»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:03 Hor|zon left 00:23 ifim` joined 00:25 ifim left 00:37 BenGoldberg joined 00:47 _slade_ joined 00:48 raiph left 00:58 ifim` left 01:10 pdcawley left, pdcawley joined 01:11 Sqirrel left 01:21 Mso150 joined 01:44 Mso150 left 01:45 Mso150 joined 01:48 pdcawley left 01:49 pdcawley joined 01:54 Mso150 left 01:58 kurahaupo joined, _slade_ left 01:59 _slade_ joined, Hor|zon joined 02:04 Hor|zon left 02:20 cognome left, chenryn joined 02:22 lucs joined
gtodd azawawi: I don't know about an API but masak did something like this for his blog posts: 02:30
m: sub REPL($line) {say "> $line"; my $result = EVAL $line; say $result if defined $result; } ; REPL "('a'.. 'd') <<~<< ^6;";
camelia rakudo-moar 8b3e8c: OUTPUT«> ('a'.. 'd') <<~<< ^6;␤a0 b1 c2 d3 a4 b5␤»
gtodd ;-)
see gist.github.com/masak/f865b7d9dd33e535b501 02:44
02:45 colomon_ joined, baest_ joined
gtodd azawawi: farabi6++ ;-) 02:46
02:47 clkao_ joined 02:48 bcode left, bcode_ joined, broquain1 joined, retupmoc1 joined 02:49 lucs_ joined, isacloud______ joined, Ulti_ joined, apejens_ joined, noganex joined, sjn_ joined, dagurval_ joined, xfix joined 02:50 bonsaikitten joined, jercos_ joined, bcode_ is now known as bcode, nine_ joined, simcop2387_ joined 02:51 pdcawley_ joined, japhb_ joined, Juerd_ joined, hoelzro_ joined, sftp_ joined 02:52 sorear_ joined, exodist_ joined 02:53 ntln- joined 02:54 pdcawley left, camelia left, Exodist left, lucs left, sftp left, apejens left, geekosaur left, sjn left, hobbs left, integral left, hoelzro left, jercos left, exodist_ is now known as Exodist, sftp_ is now known as sftp, colomon_ is now known as colomon, simcop2387_ is now known as simcop2387, Juerd_ is now known as Juerd, mtj joined 02:55 isacloud______ is now known as isacloud, clkao_ is now known as clkao 02:56 apejens_ is now known as apejens 02:57 geekosaur joined 02:58 avuserow joined 03:00 mephinet joined, lue joined 03:02 eternaleye left 03:05 vukcrni joined 03:06 eternaleye joined, [Tux] joined 03:09 camelia joined 03:10 ChanServ sets mode: +v camelia 03:11 ezra1 joined 03:14 ezra1 left 03:15 [[[j4jackj]]] joined 03:17 cibs left 03:20 mauke_ joined, mauke_ left, mauke_ joined 03:21 [[[j4jackj]]] is now known as j4jackj, dylanwh joined, oetiker joined 03:23 hobbs joined 03:24 cibs joined, chenryn left, mauke left 03:27 sorear_ left 03:28 mauke_ is now known as mauke, PZt joined 03:30 molaf left 03:33 integral joined 03:34 sorear joined 03:35 bjz joined 03:42 kaare_ joined 03:50 xenoterracide__ left 03:51 cognome joined 03:54 chenryn joined 04:01 Hor|zon joined, chenryn left 04:05 pmnexus joined, labster joined, _slade_ left, _slade joined, _slade is now known as _slade_
pmnexus Good morning #perl6 04:06
04:06 Hor|zon left 04:10 kaleem joined, jack_rabbit joined 04:15 pyrimidine joined
pyrimidine FROGGS: about? 04:17
.tell FROGGS I filed RT #122971 re: the :nl parameter in IO.open. 04:19
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122971
yoleaux pyrimidine: I'll pass your message to FROGGS.
04:20 pyrimidine left 04:21 kaare_ left 04:26 anaeem1_ joined 04:31 xenoterracide__ joined
pmnexus Afk, szg->fra 04:32
04:39 pmnexus left 04:42 BenGoldberg left 04:45 ggoebel1111119 joined
timotimo o/ from an early morning 04:57
04:57 chenryn joined 04:58 anaeem___ joined, anaeem1_ left 05:02 chenryn left, eternaleye left 05:03 eternaleye joined 05:08 kaleem left 05:15 bjz left 05:18 flaviusb joined, xenoterracide__ left, chenryn joined 05:19 kaare_ joined 05:40 chenryn left, lue left, simcop2387 left 05:46 chenryn joined, lue joined, simcop2387 joined, dj_goku joined, Util joined, ugexe joined, btyler_ joined, salv0 joined, cotto joined 05:48 anaeem___ left 05:52 anaeem1 joined 05:54 chenryn left 05:58 anaeem1 left 06:01 chenryn joined, Hor|zon joined, anaeem1 joined 06:03 anaeem1__ joined 06:04 SamuraiJack joined 06:05 gfldex joined, anaeem1 left 06:06 pmnexus joined, Hor|zon left, pepl joined
pmnexus Good morning #perl6 06:10
06:11 anaeem1__ left 06:12 baest_ is now known as baest, bartolin joined, anaeem1_ joined 06:13 mephinet left 06:15 yeahnoob joined 06:18 xfix is now known as notfix 06:19 anaeem1_ left 06:20 gfldex left, anaeem1 joined 06:24 kaleem joined 06:26 pepl left 06:32 darutoko joined
FROGGS .botsnack 06:34
yoleaux 04:19Z <pyrimidine> FROGGS: I filed RT #122971 re: the :nl parameter in IO.open.
