»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
TimToady irclog.perlgeek.de/perl6/2012-06-03#i_5678478 00:00
but comma could also usefully be outlawed there
TimToady it seems to be a very common failing to assume all() semantics in such casees 00:01
sorear I have no idea how to write that check. 00:14
dalek ecza/non-bootstrap: 154e541 | sorear++ | lib/RxOp.cs:
Port regex optimizer
 02:13
pmichaud 22:47 <felher> pmichaud: this may be and odd/noob question: In your QRPA-Code, when you resize your array, why do you only align to 4096 element boundaries for those little arrays? 02:28
"align to 4096 element boundary" is probably an incorrect phrasing. What really happens is that we size the array to the next largest multiple of 4096 elements 02:29
and it only happens for arrays larger than 8192
sorear shouldn't you be allowing 16 bytes or so for malloc headers? 02:32
dalek p/qrpa: fad75ad | pmichaud++ | src/pmc/qrpa.pmc:
QRPA: Update comment in QRPA.set_integer_native to be a little more accurate.
 02:35
pmichaud sorear: so far I'm just copying what Parrot does there; it probably makes sense to tune it a bit, yes.
pmichaud also, it's using Parrot's gc-based allocator, so there might be even more than just the malloc header (I've not researched this). 02:36
sorear ah.
dalek kudo/map: 6941eb1 | pmichaud++ | src/core/List.pm:
Temporarily revert the use of qlist in this branch until I make

back to their horribly slow state, but future commits will definitely get us the performance back.		 03:53
kudo/map: 916c59a | pmichaud++ | src/core/ (2 files):
Remove (basically unused) :$sink option to .reify(); sinks will be
dalek ecza/non-bootstrap: 0062b2c | sorear++ | lib/Utils.cs:
Port GetOptLong
 05:12
tadzik good morning #perl6 05:28
crab2313 tadzik: good morning 05:36
moritz \o 05:43
sorear good morning tadzik, moritz, crab2313, GlitchMr 05:49
crab2313 sorear: how are you doing today 05:51
sorear fine I guess. 05:54
it's always weird to be asked that
GlitchMr Just wondering, what it is? github.com/perl6/std/blob/master/STD_P5.pm6 05:57
Is it grammar for Perl 5?
sorear it is a WIP grammar for Perl 5
GlitchMr $¢.worry("Leading 0 does not indicate octal in Perl 6")
...
ok
sorear TimToady was working on it a few years ago, but has not been touched since
GlitchMr paste.uk.to/bbf2760d 05:59
I find it interesting that you can make stuff like this in few lines 06:00
GlitchMr Also, why "multi" isn't default in subs? 06:03
benabik Because there's a cost to using multi. 06:04
GlitchMr Performance? 06:05
benabik Yes, and memory. I suppose in the one candidate case the costs are minimal, but present. (more)
sorear also, sanity
benabik Also helps the compiler know if you did or didn't intend to have two subs with the same name. 06:06
sorear multis have failure modes that can be very confusing if you are expecting onlys
GlitchMr Failure modes could check if there is just one multi function 06:07
paste.uk.to/fe60a017
(and I should fix highlighting in my pastebin...) 06:08
benabik Failure modes like "I mis-spelled a named parameter and am now calling a completely different function" 06:09
Which is very unexpected if you didn't know it was a multi.
GlitchMr What about operators? It's almost required to use "multi" there (I guess that answer is no, for consistency). 06:11
benabik Most built-in operators are multis, IIRC. 06:12
GlitchMr Well, but some user may forget about it and second module might be not able to insert operator. 06:13
benabik No matter what default you choose, I bet you'll have a problem. Perhaps the first module _wanted_ a non-overrideable operator because it does something very non-standard. 06:14
sorear There is no more reason for subs named 'infix:<foo>' to be multi than there is for subs named 'foo'
benabik And there's- what sorear++ said
sorear There really isn't a conceptual difference between subs and operators
They're both subs
moritz ecept for those operators that thunk 06:16
sorear moritz: Those are just macros with funny names :)
GlitchMr And you can use prefixes to make funny function names... 06:17
sub prefix:<-e> { 'it exists' }
(of course, in real life, code would be more complex)
Well, I guess it's confusing, but makes sense 06:18
benabik We generally spell "this should be more complex" `...`. :-) 06:18
GlitchMr Right...
benabik apparently gets talkative on IRC when he has insomnia.
grondilu p6: say qx/perl -wE "print pack q(C*), 128..130"/;
GlitchMr It even protects against code like -exists()
p6eval pugs: OUTPUT«*** Unsafe function 'Pugs::Internals::runShellCommand' called under safe mode␤ at /tmp/6cE9MVisKQ line 1, column 5-46␤»
..niecza v18-2-gea3d97a: OUTPUT«���␤»
..rakudo 9d2ddd: OUTPUT«qx, qqx is disallowed in safe mode␤ in sub restricted at src/SAFE.setting:2␤ in sub QX at src/SAFE.setting:9␤ in block <anon> at /tmp/CSsuLFooAO:1␤␤»
GlitchMr n: print qx/ls/ 06:19
p6eval niecza v18-2-gea3d97a: OUTPUT«boot␤docs␤examples␤FETCH_URL␤lib␤LICENSE␤License-MS-PL.txt␤main.pl␤Makefile␤obj␤p5test␤perf␤perl5␤README.pod␤run␤simple-tests␤src␤t␤test2.pl␤test3.pl␤test.pl␤TODO␤tools␤VERSION␤»
sorear gets talkative *period* when ey can't sleep
GlitchMr ...
