»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
timotimo only a little bit of post-processing needed 00:00
manually
our PDF module on the ecosystem is very, very bare, AFAIK
i.e. wouldn't easily let you do what you want
AlexDaniel is ｢hyper｣ completely broken? “The order of elements is preserved.” but that's not what I am getting 00:08
ah, yes. And there is a bug report… submitted by me… 00:09
rt.perl.org/Public/Bug/Display.html?id=127099
azawawi Herby_: as an initial step, Excel accepts CSV as a format so you can use it for prototyping until you can find out how to ::WriteExcel 00:10
leont Inline::Perl5 to the rescue? ;-) 00:11
Herby_ azawai: hadn't even considered that... 00:13
hmm
Herby_ leont: i haven't tried using that yet... guess now is as good as time as ever 00:14
ely-se interesting 00:44
DrForr Guh, I think I'm going to have to abandon Perlmonks after just a few weeks if this nonsense keeps up. 00:59
DrForr Someone over there thinks the only way to get Perl and Scheme to share variables is to write a Scheme on top of MoarVM and rewrite the Perl 6 runtime, instead of just using the built-in Scheme callback methods that he didn't bother reading about. 01:01
timotimo i didn't know there was any reason whatsoever to spend any time whatsoever reading or even writing to perlmonks 01:06
geekosaur ^ 01:08
timotimo my recollection of perlmonks was "anonymous monk" users calling perl6-involved people names, spreading half-truths blown way out of proportion, as well as some personal attacks and other things like that 01:10
Juerd I used to spend a lot of time at perlmonks, really, really loved doing so. 01:12
But that was a long time ago.
geekosaur the name "monk" is apt. people who've been living in caves away from the world for years, with odd ideas of how things work
although plato's cave might be even more apt
Juerd The site no longer really serves a purpose now that we have stackoverflow :) 01:13
In fact, for a programming language's visibility/popularity it's quite important that questions are asked there, instead of at perlmonks.
(or some other niche site)
timotimo mhm 01:21
Herby_ dumb regex question that's beyond my knowledge... Scenario is this: if I have a line that starts with "a. ", then I want to capture the line preceding it, since I know that line contained the question 02:44
any ideas?
konobi msgx 02:45
Herby_ this is all in a text file, sorry
konobi Herby_: positive lookahead assertion
Herby_ *blank stare* 02:46
konobi Herby_: docs ftw!
Hotkeys docs.perl6.org/language/regexes#Loo...assertions
Herby_:
Herby_ hmmm! 02:47
let me take a peek!
m: say "foobar" ~~ rx{ foo <?before bar> }; 02:48
camelia rakudo-moar a5fe34: OUTPUT«｢foo｣␤»
Herby_ m: say "foo\nbar" ~~ rx{ foo <?before \nbar> }; 02:49
camelia rakudo-moar a5fe34: OUTPUT«｢foo｣␤»
Herby_ m: say "1. question here\n a. anAnswerHere" ~~ rx{ .+ <?before \n> }; 02:51
camelia rakudo-moar a5fe34: OUTPUT«｢1. question here｣␤»
Herby_ i might be onto something :)
thanks Hotkeys , konobi
Hotkeys m: "1. question here\n a. anAnswerHere" ~~ /^^ \N* <?before \n\s*'a. '>/ 02:53
camelia ( no output )
Hotkeys rude
m: say "1. question here\n a. anAnswerHere" ~~ /^^ \N* <?before \n\s*'a. '>/
camelia rakudo-moar a5fe34: OUTPUT«｢1. question here｣␤»
Hotkeys Need to remember to include the 'a. ' if that's what you looking for Herby_
Hotkeys m: "0. where's the soap?\n a. no soap, radio!\n1. question here\n a. anAnswerHere" ~~ m:g/^^ \N* <?before \n\s*'a. '>/ 02:54
camelia ( no output ) 02:55
Hotkeys m: say "0. where's the soap?\n a. no soap, radio!\n1. question here\n a. anAnswerHere" ~~ m:g/^^ \N* <?before \n\s*'a. '>/
camelia rakudo-moar a5fe34: OUTPUT«(｢0. where's the soap?｣ ｢1. question here｣)␤»
Herby_ good point, not sure why i left off the 'a. ' thanks
whats the deal with the \N*?
my regex ex in general is very weak
trying to work on it
skids Herby_: \N is not any sort of newline 02:56
Like \D is "not a number"
Herby_ ah ok
awwaiid If anyone wants to translate some Perl 5 async examples into Perl 6 that would be lovely -- github.com/plicease/dc.pm-async -- tomorrow we're having a perlmonger meetup. I was going to do some translating but have so far failed and have to go offline for the evening. so... donations welcome :) 02:57
Herby_ m: "random crap thats not a question\n 0. where's the soap?\n a. no soap, radio!\n1. question here\n a. anAnswerHere" ~~ m:g/^^ \N* <?before \n\s*'a. '>/
camelia ( no output )
Herby_ ? 02:58
m: "random crap thats not a question\n0. where's the soap?\n a. no soap, radio!\n1. question here\n a. anAnswerHere" ~~ m:g/^^ \N* <?before \n\s*'a. '>/
camelia ( no output )
Herby_ what am I missing here?
Hotkeys hmm 03:00
mspo Herby_: camelia isn't the repl; you need a say ?
Juerd Herby_: Output... :)
Hotkeys oh yeah
Herby_ ?
doh 03:01
Hotkeys you have to put 'say' before that
Herby_ m: say "random crap thats not a question\n0. where's the soap?\n a. no soap, radio!\n1. question here\n a. anAnswerHere" ~~ m:g/^^ \N* <?before \n\s*'a. '>/
camelia rakudo-moar a5fe34: OUTPUT«(｢0. where's the soap?｣ ｢1. question here｣)␤»
Herby_ man, you guys are some geniuses
works like a charm
Hotkeys lol
Herby_ let me try it on my 500 page mess
Juerd Wouldn't you want \N+ instead of \N* ? 03:02
Herby_ m: say "random crap thats not a question\n0. where's the soap?\n a. no soap, radio!\n1. question here\n a. anAnswerHere" ~~ m:g/^^ \N+ <?before \n\s*'a. '>/ 03:03
camelia rakudo-moar a5fe34: OUTPUT«(｢0. where's the soap?｣ ｢1. question here｣)␤»
Herby_ Juerd, why do you say that? 03:04
for my own knowledge
dalek kudo/2016.01-preparation: 6384b1e | coke++ | tools/build/NQP_REVISION:
now for reals.
 03:05
[Coke] . 03:08
cutting compiler releases now. 03:09
skids [Coke]++
geekosaur Herby_, \N* matches the empty string
[Coke] I suspect we'll have an R* in the next day or two. it was really just waiting for a good tagged compiler release.
