»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:11 RabidGravy left 00:17 _28_ria left 00:42 zacts left 00:47 aindilis left 00:48 aindilis joined 00:52 edehont left
dalek c: 4897e96 | (Wenzel P. P. Peppmeyer)++ | doc/Language/functions.pod6:
show my &somefix:<foo> = form
01:03
sectbot/graphs_for_benchable: 669e5c7 | (Daniel Green)++ | / (2 files):
Initial attempt at adding graphing to benchable

This works, but requires cloning the gist onto the filesystem and doing git add, git commit, git push
01:06
01:12 woolfy left, woolfy joined 01:15 kid51 joined 01:18 dwarring left 01:31 huggable left, huggable joined, FROGGS_ joined 01:33 molaf left 01:35 FROGGS left 01:46 molaf joined 01:53 kalkin- joined 01:57 kalkin-_ left 02:08 kid51 left 02:13 cooper joined, canopus left 02:20 canopus joined 02:31 mohae left 02:43 noganex joined 02:46 noganex_ left
dalek sectbot: 6fd5760 | (Daniel Green)++ | / (2 files):
Initial attempt at adding graphing to benchable
03:02
03:10 setty1 joined 03:39 mohae joined 03:44 TreyHarr1 left
dalek sectbot: 498c685 | (Daniel Green)++ | benchable.pl:
No need for a unique filename when uploading a graph
03:45
03:50 adu joined 03:55 djbkd joined 04:01 skids joined 04:03 rgrinberg joined 04:08 khw left 04:12 vimal2012 joined
vimal2012 I got the error "Cannot modify an immutable Str", when I ran this file paste.debian.net/plain/782191 How to correct this error? 04:13
04:14 MasterDuke left
AlexDaniel m: my $string = "The \n quick \n brown \n fox \n jumps \n over \n the \n lazy \n dog \n"; for $string.lines { say S/ \s // } 04:15
camelia rakudo-moar 7976fe: OUTPUT«The␤quick ␤brown ␤fox ␤jumps ␤over ␤the ␤lazy ␤dog ␤»
AlexDaniel that's one way
vimal2012 Upper case S?
AlexDaniel yea. “The lower-case version (s///) substitutes in-place, while the upper-case version (S///) leaves the original alone and returns the resultant string.” (docs.perl6.org/language/operators#...%2F%2F%2F) 04:17
Frameless vimal2012: perl6.party/post/Perl-6-S---Substit...n-Operator
m: .say for "The \n quick \n brown \n fox \n jumps \n over \n the \n lazy \n dog \n".words
camelia rakudo-moar 7976fe: OUTPUT«The␤quick␤brown␤fox␤jumps␤over␤the␤lazy␤dog␤»
AlexDaniel yea, it depends on what you actually want 04:18
Frameless m: .say for "The \n quick \n brown \n fox \n jumps \n over \n the \n lazy \n dog \n".lines».trim
camelia rakudo-moar 7976fe: OUTPUT«The␤quick␤brown␤fox␤jumps␤over␤the␤lazy␤dog␤»
04:31 vimal2012 left 04:35 Cabanossi left 04:39 Cabanossi joined 04:40 Tonik joined 04:57 kaare_ joined 05:02 holyghost joined 05:11 adu left 05:17 rgrinberg left 05:19 BenGoldberg left 05:23 rgrinberg joined 05:27 holyghost left 05:31 huggable left, huggable joined 05:39 AlexDaniel left 05:57 stmuk_ joined, setty1 left 05:58 skids left 05:59 rgrinberg left, stmuk left 06:18 holyghost joined 06:35 blah joined, blah left 06:41 Tonik left 06:43 stmuk joined 06:46 stmuk_ left 06:47 CIAvash joined 06:48 holyghost left 06:50 holyghost joined 07:13 CIAvash left 07:16 darutoko joined 07:24 labster left 07:29 aries_liuxueyang left 07:30 marcel_ joined, aries_liuxueyang joined 07:33 canopus left 07:34 canopus joined 07:35 wamba joined 07:37 firstdayonthejob joined 07:38 setty2 joined
TimToady drove from Köln to Milano yesterday, today we drive to Rome for the Curry On conference that starts tomorrow 07:38
07:39 djbkd left
FROGGS_ TimToady: happy driving :o) 07:41
07:44 setty2 left 07:45 djbkd joined
ShimmerFairy suddenly wishes $*PROGRAM-NAME were really $*PROGRAM_NAME again, grumble grumble... 07:52
07:52 tomboy64 left
FROGGS_ aye 07:54
07:55 CIAvash joined
ShimmerFairy since my US keyboard has _ as the shifted version of -, as I've mentioned before (also, it looks better) :P 07:55
07:56 djbkd left 07:57 djbkd joined 07:58 labster joined 07:59 AlexDaniel joined 08:00 tomboy64 joined 08:02 domidumont joined 08:06 domidumont left 08:07 domidumont joined 08:10 pmqs_ joined
FROGGS_ ShimmerFairy: that's what I said when the kebab-all-the-things patch landed 08:10
we should kebab methods names and variables only if the are lowercase 08:11
ShimmerFairy It's what I've said since I found out about it :) . I had to look in the documentation because I couldn't figure it out with $*EXECUTABLE_NAME, $?EXECUTABLE_NAME, $*PROGRAM_NAME, etc.