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122971
yoleaux :D
06:35 broquain1 is now known as broquaint
dalek ast: ae245a6 | usev6++ | S05-match/capturing-contexts.t:
Add test for RT #83508
 06:37
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=83508
06:37 mephinet joined
FROGGS morning o/ 06:38
bartolin good morning FROGGS \o 06:39
FROGGS morning bartolin 06:40
06:42 rindolf joined 06:58 abraxxa joined
tadzik morning 07:00
FROGGS tadzik: morning, I have missed you :o)
tadzik :)
I wasn't away that long :)
or maybe I was
maybe I slept through the last 3 months
FROGGS well, you were not in Salzburg 07:01
tadzik ah, that's true :(
07:04 pepl joined
dalek ast: 7396503 | usev6++ | S32-exceptions/misc.t:
Add test for RT #88748
 07:06
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...l?id=88748
07:06 Mso150 joined 07:09 rurban joined 07:12 azawawi joined
azawawi good morning #perl6 07:12
yoleaux 02:14Z <gtodd> azawawi: I guess it is not a REPL "API" but masak did this --> gist.github.com/masak/f865b7d9dd33e535b501 :-D
azawawi gtodd: true but does it hold state? or remember variables? 07:13
07:14 cognome left, zakharyas joined 07:15 cognome joined
timotimo o/ 07:16
pmnexus o/ 07:17
FROGGS o/
07:18 cognome_ joined
pmnexus In case people missed it in scrollback : gist.github.com/pmichaud/460bc17afcc5ca93b6b1 07:18
07:19 cognome left, lizmat joined
FROGGS ahh, it's you! :D 07:19
FROGGS is reading now
masak pmichaud++ # draft report 07:20
pmnexus Yes, I'm logged in from my phone :-)
Awaiting boarding of flight fra->dfw
masak (morgens, #perl6) 07:21
07:21 leont joined
timotimo we seem to not have any tests for proto tokens in grammars with multiple implementations 07:22
er, i mean, with multi tokens that have signatures 07:23
07:24 woolfy joined
pmnexus I'm hoping to have power outlet on plane for extended writing. dfw->fra had power outlet, so there's hope 07:24
timotimo pmnexus: i see what you did there: "people are eager for it" 07:26
07:26 FROGGS_ joined, FROGGS left
pmnexus timotimo: :-D 07:27
timotimo m: youhave: my $word and my @xe &(my &bow); 07:28
camelia ( no output )
timotimo actually, |bow seems more appropriate
m: youhave: my $word and my @xe &(my |bow);
camelia rakudo-moar 8b3e8c: OUTPUT«===SORRY!=== Error while compiling /tmp/zLkmjNDkyD␤Malformed my␤at /tmp/zLkmjNDkyD:1␤------> youhave: my $word and my @xe &(my ⏏|bow);␤ expecting any of:␤ scoped declarator␤»
timotimo hah, yeah
m: youhave: my $word and my @xe and (my |bow);
camelia rakudo-moar 8b3e8c: OUTPUT«===SORRY!=== Error while compiling /tmp/MAumbZZOAV␤Malformed my␤at /tmp/MAumbZZOAV:1␤------> youhave: my $word and my @xe and (my ⏏|bow);␤ expecting any of:␤ scoped declarator␤»
timotimo interesting, we can only get a capture in a signature list then
yeah that makes sense 07:29
m: youhave: my $word and my @xe and (my \bow);
camelia rakudo-moar 8b3e8c: OUTPUT«===SORRY!=== Error while compiling /tmp/GECODjIN_z␤Term definition requires an initializer␤at /tmp/GECODjIN_z:1␤------> ouhave: my $word and my @xe and (my \bow⏏);␤ expecting any of:␤ scoped declar…»
timotimo m: youhave: my $word and my @xe and (my \bow =0)
camelia ( no output )
07:29 FROGGS_ is now known as FROGGS
timotimo m: class You { has $word; has @xe; has &bow } 07:30
camelia ( no output )
07:31 _slade_ left
dalek rlito: ef7a08d | (Flavio S. Glock)++ | misc/typeglob/import.pl:
Perlito5 - typeglob vs. import behaviour investigation
 07:34
07:34 azawawi left
07:37 Mso150 left 07:38 virtualsue joined, FROGGS left, Mso150 joined 07:45 kjs_ joined 07:46 grondilu joined
timotimo m: grammar Test { proto token Foo(Int) {*}; token Foo:sym<a>(Int $a) { "a" }; token Foo:sym<b>(Int $a) { "b" } }; say Test.parse("a", :rule<Foo>); 07:47
camelia rakudo-moar 8b3e8c: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in regex Foo at /tmp/kAsQHX1NCU:1␤ in method parse at src/gen/m-CORE.setting:14040␤ in block <unit> at /tmp/kAsQHX1NCU:1␤␤» 07:48
timotimo ven++ stumbled upon this
masak submits rakudobug 07:51
07:53 araujo left
masak waitwait. 07:55
why would you expect any different above?
