20:51
20:51 ZoffixW left
El_Che thx everyone for the help 20:51
20:55 hudo joined
awwaiid so... psch just demonstrated that it _should_ work but doesn't? 20:56
20:57 Sqirrel left
Hotkeys m: say 2³²; say ²³²; 20:59
camelia rakudo-moar 328e95: OUTPUT«4294967296␤4294967296␤»
Hotkeys That seems like it would be easy enough to fix
Inb4 not at all easy to fix 21:00
m: say 42 + ²
camelia rakudo-moar 328e95: OUTPUT«44␤»
Hotkeys Just to confirm superscripts on their own probably shouldn't be treated as regular numbers right? 21:01
Juerd On their own they're pretty weird
m: ².WHAT 21:02
camelia ( no output )
Juerd m: ².WHAT.say
camelia rakudo-moar 328e95: OUTPUT«(Int)␤»
Hotkeys I'll take a look-see when I get home
dha How would you treat them other than as regular numbers? Aren't they just numbers being used in a particular way?
21:03 hankache left
Hotkeys m: say ³²; # just kinda weird imo 21:03
camelia rakudo-moar 328e95: OUTPUT«9␤»
Hotkeys Since we also have them used for exponentiation
21:04 leont joined
dha Hrm. Ok, I can certainly see that in that case, superscript as exponentiation is not applied consistently. 21:04
21:04 pwd_ left
dha Which generates two different responses for me. 1) "huh. weird." 2) "don't do that" 21:05
Hotkeys Don't do what?
diakopter well superscript shouldn't mean exponentation if the base is also superscript
timotimo the rule that parses that should actually give you 32 for ³² i think 21:06
Juerd I'd say that if ³ == 3, then ³² == 32 makes sense.
But I'm not sure that ³ == 3 is sane
timotimo but it seems like fater the ³ it's satisfied that it has a number and then it makes it an exponentiation
lizmat feels to me a lone ³ should be a compile time error 21:07
vendethiel agreed with lizmat :)
Hotkeys I agree too
dalek ast: 484803b | usev6++ | S (2 files):
Unfudge two passing tests for JVM
dha Hotkeys - I guess what I'm thinking is don't use a number in superscript if you don't want some kind of special behavior for it. Or maybe I think something else. I'm not sure I know anymore. 21:08
Hotkeys I agree with lizmat's idea 21:09
lizmat at news.ycombinator.com/item?id=10690499 someone is say that "say "नि".codes" gives 3
m: say "नि".codes
camelia rakudo-moar 328e95: OUTPUT«2␤»
lizmat could someone mention that they were probably using an outdated rakudo for the 3 answer ?
dha Perhaps I am unaware of something. Is there some behavior superscripting confers on a number other than exponentiation? 21:10
21:10 Skarsnik left
timotimo nah 21:10
dha If not, I'd say using a superscripted number without a non-superscripted number for it to act upon should be an error.
timotimo m: say 2²3 21:11
dha YMMV, of course.
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/_XqgVnClhK␤Two terms in a row␤at /tmp/_XqgVnClhK:1␤------> 3say 2²7⏏053␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement…»
jnthn The ligatures comment in that post is interesting, in so far as something that worked at codepoint level will also say that a ligature is one codepoint... 21:12
Hotkeys lizmat: got it
AlexDaniel m: say 6.perl
camelia rakudo-moar 328e95: OUTPUT«6␤»
Juerd m: Ⅳ.say
camelia rakudo-moar 328e95: OUTPUT«4␤»
Juerd dha: I think the full blown unicode digit support is at work here :) 21:13
jdv79 dha: thanks for talking to mr doy
Juerd m: ↂ.say
camelia rakudo-moar 328e95: OUTPUT«10000␤»
jdv79 i appreciate it
dha jdv79 - sure.
AlexDaniel .u ↂ
jnthn .tell skids Thanks for the S17 gist; I look through it tomorrow when I've (hopefully) more brane :)
yoleaux U+2182 ROMAN NUMERAL TEN THOUSAND [Nl] (ↂ)
Juerd sees interesting new opportunities for code obfuscation :)
yoleaux jnthn: I'll pass your message to skids.
lizmat m: say uniname ↂ
camelia rakudo-moar 328e95: OUTPUT«UPPER RIGHT PENCIL␤»
AlexDaniel dha: there's also a bug report
jdv79 dha++
Juerd m: 21:14
argh
lizmat m: say uniname "ↂ"
camelia rakudo-moar 328e95: OUTPUT«ROMAN NUMERAL TEN THOUSAND␤»
AlexDaniel dha: rt.perl.org/Public/Bug/Display.html?id=126732
21:14 MadcapJake joined
dha jdv79 - I get the impression that the page will probably see some updating after perl 6 actually exists and a book proposal can actually be put together. :-) 21:14
lizmat m: say uniname 10000 21:15
camelia rakudo-moar 328e95: OUTPUT«UPPER RIGHT PENCIL␤»
dalek ast: b517276 | usev6++ | S10-packages/precompilation.t:
Keep ticket number (RT #122773)
lizmat should uniname be taking ints like that ?
MadcapJake .botsnack
yoleaux :D
16:13Z <hoelzro> MadcapJake: do you need more help on your changes to Perl6::Compiler? I would be happy to help you out with it!
jnthn Yeah
AlexDaniel lizmat: why not?