Herby_ geekosaur: thanks that makes sense 03:09
[Coke] (stackoverflow) (meta) me votes on the perl vs. perl6 tags question... 03:11
dalek kudo/2016.01-preparation: f2b8c74 | coke++ | docs/release_guide.pod:
actual release date
 03:13
kudo/2016.01-preparation: 26eef79 | coke++ | docs/release_guide.pod:
update copyright
[Coke] twitter.com/henryhoffman/status/69...6440200192 03:17
(git humor) 03:18
Herby_ ok, i think i just about got this licked. On the above regex, I'm capturing the questions just fine. How do I go about splitting the captures up? 03:24
dalek kudo/2016.01-preparation: 81a0494 | coke++ | docs/announce/2016.01.md:
moar specific google
 03:25
Herby_ rather...I'm spurting the questions into a text file
but its one continuous string
i'm looking at the regex docs but i'm not sure what I need
i'd like a newline between each captured question
[Coke] blugged: perl6advent.wordpress.com/2016/02/...ement-two/ 03:27
Herby_ very cool, Coke! 03:28
mspo Herby_: you can probably print out $/ stuff 03:29
tony-o_ # | Plan // Pass | File Name 03:30
[✗] 1 | 0 // 0 | ../perl6-xml-query/t/basic.t
Test #1 - output
The spawned process exited unsuccessfully (exit code: 1)
===SORRY!===
Could not find XML::Query in:
oops, sorry #perl6
Herby_ mspo: not positive what you mean 03:31
tony-o_ [Coke]++ 03:32
mspo Herby_: when you capture stuff you can find it in $/ 03:33
Herby_ m:g/^^ \N+ <?before \n\s*'a. '>
[Coke] gotta go give someone a ride. Will finish the rakudo release when I return. (building & testing the tarball now)
Herby_ the :g throws me off, on how thats handled
tony-o_ or you can use $/ as a hash
m: "hello world" ~~ / $<preceeds>=. 'll' /; $/<preceeds>.Str.say; 03:34
camelia rakudo-moar a5fe34: OUTPUT«e␤»
[Coke] news.perlfoundation.org/2016/02/ian...plica.html - comment if you want to see pmurias get a grant to make a lot of headway on rakudo-js. 03:37
Herby_ tony-_, thanks
Herby_ I've made a ton of progress on this tonight, thanks everyone :) 03:41
AlexDaniel geez, so how do I get the order right when using hyper? 03:49
besides fixing the issue
m: my @result = ^1000 .hyper.map: * + 10; say @result ~~ @result.sort 03:54
camelia rakudo-moar a5fe34: OUTPUT«True␤»
AlexDaniel ohhh
sounds like I just have to rebuild…
timotimo Herby_: i tend to have my programs that do processing of "mostly line-based" stuff in a big for loop and keep some state from one line to the next; that'd make the "get the previous line when a line matches something specific" doable with "more than just" regex 04:03
mspo Seq.from-loop? 04:07
or just an array/stack
sevvie github doesn't like the idea of .t being a perl6 file. 04:15
Hotkeys lol
time to switch to .t6
jdv79 prove will find .t files but it won't look for .t6 last time i tried 04:17
sevvie aye, noticed that. 04:31
[Coke] is back. finishing up release now.. 04:40
.seen mojca 04:41
yoleaux I saw mojca 4 Jan 2016 19:30Z in #perl6: <mojca> but at least it kind-of-works for now
[Coke] going to cut over the wikipedia page from "preview release" to "release" 04:44
[Coke] updated. 04:46
sorry, I should be doing the minor updates ont he release to -release. :)
Herby_ timotimo: thanks for the idea. I might end up doing it that way 04:48
AlexDaniel indeed! Upgrade seems to have fixed my hyper problem. That being said, it introduced a bunch of other problems… 04:50
[Coke] 2016.01 release cut 05:08
AlexDaniel m: my @x = ^10; my @y = @x.hyper(:3batch, :5degree).map: { sleep rand / 100000; $_ + 1 }; say @y 05:16
camelia rakudo-moar a5fe34: OUTPUT«[]␤»
AlexDaniel m: my @x = ^10; my @y = @x.hyper(:3batch, :5degree).map: { sleep rand / 100000; $_ + 1 }; say @y
camelia rakudo-moar a5fe34: OUTPUT«[1 2 3 4 5 6 7 8 9 10]␤»
AlexDaniel I don't even…
Hotkeys why? 05:17
[Coke] So, how likely am I to get yelled at if I merge the prep branch back to nom? (given that the last big branch merge resulting in yelling)
Hotkeys oh are those two lines exactly the same AlexDaniel
AlexDaniel Hotkeys: they are
Hotkeys m: say (rand for ^10) 05:18
camelia rakudo-moar a5fe34: OUTPUT«(0.808720287468173 0.76150654897164 0.122863386389548 0.435090412399772 0.281497767344602 0.587686141008414 0.316046270331001 0.298272219355734 0.64243564539821 0.552020721287313)␤»
AlexDaniel Hotkeys: you can change this delay if you want
Hotkeys weird
AlexDaniel Hotkeys: it's just that with that kind of delay the results are most inconsistent
Hotkeys I just tried it locally
it fails the first time
but otherwise works afterward
AlexDaniel try it several times 05:19
you'll get it again
Hotkeys i.imgur.com/k7OBCDk.png
AlexDaniel Hotkeys: files.progarm.org/2016-02-02-07194..._scrot.png
Hotkeys weird 05:20
I can't get it to give me an empty list again
AlexDaniel Hotkeys: try without / 10000
Hotkeys empty every time
AlexDaniel not every time. You just did not run it enough times :)
Hotkeys oh
just got one
empty most times
what's going on here 05:21
AlexDaniel well, I was struggling with this before I rebuilt rakudo: rt.perl.org/Public/Bug/Display.html?id=127099 05:22
now it went away, but here's another bug! Yeah, bug exchange
perhaps this is related: rt.perl.org/Public/Bug/Display.html?id=126597 05:23
or maybe not, grep has never worked with hyper I guess 05:24
Hotkeys how am I supposed to tell people p6 has nice concurrency with these shenanigans
AlexDaniel Hotkeys: don't tell them to use hyper yet, I guess… 05:25
Hotkeys lol
AlexDaniel there are enough features besides hyper/race to get them impressed
Hotkeys: but I'm really disappointed about this one, yes
skids Yeah, I hope concurrency bugs get some attention soon... I know the focus is supposed to shift to performance, but no point in optimizing in wrong behavior after all. 05:26
AlexDaniel not sure how am I supposed to work around this problem right now though. Rerun the code if the list has 0 elements? 05:27
AlexDaniel submits rakudobug…
Hotkeys but what if you get really unlucky
and it runs forever and ever
AlexDaniel I am getting around 50% chance it real code 05:28
so it should be fine, I hope
[Coke] I would say make sure you're using the latest release, but I don't think 2016.01 has anything for concurrency bugfixes. 05:29
AlexDaniel [Coke]: I've rebuilt it 20 minutes ago, I think that it is the latest version 05:30
unless there were some improvements during these 20 minutes
AlexDaniel m: dd @(^10 .hyper(:1batch).map: { sleep rand; $_ }) # golfy 05:41
camelia rakudo-moar a5fe34: OUTPUT«()␤»
AlexDaniel m: dd @(^10 .hyper(:1batch).map: { sleep rand; $_ }) # golfy
camelia rakudo-moar a5fe34: OUTPUT«(0, 1, 2, 3, 4, 5, 6, 7, 8, 9)␤»
[Coke] AlexDaniel: (the release didn't come from nom, for a chang) 05:45
*change
[Coke] but that makes it even less likely there are fixes. 05:45
skids Maybe less breakses? 05:50
AlexDaniel m: .say for (^10).hyper(:1batch).map: { sleep rand; $_ } # let's see if we can get that on camelia
camelia rakudo-moar a5fe34: OUTPUT«0␤2␤5␤3␤1␤7␤9␤4␤6␤8␤»
AlexDaniel m: .say for (^10).hyper(:1batch).map: { sleep rand; $_ } # let's see if we can get that on camelia
camelia rakudo-moar a5fe34: OUTPUT«2␤»
AlexDaniel m: .say for (^10).hyper(:1batch).map: { sleep rand; $_ } # let's see if we can get that on camelia
camelia rakudo-moar a5fe34: OUTPUT«0␤4␤1␤2␤6␤7␤8␤3␤9␤5␤»
AlexDaniel yeah
not only it is out of order, but it is also extremely broken
done: rt.perl.org/Public/Bug/Display.html?id=127452 06:00
gn o/
CIAvash . 07:52
yoleaux 1 Feb 2016 23:50Z <Herby_> CIAvash: That was it, thanks!