.parse calls rule Foo as the entry point. it does so in the default way, with no extra arguments. 07:56
hence, since Foo takes one parameter, there are too few positionals passed.
timotimo oh, derp.
gimme a second.
m: grammar Test { token TOP { <Foo(1)> }; proto token Foo(Int) {*}; token Foo:sym<a>(Int $a) { "a" }; token Foo:sym<b>(Int $a) { "b" } }; say Test.parse("a", :rule<Foo>); 07:57
camelia rakudo-moar 8b3e8c: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in regex Foo at /tmp/gavfFhMioV:1␤ in method parse at src/gen/m-CORE.setting:14040␤ in block <unit> at /tmp/gavfFhMioV:1␤␤»
timotimo m: grammar Test { token TOP { <Foo(1)> }; proto token Foo(Int) {*}; token Foo:sym<a>(Int $a) { "a" }; token Foo:sym<b>(Int $a) { "b" } }; say Test.parse("a");
camelia rakudo-moar 8b3e8c: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in regex Foo:sym<a> at /tmp/4zOxzv1wQ9:1␤ in any !protoregex at gen/moar/stage2/QRegex.nqp:809␤ in regex Foo at /tmp/4zOxzv1wQ9:1␤ in regex TOP at /tmp/4zOxzv1wQ9:1␤ in method parse…»
masak I think if you want to call Foo with arguments, you have to do it from within the grammar.
timotimo that should work, right?
masak oh, ok.
yes. 07:59
pmnexus It's very likely to be NYI, I think
masak reporting it now.
timotimo huh
aren't we using something like that in rakudo's own grammar all the time?
also, ven++ found this: 08:00
pmnexus A proto regex with params? Could be, but v it'd be new to me
08:00 ptc_p6 joined
timotimo m: role A[::T] { subset Test of Int }; say A[Int]::Test; # is this the right way to write this? 08:01
pmnexus Also, Rakudo doesn't compile Rakudo's grammar. :-)
camelia rakudo-moar 8b3e8c: OUTPUT«===SORRY!=== Error while compiling /tmp/XqDHJFTQEu␤Two terms in a row␤at /tmp/XqDHJFTQEu:1␤------> [::T] { subset Test of Int }; say A[Int]⏏::Test; # is this the right way to write␤ expecting any of:␤ …»
timotimo pmnexus: that's true as well :)
08:01 araujo joined, abraxxa left 08:02 Hor|zon joined
masak m: role A[::T] { subset Test of Int }; say A::Test 08:02
camelia rakudo-moar 8b3e8c: OUTPUT«Could not find symbol '&Test'␤ in method <anon> at src/gen/m-CORE.setting:13622␤ in any find_method_fallback at src/gen/m-Metamodel.nqp:2725␤ in any find_method at src/gen/m-Metamodel.nqp:988␤ in block <unit> at /tmp/3cm_yHSIJR:1␤␤»
08:03 azawawi joined, zakharyas left
masak intuitively I'd expect A[Int]::Test to work. 08:04
08:04 denis_boyun_ joined, ghostlines joined 08:05 abraxxa joined
timotimo me, too. and ven, too 08:06
08:06 azawawi left, Hor|zon left
masak submits rakudobug 08:06
pmnexus They're boarding the flight 15 people at a time... This may take a while. 08:09
08:10 kjs_ left
masak .oO( decaquintuplishly ) 08:10
moritz pmnexus: congestion avoidance, probably 08:11
08:12 Hor|zon joined
pmnexus Has something to do with plane arriving late and needing cleaning, according to the announcements 08:12
tadzik heheh, I've seen that 08:13
"we're sorry, the flight will be delayed because it arrived a wee bit late"
and through the window you can see the airport stuff forcing the plane door open...
s/stuff/staff/
masak .oO( if you're the lucky one among your group of 15, you get to clean your section! whee! ) 08:14
08:17 psch joined
psch hi #perl6 o/ 08:18
08:18 lizmat left
psch i saw the proto with args thing, and thought i knew something about that 08:20
but what i know fails as well, albeit differently
m: grammar Test { proto token Foo(Int) {*}; token Foo:sym<a>(Int $a) { "a" }; token Foo:sym<b>(Int $a) { "b" } }; say Test.parse("a", :rule<Foo>, :args(5))
camelia rakudo-moar 8b3e8c: OUTPUT«Type check failed in binding $args; expected 'Parcel' but got 'Int'␤ in method parse at src/gen/m-CORE.setting:14038␤ in block <unit> at /tmp/OXbquIFgXu:1␤␤»
psch namely :args for .parse
pmnexus They let me on the plane. :-)
08:21 woolfy left, woolfy joined, lizmat joined, notfix is now known as xfix
pmnexus And yes, the seat seems to have power outlet. 08:22
timotimo \o/
masak power to the pm! \o/ 08:23
moritz power to the nexus! 08:24
08:24 FROGGS joined
FROGGS m: grammar Test { token TOP { <Foo(1)> }; proto token Foo(Int) {*}; token Foo:sym<a>(Int $a) { "a" }; token Foo:sym<b>(Int $a) { "b" } }; say Test.parse("a", :rule<Foo>, :args(1)); 08:25
camelia rakudo-moar 8b3e8c: OUTPUT«Type check failed in binding $args; expected 'Parcel' but got 'Int'␤ in method parse at src/gen/m-CORE.setting:14038␤ in block <unit> at /tmp/NCMXHAK5A5:1␤␤»
pmnexus Alas, the nexus must now be turned off. See you all l
tadzik all available power to the phasers!