Sandbox is broken
sorear GlitchMr: there is no sandbox.
grondilu qx// expects unicode. What about if I want binary?
sorear (this is not a bug. please don't file one for the "lack of a sandbox") 06:20
moritz I don't think that's supported at the moment
GlitchMr But well... I can run any command 06:21
sorear Yes.
moritz and if you do, we'll ban you
sorear We have backups and a generally friendly userbase.
GlitchMr :) 06:22
grondilu p6: multi sub talk() { say "hello" }; multi sub talk(:$shouting) { say "HELLO!!" }; talk();
p6eval rakudo 9d2ddd: OUTPUT«HELLO!!␤»
..pugs: OUTPUT«hello␤»
..niecza v18-2-gea3d97a: OUTPUT«Potential difficulties:␤ $shouting is declared but not used at /tmp/wx2648IFNO line 1:␤------> talk() { say "hello" }; multi sub talk(:⏏$shouting) { say "HELLO!!" }; talk();␤␤hello␤»
sorear diakopter and I decided the benefits of being able to demo CLR interop outweighed the inconvenience of possibly having to clean up after a griefer
OTOH, this was before the move from diakopter's own server back to feather3
so we ought to give Juerd and moritz a chance to weigh in 06:23
moritz fwiw rakudo allows nqp:: ops, but disables run/qx/qqx etc.
because you need to have a pretty good understanding of this stuff to exploit the nqp:: stuff
diakopter but inline pir..? 06:24
moritz also allowed
grondilu should I commit a rakudo bug about the failed multi dispatch above?
moritz but anybody with a bit of p5 knowledge knows qx//
sorear Anyways it's running as an unpriviliged user in a chroot on a semi-dedicated virtual machine 06:25
grondilu r: multi talk() { say "hello" }; multi talk(:$shouting) { say "HELLO!!" }; talk();
p6eval rakudo 9d2ddd: OUTPUT«HELLO!!␤»
GlitchMr thedailywtf.com/Articles/Security-by-PostIt.aspx
Good password :)
sorear With process time limits
benabik ulimit is your friend 06:26
moritz BSD::Resource
sorear Really the worst anyone could do is start sending spam from feather3
I wonder if it would be possible to throttle feather3's uplink at the Xen level
There's no reason p6eval and dalek need to be able to transmit data at 100Mbps or whatever Juerd's datacenter allows 06:27
benabik moritz: I'm fine with whatever sugar you want over ulimit(3). :-D
sorear setrlimit(2) innit? 06:28
my ulimit(3) says Warning: This routine is obsolete.
benabik Huh. Apparently, yes. Mine says "limited to maximum file size".
benabik Been a while since I've used that interface. 06:29
dalek kudo/nom: d351efc | moritz++ | src/core/ (2 files):
typed exception for Set coercion
 06:34
ast: 85f737a | moritz++ | S32-exceptions/misc.t:
test X::Set::Coerce
ecs: f29d4ce | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] general policy for what data to carry around
 06:35
ecs: d88cbf4 | moritz++ | S32-setting-library/Exception.pod:
X::IO does X::OS
ecs: 64947f6 | moritz++ | S32-setting-library/Exception.pod:
note about mixing in X::Comp
grondilu p6: say "foo" ~~ m:g/./; 06:48
p6eval rakudo 9d2ddd: OUTPUT«False␤»
..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20120203/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** Can't locate P…
..niecza v18-2-gea3d97a: OUTPUT«===SORRY!===␤␤Regex modifier g not yet implemented at /tmp/cLavGB_OUR line 1:␤------> say "foo" ~~ m:g⏏/./;␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1401 (die @ 5) ␤ at /ho…
grondilu What is this X::Comp thing about?? 06:51
grondilu has never seen any X class/role in any synopsis. 06:52
benabik X is the exception package.
grondilu ok 06:53
benabik X::Comp is a type for exceptions thrown at compile type.
*time
Juerd sorear: If things on feather3 are run as non-root, and there's a simple iptables firewall, I think that should be enough.
sorear: Local root exploits can happen but are much more likely on feather1 anyway. 06:54
moritz at some point I'd like to redo feather3 with debian stable 06:58
running unstable when it is only seldomly upgraded is not a good idea
dalek kudo/nom: cad7b6c | moritz++ | src/core/ (2 files):
another typed exception for DateTime
 07:01
ast: 7a77b2a | moritz++ | S32-exceptions/misc.t:
test X::DateTime::TimezoneClash
sorear Juerd: thank you 07:06
moritz oh, my last three patches to S32::Exception have been sent as a single email to p6l 07:35
is that a new feature of github?