nine If that's the worst flaw in Perl 6's design, I can live with it. 08:12
FROGGS_ the reason for kebab after all was to save that keystroke (shift) for typing lowercase names
so introducing a keystroke for constants and other stuff is against that principle 08:13
moritz also I hope we'll eventually have IDEs that can autocomplete such names
ShimmerFairy I think a good compromise solution would be to have both variants, for people who like kebab-case, and for the normal people :)
FROGGS_ ahh, no
:o)
ShimmerFairy moritz: I don't use autocompletion in my programming anyway, so that's not a solution.
FROGGS_: solution #2: switch to common-lisp style case-insensitive names, with _ being the "capital letter" of - . :P 08:17
moritz ah well, you'll just have to remember then :-) 08:18
08:18 djbkd left
ShimmerFairy moritz: I think my problem ultimately comes down to the fact that, to me, YELLING-KEBAB is both ugly and unintuitive. 08:20
stmuk shouldn't kebab case be localised to gyro (.de) and kebob (.ca) at least? :) 08:31
08:37 domidumont left
AlexDaniel I think that it's weird to adapt a language to some random keyboard layout 08:37
isn't it supposed to be other way round?… In a perfect world, at least
anyway, you can always change your keyboard layout, if that bothers you a lot 08:38
and while you are at it, throw in some unicode characters
08:47 jcallen_ joined 08:50 huf joined 08:51 Walex joined 08:58 AlexDaniel left
marcel_ m: Buf.new(10 +& 0xFF) 09:02
camelia ( no output )
marcel_ m: say Buf.new(10 +& 0xFF)
camelia rakudo-moar 7976fe: OUTPUT«Buf:0x<0a>␤»
09:02 kanishka joined
marcel_ m: my int $i = 10; say Buf.new($i +& 0xFF) 09:03
camelia rakudo-moar 7976fe: OUTPUT«Type check failed in initializing element #0 to Buf; expected uint8 but got Int (10)␤ in any at gen/moar/m-Metamodel.nqp line 1736␤ in block <unit> at <tmp> line 1␤␤Actually thrown at:␤ in any at gen/moar/m-Metamodel.nqp line 3055␤ in any…»
09:03 girafe joined
marcel_ m: my int $i = 10; ($i +& 0xFF).WHAT.say 09:04
camelia rakudo-moar 7976fe: OUTPUT«(Int)␤»
marcel_ so whats the problem, anyone?