RabidGravy marning! 09:05
lizmat m: say $?PERL 09:26
camelia rakudo-moar a5fe34: OUTPUT«5===SORRY!5=== Error while compiling /tmp/MDZ9r4T2ki␤Variable '$?PERL' is not declared␤at /tmp/MDZ9r4T2ki:1␤------> 3say 7⏏5$?PERL␤»
RabidGravy boom! 09:43
I don't suppose anyone has a gadget for making a "provides" from a directory of modules? 09:44
FROGGS RabidGravy: panda gen-meta . 09:52
RabidGravy didn't work last time I tried it 09:53
FROGGS :/
FROGGS ohh, perhaps it is broken because of latest compunit work 09:54
we should unbreak it then
RabidGravy for LIB in `find lib -name '*.pm'` ; do Q=`echo $LIB | sed -e 's/^/"/' -e 's/$/"/'`; MOD=`echo $LIB | sed -e 's/lib\///' -e 's/\.pm$//' -e 's/\//::/g' -e 's/^/"/' -e 's/$/"/'` ; echo "$MOD : $Q" ; done 09:55
:) 09:56
yeah, "Could not find Panda::DepTracker" 09:57
mojca Why didn't the following commit end up in 2016.01? github.com/rakudo/rakudo/commit/fb...242997878c 10:04
I get ./perl6-m tools/build/install-core-dist.pl No writeable path found, /opt/local/share/perl6 not writeable in block <unit> at tools/build/install-core-dist.pl line 12
mojca when I try to install 2016.01 10:04
mojca I blindly assume that the mentioned commit would help 10:05
timotimo oh, the deptracker got kicked out along with the recent compunit stuff 10:15
though it has been on the chopping block before that, i think
due to being a bit of a hack anyway 10:16
FROGGS gah, dont talk like that about my baby! ò.ó 10:17
FROGGS timotimo: actually, they way it work was the most correct thing one could do 10:22
worked* 10:23
guessing packages inside a file by the filename only is a hack
timotimo well it was mostly for making sure panda would precomp stuff in the right order, no? 10:24
FROGGS no, not at all 10:27
it was about finding package definitions in a file by watching what happens when running probe 10:28
so you can have Chinese class names in a package in a different named file 10:29
though, there is a chicken-egg problem
|Tux| test 22.199 10:35
test-t 12.032
brrt good * #perl6 10:47
Ven \o 10:56
fireartist can a Role access a private attribute in a Class which `does` that role? I can't figure out the syntax 11:09
gfldex m: role R { has $!priv = 42; method m() { say $!priv } }; class C does R { trusts R }; say C.new.m; 11:12
camelia rakudo-moar a5fe34: OUTPUT«42␤True␤»
gfldex m: role R { method m() { say $!priv } }; class C does R { has $!priv = 42; trusts R }; say C.new;
camelia rakudo-moar a5fe34: OUTPUT«5===SORRY!5=== Error while compiling /tmp/lC873SbL3S␤Attribute $!priv not declared in role R␤at /tmp/lC873SbL3S:1␤------> 3role R { method m() { say $!priv } }7⏏5; class C does R { has $!priv = 42; trus␤ expecting any of:␤ hor…»
gfldex m: role R { method m() { say self.("priv") } }; class C does R { has $!priv = 42; trusts R }; say C.new.m; 11:13
camelia rakudo-moar a5fe34: OUTPUT«Cannot find method 'CALL-ME'␤ in method m at /tmp/Bgr2KOT_d_ line 1␤ in block <unit> at /tmp/Bgr2KOT_d_ line 1␤␤»
gfldex you will need to create the accessor by hand 11:14
fireartist gfldex: ok, I think I understand that 11:16
gfldex: thanks
bhm p6: say Date.new(now); 11:32
camelia rakudo-moar a5fe34: OUTPUT«2016-02-02␤»
bhm p6: say Date.new(now).earlier(:1month);
camelia rakudo-moar a5fe34: OUTPUT«2016-01-02␤»
bhm p6: say Date.new(now).earlier(:2month);
camelia rakudo-moar a5fe34: OUTPUT«2015-12-02␤»
bhm ok, so I must upgrade. :)
|Tux| FWIW CSV::Parser still won't install and Inline::Perl5 still broken on long-double perl5 builds 11:34
jnthn Did long-double Perl 5 builds ever actually work with Inline::Perl5? 11:35
RabidGravy BOOM! 12:01
nine jnthn: no, they didn't. They just pretended to. 12:07
jnthn nine: ah, k :) 12:11
abraxxa1 oh, 2016.01 is already available! congrats!
k-man hi 12:24
i was just reading the perl6intro.com, and it says the Lazy List Constructor is … 12:25
as in ellipsis?