dalek ecza/non-bootstrap: 0747704 | sorear++ | / (4 files):
Port over top-level driver
 07:36
sorear non-bootstrap now parses command line options
can't actually *do* anything with it yet, though 07:37
sorear sleep& 07:46
moritz .u ェ 07:47
phenny U+30A7 KATAKANA LETTER SMALL E (ェ)
kresike morning all you happy perl6 people 08:01
dalek ecs: 087aac0 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] small nits
 08:02
moritz good morning happy kresike :-)
kresike hello moritz o/ :) 08:03
moritz std: use ::; 08:38
p6eval std f179a1b: OUTPUT«Modification of non-creatable array value attempted, subscript -1 at CursorBase.pm line 307.␤FAILED 00:00 40m␤»
moritz wow, I don't think I've ever seen that error message from p5 before 08:42
dalek ecs: 4d6a980 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] first rough import of exception classes from Rakudo

needs more polishing, and some classes also need descriptions		 08:51
Ulti moritz: can you create a file just called .pm and have that work? 09:04
so like use Fun:: where Fun/.pm exists
moritz Ulti: I hope not 09:05
r: use ::;
p6eval rakudo cad7b6: OUTPUT«===SORRY!===␤Could not find :: in any of: /home/p6eval/.perl6/lib, /home/p6eval/nom-inst2/lib/parrot/4.4.0-devel/languages/perl6/lib␤»
moritz eeks
Ulti moritz: you totally can 09:06
I just did it
heh
neat but also wtf
moritz we'll disallow that in time
Ulti this is p5 not rakudo though
moritz waits for ACME::Empty 09:07
Ulti lol
moritz I should release that later today :-)
Ulti it complained about a lack of 1; at first :3 09:08
moritz let's see how well the CPAN tool chain can handle empty module names :-)
Ulti because thats the problem with that module
might be useful as some kind of meta package 09:09
so Fun::; describes everything inside Fun
moritz in p6, Fun:: is the symbol table of Fun
Ulti yeah so thats like the most sane thing to expect
moritz eval: package Fun::::; __PACKAGE__ 09:10
buubot_backup moritz: Fun::::
dalek ecs: 4ef9759 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] elaborate on a few more exception classes
 09:41
dalek ecs: 6789051 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] elaborate on a few more exception classes
 10:06
ecs: ba2386f | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] finish describing all exception types
 10:12
moritz now that was quite a bit of work.
dalek ecs: a36af6c | gerdr++ | S32-setting-library/Exception.pod:
[S32-setting-library/Exception.pod] Fix some typos in
 10:39
ecs: 11998de | (Gerhard R)++ | S32-setting-library/Exception.pod:
[S32::Exception] fix some typos
 10:44
not_gerd rewrites public history 10:45
not_gerd--
sergot hi o/ 11:07
moritz not_gerd++ # fixing my stuff 11:27
pmichaud good morning, #perl6 12:35
[Coke] GOOD MORNING, LOWERCASE PERL6 USER! 12:37
Ulti lol 12:41
pmichaud averts his eyes from the too-early-in-the-morning uppercase shouting on the channel
[Coke] has a friend who types in all caps all the time because trying to read her own lowercase typing makes her migraines activate sooner. 12:42
frettled [Coke]'s friend should experiment more with different fonts and sizes. :) 12:43
Ulti 00[08C04o09k00e09] 11Y09O11U 08S04H11O00U12L11D 08T00A12L04K 09I04N 00R09A12I13N12B00O04W12S 09T00O08O
though thats not going to help anyones migraines
crab2313 Ulti: Wow. 12:51
JimmyZ r: $_ = 2, 3, 5; say ~$_ 12:53
p6eval rakudo cad7b6: OUTPUT«2␤»
JimmyZ is it right?
crab2313 r: $_ = (2,3,5); say $_; 12:56
p6eval rakudo cad7b6: OUTPUT«2 3 5␤»
Ulti crab2313: dont worry it was scripted rather than hand tpyed ;) 12:57
crab2313 r: $_ = (2,3,5); say ~$_; 12:59
p6eval rakudo cad7b6: OUTPUT«2 3 5␤»
crab2313 Ulti: creative 13:00
crab2313 n: $_ = (2,3,5); say ~$_; 13:06
p6eval niecza v18-2-gea3d97a: OUTPUT«2 3 5␤»
crab2313 n: $_ = 2,3,5; say ~$_;
p6eval niecza v18-2-gea3d97a: OUTPUT«2␤»
pmichaud afk, kid to camp 13:12
felher pmichaud: @QRPA: I see, thnx :) 13:31
[Coke] r: say 604-105 14:02
p6eval rakudo cad7b6: OUTPUT«499␤»
[Coke] < 500 open tickets!
adu wow 14:03
can I help?
[Coke] rakudo: say 1.++() # RT #73156 14:04
p6eval rakudo cad7b6: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&postfix:<.>' called (line 1)␤»
[Coke] adu: sure. if you want, go to:
rakudo.org/rt/testsneeded - any tickets here need tests added to roast before they can be closed. 14:05
diakopter added to (or found already in) roast
[Coke] alternatively, go to rt.perl.org/rt3 , perl6 queue, and see if any of the errors reported in those tickets (that are NOT marked testneeded) now work.