masak m: say Int ~~ uint8 09:05
camelia rakudo-moar 7976fe: OUTPUT«Cannot unbox a type object␤ in block <unit> at <tmp> line 1␤␤»
marcel_ It is introduced in the latest few rakudo versions, before it was ok 09:06
nine bisectable: my int $i = 10; say Buf.new($i +& 0xFF) 09:20
bisectable nine: No build for 'bad' revision
nine bisectable: help 09:21
bisectable nine: Like this: bisectable: good=2015.12 bad=HEAD exit 1 if (^∞).grep({ last })[5] // 0 == 4 # RT128181
nine bisectable: good=2015.12 bad=HEAD my int $i = 10; say Buf.new($i +& 0xFF)
bisectable nine: No build for 'bad' revision
nine bisectable: good=2015.12 bad=b06a99ebdd79393ce109cc214a6019ada5a0218a my int $i = 10; say Buf.new($i +& 0xFF)
bisectable nine: bisect log: gist.github.com/58e1921c8d812add97...d8fe30663b
nine: (2016-07-11) github.com/rakudo/rakudo/commit/de5d9e7
nine lizmat: ^^^ 09:22
09:28 iH2O joined
iH2O hi folks 09:28
09:28 kaare_ left
holyghost Hello 09:28
09:29 b2gills left
iH2O hello holly 09:31
09:31 huggable left, huggable joined 09:35 Actualeyes joined 09:37 djbkd joined
holyghost I am going to install strawberry perl one o these days, on a core duo 2 09:44
I'd really like something as activestate or strawberry for perl6
It's just a no-jit rakudo now 09:45
2016.01 or something
I am more of an end-user as I don't fully grok the source code of perl6, it's difficult enough for some people, except maybe C++ folks 09:46
stmuk there is a 64 bit jited 2016.04 msi
holyghost I see 09:47
right I forgot about that
I only have mac mini as a iX processor system
an i5 09:48
iH2O srry to hear that
holyghost It's loaded with mac os x el capitan
stmuk the 64 bit 2016.07 dmg then
holyghost sure I am going to send in a program to nintendo, a metroid game then I'll have some money
maybe I'll buy an amiga x1000 or x5000 09:49
an amigo os 4.1 PPC machine
s/amigo/amiga
They went more or less the apple way with their processors
stmuk an i5 is 64 bit 09:50
holyghost I know
stmuk I don't understand your problem then 09:51
holyghost well strawberry and activestate is win32
e.g. they have COM bindings
stmuk you are using Windows?
which OS? 09:52
holyghost the interface of an instance in an array
7
stmuk on what processor
holyghost core duo 2
stmuk well that runs Win 7 64 bit OK then
holyghost I forgot if you can load a 64bit on an amd64 arch
I though not 09:53
stmuk is whether your OS is 32 bit or 64 bit
holyghost s/though/thought
It's 32 bit
stmuk I think you can still download the 64 bit for free until the end of the month 09:54
for Win 10 I mean
holyghost ok good point
stmuk you could also probably build 32 bit rakudo OK with the gcc in strawberry perl 09:55
but there would be no jit and some of the NativeCall modules might not work 09:56
holyghost To download win 10, do I just click the microsoft.com link in google ?
stmuk I forget but there was a link on the MS site for the ISO 09:58
holyghost start a download
10:02 djbkd left 10:03 iH2O left
holyghost ok I hope that works out 10:04
thanks
10:04 RabidGravy joined
stmuk np 10:05
10:07 jeek joined
holyghost If you install win 7 from a 6 year old DVD-R, it installs 32 bit right ? 10:16
I don't know much of windows, there's no 'uname -a' for example
timotimo you should be able to right-click on "my computer" on the desktop to figur eout if it's 32bit or 64bit 10:17
holyghost ok
El_Che if I remember correctly, there where 64-bit win7 iso's also 6 year ago 10:18
10:19 domidumont joined
timotimo who is hack user emeric? 10:19
El_Che on my laptop (I am a Linux user) I have a w32 virtual box vm, and a 64 win7 parition for games (I boot that once a year for steam sales :) )
timotimo they have a screen -r irc that's consuming a crapton of cpu time
one is already up to 106 hours, the other is at 107 hours 10:20
i'm pretty sure it isn't supposed to be like that?
psch the 50% cpu for each is the more interesting bit i think
well, unless the 106h is cpu time too, maybe? 10:21
timotimo i was just wondering what might be causing the noticable amount of cpu usage on collect 10:22
for one week, the usage has been pretty much constant
one part is of course your development, but i don't think you're developing for actually 24 hours a day without pause :) 10:23
i kind of wish i could ask collectd for graphs 2x as wide for the same amount of time
i mean collectd-web
psch nah, i kinda like sleep occasionally :)
timotimo grafana is so much nicer user-interface wise, i think
w says emeric has been idle for 7 days now 10:24
10:24 holyghost left
timotimo i can just kill the weechat process and see who leaves %) 10:25
timotimo read too many BOFH stories
jeek /kill 'em all, root'll sort 'em out
timotimo %) 10:26
10:28 domidumont left 10:29 holyghost joined
nine timotimo: you can actually. Open the image and add &width=2000 to the URL 10:30
timotimo well, it'd be nice if the web interface'd allow me to do something like that for all graphs, and keep that setting 10:31
10:32 labster left
nine There's also a GraphWidth setting in the config file 10:32
timotimo oh? neat.