Ven it's wrong 12:26
jnthn That's the sequence operator
m: say 1, 3, 5 ... 21
camelia rakudo-moar a5fe34: OUTPUT«(1 3 5 7 9 11 13 15 17 19 21)␤»
jnthn m: say 1, 3, 5 … 21
camelia rakudo-moar a5fe34: OUTPUT«(1 3 5 7 9 11 13 15 17 19 21)␤»
k-man 3 full stops
jnthn 3 full stops is the usual form, but you can write it as ellipsis too 12:27
Skarsnik Hello
k-man oh i see, interesting
thanks
abraxxa has there been a 2016.01 release announcement? 12:30
I don't see one on perl6.org
rakudo.org doesn't have one either ;( 12:31
moritz typically we only announce Star releases 12:32
abraxxa moritz: :( 13:05
dalek Iish: da3cc93 | (Sylvain Colinet)++ | .travis.yml:
Update travis to install panda and dbiish deps
 13:12
[Coke] we haven't really settled on how to announce a post-christmas release. 13:17
email went out. wikipedia was updated. tarballs were uploaded.
[Coke] (missed patch) this is why we did an RC. 13:19
If you have something you want in the release, you have to review them.
[Coke] guess I'll hold off on announcing 2016.01 on blogs.perl.org just yet. 13:23
Skarsnik abraxxa, there is still conflict in your pr for DBIish, I am not sure why 13:24
abraxxa Skarsnik: really? weird
Skarsnik: where there commit to the main repo after my last rebase?
abraxxa Skarsnik: this commit is included in my PR: github.com/perl6/DBIish/commit/da3cc933a3 13:25
but install is misspelled in this commit
[Coke] anyone interested in a 2016.01.1, please join #perl6-release
dalek kudo/2016.01-preparation: 206148c | (Anthony Parsons)++ | tools/build/ (3 files):
Fix `make install` when nqp is distro-installed

During the installation process, tools/build/install-core-dist.pl is called to precompile the bundled non-CORE.setting libraries such as Test.pm. The way this was initially coded, it pulled the install path from NQP's build-time configuration, which will be a root-owned filesystem path if NQP was installed via a system package manager.
Other parts of the CompUnit code will see these paths don't exist, and attempt to helpfully call `mkdir` on them which brings the whole process to a messy end. This part usually works fine when building everything from a git checkout, because Moar/NQP/Rakudo all default to an installation prefix the user has write access to.
This patch fixes that step of the build process by passing the makefile's path prefixes to install-core-dist.pl (ensuring precompilation output goes to the right place), and modifying it to *only* use that path (ensuring it doesn't try to mkdir where it shouldn't). This isn't a 100% solution, but it fixes enough to unblock packaging work on a few distros.
Thanks to mst++ for doing almost all of the legwork here, crux++ for an initial patch that gave us a few pointers to work from, and nine++ for giving us some Boxing Day fun ☺		 13:27
dalek kudo/2016.01-preparation: 48a3af9 | coke++ | VERSION:
[release] bump VERSION
 13:31
Iish: eedf43c | (Sylvain Colinet)++ | .travis.yml:
fix type on install
 13:35
Skarsnik abraxxa, it's not about travis, there is a conflict somewhere x) 13:38
it missing perl 6 github.com/mame/quine-relay 13:41
abraxxa Skarsnik: of course because you've added a commit that changes the same file (.travis.yml) my commit changes 13:43
Skarsnik abraxxa, it was not confliting already? 14:03
abraxxa Skarsnik: not when I created the pr 14:03
Skarsnik try to merge with the origin repo on your side to see where the conflit is? I commit these to see if you pR could pass travis 14:05
dalek c: fc4ec54 | raiph++ | doc/Language/list.pod:
Typos and fix a link
 14:35
abraxxa Skarsnik: travis failed to install Data::Dump if you look at the build linked in my PR 15:00
dalek kudo/2016.01-preparation: f033f2b | coke++ | docs/release_guide.pod:
mention point release
 15:04
tony-o_ abraxxa: ? 15:05
abraxxa: i may have screwed up the META6 thing, i'll fix it rn 15:06
abraxxa tony-o_: the Data::Dump one? I don't think so as it installed fine here but not on Travis
dalek osystem: a60e107 | tony-o++ | META.list:
Updating my META links.
 15:07
tony-o_ abraxxa: ah, just found the PR. fixing the META problem won't solve that issue :-p 15:09
abraxxa tony-o_: are you talking about Travis not working or the merge conflict? 15:13
tony-o_ travis not running the installdepends 15:14
skids www.infoworld.com/article/3027100/m...tml?page=2 # A couple of those look... familiar. 15:20
The "defer" one is neat though, always wanted something like that. 15:21
jnthn nine: rt.perl.org/Ticket/Display.html?id=127454 looks like it may be an I::P5 question/issue/bug 15:26
RabidGravy tadzik, if you get a minute I'd love you long time if you looked at github.com/tadzik/JSON-Unmarshal/pull/16 15:29
tadzik RabidGravy: looking 15:30
RabidGravy only tiny
nine jnthn: yes it is. One may argue that it even is a bug in Net::WebSocket::Server as that explicitly checks "ref $value eq 'CODE'" while any callable would work 15:31
tadzik RabidGravy: merged, thanks :)
RabidGravy :-* 15:32
nine But maybe there's even a way around that by using magic instead of overloading
ugexe is there anything like perl6-gdb for windows? i have a piece of code that immediately ends the process with no message (just a `(try require ::($xxx)) ~~ Nil) but *only* for a specific $xxx. and that $xxx works find by itself and gets handled fine on linux 15:35
jnthn ugexe: What does `echo %ERRORLEVEL%` output?
But no, you'd probably need to do a debug Moar build and run it under the debugger 15:36
ugexe jnthn: 0
jnthn ...
Immediately after the crash?
That's odd
ugexe oh, no. i just ran it in cmd.exe
jnthn Ah
You'd need to run the thing that breaks from the same window 15:37
ugexe ill try doing it immediately after the crash
jnthn Then do itimmediately afterwards
It's a shell variable, iirc
Gets the exit code
ugexe jnthn: -1073740940 15:39
jnthn Pretty sure that's SIGSEGV
ugexe what causes seg faults to sometimes ay Segmentation Fault and other times not? 15:40
arnsholt The "Segmentation Fault" message is printed by the shell, at least 15:43
ugexe i seem to have narrowed it down a bit though. I have Zef::Shell and Zef::PowerShell::Shell is Shell. Then i have Zef::PowerShell::download is PowerShell and Zef::PowerShell::unzip is PowerShell. Whichever one is reached first by the plugin loader loads fine, but the next one always causes the immediately end 15:46
(only on windows, linux is able to load them) 15:50
oh wow, removing a `my constant XXX = q:to/END/;xxx END` fixes it 15:52
github.com/ugexe/zef/blob/master/l...zip.pm6#L5 # this `constant` is technically ok where it is right? 15:53
jnthn ugexe: I've never seen it output Segmentation Fault on Windows...