Juerd m: say ↈ/ↅ²*६
camelia rakudo-moar 328e95: OUTPUT«16666.666667␤»
jnthn It's got a codepoint and string form
as a multi iirc
dha AlexDaniel - I would probably agree with that bug report. But I'm not that strong on unicode. It's possible something is relevant here that I don't know about. *shrug*
lizmat m: say uniname ↂ # was a WAT to me 21:16
camelia rakudo-moar 328e95: OUTPUT«UPPER RIGHT PENCIL␤»
AlexDaniel dha: nah it's just that perl 6 treats all number-like things as real numbers, and superscript stuff acts like a postfix or something
21:16 rindolf left
AlexDaniel dha: so since it does not expect a postfix there, it treats the first digit as just a number… 21:16
Juerd AlexDaniel: Yes, &postfix:sym<ⁿ>
AlexDaniel dha: and only then it's like “oh, here's a superscript thing!” 21:17
21:17 dha left, dha joined
MadcapJake hoelzro: Yeah I wasn't able to get the change to work, it just keeps falling back to the CATCH block, It was giving me "cannot invoke this object", iirc and it was a NQPMu. Not sure where to go from there but I would definitely appreciate the help! 21:17
some part of me feels like it's a non-issue if panda is installing .so files now (Thought I read that it was again) 21:18
AlexDaniel m: say ½
camelia rakudo-moar 328e95: OUTPUT«0.5␤»
21:18 virtualsue left
Juerd mQ say ₃⁴ 21:18
AlexDaniel m: say ½²
camelia rakudo-moar 328e95: OUTPUT«0.25␤»
Juerd m: say ₃⁴ 21:19
camelia rakudo-moar 328e95: OUTPUT«81␤»
jdv79 sehr gut herr adler. danke.
dha bitte
AlexDaniel m: say ₃₃
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/HtdTpOZaDQ␤Bogus postfix␤at /tmp/HtdTpOZaDQ:1␤------> 3say ₃7⏏5₃␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement mod…»
hoelzro MadcapJake: panda can handle that now, but I think it wouldn't be a bad thing to be robust
lizmat wonders why jdv79 starts talking german to dha ?
hoelzro otherwise, a bad Linenoise install busts the REPL
RabidGravy MadcapJake, yes it is installing .so files, you need to put a resources key in the META.info
Juerd Given that ₃₃ == 33, shouldn't ³³ == 33 for consistency, instead of 27 or a compile time error?
jdv79 im in berlin. osmosis? 21:20
idk...
MadcapJake RabidGravy: cool, good to know!
AlexDaniel Juerd: but ₃₃ is not 33
timotimo but it should be!
21:20 rurban joined
Juerd Oh, it's not 21:20
Why not?
₃ == 3
AlexDaniel I don't think that it should be
timotimo huh. 21:21
Juerd TBH I don't really understand why all these things are supported anyway.
AlexDaniel what if you write ½½ – what would that mean?
MadcapJake hoelzro: I don't mind following through, I'm just not sure what the next course of action would be. To me, it seems that really the problem is actually not being caught further up when it was getting obtained via that `eval` call (but I'm a Perl6 noob and a NQP n00b!)
RabidGravy I'm quite happy that I can't type these superscript numbers anyway so it'll probably never trouble me
Juerd I don't really see use cases for ⅺ
AlexDaniel Juerd: one reason to support weird digits is because ½ was requested so many times
Juerd m: Ⅿ.say
camelia rakudo-moar 328e95: OUTPUT«1000␤»
AlexDaniel Juerd: and if support that, why not the whole unicode? :) 21:22
21:22 morlock joined, molaf left
TimToady that it can be abused is not a strong argument here 21:22
Juerd AlexDaniel: If you support ½ then supporting ¹⁄₂ makes equal sense 21:23
There's no end to that :)
dalek ast: 9800a63 | usev6++ | S32-str/lines.t:
Fudge some more tests for \r\n for JVM
hoelzro MadcapJake: if we could accomplish that, I would prefer we test Linenoise for functionality in the eval
perhaps history loading
AlexDaniel Juerd: well, you are right, we could support that too :) 21:24
Juerd I don't understand what Ⅿ is, by the way. Is it a digit or a constant?
hudo i like to have in a hash or an array only the newest version of each file in a directory, the files look like 2015-12-07_105935_user3.csv
21:24 mohae_ is now known as mohae
timotimo Juerd: it's the digit 21:24
lizmat Juerd: well, I guess it's really a number
AlexDaniel timotimo: then why can't you write Ⅿ42?
Juerd m: ⅭⅯ 21:25
camelia rakudo-moar 328e95: OUTPUT«5===SORRY!5=== Error while compiling /tmp/CIItF7XtO_␤Bogus postfix␤at /tmp/CIItF7XtO_:1␤------> 3Ⅽ7⏏5Ⅿ␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ statem…»
Juerd I'd expect digits of the same class to be composable, like how 4 and 2 make 42
lizmat but the composing rules for roman numerals are different
Juerd Yep.
timotimo now that i think about that stuff, it makes sense that we can only ever put one single "special" digit in a row 21:26
Juerd I'm not saying I like the support for roman numerals to begin with
I like it because it's quirky but it doesn't seem worth any CPU cycles or memory to begin with, it being kind of outdated and all :)
lizmat I think pre 6.c roman numerals should be considered a number, so not composable
Juerd Why have Roman numerals at all? 21:27
lizmat after 6.c, we could consider multi-char roman numerals to do the right thing
Juerd Or subscript digits
lizmat it came with the rest of the numeric codepoints
afaik
AlexDaniel hudo: Hmm, do you expect it to be a built-in feature? :)
hudo: I'm just thinking that you can just code it traditionally 21:28
hudo: like loop over the files, parse the filename, see if that file already exists… if not, add it. If it exists, then what you are seeing now is an older version
Juerd Serious question: do these things (supporting all those weird Unicode digits) cause actual performance slowdowns, and if so, is it something that's ever going to be optimizable? 21:29
AlexDaniel hudo: given that your file format is sortable, and it is!
Juerd Or are they practically free to support?
AlexDaniel Juerd: I don't think that it causes any slowdown if you don't use these things
Juerd: in fact, even if you use them, it is probably converted just once during the compilation 21:30
Juerd Does it noticeably affect parsing speed?
lizmat Juerd: I would think they only cause compilation slowdown, which would be unnoticeable when precomped
AlexDaniel at least that would make sense, I don't know the implementation
21:30 kaare_ left
AlexDaniel and even if it causes any compilation slowdown, then why not get rid of everything else… like fancy quotes, texas versions and so on 21:31
Juerd I would not oppose that
timotimo Juerd: i think we're just looking for the unicode property, so it's not very expensive
there is, of course, no run-time performance impact
Juerd I don't know if it's a package deal where if you support fancy quotes you get fancy digits for free, or if it's perhaps no performance hit at all...