but that's not per-user, right? 10:33
nine no, it isn't
timotimo i don't want to make it unusable for the other users
holyghost Is there going to be 32 bit support for perl6 in the future ? 10:42
timotimo uh oh, what's it now. ssh hack.p6c.org is taking a long time
it hasn't crashed again, has it? >:(
psch uptime is <2 10:43
well, load :)
timotimo nope, it's fine
it just took a long time for some random reason 10:44
stracing the screen session emeric has tells me there's something rather wrong going on 10:46
it seems to be stuck in a while loop trying to attach or something
i killed that one that was spinning its wheels, that seems to have helped 10:47
10:51 domidumont joined 10:58 Ven joined 11:01 rindolf joined 11:09 kid51 joined 11:18 domidumont left 11:20 kaare_ joined
nine holyghost: rakudo should already run on 32 bit systems? 11:21
lizmat nine: looks like:
# nqp::setelems(to, $j + $elems); # presize for efficiency
fixes the problem ??!!
oops, no, my bad 11:22
11:23 rgrinberg joined 11:26 marcel_ left 11:31 rgrinberg left 11:42 iH2O joined
holyghost nine : no-jit and 2016.01 11:44
nine : I'll try the strawberry perl to install a later version tomorrow
I just got back from the pub 11:46
I drank a good coffee
11:51 domidumont joined 11:57 silug left 12:12 silug joined 12:13 iH2O left 12:29 BenGoldberg joined 12:31 Ben_Goldberg joined 12:34 BenGoldberg left 12:55 kid51 left, _28_ria joined 13:10 Actualeyes left 13:26 bob777 left 13:27 Actualeyes joined, bob777 joined 13:31 huggable left, huggable joined 13:37 Zoffix joined
Zoffix holyghost, Strawberry Perl is a sister language, not Perl 6. You should build Rakudo from sauce. Latest version is only a day old: rakudo.org/downloads/rakudo/ 13:38
Though I don't think there's JIT on 32-bit versions
stmuk you could also probably build 32 bit rakudo OK with the gcc in strawberry perl 13:39
holyghost ok 13:40
I'll give it a go tomorrow 13:41
I'm taking the day off
I never used strawberry perl, I only know some activestate
mst they're both perl5 13:42
holyghost The shell is not a problem though
mst strawberry's the (more recent) community one, activestate is a commercial one
they're the same language. perl6 is not.
holyghost ah I didn't know that
that I know
13:43 TreyHarr1 joined
holyghost I need the perl5 for building perl6 I thought 13:43
I won't meddle in perl5 anymore though, the language is really the swiss army knife but maybe the OOP is lacking, though I try to use new languages except for C/C++ 13:44
stmuk you can build rakudo with strawberry perl and git on windows
Zoffix holyghost, yeah, and strawberry also has C compiler and make tool that you'll need. So do install it. I was just clarifying that you won't get Perl 6 by just installing Strawberry alone
holyghost ok I see
mst holyghost: perl5's OOP is better than python or ruby or javascript
Zoffix holyghost, have you experimented with Moose in Perl 5? That's the Perl 5's OO
stmuk and windows git
mst holyghost: that's a lot of why I use it so much 13:45
holyghost: ruby is much prettier but ruby's OO is epically terrible in comparison
13:45 xebra joined, xebra left
holyghost I am not too good at it, but roles are actor-language things ? 13:46
It's not the C++11 or something but we need to define a whole OOP system in scheme macros for school 13:47
mst holyghost: perl5 and perl6 are the only languages with a decent 'has' and support for roles (traits)
holyghost s/need/needed
mst though traits come from smalltalk land
'has' and before/after/around are more CLOSish
holyghost I am on the basic level of such things
timotimo role isn't actor-related
mst no, but actors and traits came from, I think, similar groups of researches 13:48
because smalltalk ;)
holyghost st, I wrote a CGI script in GNU st once
timotimo probably, yeah
GNUstep?
holyghost GNU smalltalk
timotimo ah 13:49
holyghost it has a command line #!