pmnexus See you all later from the other side of the Atlantic. 08:26
lizmat safe travels, pmnexus
timotimo safe travels!
i hopei hope the atlanteans don't shoot down the plane with their
dalek ast: c855d94 | usev6++ | S04-declarations/constant.t:
Add test for RT #111734
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=111734
timotimo lightning-powered tridents
08:27 FROGGS_ joined 08:30 dakkar joined, FROGGS left
FROGGS_ m: grammar Test { token TOP { <Foo(1)> }; token Foo(Int $a) { "a" } }; say Test.parse("a", :rule<Foo>, :args(1,)); 08:30
camelia rakudo-moar 8b3e8c: OUTPUT«｢a｣␤␤»
FROGGS_ timotimo / masak: you can pass args like that, it just does not seem to work with protos 08:31
see:
m: grammar Test { token TOP { <Foo(1)> }; proto token Foo(Int) {*}; token Foo:sym<a>(Int $a) { "a" }; token Foo:sym<b>(Int $a) { "b" } }; say Test.parse("a", :rule<Foo>, :args(1,));
camelia rakudo-moar 8b3e8c: OUTPUT«Too few positionals passed; expected 2 arguments but got 1␤ in regex Foo:sym<a> at /tmp/zhNVOed8KH:1␤ in any !protoregex at gen/moar/stage2/QRegex.nqp:808␤ in regex Foo at /tmp/zhNVOed8KH:1␤ in method parse at src/gen/m-CORE.setting:14040␤ i…»
pmnexus Yes, nyi
FROGGS_ masak: very important: irclog.perlgeek.de/perl6/2014-09-14#i_9356238 08:32
psch i had used that when fiddling with the cliopt stuff with an array, which explains why i didn't encounter the "expected Parcel" error i guess
as in, passing an array into :args
FROGGS_ pmnexus: I am just saying that :args in .parse *was* implemented lately, besides that issue with protos of course 08:33
masak maybe try :args with protos? 08:34
pmnexus Froggs: yes, I meant proto'd parameters nyi. We're in agreement.
Actually, I'm in a plane. :-) 08:35
08:36 cognome_ left, jack_rabbit left
pmnexus ... And I'll bbl. 08:36
08:36 pmnexus left
FROGGS_ k :o) 08:36
08:38 Hor|zon left 08:39 Ulti_ is now known as Ulti 08:40 cognome joined 08:42 cognome left 08:44 flaviusb left 08:45 flaviusb joined 08:46 kjs_ joined 08:48 zakharyas joined
timotimo FROGGS_: it doesn't work with protos or regular multis either 08:49
08:50 ptc_p6 left
timotimo GTK::Simple::TextView::⏏$!gtk_widget - this is not how i access the private attribute of a class that trusts mine? 08:50
08:51 ptc_p6 joined
timotimo that's TTIAR 08:51
moritz timotimo: uhm, classes don't have attributes, object have
FROGGS_ the sigil is never in between
timotimo mor er, yes. 08:52
with the sigil in front it works
how do i import our-scoped subs of my parent class into my current class to use them without any prefixes?
masak m: constant True = 42; say True
camelia rakudo-moar 8b3e8c: OUTPUT«42␤»
timotimo i tried "import My::Super::Class;"
moritz constant &mything = &My::Super::class::mything 08:53
masak \o/
psch .tell TimToady irclog.perlgeek.de/perl6/2014-10-13#i_9499127 might have been irclog.perlgeek.de/perl6/2014-08-15#i_9188841 08:54
yoleaux psch: I'll pass your message to TimToady.
psch also, seeing as LSM in .trans was a temporal solution i could probably close my PR for the adverbs 08:55
08:55 Hor|zon joined
timotimo i think i'll want to decouple the gtk structs from the GTK::Simple classes 08:55
well, "the gtk pointers" really
bleh how can it be so hard to "is native" a library 09:02
Cannot locate native library 'libgtksourceview-3.0.so.1.so'
09:03 Ven joined
Ven o/ timotimo 09:03
09:03 virtualsue left, Mso150 left
Ven It's nice I'm able to talk to you from IRC, considering you're sitting in front of me! 09:03
timotimo yes
Cannot locate native library 'libgtksourceview-3.0.so.1.so'
what do i need to type to make this crap work? :P
Ven timotimo: talk to it more nicely 09:05
timotimo srsly though. 09:06
Ven github.com/EnterpriseQualityCoding...vider.java <- this is why tim's idea of java2perl6 might not be what we want :P 09:07
timotimo Cannot locate native library '/usr/lib64/gtksourceview-3.0.so.1.1.0.so' 09:09
damn you, nativecall
Ven ... he went back to sleep :P
psch timotimo: you could hack around NativeCall by setting $*VM.config<load_ext> i guess? although that's kind of horrible 09:11
timotimo ugh :)
yes, it is 09:12
thank you
psch improving &guess_library_name would probably the sensible way...