I understand it's not relevant for runtime 21:32
AlexDaniel I think that it is not something you should worry about
_nadim is there a place where any, the junction, is examplified. wat I'd like to write is something like if any(@list) .name ¨¨ 'something'
Juerd What's ¨¨? 21:33
21:33 morlock left
jnthn char classes are things you can resolve in log(n) where n is the number of things in the char class, so the effect of adding more things that could be matches is increasingly less costly per thing :) 21:33
Juerd jnthn: Great explanation, thanks :)
RabidGravy you mean @list.grep( { $_.name eq 'something'}) 21:34
hudo paste.ee/p/CrpNj
Juerd _nadim: any(@list».name), perhaps?
hudo AlexDaniel, i did it somehow, but im sure there are better ways
dalek kudo/nom: e2e23fb | TimToady++ | src/ (3 files):
use only lc thunk types to allow uc for listy ops
RabidGravy let me take you down to thunky town 21:35
AlexDaniel hudo: hmm this does not look like Perl 6
_nadim I'll try
hudo no it is not
RabidGravy nope definitely Perl 5 21:36
hudo wrong channel ?
AlexDaniel hudo: depends!
RabidGravy if your question is a Perl 5 question then it's the wrong channel 21:37
AlexDaniel hudo: maybe you want to try writing the same thing in Perl 6 and see if it ends up being much shorter, who knows!
jdv79 #perl is goid for p5
good
dha No, actually, it really isn't. :-)
#perl-help maybe. 21:38
21:38 bjz_ left
hudo hm, i was never occupied with perl6 21:38
RabidGravy I wouldn't have thought so either but hey what do I now
hudo what would be an appropiate start ?
AlexDaniel RabidGravy: actually, it would be fun to answer #python questions here :)
hudo: download and install rakudo!
jdv79 on freenode? 21:39
timotimo p6weekly.wordpress.com/2015/12/07/...commences/ - with the password "hello#perl6", you can give me feedback on the weekly before i publish it!
dha AlexDaniel - because we would answer them wrongly? :-)
RabidGravy the docs.perl6.org
jdv79 #perl on freenode is cool.
AlexDaniel hudo: perl6.org/ there is a download button
dha: well, it might not compile for them but it is definitely not wrong :) 21:40
dha :-)
21:41 ZoffixMobile joined
[Coke] timotimo: weird font issue on some bullet points. 21:41
starting with "on opening files"
timotimo ah, probably things i pasted got a bad font
ZoffixMobile timotimo, lol, don't use Peter Evans for me
timotimo ZoffixMobile: lizmat did :)
will change. 21:42
[Coke]: huh, could it be the rel="bookmark" attribute that does that?!
ZoffixMobile heh, not sure how that percolated from Facebook lol :)
[Coke] ShimmerFairy: the say "नि".codes; error mentioned on ycombinator is copied from your advent post. Can you update it?
timotimo pushed an update 21:43
[Coke] actually, any advent author could probably update that.
leont nine: did you do more work on the loading since you fixed that bug? (wondering if I should recompile and test further)
timotimo ARGH
why are the texts so gigantic
21:43 bjz joined
ZoffixMobile timotimo, alsp, the last two posts seem to have an issue: m.imgur.com/CKiOrQ0 21:44
AlexDaniel timotimo: hmmm, maybe you can add something for those who are experiencing post-curli regressions. Like “if you still see any bugs, please report them”… otherwise it sounds like everything is running smoothly right now, which is not
ZoffixMobile and thanks for updating :)
timotimo OK, fixed-ish
AlexDaniel: sure 21:45
jnthn 'night, #perl6 21:46
nine leont: err...which bug?
Juerd Ciao :)
lizmat good night, jnthn
timotimo gnite jnthn
21:47 ZoffixMobile left
timotimo oh god damn it 21:47
did wordpress send out the post via mail to all my readers?
leont «Missing or wrong version of dependency '/home/leon/Code/rakudo/Foo.pm' (from '/home/leon/Code/rakudo/Foo.pm')»
lizmat didn't see anything yet in my email
leont should probably rakudobug it 21:48
21:48 bjz left
lizmat ++leont 21:48
21:49 labster left
nine leont: I guess I didn't 21:49
leont You fixed it from hanging (which is very much improvement)
lizmat timotimo: perhaps mention the HN discussion about ShimmerFairy's advent blog post ? 21:50
news.ycombinator.com/item?id=10690499
timotimo ah, sure
i was only 3/4 finished with the post, so it'd be bad if it'd appear in feed readers and such and people would just see the password prompt 21:52
21:52 bjz joined
lizmat m: say "\x1f1f3\x1f1f1" # funny 21:53
camelia rakudo-moar e2e23f: OUTPUT«🇳🇱␤»
leont Sent
lizmat m: say "\x1f1f3\x1f1f1".chars 21:54
camelia rakudo-moar e2e23f: OUTPUT«1␤»
leont #126839 it is 21:55
timotimo i think i should link to some peoples' advent of code perl6 stuff
AlexDaniel lizmat: huh! :D 21:56
timotimo i don't think anybody posted it
21:57 bjz left
lizmat AlexDaniel: something I picked up from www.mikeash.com/pyblog/friday-qa-2...-hard.html 21:57
I think I'm going to like Perl 6's grapheme support better :-) 21:58
diakopter at least until Unicode goes 64-bit 21:59
22:00 hankache joined
hoelzro timotimo: posted their solutions, or posted links to solutions? 22:00
timotimo none of us did, it seems
lizmat timotimo: is it correct that the P6W is now published ? 22:01
timotimo no
it's set to "password protected" because i want feedback from people
lizmat hmmm... maybe because I'm logged in already with pw
Juerd lizmat: Probably the most interesting sentences in that article: "I think that having different conceptual kinds of strings be distinct types would eliminate a lot of bugs. I'm not aware of any language or standard library that does this, though." :)
lizmat: Perl 6 has everything that make that happen already present. I wonder what the first application for it will be :) 22:02
_nadim how do I transform an object into one of its base class objects? In this case I want to call a multi with the object as an Any.