diakopter++
diakopter though I very recently went through them all, doing that 14:06
but didn't comment on the ones that were stlil broken unless the behavior changed 14:07
slinkylinda hello 14:08
diakopter yay, my jury duty today was placed on phone standby 14:11
adu diakopter: it probably means you're number 13 14:18
diakopter well, I hadn't even gone through the prospective juror process 14:19
you have to do that first, I think 14:20
adu well, if numbers 1..12 all have heartbeats, then you might never get a call 14:21
kresike and if one of them dies from unnatural causes, then run :)
kresike bye all 14:38
sorear good * #perl6 14:55
diakopter howdy sorear
crab2313 diakopter: If a test on rakudo.org/rt/testsneeded failed, should I report it? 15:00
sorear moritz: should I read your individual patches to S32/Exception or wait for the final result? 15:02
diakopter crab2313: which one 15:07
crab2313: sure, just mention it here
crab2313 diakopter: #104514
diakopter r: my @a = -> { say 1 }, -> { say 2 }; @a>>.() 15:08
p6eval rakudo cad7b6: OUTPUT«1␤2␤»
diakopter seems to work okay
crab2313 my $a; $a>>.() 15:10
moritz sorear: wait a bit, though feedback wouldn't hurt
crab2313 my $a; $a>>.(); will cause a segfault 15:11
diakopter r: my $a; $a>>.(); 15:12
p6eval rakudo cad7b6: OUTPUT«No such method 'Any' for invocant of type 'Parcel'␤ in <anon> at src/gen/BOOTSTRAP.pm:804␤ in <anon> at src/gen/BOOTSTRAP.pm:801␤ in any <anon> at src/gen/BOOTSTRAP.pm:796␤ in sub METAOP_HYPER_CALL at src/gen/CORE.setting:11588␤ in block <anon> at /tmp/orBoNPu…
crab2313 diakopter: gist.github.com/2868992 15:14
diakopter crab2313: it might be different on the REPL 15:21
which build of rakudo do you have
crab2313 diakopter: This is perl6 version 2012.05-167-gcad7b6c built on parrot 4.4.0 revision RELEASE_4_4_0-90-g22d7d59
diakopter well hm 15:22
diakopter it doesn't segfault for me on windows in the repl; lemme try linux 15:23
nor on linux 15:24
oh!
but if I hit Ctrl-d at the REPL prompt, it segfaults :)
after typing that 15:25
crab2313 diakopter: yes, it is.
diakopter it exits with error on windows if I Ctrl-Z->Enter, too
diakopter there, updated 15:28
crab2313++
crab2313 diakopter++ 15:29
diakopter looks for an MIT-ish licensed codepoint->byte utf8 encoder in C 15:30
JimmyZ nfg? 15:31
diakopter just normal codepoints to utf8 15:31
colomon diakopter: they're actually very easy to write. 15:32
diakopter: hold on a moment, I did this myself a few weeks ago.
diakopter: my version was based on this: unicode.org/resources/utf8.html (BSD license)
diakopter: though I rewrote it to switch from wchar_t to unsigned int for the code points 15:33
diakopter: (it got simpler in the process)
arnsholt It's mostly a question of figuring out how many bytes you need and doing the bit-shuffling, isn't it? 15:38
moritz what was the difference BSD and MIT licensing again? 15:39
one east coast, one west coast?
tadzik BSD is longer
moritz but both allow basically everything, as long as you include the license, no? 15:40
daxim www.gnu.org/licenses/license-list.html
tadzik I think so
diakopter BSD says you can't use the author/holder's name
TimToady early BSD had an advertising clause, and language specific to the Regents of California
but then they generalized it
daxim "[…] MIT license, but that term is misleading, since MIT has used many licenses for software." 15:41
better say expat license, x11 license etc. 15:42
not_gerd also has an UTF-8 encoder lying around 15:45
diakopter: do you need codepoint verification (exclude non-chars/surrogates), and/or modified UTF-8 (ie encode NUL as an overlong 2-byte sequence)? 15:46
I'm happy to donate my code under whatever license you like
moritz jnthn: I want to implement two module loading related things on which I'd like to get your input 15:50
jnthn: first, some kind of debugging mode for module loading which tells us which modules are actually loaded (ie which path, if .pir or .pbc etc.)
how should I trigger that? env variable? command line switch? 15:51
diakopter not_gerd: see privmsg
moritz second, I want to get some module loading traces 15:52
not_gerd digs out the code
moritz ie if I get a a compilation error from C.pm, I want to have a backtrace like "used from B.pm line 2\nused from A.pm line 5; use from script.pl line 22" 15:53
s/if/when/
the question is, where should I stick that information? 15:54
sergot perl6.org/community/ : November Wiki and Perl.net.au wiki doesn't work. Should I remove them from there? 15:56
tadzik I think so 15:57
moritz sergot: yes please
sergot perl6.org/documentation/ : Perl 6 Now: The Core Ideas Illustrated with Perl 5 (Apress book) (2004) - doesn't work too, what to do with this?
:)
moritz sergot: remove
sergot ok
moritz sergot: perl6.org was always meant to contain up-to-date information, rather than comprehensive information 15:58
diakopter not_gerd: fyi, I ended up using the decoder here: bjoern.hoehrmann.de/utf-8/decoder/dfa/ 16:01
moritz (I'd also appreciate input from others on the two design questions above) 16:02
diakopter not_gerd: (but I need a time-tested encoder) :) 16:02
dalek href="https://perl6.org:">perl6.org: 92cfd27 | sergot++ | source/ (2 files):
Broken links removed.