timotimo changing it to "" gives me Cannot locate native library 'libgtk-3' 09:13
Ven is there a RT for the grammar bug thingie?
oooh, we actually have "earliest" now :P
timotimo instead
Ven m: anon class is native {}
camelia rakudo-moar 8b3e8c: OUTPUT«===SORRY!=== Error while compiling /tmp/U8LXcbVUCh␤Unable to parse class definition␤at /tmp/U8LXcbVUCh:1␤------> anon class is ⏏native {}␤ expecting any of:␤ scoped declarator␤ generic rol…»
timotimo bleh.
need star: use NativeCall
psch timotimo: looking at &guess_library_name and assuming you want to link 'libgtksourceview-3.0.so.1' i'd say pass 'libgtksourceview-3.0.so.' and set $*VM.config<load_ext> = 1 09:14
FROGGS_ star-m: anon class is native {}
camelia star-m 2014.08: OUTPUT«===SORRY!=== Error while compiling /tmp/aSEThRLyOJ␤Unable to parse class definition␤at /tmp/aSEThRLyOJ:1␤------> anon class is ⏏native {}␤ expecting any of:␤ scoped declarator␤ generic role␤» 09:15
FROGGS_ star-m: use NativeCall; anon class is native {}
camelia star-m 2014.08: OUTPUT«===SORRY!=== Error while compiling /tmp/hBuIGVq8jE␤Unable to parse class definition␤at /tmp/hBuIGVq8jE:1␤------> use NativeCall; anon class is ⏏native {}␤ expecting any of:␤ scoped declarator␤ …»
psch ...which is definitely all kinds of internals-leaking horrible...
FROGGS_ timotimo: I guess it should not append load_ext when the libname already contains load_ext (followed by optional digits and dots) 09:16
psch m: say $*VM.config<load_ext>; # but load_ext isn't neccessarily set 09:17
camelia rakudo-moar 8b3e8c: OUTPUT«(Any)␤»
09:17 rmgk_ joined, rmgk_ is now known as rmgk
FROGGS_ psch: how is that possible? 09:18
psch FROGGS_: i don't know. it's the same on my local install
i haven't look into it
09:18 Ven left
FROGGS_ m: say $*VM; say $*VM.config<load_ext>; 09:18
camelia rakudo-moar 8b3e8c: OUTPUT«moar (2014.9.54.g.3.ac.9.a.7)␤(Any)␤»
FROGGS_ m: say $*VM; say $*VM.config.keys;
camelia rakudo-moar 8b3e8c: OUTPUT«moar (2014.9.54.g.3.ac.9.a.7)␤crossconf staticdir defs[0] defs[1] dcbclean asmout dcclean lnclean auxclean shaincludedir dcblib exe dlllocal name ar versionmajor dclib jit lnlib laolib ccdef cppswitch dynasmlua ccinstflags dcrule ldsys cincludes asmswitc…»
psch fwiw, an old R* install on windows (where $*VM is still a hash.. i should upgrade) has .dll in there 09:19
maybe it broke during migration, no idea
timotimo FROGGS_: ŷou need "anon class :: is native" 09:21
it says so in the specs
FROGGS_ yeah, load_ext is gone somehow
psch m: say grep /^l/, $*VM.config.keys
camelia rakudo-moar 8b3e8c: OUTPUT«lnclean lnlib laolib ldsys ld ldoptiflags lnobjects lddebugflags lib lddir ldinstflags ldmiscflags ldflags lnrule ldimp ldout laoclean ldrpath laorule lua ldlibs laoobjects ldshared ldusr␤»
FROGGS_ psch: I have nothing with '.so' as a value in there 09:22
that might be a parrot thing?
p: say $*VM; say $*VM.config<load_ext>;
camelia rakudo-parrot 8b3e8c: OUTPUT«parrot (6.8.0)␤.so␤»
FROGGS_ seems so 09:23
psch right
the windows R* i have is parrot as well 09:24
maybe moar never had load_ext
FROGGS_ right, that's what I am thinking
09:27 Ven joined 09:28 yeahnoob left, denis_boyun_ left 09:30 pdcawley_ left, Ven left, pdcawley joined
timotimo so i thought hooking up gtksourceview to my live coding environment would only take five minutes or so 09:31
... i at least have a gtksourceview in my window now. but it looks exactly like a regular gtk text view
now i'll have to bind a whole lot of crap to make syntax highlighting and font selection work
psch m: say $*VM.config<dll> 09:36
camelia rakudo-moar 8b3e8c: OUTPUT«lib%s.so␤»
psch right, so that's where the '.so' comes from, not the else in &guess_library_name 09:37
makes me wonder how sensible and standardized the numeric postfixes for shared objects really are 09:38
i'm far from a sysadmin though, and it's probably documented somewhere why it's done
timotimo: on debian, the package libgtksourceview-dev contains a libgtksourceview-3.0.so, which is probably a symlink, which should make NativeCall find it i think? 09:42
timotimo nativecall is awesome, but this is painful.