ugexe: If changing something tiny and random fixes it then it's almost always memory corruption, which is often (especially if no NativeCall is being used) a sign of a GC related issue 15:54
xenu "segmentation fault"? 15:55
ugexe jnthn: but i've changed it to be loaded at different times in the parent application and its always the same
xenu under windows it would be "the application has stopped working" i think
ugexe the first module/plugin that uses `my constant` will always work, and any subsequent one will fail 15:56
jnthn xenu: Yeah, depends on your configuration. 15:58
xenu: But Windows never calls it segfault
ugexe: Hm, OK. It's hard to guess more without seeing the C-level stack trace from the debugger.
xenu you can run rakudo under visual studio (i don't like it) 15:59
or
RabidGravy well it used to be "general protection fault", dunno what the call it now
xenu windbg (i recommend it)
it's really nice
jnthn doesn't code MoarVM in Visual Studio, but does like/use its debugger :)
xenu windbg has more gdb-like experience 16:00
also is jit enabled? i would guess it can cause problems sometimes 16:02
AlexDaniel has anybody looked at rt.perl.org/Public/Bug/Display.html?id=127452 yet? Is there anything I can do to help? E.g. I think that it's a rather recent problem, perhaps it will make sense to bisect it? 16:04
ugexe hmm changing from a heredoc to a plain string seems to fix the behavior as well 16:05
xenu: JIT is enabled 16:13
[Coke] anyone interested in seeing a 2016.01.1, please join #perl6-release. 16:33
abraxxa [Coke]: can I help? 16:35
musiKk What is the purpose of a colon after a type in a method signature? 17:21
musiKk It's similar to an invocant declaration but the parameter has to be provided explicitly and can even be another instance. 17:21
MadcapJake got the evalbot working on slack thanks to shoichikaji++ 17:26
geekosaur musiKk, not sure what you are asking about. example? 17:32
musiKk Not so sure myself. The docs have it a lot, e.g, doc.perl6.org/type/Signature#method_params 17:33
However, the actual implementation does not have it.
geekosaur I think that's just an invocant marker, saying that it has to be called on a defined object (not the type object)? 17:35
musiKk But one is allowed to declare a signature that has a colon behind the type and this is only allowed for the first parameter.
geekosaur implementations generally won't have it, it's just the documentation being clear
musiKk And this parameter has to be passed explicitly.
jnthn It's talking about the self parameter 17:36
musiKk "method(Foo: $foo)" can be called $inst.method($inst) or $inst.method($other_inst) but $inst.method() won't compile. 17:37
jnthn m: class C { method m(C:) { } }; C.m 17:37
camelia ( no output )
jnthn Looks fine to me?
musiKk O_o
geekosaur m: class C { method m(C:D:) { } }; C.m
camelia rakudo-moar a5fe34: OUTPUT«Invocant requires an instance of type C, but a type object was passed. Did you forget a .new?␤ in method m at /tmp/pvaDy1_Lb0 line 1␤ in block <unit> at /tmp/pvaDy1_Lb0 line 1␤␤»
jnthn It's most often used for what geekosaur just showed 17:38
geekosaur m: class C { method m(C:D:) { } }; my C $x .= new; $x.m
camelia ( no output )
musiKk I get "Too few positionals passed; expected 2 arguments but got 1".
geekosaur note that jnthn's passed the type object, which fails :D
what version of rakudo are you running?
musiKk "This is Rakudo version 2015.12-164-g79f77f7 built on MoarVM version 2015.12-29-g8079ca5" 17:39
jnthn Another possible use is writing methods that can only be called conditional on some state
m: class C { has $.state is rw; method m(C $ where $!state == 2:) { } }; my $x = C.new; $x.state = 1; $x.m; 17:40
camelia rakudo-moar a5fe34: OUTPUT«Constraint type check failed for parameter '<anon>'␤ in method m at /tmp/0OP6tSEER2 line 1␤ in block <unit> at /tmp/0OP6tSEER2 line 1␤␤»
jnthn m: class C { has $.state is rw; method m(C $ where $!state == 2:) { } }; my $x = C.new; $x.state = 2; $x.m; 17:40
camelia ( no output )
jnthn Though not sure I've seen anyone do that in the wild...yet :)
musiKk Well, the difference is that I provided a variable name which is wrong it seems.
musiKk But confusing that it compiles. 17:41
Maybe.
timotimo oooh, *of course* you can access $!foo inside the where clause on the invocant parameter 17:49
musiKk: can you paste your code?
geekosaur might want to use :D on the invocant type though
and yes, would like to see the actual code. possibly there is an LTA here or something 17:50
musiKk timotimo: I don't have it anymore but it's pretty much what I showed earlier.
this compiles "class Foo { method foo(Foo: $foo) {} }; my $foo = Foo.new; $foo.foo($foo)" 17:51
timotimo mhm
yeah, that's not surprising 17:52
musiKk :D
timotimo given you declare an invocant as well as a first positional parameter
timotimo perhaps you thought you were actually doing "method foo(Foo $foo:) {}"? 17:52
musiKk I know "$foo:". It's just the documentation that shows a colon after the type. 17:53
I'm still not clear on how that works.
musiKk for example doc.perl6.org/type/Metamodel%3A%3AM...add_method 17:53
timotimo in the docs it's mostly there to make clear what the method belongs to 17:54
but you can literally write it like that if you want
musiKk add_method shows that it receives a "Metamodel::MethodContainer:". The source code has $obj without a type and at runtime it's a Mu.
no, wait
grepped the pod
geekosaur as I mentioned before, the documentation is just being clear about it. if you don't specify a type in the implementation, it will be the type of the class. if you don't specify it at all, it's implicitly handled. 17:55
geekosaur in other words, if you need to, you can explicitly specify the invocant and provide a partial or full specification; if you don't need to, you can let perl6 handle it. but for documentation, it's good to be explicit about the invocant 17:56
musiKk Sorry, I still don't get it completely. The definition of add_method throws me off. 17:57
geekosaur add_method operates on the metamodel 17:58
musiKk It's declared as "method add_method($obj, $name, $code_obj)". How can the documentation add the type to $obj?
timotimo it doesn't 17:59
"Metamodel::MethodContainer: $obj" doesn't add a type to $obj
musiKk oh
timotimo it's the invocant marker. it talks about the type of the invocant.
the invocant is the thing you're calling the method on 18:00
if you add a type to $obj, it doesn't have a colon in between
and if you add a name to the invocant, it'll be before the colon
that should cover about everything, i think? 18:01
musiKk Foiled by the syntax then. That was quite unexpected but yeah, you cleared everything up.
thanks!
timotimo i'm glad :) 18:07
Skarsnik MadcapJake, damn nice x) 18:40
[Coke] nine: rt.perl.org/Ticket/Display.html?id=127454 might be fore you. 18:41
*for
nine [Coke]: yes, jnthn++ already noticed that 18:44
ely-se .tell masak I got the type checker to work well now, in Haskell :) 18:47
yoleaux ely-se: I'll pass your message to masak.