lizmat so that's one thing Perl 6 does right :-)
Juerd s/that/that can/
timotimo _nadim: you can $myObject.Any::foobar()
_nadim can I $o ; some_sub($o.Any) ? 22:04
hoelzro timotimo: if you're looking for examples of solutions, I've been keeping mine in a gist
the code may be LTA, though =/
leont _nadim: /achieve? 22:06
22:06 ChristopherBotto joined
_nadim leont: got stuck on Error message for fun and profit: No appropriate parametric role variant available for 'name' 22:06
ChristopherBotto m: my $today = Date.new(now); $today.say; my $two_weeks_later = $today + 14; $two_weeks_later.say; 22:07
camelia rakudo-moar e2e23f: OUTPUT«2015-12-07␤2015-12-21␤»
MadcapJake hoelzro: the whole block at github.com/MadcapJake/rakudo/blob/...r.nqp#L130 isn't running.
hoelzro MadcapJake: in the case of Linenoise.pm but no .so?
MadcapJake hoelzro: yeah
hoelzro hmm 22:08
how odd
try throwing some printing into the CATCH block on L127
see what the error is
leont got an error message when using Proc::Async «Unhandled exception in code scheduled on thread 6nCan only perform big integer operations on concrete objects», anyone a clue 22:09
ChristopherBotto camelia: help 22:10
camelia ChristopherBotto: Usage:
..<(nqp-js|star-j|rakudo-MOAR|niecza|nqp-parrot|rakudo-moar|p5-to-p6|debug-cat|pugs|nqp-jvm|nqp-moarvm|star-m|prof-m|std|rakudo-jvm|rPn|nPr|r-jvm|rj|rn|nrP|r-j|nqp|perl6|rakudo|Prn|star|n|nqp-m|nr|sm|nom|rnP|nqp-mvm|P|M|p56|Pnr|m|r|rm|sj|p6|nqp-q|r-m|nqp-p|nqp-j|j)(?^::\s(?!OUTPUT)) $perl6_program>
diakopter leont: have you tried the exact same code running in just one thread?
leont Doing that right now
flussence m: use newline :crlf; # did I miss the memo, was this pragma just a temporary workaround? 22:11
camelia rakudo-moar e2e23f: OUTPUT«===SORRY!===␤Could not find newline:ver<True>:auth<True>:api<True> in:␤ /home/camelia/.perl6/2015.11-417-ge2e23fb␤ /home/camelia/rakudo-m-inst-2/share/perl6/site␤ /home/camelia/rakudo-m-inst-2/share/perl6/vendor␤ /home/camelia/rakudo-…»
leont Or actually, yes, depending on what you mean
eternaleye There seems to be a formatting error in today's Perl 6 Advent Calendar post: "will match multiple alphabetic characters ( will do almost the same thing, just with the addition of matching underscore)"
timotimo hoelzro: no, we shall never show people your ugly code! :P :P
leont I have a non Proc::Async codepath, and a Proc::Async codepath
eternaleye Nothing following the open paren
leont The former works fine
The latter give the above error with two concurrent PA's
With one Proc::Async, it doesn't give that error, but seems to hang
hoelzro =P
diakopter sounds blocker-worthy to me
eternaleye ShimmerFairy : ^^
flussence eternaleye: that *was* <alpha> last time I read it... 22:12
MadcapJake where will nqp look for my so file? I'm actually not getting a different response whether it's in /usr/lib or not
flussence throws rotten fruit at wordpress
diakopter blocker, as in, not as if it would push back some milestone, but "feels pretty important"
eternaleye Clearly, we need a POD6-based static blog engine! 22:13
leont It's github.com/rakudo/rakudo/pull/578, when running TEST_JOBS=2 ./perl6-m --ll-exception -Ilib t/harness6 t/01-sanity
22:13 labster joined
_nadim type casting, is it possible to cast an object to Any (the reason why is that there are handler for different types and I want an object of type xxx to be handled like an Any) 22:14
flussence (wordpress's php heritage is really showing: parsing something foreign in angle brackets? don't escape it, just silently truncate!) 22:15
timotimo _nadim: we don't have something for that built-in, no
22:16 skids left
MadcapJake hoelzro: I'm a bit confused, I have only done `make` and not `make install`. The panda I used to install the linenoise module is inside `~/.rakudobrew/bin`, is it looking for my module/so files in the wrong place? 22:16
_nadim timotimo: and not built-in? worth digging in or better to wait
hoelzro MadcapJake: it looks in all of the standard library locations, plus the MoarVM rpath 22:17
timotimo _nadim: you can create a class that forwards all method calls via FALLBACK to $!object."Any::$methodname"(capture)
_nadim timotimo: what happens if I say my Any $any = $some_class? is $any of some_class? 22:18
timotimo yes, it'd be of some_class
dj_goku timotimo: p6weekly goods to me. /me I still have a lot to learn in P6 land.
timotimo that's just a type check to see if $some_class derives from Any or not
dj_goku: thank you!
i just published it ... hopefully
22:18 mr_ron joined
[Coke] chuckles that the commit message "avoid using C<first> for better performance" showed up in his feed as "Avoid using C fo..." 22:19
MadcapJake hoelzro: is it looking in my $PATH?