hasbang
hashbang I mean
Then I learned some squeak, pharo and seaside smalltalk 13:50
It was based on the book ST-80 by adele 13:51
goldberg
mst ok, so, if you learn traits from smalltalk 13:52
those apply to perl6 and to M* OO in perl5
holyghost I misparsed Moose
I thought it was the MUD perl5 system 13:53
MOO, it is called I think
perl5 OO was full of references and such in a program I wrote 13:54
In the days of perlyroids 13:55
about 13 years ago
perlyroids was a perl GUI/adobe module based (only for IE) game written in perl5, you could run it in your browser such as hotjava in those days 13:56
It's long gone though
I had a friend back at worldforge.org, bryce who was a perl wizard, he brought me to perl5 13:57
I adapted OO of perl5 then 13:58
It was too difficult for me to fully understand
@ISA for example has arcane features for example
like the rest of perl5 :-)
Anyway, I hope I can compile the thing tomorrow 13:59
I am working on a game for perl6, I'll post it to perlmonks.org
SDL2::Raw based 14:00
I only need to make an xpm parser to not swap buffers
e.g. png buffer to SDL2 texture
SDL2 is slow in windows though, the linux version is faster, I want to make it portable 14:01
SDL2::Raw I mean
timotimo why is the linux version faster?
that makes no sense to me
holyghost I have no idea
mst holyghost: ah, you got taught the obsolete 2000-era perl5 OO 14:02
holyghost: p3rl.org/Moose is what you'd use these days
holyghost: example: git.shadowcat.co.uk/gitweb/gitweb.c...be;hb=HEAD 14:03
Juerd Honestly, I don't even consider Moose because my code needs to run fast. 14:04
holyghost Moose looks good, it's like our OOP scheme classes
stmuk I've used SDL2::Raw on Mac and Linux and noticed no difference in speed 14:05
nine Juerd: but Moose is reasonably fast? It just comes with a huge startup penalty.
Juerd nine: Compared to Perl 6 it's fast :P
14:07 Zoffix left
holyghost Is the syntax of the perl6 OO system finished or may it be extended without meta-thingy ? 14:08
mst Juerd: I'm sorry? please explain how Moose is slower than the equivalent raw perl5 OO
holyghost metasyntactic evauluation I mean
mst holyghost: roughly, Moose was a port to perl5 of the original perl6 metamodel, and the current perl6 metamodel is a port back to perl6 of Moose ;)
holyghost *lol* 14:09
mst note that Moose does have significant startup overhead, which is why for fast starting scripts I wrote Moo
14:09 spider-mario joined, canopus left
Juerd mst: It was when I benchmarked it (2 years ago), mostly when it came to accessors. 14:09
If you say I did something wrong because it's easily as fast as classical Perl 5 OO, I'll take your word for it. 14:10
mst Juerd: the accessors are normally as fast as they would be if hand-written
nine Juerd: with or without __PACKAGE__->meta->make_immutable?
mst now, admittedly, perl5 sub call overhead means that's not *amazingly* fast anyway 14:11
nine: irrelevant
Juerd nine: Without
mst oh, Moose without ->make_immutable is "please make my program really slow"
however the accessors are the one thing I *wouldn't* expect to be affected by that
but if you didn't even get as far as ->make_immutable, you definitely didn't learn enough Moose to produce valid benchmarks, I'm afraid
timotimo if you have to reach a double-underscored allcaps thing in order to make your program behave sensibly, that sounds bad :P 14:12
mst timotimo: __PACKAGE__ is standard perl5 thing, blame larry for the name :P
14:12 canopus joined
Juerd mst: If that's an important thing, I'm confused that it didn't make it to perldoc Moose :) 14:13
nine Juerd: it's in Moose::Manual and explained in Moose::Manual::BestPractices
Juerd timotimo: metacpan.org/pod/CLASS :) 14:14
mst Juerd: perldoc Moose has a 'new to Moose' section which says you must start with Moose::Manual
Juerd: which shows it in the examples
Juerd I see 14:15
mst Juerd: but, yeah, a Moose accessor will be basically the same speed as a hand written one, and if it isn't, either you did something silly that pessimised it, or there's a bug
either way, you should've come to #moose on irc.perl.org and asked
rather than just spreading FUD about perl :(
timotimo pff, having to put a million use statements at the beginning of every class :P
mst timotimo: p3rl.org/Import::Into solves that
nine timotimo: I think that's why Perl 6 was started ;)
Juerd mst: I've been on IRC too long to *ever* join a channel just to complain about something.
mst Juerd: so instead you've spent two years telling people to avoid one of the best parts of perl5 because it's 'slow'. I don't consider that a net improvement :P 14:16
Juerd mst: Hold your assumptions please. I believe this is the first time I ever said anything about this outside the project group.