FROGGS[mobile] o/ 18:54
hankache hola #perl6 18:56
vendethiel ely-se: hah, you ended up switching to haskell?I guessed it 19:10
ely-se :) it works very well
stmuk blogs.perl.org/users/steve_mynott/2...-2016.html 19:38
jnthn stmuk++ 19:41
avuserow is there a cute perl6 way to take an array (say `my @a = "a".."d"`) and get all head/tail partitions of it (e.g. (["a"], <b c d>), (<a b>, <c d>), (<a b c>, ["d"]))? 19:57
mspo map to splice or something? 19:58
moritz m: my @a = <a b c>; say @a.keys.map({[ @a[0..$_], @a[$_+1 ..*]]}.perl 19:59
camelia rakudo-moar a5fe34: OUTPUT«5===SORRY!5=== Error while compiling /tmp/RwS90YfIvy␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/RwS90YfIvy:1␤------> 3ys.map({[ @a[0..$_], @a[$_+1 ..*]]}.perl7⏏5<EOL>␤»
dalek href="https://perl6.org:">perl6.org: 396dcb8 | (Steve Mynott)++ | source/downloads/index.html:
bump downloads page to use 2016.01.1
moritz m: my @a = <a b c>; say @a.keys.map({[ @a[0..$_], @a[($_+1) ..*]]}).perl
camelia rakudo-moar a5fe34: OUTPUT«([("a",), ("b", "c")], [("a", "b"), ("c",)], [("a", "b", "c"), ()]).Seq␤»
moritz avuserow: something like that? (maybe with the last element removed)
mspo moritz: nice use of * 20:00
avuserow yeah, that'd work. trying to wrap my head around it :)
moritz m: my @a = <a b c d>; say @a.keys.map({[ @a[0..$_], @a[($_+1) ..*]] if $_ < @a.end}).perl
camelia rakudo-moar a5fe34: OUTPUT«([("a",), ("b", "c", "d")], [("a", "b"), ("c", "d")], [("a", "b", "c"), ("d",)]).Seq␤»
moritz avuserow: it just iterates over all the array indexes, and splits the array there using slices 20:01
skids my @a = "a".."c"; @a[((0,),(1,2)),((0,1),(2,))].perl.say; # just need to figure out how to gen the indices.
avuserow m: my @a = <a b c>; say @a.keys.map({[ @a.head($_), @a.tail(*-$_)}).perl # wonder if .head/.tail are useful?
camelia rakudo-moar a5fe34: OUTPUT«5===SORRY!5=== Error while compiling /tmp/vIoKFoc3xw␤Unable to parse expression in array composer; couldn't find final ']' ␤at /tmp/vIoKFoc3xw:1␤------> 3a.keys.map({[ @a.head($_), @a.tail(*-$_)7⏏5}).perl # wonder if .head/.tail are usef␤…»
skids m: my @a = "a".."c"; @a[((0,),(1,2)),((0,1),(2,))].perl.say; # just need to figure out how to gen the indices.
camelia rakudo-moar a5fe34: OUTPUT«((("a",), ("b", "c")), (("a", "b"), ("c",)))␤»
avuserow m: my @a = <a b c>; say @a.keys.map({[ @a.head($_), @a.tail(*-$_)]}).perl # wonder if .head/.tail are useful?
camelia rakudo-moar a5fe34: OUTPUT«Cannot call tail(Array: WhateverCode); none of these signatures match:␤ (Any:D $: Cool $n = { ... }, *%_)␤ in block <unit> at /tmp/2op8PwpoxC line 1␤␤»
avuserow m: my @a = <a b c>; say @a.keys.map({[ @a.head($_), @a.tail(@a.end-$_)]}).perl
camelia rakudo-moar a5fe34: OUTPUT«([(), ("b", "c").Seq], [("a",).Seq, ("c",).Seq], [("a", "b").Seq, ()]).Seq␤»
moritz m: my @a = <a b c>; say @a.keys.map({[ @a.head($_), @a.tail(@a-$_)]}).perl 20:02
camelia rakudo-moar a5fe34: OUTPUT«([(), ("a", "b", "c").Seq], [("a",).Seq, ("b", "c").Seq], [("a", "b").Seq, ("c",).Seq]).Seq␤»
moritz m: my @a = <a b c>; say @a.keys.map({[ @a.head($_), @a.tail(@a-$_)] if $_}).perl
camelia rakudo-moar a5fe34: OUTPUT«([("a",).Seq, ("b", "c").Seq], [("a", "b").Seq, ("c",).Seq]).Seq␤»
avuserow I can handle empty elements since I have a multi candidate for that. :) 20:02
nine Oh my. Fooling Perl 5 code that checks $ref eq 'CODE' will be awful. Seems like I really need to generate a P5 code reference at runtime that forwards to the P6 callable. 20:08
No overloading or magic will help
hankache why doesn't "rakudobrew build moar" build 2016.01.1? 20:09
geekosaur nine, isn't that sust a sub {}? 20:10
hankache it is building 2015.12-232-ga5fe347
geekosaur *just
stmuk hankache: it's building nom not 2016.01.1 20:12
rakudobrew build moar 2016.01.1
nine geekosaur: currently I create a Perl6::Callable object that overrides &{} to make calls on it just work. But apparently many people don't check for callability but for genuine CODE references.
hankache stmuk: i thought nom would always have the latest commits 20:13
stmuk hankache: not at the moment
hankache stmuk thanks for clarifying 20:14
jnthn hankache: The plans is to cherry-pick those bits into nom; because 2016.01 was built out of cherry-picks, it's non-trivial to simply merge the release branch back.
[Coke] hankache, stmuk, added github.com/tadzik/rakudobrew/issues/80 20:16
lizmat is looking forward to commit to nom or any other branch again
hankache jnthn ok 20:17
hankache [Coke]++ #github.com/tadzik/rakudobrew/issues/80 20:18
stmuk [Coke]: I had a PR which did something similar github.com/tadzik/rakudobrew/pull/75 20:19
stmuk that was added as a new rakudobrew build target rather than default behaviour 20:20
skids m: my @a = "a".."f"; say ([\,] @a) Z (([\,] @a[1..*].reverse)>>.reverse).reverse; # avuserow :-) 20:21
camelia rakudo-moar a5fe34: OUTPUT«(((a) (b c d e f)) ((a b) (c d e f)) ((a b c) (d e f)) ((a b c d) (e f)) ((a b c d e) (f)))␤»
dalek ar/release: b942f0f | moritz++ | / (2 files):
Bump some versions, update README
moritz better wording for the README welcome :-) 20:21
stmuk hmmm actually nom:VERSION doesn't work anymore :/ 20:22
moritz what would that even mean? 20:23
either you want a branch *or* you want a version
stmuk moritz: git show nom:VERSION used to show the last release branch but clearly won't anymore 20:25
moritz oh, as a git command 20:26
stmuk: it'll work again once the commits from the release branch have been cherry-picked back to nom
stmuk ah 20:27
stmuk so the VERSION in the dev branch isn't actually the dev VERSION but refers with a time lag to the last release version :) 20:29
stmuk is it a Rakudo Star 2016.01 or 2016.01.1? the latter might be less confusing? 20:32
moritz 2016.01
stmuk its confusing since R* 2016.01 is based on R 2016.01.1 whereas R* 2014.12.1 was based on R 2014.01 20:34
stmuk R* 2014.12.1 was based on R 2014.12.1 I mean 20:34
its also confusing if R 2016.01.2 is ever created 20:35
[Coke] what do you mean "was based on" ? 20:38
oh, R* vs. R. nevermind! 20:39
I thought you were referring to R itself in each case.