really, really painful.
psch i don't know if you're on debian though :)
timotimo i'm on fus-roh-da 20 09:43
09:43 bjz joined
FROGGS_ psch: yes, the patterns of such a library name have some sort of rules across unixes... you sometimes see patterns showing up in configure scripts that locate libs 09:44
that's from a P5 module I was involved with: 09:46
($found_dll) = find_file($ld, qr/[\/\\]lib\Q$lib\E[\-\d\.]*\.($dlext[\d\.]*|so|dll)$/);
$found_dll = $1 if $found_dll && $found_dll =~/^(.+($dlext|so|dll))/ && -e $1;
($found_lib) = find_file($ld, qr/[\/\\]lib\Q$lib\E[\-\d\.]*\.($dlext[\d\.]*|a|dll.a)$/);
timotimo blergh
psch timotimo: rpm.pbone.net/index.php3/stat/6/idp...x86_64.rpm has a libgtksourceview-3.0.so 09:47
FROGGS_ $lib is something like 'gtk' if you had a libgtk.so, and $dlext is something like 'la' or 'so' or 'so|dylib|bundle'
timotimo timo@schmetterling ~/p/e/gtk_cairo_shooter (master)> rpm -ql gtksourceview3 | grep \\.so 09:48
/usr/lib64/libgtksourceview-3.0.so.1
/usr/lib64/libgtksourceview-3.0.so.1.1.0
btw, we ought to check out /etc/ld.so.conf (.d/*)
FROGGS_ we have to do that on our own? that sounds unlikelyish
timotimo right. maybe ld has something for that. 09:53
FROGGS_ dlopen should find your lib when just feed it 'gtksourceview', no? 09:54
I guess we just mess with the libname in the wrong way before passing it off
psch fwiu, we shouldn't have to mess with the libname at all on *nixes, exactly because ld and such exist, no? 09:55
oh wait, we're not linking in the usual sense of the word, we're just loading the .so 09:56
timotimo just a day after APW2014, someone started the Anty Poverty Week 2014
FROGGS_ timotimo: but there is no angry mob yet in Salzburg? 09:57
timotimo that's not where they started that 09:58
FROGGS_ k, that gives room for hope
timotimo ven is off the train again, won't have mobile network connection until he reaches france
09:58 virtualsue joined
tadzik Anti-poverty week? Angry mob? 09:59
10:01 SamuraiJack left
FROGGS_ tadzik: jnthn and/or masak cause trouble when they leave from somewhere 10:01
tadzik heh
FROGGS_ tadzik: and when there is no outrage this time in Salzburg, then probably masak is to blame :o) 10:02
so, we'll know more in about a week
timotimo tadzik: have you checked out my game yet?
tadzik timotimo: not really :( I couldn't figure out what my system is missing, and now I got caught up with $work 10:03
timotimo all your system needed was cairo and gtk3
tadzik hm
will see
but all this confusion made me think that perhaps we want something that will make it both easier for developers to write nativecall code, so they don't have to write all this "libglib-2.0-dev.so.42", and for the users looking for stuff they have or don't have installed 10:04
timotimo yes 10:05
of course we do. 10:06
at some time someone suggested we'd have something like a module configuration class that you can derive from and pass to "native" or something like that 10:07
10:08 Hor|zon left
FROGGS_ that might also help us to define fallbacks 10:09
i.e. fall back to a pure perl implementation when a C lib is not available
tadzik or fetch and compile that C lib, like Alient 10:10
timotimo that sounds like a ... great idea 10:11
tadzik and lots of work :) 10:12
psch apparently dlopen needs the full libraryname, while e.g. ld would construct it with prepending 'lib' and $ext and potentially versioning. which means we have to figure out the whole name, which means we have to know what we need :/
unless i'm misunderstanding something that means we have to ask the ldconfig cache or duplicate it's functionality..? 10:13
10:14 Hor|zon joined, leont left
psch has a feeling he's articulating intermediate steps that have been thought further by tadzik++ and FROGGS_++ and resulted in thing said by those two shortly before he wrote them 10:17
also i gotta run, later #perl6 o/
10:17 yeahnoob joined
FROGGS_ psch: see you :o) 10:17
10:21 psch left 10:24 anaeem1 left, anaeem1 joined
moritz (re library names) we aren't the only one who have those problems, right? How do other dyncall users solve this? 10:30
FROGGS_ I have no answer (yet) :o) 10:32
I might have time for that in the next days, but more likely at the weekend
damn it, I've got somebody sitting right next to me that is mumbling while reading his newspaper 10:40
tadzik reply to him
"oh, interestin!"
"hmm, really?"