 16:03
jnthn evenin' 16:05
not_gerd diakopter: needed some time to rip out and sanitize the code (as well as locate the PDCLIB license file) 16:06
for what it'S worth, here it is: gist.github.com/2869236
diakopter I need some kind of author/copyright attribution so it's known who issued the license - privmsg me your name or website or public email or irc handle if you want to use that :) 16:08
not_gerd: ^
not_gerd: thanks! :) 16:13
spider-mario .u ⏏ 16:22
phenny U+23CF EJECT SYMBOL (⏏)
TimToady that's when the umpire ejects you from the ballgame 16:25
sergot does "andthen" still exist? 16:28
GlitchMr n: use Threads; Thread.new: * + * 16:29
p6eval niecza v18-2-gea3d97a: OUTPUT«␤Unhandled Exception: System.NullReferenceException: Object reference not set to an instance of an object␤ at Niecza.Frame.MakeChild (Niecza.Frame outer, Niecza.SubInfo info, Niecza.P6any sub) [0x00000] in <filename unknown>:0 ␤ at Niecza.Kernel.GetInferi…
[Coke] sergot: NO ANDTHEN! 16:31
flussence Didn't we talk about something almost identical to this? perlmonks.org/?node_id=973015 16:37
TimToady phenny tell JimmyZ irclog.perlgeek.de/perl6/2012-06-04#i_5680738 is right behavior, except that the compiler should really warn about "Useless use of 3 in a sink context"
phenny: tell JimmyZ irclog.perlgeek.de/perl6/2012-06-04#i_5680738 is right behavior, except that the compiler should really warn about "Useless use of 3 in a sink context" 16:38
phenny TimToady: I'll pass that on when JimmyZ is around.
sergot [Coke]: What do you mean? :) 16:41
[Coke] sergot: from _Dude, Where's my Car_ 16:49
[Coke] # 06/04/2012 - rakudo++ (22405); niecza (91.3%); pugs (34.58%) 17:46
pmichaud back again
jnthn o/ pmichaud
tadzik hello hello 17:47
sergot o/
dalek ecs: 0fc2c74 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] big reordering

group syntax errors and compilation errors together		 17:49
[Coke] pmichaud: ~~ 17:55
dalek ecs: aa0b910 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] bump version and modification date
 17:58
dalek ecs: b469029 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] fix a heading level
 18:03
ecs: cbf1853 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] fix role name
 18:06
dalek ecs: 47b29a8 | moritz++ | S32-setting-library/Exception.pod:
group dispatch errors
 18:12
dalek ecs: 7ceaeb9 | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] notes on the general naming scheme
 18:16
[Coke] moritz: I would like to register a small complaint about core exceptions being in the "X" namespace - that reads "user extension" to me. 18:19
pmichaud I think "user extension" should be the "X-" prefix :-) 18:20
(istr making a similar complaint when "X" was proposed, though. :-)
TimToady I think everything is a user extension that isn't builtin 18:21
moritz I've heard complaints about X::, but no better suggestions so far
would you prever EX:: or Ex::?
pmichaud okay, call "X-" the "draft extension prefix" then. :) 18:22
TimToady the X was my decision, and I haven't changed my mind :)
pmichaud yes, if you were still uncertain about it, it'd be "X-X" :-P
moritz fwiw I like X too
tadzik I like X
moritz
.oO( X-rated objects )
pmichaud you can find your errors because they're all marked with X's
just like in school
TimToady I think people will see exceptions much more often than experimental extensions 18:23
and if something is experimental is should be reflected in the version, not the name
*it
PerlJam TimToady: what does an experiment version look like? 18:24
er, experimental
TimToady usually v0.0.1 or so :) 18:25
PerlJam :-)
PerlJam In Moose-land, the X is a suffix (MooX, MooseX, etc.), for exceptions, it's a prefix. I'm pretty sure people can make the distinction if necessary 18:26
PerlJam (If Perl 6 grows a set of experimental extensions that is) 18:27
moritz as long as the exceptions aren't experimental :-) 18:35
colomon XfooX
pmichaud one would want to be careful with the experimental OXO exception type, then :-) 18:37
TimToady I'm reminded of the tee-shirt my wife's sister owns, that shows a princess throwing a frog into a pile of frogs, with the legend "Before you find your prince you have to kiss a lot of frogs." 18:42
moritz :-) 18:43
.oO( people always want a fairtale wedding, but when I then show up and curse their first-born, they all freak out ) # paraphrased from someone on twitter
 18:44
colomon nice 18:45
geekosaur *snrk* 18:47
krunen pmichaud: Hi. sjn asked me to contact you about the p5-mop hackaton after YAPC. I'll try to get funding if you want to come. 18:53
pmichaud krunen: I'd love to come, yes.
I think that hackathon will be something really special, plus jnthn++ thinks that would be a good time for he and I to meet/plan again (so do I) 18:54
pmichaud krunen: let me know if I can provide any details to help with funding; I probably cannot make it by relying on my own resources. 18:57
krunen pmichau: Great. sjn talked about a possibility for a Hague grant. I can send an application, if you think there's a chance / time enough. I'd need a bit of help with the details. 18:59
pmichaud I think there's a chance and definitely time enough. perhaps contact me privately and I can give some details? 19:03
email is [email@hidden.address] or we can /privmsg here on irc
krunen i'll send you a quick email, I don't trust my scrollback 19:06
dalek ok: e10953e | alvis++ | authors.pod:
This fixes the Pod error: '=item' outside of any '=over'.