_nadim timotimo: that's how I understood it too. There are other ways to achieve what I want, I just notice a bigger error in my module :)
22:19 mr_ron left
hoelzro MadcapJake: if you didn't install Linenoise recently (since Friday night), the so file isn't installed 22:19
MadcapJake: nope
timotimo [Coke]: in mine, too! LOL. now i get it! :)
MadcapJake right, i haven't; however I did build and install the linenoise repo, so it's available inside `/usr/lib`
will it not look there for the .so, is my question. 22:20
hoelzro it should
AlexDaniel m: my $today = Date.new(now); $today.say; my $t = $today.later(weeks => 2); $t.say
camelia rakudo-moar e2e23f: OUTPUT«2015-12-07␤2015-12-21␤»
[Coke] For a moment, I was thinking that it was related to nativecall, and that passing things back and forth between C & perl6 was a problem. (I've problems like that in the past with going between, say, sql & pl/sql)
hoelzro by that I mean, I would be awfully surprised if it didn't =)
AlexDaniel ChristopherBotto: ^
ChristopherBotto AlexDaniel: Thanks! 22:21
MadcapJake hoelzro: ok :) then I'm again at a loss for what's going on. linenoise the p6 module is installed, linenoise the shared object is available in /usr/lib, and yet Compiler.nqp fails to set @symbols. 22:22
hoelzro MadcapJake: did you print $_ in the CATCH block?
to see what the issue is?
leont Is there a reason why "Unhandled exception in code scheduled on thread X" doesn't give a backtrace?
MadcapJake compiling now
AlexDaniel ChristopherBotto: you might also like Date.today instead of Date.new(now), but these two are different
flussence logs in and fixes that one day 6 thing so ShimmerFairy doesn't have to again 22:23
AlexDaniel leont: perhaps, but it is still less than awesome. So maybe it will make sense to submit it as LTA 22:24
ChristopherBotto AlexDaniel: AlexDaniel++. Thanks! I like Date.today better. 22:25
AlexDaniel ChristopherBotto: keep in mind though that Date.today will give you your local time 22:26
ChristopherBotto: while Date.new(now) seems to be based on UTC
MadcapJake `panda list installed` doesn't work
AlexDaniel m: say Date.new(now)
camelia rakudo-moar e2e23f: OUTPUT«2015-12-07␤»
AlexDaniel m: say Date.today 22:27
camelia rakudo-moar e2e23f: OUTPUT«2015-12-07␤»
MadcapJake that is `panda list --installed`
AlexDaniel m: say now
camelia rakudo-moar e2e23f: OUTPUT«Instant:1449527281.988911␤»
AlexDaniel hm
zengargoylew panda --installed list
timotimo m: say nom
camelia rakudo-moar e2e23f: OUTPUT«5===SORRY!5=== Error while compiling /tmp/VYDw44aYJs␤Undeclared routine:␤ nom used at line 1. Did you mean 'not'?␤␤»
zengargoylew unless that got fixed...
AlexDaniel m: say DateTime.now
camelia rakudo-moar e2e23f: OUTPUT«2015-12-07T23:27:50+01:00␤»
timotimo ^- should perhaps also know about terms?
m: say DateTime.new
camelia rakudo-moar e2e23f: OUTPUT«Required named parameter 'year' is missing for DateTime.new()␤ in block <unit> at /tmp/dwxrmgFNEn:1␤␤Actually thrown at:␤ in block <unit> at /tmp/dwxrmgFNEn:1␤␤»
MadcapJake zengargoylew: thanks that worked 22:28
ChristopherBotto AlexDaniel: Thanks
AlexDaniel ChristopherBotto: in other words, Camelia will start giving different results in 33 minutes :)
ChristopherBotto timotimo: Thanks
flussence oh ffs, wordpress absolutely butchered that post and lied to me on the preview. thank god there's a revert button. 22:29
dalek ast: ad7263a | lizmat++ | S17-supply/throttle.t:
Some more Supply.throttle tests
flussence never mind, I ain't touching this thing with a 12 parsec pole.
22:30 uruwi left
|Tux| www.perlmonks.org/?node_id=1149576 :( 22:33
MadcapJake where are perl6 modules installed? 22:34
22:35 obra_ is now known as obra
AlexDaniel MadcapJake: perl6 -e 'use blergh' 22:35
and it will tell
ChristopherBotto AlexDaniel: Thanks for all of your help. I'm going to read design.perl6.org/S32/Temporal.html now.
AlexDaniel ChristopherBotto: actually, I'll recommend the docs 22:36
ChristopherBotto: e.g. doc.perl6.org/type/DateTime
timotimo [Tux]: "oh no, a butterfly!" ... so what?
diakopter |Tux|: is that person influential?
MadcapJake is there first one where panda installed them to?
i.e., `~/.perl6/2015.11-347-g47d2daf` 22:37
ChristopherBotto AlexDaniel: Thanks again!
MadcapJake oi, I have so many versions in here and the one that perl6 is looking for isn't the newest one o_O 22:38
AlexDaniel MadcapJake: It's more like install/share/perl6/
22:38 espadrine left
MadcapJake but what is this directory for? 22:38
hoelzro |Tux|: btw, I was looking at incorporating your changes for IO::String, but panda-build and panda-test are broken atm =/
AlexDaniel MadcapJake: though the folder that you've mentioned contains some precomp stuff. It is a bit complicated :/
timotimo [Tux]: note also the little piece that basically says "god damn you, perl6 people, for taking away all the good people who used to work on perl5!" 22:39
hoelzro is it [Tux] or |Tux| ?
AlexDaniel MadcapJake: also, it looks like the filenames are now mangled… honestly, I don't know what would be the right answer for your question
22:39 jkva left
diakopter timotimo: yeah, it's absurd and ridiculous and uninformed, as rants go 22:39
MadcapJake AlexDaniel: but the other three paths given don't seem to contain any modules 22:40
/home/jrusso/.rakudobrew/moar-nom/install/share/perl6, /home/jrusso/.rakudobrew/moar-nom/install/share/perl6/vendor, /home/jrusso/.rakudobrew/moar-nom/install/share/perl6/site
22:41 ChristopherBotto left
timotimo diakopter: and i'm glad about that fact. 22:41
diakopter: Jun 13, 2015 at 07:00 UTC 22:42
diakopter: is when that person joined perlmonks 22:43
diakopter the site was static for a long time, I think many people lost their logins
timotimo oh
i did not know that
MadcapJake ok maybe actually the problem is, linenoise was never installed, panda thinks it was, but i can't locate it anywhere! just the so file.