FROGGS_ and when I have bad luck I have to cope with that for another 4+ hours
tadzik: no, I fear he and one person sitting in front to me are mentally ill, so I probably just wince quietly 10:41
tadzik :o 10:43
10:44 SamuraiJack joined, denis_boyun_ joined 10:46 rindolf left
dalek ast: 09ff536 | usev6++ | S02-lexical-conventions/sub-block-parsing.t:
Add test for RT #114456
 10:48
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=114456
10:49 abraxxa left
timotimo i feel a little down... getting up so early is not for me... 10:55
ptc_p6 timotimo: what's early for you?
normally, I mean ... 10:59
timotimo heh.
like 9
ptc_p6 ah, ok. :-) 11:00
timotimo twitter.com/aminize/status/521426254902206464 - discuss
tadzik "u wot m8" 11:03
moritz ... 11:06
11:06 Hor|zon left
timotimo why can't we have data structures that have other data structures in them? :( 11:07
seems like that's a rather important feature for programming languages
still perl6 has nothing like that
that's sad
tadzik well, we just got rid of one (Parcels), right? :P
I think what aminize means is "lists flatten and arrayrefs don't and it's confusing" 11:08
compared to python that only has dicts and lists, right? No confusing tuples or anyting 11:09
timotimo in perl5?
tadzik timotimo: yeah
timotimo ah
11:10 FROGGS_ is now known as FROGGS
FROGGS .tell leont: so, did I understand you right that I would need to look deeper into how IO::Uncompress::Gunzip achieves platformability in P5? 11:10
yoleaux FROGGS: What kind of a name is "leont:"?!
FROGGS .tell leont so, did I understand you right that I would need to look deeper into how IO::Uncompress::Gunzip achieves platformability in P5?
yoleaux FROGGS: I'll pass your message to leont.
FROGGS meh
11:10 Sqirrel joined 11:11 chenryn left
colomon pmichaud++ # draft report 11:11
11:11 pecastro left
colomon still thinks there probably should be a list concatenation operator. 11:12
something that distinguishes between @a, @b (list of two array refs) and @a SOMETHING @b (list is @a then @b) 11:13
moritz colomon: +1
11:14 Hor|zon joined
colomon probably should be lazy, too. 11:14
moritz maybe infix:<,,> ?
colomon infix:<++> ?
colomon has no feel on what it should look like.
11:15 kjs_ left, kurahaupo left
colomon though it's possible my thinking is backwards, and @a, @b is now @a then @b, and you need to say $@a, $@b to get a list of two array refs. 11:16
11:18 pecastro joined
FROGGS colomon: as I understand it it is about the context in which @a, @b are used 11:20
the comma does nothing on its own
[@a, @b] is an array constructor, and thus it will flatten one level, aka joining them
(@a, @b) is equal to @a, @b and does nothing 11:21
timotimo aye 11:22
assigning to a list will also flatten, like @c = @a, @b 11:23
FROGGS which is like constructing an array :o)
timotimo yes 11:24
and $a = @b, @c ?
sounds like that will just create a list that will wait for a context to flatten or itemize 11:25
FROGGS you just drop the @c on the floor there 11:26
timotimo mhm 11:27
I thought that says for binding
was*
11:28 ptc_p6 left 11:31 ptc_p6 joined 11:35 abraxxa joined, pepl left
FROGGS $a = (@b, @c) probably does what you meant 11:36
timotimo but parenthesis is supposed to not make a difference! :P 11:37
11:47 abraxxa left 11:48 abraxxa joined
tadzik what was that term for "inventing things before they're needed"? 11:51
I think I learned it here
FROGGS timotimo: yay, let's get rid of precedence and let it just dwim instead :P 11:53
12:11 kaleem left 12:12 pdcawley left, pdcawley_ joined
masak re twitter.com/aminize/status/521426254902206464 -- flattening is a fairly unconventional design decision, yes. but it's not 100% bad -- a lot of nice things that we like fall out of it too. 12:13
and concluding that we can't have nested things because of it is just... misinformed.
12:15 Hor|zon left
nine_ I often miss flattening in Python code and even in Perl 6. 12:17
masak you miss flattening in Perl 6? 12:18
nine_ masak: easy flattening :)
masak m: my @a = 1, 2, 3; my @b = 4, 5; my @c = @a, @b; say @c.perl 12:19
camelia rakudo-moar 8b3e8c: OUTPUT«Array.new(1, 2, 3, 4, 5)␤»
masak look, it's easy.
it's easy, and it flattens. it's easy flattening.
nine_ masak: then there's three more: 12:21
-> \self, |args { $.parent.perl5.invoke($package, $.parent.ptr, $name, self, args.list); }
my $av = p5_call_package_method($!p5, $package, $function, |self!setup_arguments(@args));
multi method invoke(OpaquePointer $obj, Str $function, *@args) { self.invoke(Str, $obj, $function, @args.list); }
Three different ways to make it flatten where in Perl 5 there's really only one. 12:22
masak well, when you put it like that...
dalek kudo-star-daily: 5d4a377 | coke++ | log/MoarVM-version.log:
today (automated commit)
 12:23
rl6-roast-data: 6f41da6 | coke++ | / (4 files):
today (automated commit)
masak it would help if we could at least have the same symbol for all those three cases. 12:24
once upon a time we did (`*`), but then we got two.
nine_ This would really belong into a 5 to 6 guide.
masak I never quite managed to internalize why. Python makes do with just one (for positional flattening on both caller and callee sides). 12:25
nine_ Oh and then there's method postcircumfix:<( )>(\args) { $.perl5.execute($.ptr, |args); }
In Inline::Perl5 alone I use three different ways to pass on arguments to a function. 12:26
And not by choice :) It's somehow required. Though I'm not entirely sure why.