 19:07
ok: aac7aa1 | moritz++ | authors.pod:
Merge pull request #69 from ayardley/ayardley/fixup_authors.pod

This fixes the Pod error: '=item' outside of any '=over'.
timotimo allow me a stupid question: how similar/dissimilar is perl6's handling of destructors to the RAII model? 19:08
pmichaud krunen: works for me, I can suggest a process to go forward from there. Thanks! 19:09
sorear timotimo: very dissimilar. perl6 destructors are closer to Java finalizers
timotimo: there's no guarantee they will be run in a manner at all timely.
timotimo okay
benabik RAII is tied to C++'s destruction of stack objects when you leave a scope. Finalizers in a GC system are similar, but the timing is very very different. 19:10
pmichaud afk for a while 19:15
dalek ecs: 54584ca | moritz++ | S32-setting-library/Exception.pod:
[S32::Exception] remove redundant error; harmonize error messages to be easier copy&pastable
 19:20
zby_home_ in a role I need to check if something isa WebNano6::Controller 19:36
but I use that role in WebNano6::Controller
in result it hangs 19:37
benabik Predeclare either the role or the controller?
r: role Foo {...}; class Controller { method bar() { Foo.new.baz() } }; role Foo { method baz() { say 'hi' } }; Controller.bar 19:38
p6eval rakudo cad7b6: OUTPUT«hi␤»
zby_home_ thanks!
GlitchMr nr: sub infix:[] { $^a, $^b }; my @no-commas = 1 3 6 .. 8; print @no-commas.perl 19:48
p6eval niecza v18-2-gea3d97a: OUTPUT«Unhandled exception: cannot increment a value of type Parcel␤ at /home/p6eval/niecza/lib/CORE.setting line 1640 (List.gist @ 5) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2895 (RangeIter.reify @ 7) ␤ at <unkno…
..rakudo cad7b6: OUTPUT«No such method 'succ' for invocant of type 'Parcel'␤ in sub postfix:<++> at src/gen/CORE.setting:1318␤ in method reify at src/gen/CORE.setting:4825␤ in method reify at src/gen/CORE.setting:4946␤ in method gimme at src/gen/CORE.setting:5333␤ in method eager at …
GlitchMr nr: sub infix:[] { @($^a), @($^b) }; my @no-commas = 1 3 6 .. 8; print @no-commas.perl 19:50
p6eval niecza v18-2-gea3d97a: OUTPUT«Unhandled exception: cannot increment a value of type Parcel␤ at /home/p6eval/niecza/lib/CORE.setting line 1640 (List.gist @ 5) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/lib/CORE.setting line 2895 (RangeIter.reify @ 7) ␤ at <unkno…
..rakudo cad7b6: OUTPUT«No such method 'succ' for invocant of type 'Parcel'␤ in sub postfix:<++> at src/gen/CORE.setting:1318␤ in method reify at src/gen/CORE.setting:4825␤ in method reify at src/gen/CORE.setting:4946␤ in method gimme at src/gen/CORE.setting:5333␤ in method eager at …
GlitchMr :(
nr: sub infix:[] { @($^a), @($^b) }; my @no-commas = 1 3 (6 .. 8); print @no-commas.perl
colomon errr... you're trying to create a zero-length infix operator?
p6eval niecza v18-2-gea3d97a: OUTPUT«[1, 3, 6, 7, 8].list»
..rakudo cad7b6: OUTPUT«Array.new(1, 3, 6, 7, 8)»
GlitchMr colomon, yes
What's wrong with it 19:51
colomon how do you know when you are using it?
GlitchMr colomon, as long it works I don't care :P
How it works... I don't know 19:52
But it seems more natural this way
benabik Sadly, the compiler doesn't know how that works either. 19:52
GlitchMr Removing semicolons would be next step, but it would remove adding automatic semicolon insertion, just like in Python 19:53
benabik If every place a null regex could match is an operator, that's an awful lot of work.
GlitchMr s/remove/require/
flussence nr: sub infix:["\c[ZERO WIDTH NON-JOINER]"] { @($^a), @($^b) };
p6eval rakudo cad7b6: ( no output )
..niecza v18-2-gea3d97a: OUTPUT«Potential difficulties:␤ &infix:<‌> is declared but not used at /tmp/1suziu8f6X line 1:␤------> sub infix:["\c[ZERO WIDTH NON-JOINER]"] ⏏{ @($^a), @($^b) };␤␤»
GlitchMr But both rakudo and niecza implement this 19:54
It cannot be just illusion
benabik Implement what?