hoelzro ah ha
22:44 hudo left
MadcapJake weird, i wonder why panda would say it's installed 22:44
lizmat sometimes I think Camelia acts as a brown M&M: www.snopes.com/music/artists/vanhalen.asp
MadcapJake wait, now it's actually crashing on install (Linenoise already installed) 22:45
hoelzro o_O
MadcapJake: what's the exact error message you get in the CATCH block?
lizmat good night, #perl6! 22:46
MadcapJake Here's the error that my rakudo fork gives:
Could not find Linenoise:ver<True>:auth<True>:api<True> in:
/home/jrusso/.perl6/2015.11-423-g3a798bd
/home/jrusso/github/rakudo/install/share/perl6/site
/home/jrusso/github/rakudo/install/share/perl6/vendor
/home/jrusso/github/rakudo/install/share/perl6
hoelzro eesh
MadcapJake is there a panda uninstall? 22:47
Oh I suppose I could force it
hoelzro I don't think so =/
give that a shot
MadcapJake interestingly, liblinenoise was now installed: /home/jrusso/.rakudobrew/moar-nom/install/share/perl6/site/lib/liblinenoise.so but still no Linenoise.pm o_O 22:48
hoelzro I don't know if Linenoise.pm is installed into .perl6
22:49 vendethiel left
hoelzro maybe just the pre-comp'd version with a SHA-1 sum as its filename 22:49
RabidGravy |Tux|, timotimo that article and most of the responses is an examplar of why I stopped reading perlmonks years ago 22:50
infact why I largely ignore the "perl community" 22:51
MadcapJake well it really has to be somewhere because my main perl6 install uses it and it works :P
timotimo RabidGravy: lol. for a really short moment when i saw the line end in "I stopped reading" my brain connected that with perl6weekly
RabidGravy well that too but hey ;-)
hoelzro ¯\_(ツ)_/¯ 22:52
MadcapJake i wish modules were installed in a more obvious, accessible way
hoelzro MadcapJake: I agree, but I think it's partially growing pains of the CURLI implementation and us getting used to the new deal 22:53
|Tux| hoelzro, [Tux] is my home-PC, |Tux| is my $work PC 22:54
hoelzro ahhh
|Tux| FWIW my IO::String implementation is byfar perfect (yet) 22:55
RabidGravy "far from perfect"?
|Tux| but for now it serves its purpose. Thanks for providing me the basics to dare to start
RabidGravy, yes
RabidGravy :) 22:56
|Tux| shows it is (too) late. Just back from badmintoncompetition and ready to drive home
sleep tight all, hope to be back tomorrow 22:57
RabidGravy have fun
22:57 ChristopherBotto joined
hoelzro good night 22:58
timotimo RabidGravy: i don't think i understand your last reply :\
hankache www.perlmonks.org/?node_id=1149576 seriously!! 22:59
ChristopherBotto m: Date.today
camelia ( no output )
timotimo hankache: tux just linked that here :)
hankache all that hate?!
22:59 doublec left
hankache man people are sick 23:00
ChristopherBotto m: Date.today();
camelia ( no output )
RabidGravy which last reply, I usually dob't make sense so I'm not surprised
flussence hankache: hilarious, isn't it? that's the sort of toxic, superficial trolling we're leaving behind for good thanks to one simple image :)
RabidGravy ChristopherBotto, say Date.today
timotimo hankache: who cares, let that person alone with their misery
hankache all that hate for an image!!
timotimo and what the image represents 23:01
dalek kudo/nom: 97ef742 | TimToady++ | src/Perl6/Actions.nqp:
extend meta-thunking to comma-listy X and Z args
hankache if people hate us then we're doing the right thing
RabidGravy I *like* camelia, it's cheerful
Juerd I love the weird words that keep popping up. But... is there a dictionary somewhere? I'd love to look up terms like 'meta-thunking' :) 23:02
hankache RabidGravy camelia is waaaay better than an onion :)
ChristopherBotto M: Date.today
camelia ( no output )
ChristopherBotto m: say Date.today 23:03
camelia rakudo-moar e2e23f: OUTPUT«2015-12-08␤»
timotimo Juerd: we have S99, for one
flussence Juerd: "meta" would be the X/Z and "thunking" would be... using them on xx I guess?
ChristopherBotto RabidGravy: Thanks. I guess I'm too used to the REPL.
RabidGravy isn't some Sly and the Family Stone thing? (thunking that is)
Juerd Ah, S99 is great. Thanks :)
flussence looking at that commit, I think I guessed right 23:04
lucs What's the "@0>>.lc" in "if $l ~~ / :s ^ '[' (\S+) + ']' $/ { @foo = @0>>.lc; }"? 23:05
timotimo puts all contents of the first capture group into @foo upper-cased 23:06
Juerd lucs: @0 is shorthand for @( $0 )
lucs Aha, thanks guys.
Juerd lucs: @foo>>.lc applies .lc on each element of @foo
timotimo: Upper? :)
timotimo er. yes, lower case 23:07
sorry
lucs The other upper.
Thanks
timotimo no, that's foldcase!
Juerd I don't know what an upper is, but an up-goer is a rocket: www.youtube.com/watch?v=2p_8gx-XHJo 23:08
lucs Okay, okay, my problem was with the @0 actually :)
23:08 ZoffixGaming joined
lucs And come to think of it, with the >> too. 23:09
TimToady m: say 4,5,6 Zxx 1,2,3
camelia rakudo-moar e2e23f: OUTPUT«((4) (5 5) (6 6 6))␤»
TimToady m: say 4,5,6 RZxx 1,2,3
camelia rakudo-moar e2e23f: OUTPUT«((1 1 1 1) (2 2 2 2 2) (3 3 3 3 3 3))␤»
Juerd lucs: $0 is the first capture, like $1 in several other languages.