I'm not really involved in Perl that much anymore. Haven't been to a YAPC or national workshop in years, etc. 14:17
So no, I have not been spreading FUD. If I had, you'd have noticed that :)
timotimo mst: pff, having a million modules to do exactly what you need!
Juerd mst: About irc.perl.org: my IRC box is ipv6 only. 14:19
mst Juerd: I'd still rather have fixed your mistake two years ago
ah, hrm, I'm not sure which of the nodes currently do ipv6
Juerd I used to be on irc.perl.org. Out of a dozen networks it's the only one I had to ditch when I moved my client to an ipv6 only box... :)
mst it was more of them, but then the upstream v6 transport for some of them started being flaky 14:20
Juerd Although I used to be in #amsterdam.pm only
mst and thereby were causing more harm than good
Juerd: well, if you ever write perl5 again, come onto #perl on this network and we can help you do OO competently :) 14:21
timotimo oh no, not the mst-ruled channel with all of mst's goons!
Juerd mst: I'm perfectly fine with my competence regarding classic Perl OO.
timotimo on an mst-controlled network!
mst Juerd: yeah, but that's way too much typing and opportunity to typo things :)
Juerd Too much typing? Nah.
mst I mean, the first thing that comes to my mind is - if you didn't like 'has', why are you even in here? surely you'd be happier in #banging-rocks-together ? :D 14:22
Juerd I like "has $.foo is rw;"
mst seriously, I spent like four years with no Moose to figure out what the feature set for Moo needed to be and it was horrible
I don't wanna go back there ever again
Juerd I dislike "has 'foo' => ( is => 'rw' );" (if I recall the Moose syntax correctly)
mst hence why Moo fatpacks so I don't even have to treat it as a dep
has foo => (is => 'rw'); # is how I'd spell it, but yes 14:23
Juerd I'd actually spell that "has qw(foo is rw);"
mst personally, I mostly use shadow.cat/blog/matt-s-trout/do-you-copy/ 14:24
nine That's actually only 2 characters more than Perl 6
Juerd In any event I don't think I'll ever write the "modern" Perl 5, if that means Moose, unless I happen to get involved in a project that already does. 14:26
spider-mario weird,
running make install after building rakudo, I get:
Could not find CompUnit::Repository::Staging
Juerd I'd rather wait for Perl 6 to become production ready
spider-mario is that a known issue?
nine spider-mario: it isn't
spider-mario: what rakudo version? 14:27
spider-mario 2016.07
2016.06 was fine
I can try and bisect the issue
14:27 setty1 joined
mst Juerd: that's fair enough. I've been addicted to it for years now so 14:27
timotimo spider-mario: you probably need to run configure.pl again
nine spider-mario: do you have a lib/CompUnit::Repository::Staging?
spider-mario timotimo: I did, it’s a fresh build 14:28
mst Juerd: oh, minor note, Moo uses Class::XSAccessor if installed where possible
Juerd: so "has foo => (is => 'rw');" gets you an XS accessor and is actually rather more performant than writing your own
spider-mario nine: there is one in the source directory, but it seems to look for it in /usr/share/perl6
Juerd Oh, that's another thing. I write perl5 for tiny perls where even 'strict.pm' is not shipped by default.
mst aha
Juerd So I have to at least keep up with raw perl skills :D 14:29
nine spider-mario: can you tell me more about how exactly you install rakudo?
Juerd (Note that this is why in Net::MQTT::Simple, use strict is commented out!)
spider-mario I am actually packaging it
the build script runs Configure.pl --prefix=/usr --backends=moar
then make,
Juerd mst: Haven't tried Moo at all yet.