GlitchMr Ok, it's more like unintentional side-effect 19:55
colomon you mean, it works for you in incredibly limited test cases
colomon have you ever gotten a "two terms in a row" error message? 19:55
std: sub infix:[] { @($^a), @($^b) }; my @no-commas = 1 3 (6 .. 8); print @no-commas.perl
p6eval std f179a1b: OUTPUT«Use of uninitialized value $x in pattern match (m//) at STD.pm line 66577.␤Use of uninitialized value $x in concatenation (.) or string at STD.pm line 66616.␤===SORRY!===␤Null operator is not allowed at /tmp/8KOrpwYIyg line 1:␤------> sub infix…
benabik Ignoring the warnings, the key is towards the end: "Null operator is not allowed" 19:56
colomon Note there: "Null operator is not allowed "
GlitchMr But why it works?
benabik The bug here is that neither niezca or rakudo check for that.
colomon what you've done is found a bug
moritz bug
GlitchMr But I like this bug :P
TimToady no you don't, you just think you do
colomon TimToady++
benabik null or whitespace operators are far more likely to be invoked by accident than intention. 19:57
TimToady great gobs of Perl (both 5 and 6) depend on knowing whether a term or operator is coming
flussence unless you're using really obscure unicode whitespace...
TimToady is / division or the start of a regex?
GlitchMr I'm going to use older versions of Perl 6 if it will be fixed (not really)
TimToady is ^ the "upto" prefix or the xor operator? 19:58
benabik Even unicode whitespace should be eaten by <ws>... Null is the only meaningful version, but it's a little crazy.
TimToady is & the and operator or the start of a function name?
spider-mario .. is one of my favorites
TimToady is * a Whatever or a multipy?
*ply 19:59
spider-mario or a typo announcer?
benabik r: say (* * *)(2, 2)
p6eval rakudo cad7b6: OUTPUT«4␤»
TimToady is < less than or the start of a <a b c>?
is the next +, -, or ~ a unary op or a binary op? 20:00
you really, really, really, really, really, really, really, really don't want this
moritz fully agrees 20:01
geekosaur likewise; the whole idea screams that it's the sort of hack that comes to bite you very, very hard in the ass when you least expect it
colomon TimToady, as usual, is righter than right
actually, it's likely the sort of hack that will bite you the first time you try using it in a program with more than two lines of code. 20:02
GlitchMr Somehow add automatic semicolon insertion to Perl and it works nicely.
(ok, it won't be Perl anymore)
geekosaur and now I'm reminded of the brace inserion rule in haskell98
moritz automatic semicolon insertion has been a mistake in any language that has doen it
it's a PITA in javascript and nasty in Go 20:03
GlitchMr What about Ruby and Python?
moritz python doesn't use semicolon at all
and I don't know enough ruby to comment on it 20:04
sorear o/ colomon
colomon \o sorear
GlitchMr >>> 2; 3
2
3
But Python uses semicolon
geekosaur python uses layout instead of semicolons; not quite the same thing, as there is still no syntactic ambiguity 20:05
moritz right, what geekosaur++ said
TimToady pythong *allows* semicolons in the middle, as GlitchMr just illustrated
*on
geekosaur yes, you can use semis if you want, but omitting one isn't ambiguous because a continued expression must be indented further
GlitchMr Perl 6 has mutable grammars 20:07
But are those mutable enough to add automatic semicolon insertion?
TimToady there are always tradeoffs
benabik You can probably add it. No certainty it would be easy though. 20:08
TimToady you can turn Perl 6 into any other language, but then you're stuck with that other language..
TimToady Perl 6 has removed many of the things that were mandatory in Perl 5; as a result, those remaining things that are mandatory are going to be fairly heavily justified, or at least justifiable. 20:10
sorear Why doesn't statement_control:if take a <blast>?
sorear ducks 20:10
TimToady it does in the C dialect of Perl 6 :P 20:11
but then you have C 20:12
flussence (I just noticed the package manager I'm using seems to have typed exceptions - I just wish they'd add useful error strings for them instead of blurting out huge stacktraces...) 20:14
dalek ecza/non-bootstrap: 4ba1fbc | sorear++ | lib/ (3 files):
Port OpHelpers
 20:19
ecza/non-bootstrap: 7cb2ef1 | sorear++ | lib/CompMgr.cs:
Port OptBeta
 20:24
kboga hi #perl6 20:40
[Coke] kboga: hio
kboga hi [Coke]
[Coke] ponders setting up a bot to try to get a []less nick. 20:43
kboga ponders building rakudo with a forked nqp 20:46
spider-mario is there anything like Haskell’s Data.List.findIndices in Perl 6? 20:54
it does ~ what grep does but returns the index of the elements that satisfy the condition
instead of the elements themselves
tadzik we were once golfing something like this 20:55
I don't think there's a built-in
spider-mario actually, it’s Perl5’s List::MoreUtils::indexes
ok, thanks
currently, I grep on .keys
spider-mario it’d just be a convenient notation, it is by no mean required 20:56