TimToady m: say rand Xxx 1,2,3
camelia rakudo-moar e2e23f: OUTPUT«((0.526625786782889) (0.526625786782889 0.526625786782889) (0.526625786782889 0.526625786782889 0.526625786782889))␤»
TimToady m: say (rand,) Xxx 1,2,3
camelia rakudo-moar e2e23f: OUTPUT«((0.442517361025756) (0.442517361025756 0.442517361025756) (0.442517361025756 0.442517361025756 0.442517361025756))␤»
TimToady that last one will fix in 2 minutes
23:09 b7j0c joined
ZoffixGaming hanekomu: heh, I love that those rants always go with "You want to attract serious Perl programmers with this"... well, maybe I don't! I want the quirky and fun and non-serious programmers :P 23:10
timotimo: what was the thing you wanted on the Advent added called?
"table of posts"? "archive"? (/me wants to know what to google for)
timotimo ZoffixGaming: you mean somethin glike an "aside"?
Juerd lucs: So here, @0 is a list of the several (\S+) that were matched (note the + after the capture group)
ZoffixGaming looks that up
timotimo "Table of Contents" is the search term that gives you the posts we want, but that's just because that's the title these posts have
lucs Juerd: Ah, I see. Nice. 23:11
b7j0c in a CATCH block, is there a way to get the message for a caught Exception type...i.e., where i normally have a `when` clause, i want to actually be able to access the members of the Exception i caught instead of simply matching its type
hankache ZoffixGaming?!
TimToady m: say (rand,) Xxx 1,2,3
camelia rakudo-moar 97ef74: OUTPUT«((0.578100359876706) (0.585752122516995 0.834068598028611) (0.816395699874 0.802171392461199 0.424036099637496))␤»
TimToady there we go
hankache how many nicks do u have? lol
timotimo hankache: nicks are cheap
TimToady next step, thunking reductions correctly...
ZoffixGaming Jesus... I've used 10% of this ramen spice packet and my mouth is burning >_< 23:12
Juerd lucs: » is a meta-operator, it changes the way the . works. Specifially, in ». it lets the method be called on each element of the list. @foo».bar is somwhat like @foo.map(*.bar)
ZoffixGaming: I'm not sure if Jesus reads IRC.
ZoffixGaming Juerd++
23:12 patrickz left
timotimo it's similar to .map, but it may happen in parallel, so it mixes up the order of execution (but the order of results stays the same) 23:12
23:12 mr_ron joined
Juerd timotimo: Is that actually implemented yet? :) 23:13
hankache he doesn't? lets turn IRC into a brothel
leont uses ». all the time, but more as a short map than ass a hyper per-se
ZoffixGaming hankache: ZoffixW for $work box, ZoffixMobile for phone, Zoffix for my main box, ZoffixLappy, for my laptop, ZoffixGaming for when I'm booted into Windows on my main box :) I also have some exotic nicknames registered, like sexy-coder-girl, for when I wanna troll people :P
timotimo m: <a b c d e>>>.say
camelia rakudo-moar 97ef74: OUTPUT«a␤b␤c␤d␤e␤»
timotimo we used to intentionally mix up the order. seems like we don't do that any more
Juerd Sounds like a good idea to randomize the order
hankache ZOffixGaming #sexy-coder-girl hahahahahahahaha
Juerd Or at least swap the first two, if a full shuffle is too much work :) 23:14
mr_ron m: class URI { has $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI.new; $u.scheme = '\\//'; $u.host = '?'; say $u
camelia rakudo-moar 97ef74: OUTPUT«\//://?␤»
23:14 dha left
lucs Juerd: Thanks for spelling it out for me (on doc.perl6.org, I didn't find the >> documentation (apart from regex word boundary matching). 23:14
ZoffixGaming waves to mr_ron
mr_ron ZoffixGaming: your idea for URI sort of works but not quite right
timotimo lucs: yeah, hyper operators is in the WANTED :(
ZoffixGaming mr_ron: oh, what is not right?
mr_ron m: class URI { has $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI.new; $u.scheme = '\\//'; $u.host = '?'; say $u 23:15
camelia rakudo-moar 97ef74: OUTPUT«\//://?␤»
lucs timotimo: Oh :)
masak 'night, #perl6
ZoffixGaming mr_ron: if that code sample is meant to illustrate what's not right, then I'm not following...
mr_ron with real mutators I could check the scheme for validity and throw a exception in a case like my example
ZoffixGaming Ahh
m: class URI { has $.scheme is rw; method scheme (Str $s) { $s ~~ /\\/ and fail "You shall not pass!"} has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI.new; $u.scheme = '\\//'; $u.host = '?'; say $u 23:17
camelia rakudo-moar 97ef74: OUTPUT«5===SORRY!5=== Error while compiling /tmp/UHDwyAkq32␤Strange text after block (missing semicolon or comma?)␤at /tmp/UHDwyAkq32:1␤------> 3 ~~ /\\/ and fail "You shall not pass!"}7⏏5 has $.host is rw; method gist { "$!sche␤ expecting an…»
ZoffixGaming :(
mr_ron: IMO that can be done when stringification (or full URL) is requested. It doesn't matter if scheme is wrong until you get the actual URL. 23:18
m: 0/0; say "eh, whatever"; # sorta like this
camelia rakudo-moar 97ef74: OUTPUT«eh, whatever␤»
23:18 uruwi joined, b7j0c left, kid51 joined
timotimo ZoffixGaming: just a ; that's missing 23:19
ZoffixGaming timotimo: but i won't be able to assign to a method, would I?