[Coke] stmuk: if people want to not be confused by versions, they are coming to the wrong project here. :) 20:40
stmuk I'm arguing star should always be versioned as the R it is based on
[Coke] I don't think that's worth it.
hankache i agree with stmuk
[Coke] "where's R* 2016.01 ??" 20:40
hankache ah ok you have a point 20:41
perlpilot agrees with Coke ;)
(only because he's right! :)
jnthn More broadly, we may want to be thinking through R*'s role over the coming months: who does it serve, what should its release cadence be, etc. 20:43
stmuk I think where is 2016.01 is more easily explained than why is R* 2016.01.1 based on R 2016.01.2?
stmuk but perl versioning has always been confusing and this is something perl 6 hasn't fixed :) 20:44
moritz stmuk: I don't think people care too much about the compiler versions. We just tell them "download and install R* 2016.01, it implements the stable release of Perl 6 that you all celebrated on Christmas"
hankache jnthn R* is the only way of getting an MSI
i am no big fan of batteries included, but i like MSI's for Windows 20:45
jnthn hankache: Well, it's the only thing we make an MSI for. ;) But yes, "get a batteries included Perl 6 distribution in a single download on Windows" is an important role it serves.
spider-mario hey, so this hangman implementation used to work (don’t remember when exactly, though): paste.awesom.eu/stXJ
but now, it prints this instead: paste.awesom.eu/bPVr
(“now” = rakudo 2015.12) 20:46
why does it think that draw expects two arguments?
El_Che The 6 in Perl 6 kind of broke Perl 5 versioning and confused everyone outside the echochamber. A good versioning scheme would be golden for adoption
moritz spider-mario: did it ever work post-GLR? 20:47
spider-mario when was that?
jnthn for @drawing Z @mask -> $d, $m {
I think it wants to be
for flat @drawing Z @mask -> $d, $m {
Which afaict is line 30
It's the block that's getting 1 arg instead of 2, not draw itself 20:48
spider-mario oh
that would make sense
jnthn 'cus Z produces sublists and those don't flatten.
spider-mario (although the error message is LTA, to be honest)
hm, the error message seems to stay the same
oh, not the same, actually 20:49
moritz or for @drawing Z @mask -> ($d, $m) {
spider-mario it now points to a different line that uses Z as well, I’ll try and fix it too
[Coke] el_che; well, it's always going to be a 6. that ship has sailed. 20:50
El_Che [Coke]: of course. I am talking about Perl 6 itself
spider-mario ah, it works now :)
thank you, moritz and jnthn
El_Che [Coke]: if there is a lesson in bygones, it's that clear versioning is important 20:51
[Coke] El_Che: are you familiar with the plan going forward?
for v6.c, 6.d, 6.d.a, 6.c.1, etc?
El_Che [Coke]: the spec, I think so. 20:52
stmuk I assume pure 6.c monthly releases are less likely in the future?
El_Che talked with liz and nine about it at fosdem
hankache 6.d.a, 6.c.1?? 20:53
dalek ar/release: e0eb876 | moritz++ | tools/build/Makefile.in:
Bump another version
 20:54
[Coke] stmuk: there will no doubt be a 6.c.1 this year. not planned, based on what comes up. 21:01
feb release will probably also support 6.d.a
(I think that's the right version. basically: the beginnings of 6.d)
moritz hack.p6c.org/~moritz/rakudo-star-20...RC0.tar.gz # please test 21:02
[Coke] OH NOES I started from 1 when I did it!
stmuk :D 21:04
moritz next time I'll start from -0.5 :-)
stmuk its actually RC2
moritz my mind seems too small for two options 21:05
whenever I specify --prefix in Configure.pl, I forget --gen-moar
AlexDaniel again, why not go the browser path and have Perl 42 next year? :) 21:06
moritz AlexDaniel: because we don't want to buy that many domain names
stmuk there was an rakudo-star-2016.01-RC0.tar.gz released on 2016-01-07
jnthn For those wondering about Rakudo versioning etc., there's some more details at gist.github.com/jnthn/c10742f9d51d...-rakudo-md - though note that to full appreciate that short text you may want to read the much longer text above it. :)
stmuk see irclog.perlgeek.de/perl6-toolchain...01-07/text 21:07
moritz stmuk: ah right, but that wasn't based on an actual release :-)
so, it like, doesn't count
stmuk I give up! :)
AlexDaniel moritz: buying another domain is as expensive as renewing the old ones, so let's just keep 3 or 4 latest ones
moritz anyway, feedback for *my* R* RC0 very welcome
stmuk . o O ( rakudo-star-2016.01-RD0.tar.gz ) 21:09
AlexDaniel … at least such approach will make sure that Perl 5 folks are not going to jump to Perl 7 :D 21:10
moritz I'd have no problem with that
packagers may, though :-)
xenu i have an idea how to make perl 5 guys feel good. new versioning scheme: from now on, perl 6 uses even numbers (so perl 6, 8, 10, 12...) and perl 5 uneven (5, 7, 9). perl 6 has releases on december and perl 5 in june ;) 21:13
xenu s/uneven/odd/ 21:14
moritz
.oO( new channel: #perl6-version-trolling )
stmuk Koremutake Memorable Random Strings based on git hashes :)
AlexDaniel xenu: this does not provide enough room for another perl 21:20
revhippie uneven is for you can't even 21:26
revhippie ...that sentence was hard enough to parse *before* I left out the word "when" in the middle. 21:28
El_Che version-trolling? perl6 uses prime numbers :) 21:33
sortiz \o #perl6 21:36
ely-se hi sortiz 21:41
perlpilot moritz: almost done testing RC0 now and it looks good so far 21:43
moritz perlpilot: \o/ and thanks for testing 21:44
stmuk moritz: maybe not worth it but I think the JSON::Fast needs a bump to get rid of the leading v in meta info warning 21:44
moritz if somebody wants to help: please contribute a release announcement in the "release" branch of rakudo/star 21:45
moritz stmuk: seems to be at the right version 21:46
moritz stmuk: did you check the "release" branch of star? 21:46
stmuk I just installed your tarball and saw the v warning 21:47
which appears fixed in github.com/timo/json_fast/commit/6...97772c615a
I'll look further 21:48
stmuk oops your version is right 21:50
perlpilot moritz: When I run make rakudo-spectest, I get t/spec/S17-supply/syntax.t .................................... Failed 3/60 subtests 21:55
moritz: when I run that test by itself, I get no failures.