jnthn r: my @a = 1, -2, 5, -7; say @r.map: { .key if .value < 0 } 20:57
p6eval rakudo cad7b6: OUTPUT«===SORRY!===␤Variable @r is not declared␤at /tmp/8qsjwyR5nB:1␤»
jnthn r: my @a = 1, -2, 5, -7; say @a.map: { .key if .value < 0 }
p6eval rakudo cad7b6: OUTPUT«No such method 'value' for invocant of type 'Int'␤ in block <anon> at /tmp/7yAthpeBBH:1␤␤»
jnthn r: my @a = 1, -2, 5, -7; say @a..pairs.map: { .key if .value < 0 } 20:58
p6eval rakudo cad7b6: OUTPUT«===SORRY!===␤CHECK FAILED:␤Calling 'pairs' will never work with no arguments (line 1)␤ Expected any of:␤ :($x)␤»
jnthn grr
r: my @a = 1, -2, 5, -7; say @a.pairs.map: { .key if .value < 0 }
p6eval rakudo cad7b6: OUTPUT«1 3␤»
tadzik hands jnthn a cup-o-coffe
* cofee
grr
coffee
spider-mario thanks
jnthn tadzik: Yeah...I got 3 hours sleep last night or something :/ 20:58
spider-mario here is what I did:
eiro hello
tadzik oh hai eiro
spider-mario r: my @a = 1, -2, 5, -7; say @a.keys.grep: {@a[$^i] < 0}
p6eval rakudo cad7b6: OUTPUT«1 3␤»
jnthn spider-mario: That also works 20:59
spider-mario yours might be more readable 20:59
thanks for mentioning it 21:00
diakopter___ jnthn: if you're there, see if you can privmsg this nick; if I don't reply, it's b/'c I don't see it 21:02
spider-mario I fail to understand why ... is not always lazy
well, it doesn’t seem to be
r: say (1, 7 ... *).first(* > 50)
p6eval rakudo cad7b6: OUTPUT«55␤»
spider-mario r: say (1, 7 ... * > 100).first(* > 50)
p6eval rakudo cad7b6: OUTPUT«55␤»
spider-mario r: say (1, 7 ... 100).first(* > 50)
p6eval rakudo cad7b6: OUTPUT«(timeout)»
spider-mario and there, it tries to generate the whole list, it seems
(100 − 1 !%% 6) 21:03
(I mean (100 − 1) !%% 6) 21:04
colomon n: say (1, 7 ... 100).first(* > 50) 21:13
p6eval niecza v18-2-gea3d97a: OUTPUT«55␤»
colomon spider-mario: looks like a rakudobug to me
spider-mario I suspected so
colomon though I guess things can be fuzzy where lazy lists are concerned. 21:14
jnthn That one may well work with pmichaud++'s current efforts
[Coke] r: say (1, 7 ... 100).first(* >= 100) 21:59
p6eval rakudo cad7b6: OUTPUT«(timeout)» 22:00
[Coke] r: say (1, 7 ... * >= 100)
p6eval rakudo cad7b6: OUTPUT«1 7 13 19 25 31 37 43 49 55 61 67 73 79 85 91 97 103␤»
[Coke] jnthn: do you really think that should work? 22:08
sorear I think it ought to work
with a warning because the semantics of 1,7 ... 100 are somewhat counterintuitive 22:09
pmichaud back 22:11
jnthn [Coke]: Yeah, I think if things are lazy enough it can work. .first only needs one result. 22:14
pmichaud 21:02 <spider-mario> r: say (1, 7 ... 100).first(* > 50) 22:15
1, 7 ... 100 isn't lazy because it's not infinite.
I mean, because it's not signalled as infinite. It has a finite end value, so it assumes infinite
jnthn ...assumes finite? 22:16
pmichaud sorry, assumes finite # jnthn++
jnthn pmichaud: My guess was that this one was vaguely related to MapIter tring to munch a load of stuff though.
pmichaud and since .map is currently over-eager, we end up eating a lot of stuff.
jnthn Right.
I'm guessing .first is implemetned in terms of MapIter.
pmichaud in terms of map, yes. 22:17
jnthn Yes, that.
pmichaud it's still over-eager in my branch, but that will likely be fixed very very soon.
jnthn yay 22:18
jnthn should have some tuits for his Rakudo tasks on Wed. 22:18
pmichaud prepares to rewrite major sections of MapIter 22:19
jnthn hands pmichaud a Dr Pepper
jnthn Time to try and sleep... 22:43
jnthn guesses he shouldn't miss the momentary darkness :)
o/
kboga jnthn: good night 22:45
[Coke] consumes a dr. pepper freeze pop 22:53
sergot good night! o/ 23:20
skids How does one syntactically get at the Routine object of a method, to do wrap/assuming on it? 23:45
(of the dispatcher, of course) 23:46
sunfl0wer hi. I am trying to compile rakudo star on windows7 with strawberry perl and it is giving me errors like libdyncall_s.a undefined reference to dcCall_x86... 23:49
can anyone help?
did someone had this error before? 23:50
colomon I don't ever compile Rakudo on windows, but libdyncall_s.a sounds like a linux library rather than a Windows one. 23:51
what steps did you take to build Rakudo?
sunfl0wer I downloaded rakudo star from git. then installed strawberry perl then opened strawberry perl shell and entered perl Configure.pl --gen-parrot 23:52
it builds a lot of stuff and then stops at this library 23:53
I did not installed icu library
and my english is bad sorry 23:54
geekosaur sounds like strawberry perl is based on cygwin or msys/mingw 23:55
sunfl0wer it comes with its own gcc version 4.4.7 20111023
colomon do you still need --gen-nqp in the config line? 23:56
(not directed at sunfl0wer) 23:57
sunfl0wer I tried with that and without that still the same error
ah sorry
colomon well, if you tried both ways then that's not the problem.
skids irclog.perlgeek.de/perl6/2012-02-12#i_5138759 23:58
colomon skids++ 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!