OTOH:
mr_ron I could parse the whole URI and throw an exception when .gist is called and the URI doesn't parse - that is true and sort of works ... 23:20
ZoffixGaming m: subset ValidScheme of Str where !/\\/; class URI { has ValidScheme $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI.new; $u.scheme = '\\//'; $u.host = '?'; say $u
camelia rakudo-moar 97ef74: OUTPUT«Type check failed in assignment to $!scheme; expected ValidScheme but got Str␤ in block <unit> at /tmp/dXV2ilFvGP:1␤␤»
ZoffixGaming Tada
timotimo that works, yeah
if you want to assign to a method, you'll mark the method "is rw" and put the container as the last thing 23:21
23:21 adu joined
timotimo alternatively use "return-rw" 23:21
mr_ron token scheme { <.uri-alpha> <[\-+.] +uri-alpha +digit>* }; - from github.com/perl6-community-modules...mar/URI.pm 23:22
ZoffixGaming timotimo: ah, neat
timotimo: don't see how to add that thing back. All the things I have on a the only off-wordpress.com-hosted wordpress site aren't there on the Advent. 23:24
Maybe someone who's more experienced with Wordpress could add it back. Or someone with enough patience could find a different decent-looking theme that has that posts list included 23:25
timotimo right. 23:26
ZoffixGaming :( 23:29
23:31 virtualsue joined
RabidGravy right that's me done for the day 23:33
catch y'all later
ZoffixGaming \o
23:35 rurban left
mr_ron timotimo: if I put the container as the last thing or return-rw how do I validate before assigning? 23:39
timotimo mr_ron: in that case you need a Proxy 23:40
23:40 RabidGravy left
mr_ron that's what I explained to ZoffixGaming here: github.com/perl6-community-modules.../issues/27 23:40
timotimo but with a subset, that'd also work 23:41
Hotkeys Sometimes I forget to disconnect from my other irc clients and then i get on my computer and have no context (ZNC gives me up to a 50 line history when I connect the first client) 23:42
and it makes me sad 23:43
I could make znc always tell me but that would get annoying with unintentional disconnects
23:43 hankache left
timotimo only 50 lines? that's pretty weak 23:43
Hotkeys I mean
I can make it more
23:43 spider-mario left
Hotkeys 50 is usually enough 23:43
ZoffixGaming mr_ron: I do thinking validating each component individually is over-engineering things a bit. Just validate when the full URL is requested.
my two cents
ZoffixGaming leaves to play Fallout 4
night all \o 23:44
23:44 ZoffixGaming left
tony-o screen + irssi 23:45
forever history 23:46
timotimo no. screen + weechat
actually
tmux + weechat
+ weechat remote protocol
+ glowing bear
tony-o tmux is nice but i have an uptime to maintain
timotimo tmux has never crashed on me
23:46 virtualsue left
timotimo unless you mean you don't want to switch between screen and tmux becasue that'd mean you'd have to close irssi once 23:46
tony-o nor me, i just started using screen before i actually experienced the full power of tmux
timotimo i wonder if you can reptyr a program from a screen session into a tmux session ... 23:47
flussence hasn't crashed on me, but it has broken the pipe protocol between updates more than once
23:47 csd_ left
timotimo that's true. 23:47
tony-o [15:47] Uptime: 404d 4h 39m 48s
i have that going on
timotimo i bet your kernel is still super secure, though
tony-o probs
this thing isn't a productive server, just something i screw around on 23:48
timotimo :)
diakopter m: say (rand,) Xxx ^99**99
camelia rakudo-moar 97ef74: OUTPUT«(() (0.790520055025987) (0.8513212235925 0.491482906034424) (0.600896884200228 0.799362530141523 0.298255295298534) (0.484338858622959 0.99495079214283 0.234830354056885 0.766573864783707) (0.859617055848472 0.612114303608827 0.866768992417288 0.3844211538…»
tony-o probably still has the ssh bug
timotimo which one? :)
insecure debian keys or heartbleed?
mr_ron m-star: subset ValidScheme of Str where IETF::RFC_Grammar::URI::scheme; class URI2 { has ValidScheme $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI2.new; $u.scheme = q!\\//!; $u.host = "?"; say $u
timotimo i think it's gotta be star-m:
mr_ron star-m: subset ValidScheme of Str where IETF::RFC_Grammar::URI::scheme; class URI2 { has ValidScheme $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI2.new; $u.scheme = q!\\//!; $u.host = "?"; say $u 23:49
camelia star-m 2015.09: OUTPUT«Could not find symbol '&scheme'␤ in any accepts_type at gen/moar/m-Metamodel.nqp:3237␤ in block <unit> at /tmp/BAt_7xvfPi:1␤␤Actually thrown at:␤ in any accepts_type at gen/moar/m-Metamodel.nqp:3237␤ in block <unit> at /tmp/BAt_7xvfPi:1␤␤»
timotimo yeah, you'd need to use IETF::RFC_Grammar::URI first
tony-o hearbleed 23:50
mr_ron star-m: use IETF::RFC_Grammar::URI; subset ValidScheme of Str where IETF::RFC_Grammar::URI::scheme; class URI2 { has ValidScheme $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI2.new; $u.scheme = q!\\//!; $u.host = "?"; say $u
camelia star-m 2015.09: OUTPUT«Could not find symbol '&scheme'␤ in any accepts_type at gen/moar/m-Metamodel.nqp:3237␤ in block <unit> at /tmp/sNWeCaUGzY:1␤␤Actually thrown at:␤ in any accepts_type at gen/moar/m-Metamodel.nqp:3237␤ in block <unit> at /tmp/sNWeCaUGzY:1␤␤»
23:51 abaugher_ left
mr_ron I'm interested if someone has a way to make the subset work ... 23:52
timotimo i don.t know what ::URI looks like and what "scheme" is inside that 23:53
23:53 cdg left
timotimo you may have to do it like / <IETF::Blah::scheme> / perhaps. 23:53
mr_ron token scheme { <.uri-alpha> <[\-+.] +uri-alpha +digit>* }; - from github.com/perl6-community-modules...mar/URI.pm
star-m: use IETF::RFC_Grammar::URI; subset ValidScheme of Str where /<IETF::RFC_Grammar::URI::scheme>/; class URI2 { has ValidScheme $.scheme is rw; has $.host is rw; method gist { "$!scheme://$!host" } }; my $u = URI2.new; $u.scheme = q!\\//!; $u.host = "?"; say $u 23:54
camelia star-m 2015.09: OUTPUT«Type check failed in assignment to '$!scheme'; expected 'ValidScheme' but got 'Str'␤ in block <unit> at /tmp/EiEhZYitx4:1␤␤»
mr_ron fantastic ... it worked ...
timotimo you can always count on me!*
mr_ron timotimo: thanks much 23:55
timotimo (not a guarantee. some terms & conditions apply. only valid during official office hours)