avuserow skids++
the context of my question is actually kind of interesting. I was helping a friend work through www.reddit.com/r/dailyprogrammer/c...er_splits/
and after making a boring python one, I made a slightly less boring p6 one, and then thought of a nice way to make it work with partitions. but looking at it, smls++ has the top-voted solution with an amazing regex hack 21:56
stmuk its a pity all the R* modules aren't under perl6-community-modules so PRs like github.com/daleevans/p6-Template-M...427a652840 get merged
ahhh its the JSON::Fast specified by panda which is out of date 21:57
embedded in I mean 21:58
moritz perlpilot: ah well, I chose to ignore that :-) 22:05
stmuk I don't see panda in /opt/rakudo-star-2016.01-RC0/bin 22:08
like it used to be with R* .. I guess that's expected? 22:09
Zero_Dogg can't () be omitted on function calls in classes on self? self.printd "Blabla"; gives a "two terms in a row" error 22:13
ugexe printd: "Blablah";
(note the colon) 22:14
Zero_Dogg interesting, did not know about the colon. What does it do, just start a list of parameters for whatever preceeds it?
ugexe m: (^5).map: {.say} 22:16
camelia rakudo-moar a5fe34: OUTPUT«0␤1␤2␤3␤4␤»
jnthn Zero_Dogg: Pretty much that. Means that the precedence after it is dropped low enough to parse a comma-separated argument list. 22:17
Zero_Dogg jnthn, ugexe: Thanks!
sortiz moritz, The relevant output when building and installing your R* RC0 in FC23: gist.github.com/salortiz/1087df9fb74b34f576ca 22:25
dalek c: ce4f4b2 | (Wenzel P. P. Peppmeyer)++ | doc/Language/functions.pod:
tell how arguments are written down
 22:32
Zero_Dogg When is BUILD actually executed? Will Mu's default initialization have been performed before it reaches BUILD (ie. can BUILD already access class variables that the default constructor would initialize?). Given obj.new(debug => True); will BUILD on obj be able to check $.debug? 22:34
Zero_Dogg When is BUILD actually executed? Will Mu's default initialization have been performed before it reaches BUILD (ie. can BUILD already access class variables that the default constructor would initialize?). Given obj.new(debug => True); will BUILD on obj be able to check $.debug? 22:37
sorry, wrong terminal, didn't mean to spam.
dalek c: 8a5200c | (Wenzel P. P. Peppmeyer)++ | doc/Type/Block.pod:
show that Block got a return value
 22:40
timotimo sortiz: the "implicit declaration of pthread_yield" has been known for at least a year :) 22:42
the "please remove initial v from meta" in JSON::Fast has been fixed some time ago, but it apparently has to be pulled into some other repositories 22:44
stmuk_ github.com/tadzik/panda/tree/maste...JSON__Fast 22:46
timotimo oh 22:47
yeah, that patch i meant is only 7 days old
sortiz timotimo, Ok. BTW, can you take a look at PR#701?, I appreciate some feedback and guidance 22:49
timotimo what repository is that for? 22:49
sortiz rakudo 22:50
timotimo oh, oof 22:51
timotimo i don't think i'm comfortable making that call 22:51
between adding the shortname method or using .can
sortiz You was the instigator ;)
timotimo i recall 22:52
Zero_Dogg sorry, wrong terminal, didn't mean to spam.
GAH, I'll just close this terminal now, before I end up pasting my entire sourcecode here
sortiz Well, I'll be waiting for some other opinions, thank you. 22:57
donaldh moritz: R* RC0 builds and installs on OSX El Cap 22:58
moritz: 1 spectest failure - t/spec/S17-supply/watch-path.t - which looks OSX specific 22:59
perlawhirl avuserow: an attempt using head and tail... if a little verbose 23:36
m: my @a = 'a'..'c'; for 1..@a.elems -> $i { say @a.head($i), @a.tail(@a.elems - $i+1) }
camelia rakudo-moar a5fe34: OUTPUT«(a)(a b c)␤(a b)(b c)␤(a b c)(c)␤»
perlawhirl htm
hrm
m: my @a = 'a'..'e'; for 0..@a.elems -> $i { say @a.head($i), @a.tail(@a.elems - $i) }
camelia rakudo-moar a5fe34: OUTPUT«()(a b c d e)␤(a)(b c d e)␤(a b)(c d e)␤(a b c)(d e)␤(a b c d)(e)␤(a b c d e)()␤»
perlawhirl m: my @a = 'a'..'d'; for 1..^@a.elems -> $i { say @a.head($i), @a.tail(@a.elems - $i) } 23:37
camelia rakudo-moar a5fe34: OUTPUT«(a)(b c d)␤(a b)(c d)␤(a b c)(d)␤»
perlawhirl ahh, that's better
choroba hi, I noticed blogs.perl.org/users/coke/2016/02/r...eased.html 23:46
does it mean I can do something to get the new version into my rakudobrew?
or do I have to wait for the "Rakudo Star release"
perlawhirl rakudobrew build moar 2016.01 23:47
then... rakudobrew global moar-2016.01
if you are using rakudobrew... which you should
captain-adequate Nice... Was gonna ask the same.
Thanks perlawhirl
perlawhirl rakudobrew, like other language *brew ( or *env ) segregates modules, so after you switch versions, rebuild panda, and reinstall modules 23:48
captain-adequate Just like npm for node. Got it.
Or nvm I should say. 23:49
Node version manager.
Glad to see the same for perl6
AlexDaniel m: my @a = ‘a’..‘e’; say [@a[0..$_], @a[$_^..*] for ^(@a - 1)] 23:53
camelia rakudo-moar a5fe34: OUTPUT«[((a) (b c d e)) ((a b) (c d e)) ((a b c) (d e)) ((a b c d) (e))]␤»
AlexDaniel avuserow: or that ↑ 23:56
by the way, is it a bug or not:
m: my @a = <a b c d>; say @a[0..*]
camelia rakudo-moar a5fe34: OUTPUT«(a b c d)␤»
AlexDaniel m: my @a = <a b c d>; say @a[0..^*]
camelia rakudo-moar a5fe34: OUTPUT«(a b c d)␤»
perlawhirl many ways to skin a cat. I'm sure i did something like that by accident with the Z op, but can't recall exactly. i may be mis-remembering 23:57
gfldex m: my $r = 0..^*; dd $r 23:59
camelia rakudo-moar a5fe34: OUTPUT«Range $r = (0, 1, 2, 3, 4, 5, 6, 7, 8, 9...Inf)␤»
gfldex AlexDaniel: what is Inf-1 in your eyes?

Powered by Cro and the Raku® Programming Language.

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

Thank you!