spider-mario and finally make install DESTDIR=(some packaging directory)
mst Juerd: lots of my code is designed to run given only a full basic perl5 install and a single script, hence Moo and App::FatPacker
but I don't generally have to go smaller than that 14:30
Moo is basically the Moose surface syntax with no metaprotocol
Juerd Sounds perfectly usable
mst but it fakes having been a Moose class all along if you point Moose at it
dalek c: d9a902c | Altai-man++ | doc/Language/operators.pod6:
Make hyper operators searchable
mst which means you can port thing Moose -> Moo and things subclassing it using Moose don't even notice
I was kinda proud of making that work ;)
14:31 pmurias joined 14:32 kurahaupo joined, kurahaupo left 14:35 kurahaupo joined 14:36 kurahaupo left
spider-mario okay, bisect in progress 14:37
nine spider-mario: no need to
spider-mario ah 14:38
mst Juerd: anyway, sorry for my slightly exothermic response, but I've been hearing 'moose is slow' for a lot of years and it's almost always (1) startup time (2) forgot make_immutable (3) did something that would've been slow if you'd done it by hand but hadn't thought about it
Juerd: and there was lots of actual FUD for quite a while, so it makes me twitch
nine spider-mario: trying to reproduce it here as I actually thought I had fixed this 14:39
14:40 khw joined
nine spider-mario: ok, I see it here 14:41
spider-mario cool
14:49 kurahaupo joined 14:55 kurahaupo left
nine This is most frustrating :( I've already had a fix for the Staging issue. But when I cleaned up my changes and put them into distinct commits it looked like I don't actually need this particular change, so I removed it. 14:55
I've got no idea why it did not explode in my tests as I did exactly the same as the failing test now.
14:57 kurahaupo joined 14:58 pmurias_ joined
nine spider-mario: please try again with this patch: github.com/rakudo/rakudo/commit/d1...232ee64341 15:00
15:01 pmurias left
spider-mario the patch applies cleanly but it still doesn’t seem to work 15:03
wait 15:04
not sure I applied it correctly
let me try again
(sorry)
holyghost is going to prepare himself a noodles-provencale (own mix) 15:05
spider-mario cool, it worked 15:07
thanks, nine :)
15:17 kurahaupo left 15:21 Xliff left 15:30 vimal2012 joined 15:32 vimal2012 left
ugexe m: sub foo($message, Str :@bar) { say $message }; foo("xxx") # what is the rational for performing a type check on an optional parameter that was not passed? 15:53
camelia rakudo-moar d1604b: OUTPUT«Type check failed in binding @bar; expected Positional[Str] but got Array ($[])␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
gfldex m: sub foo($message, Str @bar?) { say $message }; foo("xxx") 15:56
camelia rakudo-moar d1604b: OUTPUT«Type check failed in binding @bar; expected Positional[Str] but got Array ($[])␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
gfldex ugexe: that's a bug 15:57
15:57 Tonik joined
ugexe m: my Array[Empty] @a; 15:57
camelia rakudo-moar d1604b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤An exception occurred while parameterizing Array␤at <tmp>:1␤Exception details:␤ 5===SORRY!5=== Error while compiling ␤ Parameter 't' requires a type object of type Slip, but an object instance …»
ugexe or just `say Array[Empty]` 15:58
gfldex: do you know if its a known bug? 15:59
15:59 marcel_ joined 16:02 Xliff joined
moritz Empty is an instance, not a type 16:02
so the error message is correct, IMHO 16:03
m: say Empty.^name
camelia rakudo-moar d1604b: OUTPUT«Slip␤»
gfldex ugexe: i would have to query RT 16:06
16:11 j joined, j is now known as Guest24412 16:12 holyghost left 16:13 TreyHarr1 left 16:14 bob777 left 16:20 holyghost joined 16:22 holyghost left, adu joined 16:23 holyghost joined 16:24 jack_rabbit left 16:25 Ven left 16:26 canopus left 16:27 canopus joined 16:30 Ven joined 16:31 RabidGravy left 16:35 TreyHarr1 joined 16:38 rindolf left, dvinciguerra_ joined 16:40 Ven left 16:44 rindolf joined
gfldex löliblöggedagain gfldex.wordpress.com/2016/07/17/di...g-subsets/ 16:46
16:46 RabidGravy joined 16:50 pmurias_ left, Ven joined 16:56 Ven left
stmuk pl6anet.org/drop/rakudo-star-2016.07-RC0.tar.gz 17:02
17:05 omegaz left 17:10 Ven joined 17:11 Ven left 17:13 captian-adequate joined 17:15 iH2O joined 17:17 pmqs_ left 17:25 canopus left
ugexe hmm why doesnt bisectable work with private messages? 17:27
17:28 RabidGravy left
ugexe bisect: gist.github.com/ugexe/7ce08d3ad4cf...7ba9713471 17:28