12:31 Mouq joined
dalek ast: 1c4c535 | usev6++ | S04-statements/terminator.t:
Add tests for RT #115842
 12:33
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=115842
12:36 immortal joined, immortal left, immortal joined 12:42 Hor|zon joined, laben joined
laben hello #perl6'ers, how was APW hackaton? 12:43
FROGGS very enjoyable :o) 12:48
and for some even productive I guess
I was only semi produtive
+c
12:49 kjs_ joined
nine_ I was even productive after going home :) 12:54
Conference Driven Development FTW ;) 12:55
12:55 carlin joined
laben nine_: can i ask you something about Inline::Python? 12:55
nine_ laben: go on 12:58
laben to build the module i do perl6 configure.pl6 and then test with make test, is that right? or should i use panda? 12:59
12:59 anaeem___ joined
nine_ laben: that's exactly what I do when developing 13:00
13:01 xenoterracide__ joined, bbkr_ joined, anaeem1 left, kaleem joined
laben nine_: thanks, that reassures me. I saw there was a Build.pm file so i thought there might be some panda action there i was missing 13:01
nine_ laben: btw. your interest boosts my motivation to continue developing :) 13:02
13:03 pepl joined
laben nine_: i get then a test failure in call.t at test 8 13:03
13:03 kaare_ left
laben No such method 'test' for invocant of type 'Any' at line 107 13:04
13:04 guru joined, guru is now known as ajr_
nine_ laben: that's expected. Supporting objects is next on my list to implement. I ported the tests from Inline::Perl5. That's why some still fail. But I'll make sure master always has passing tests from now on. 13:05
13:06 flaviusb left
laben nine_: honestly i think that I::P5 is a good way to backward compatibility, while I::Py is the way to conquer a good share 13:06
gtodd nine_: Inline::Perl5 is fantastic :-D
nine_ gtodd: :)
13:07 pepl left
gtodd If this was a private corporation >50% "compatibility" of scripts would mean: "runs the majority of perl5 scripts" 13:09
just sayin' :-)
laben gtodd: only 50%? what are the current limitations of I::P5? i thought only modules with stranger things (like source filters) would be problematic 13:10
13:11 kaleem left 13:13 bjz left 13:15 FROGGS left, xenoterracide__ left 13:22 pepl joined
nine_ laben: I guess modules named Perl6::* will not work ;) 13:22
13:23 Hor|zon left 13:24 bjz joined 13:25 laben left
gtodd laben: it is higher here ... I'm just using silly marketing terminology :-) ... some method for getting perl6 to autoload I::P5 v5 or whatever when it "discovers" you're trying to load a script with perl5 code would allow for lots of real world tests and possibly a large number of bug reports :-D 13:27
13:28 FROGGS joined
nine_ gtodd: that's what S01:108 wants 13:33
synopsebot Link: perlcabal.org/syn/S01.html#line_108
13:34 chenryn joined 13:35 laben joined, FROGGS left
masak m: $_ = 2; when 2 { say "OH HAI" }; say "alive?" 13:38
camelia rakudo-moar 8b3e8c: OUTPUT«OH HAI␤»
masak should the mainline block allow `when` statements like that? 13:39
is it reasonable for execution to end when a `when` block in the mainline succeeds?
S04: "If the smart match succeeds, C<when>'s 13:40
associated block is executed, and the innermost surrounding block
that has C<$_> as one of its formal parameters (either explicit
laben masak: i'm but a simple user, but i wouldn't expect it to stop there
masak or implicit) is automatically broken out of."
I think it comes down to "does the mainline have a $_ formal parameter?" 13:41
13:41 rurban left
masak and as far as I know, it doesn't. so it shouldn't work at all. 13:41
it's just that we currently don't have static detection of a `when` outside of a topicalizer block. there's nothing in theory preventing us from having it. 13:42
I would dearly like a second opinion, though :)
laben i see it as a code of block, just like an if or if chain. i wouldn't expect "when" to end the execution there, breaking out of the surrounding block 13:43
masak m: given 42 { when 1 { say "one" }; when 42 { say "forty-two" }; when * { say "ANYTHING!!!" } } 13:44
camelia rakudo-moar 8b3e8c: OUTPUT«forty-two␤»
masak laben: but you are familiar with the above behavior, right? when a `when` succeeds, it stops looking inside the `given`. 13:45
(otherwise it would have matched that last `when *`)
laben masak: yes, but that's because of the given 13:46
masak nope.
it's because of the `when` succeeding.
laben i'm all right with taht, no falling down like C switch
masak which also explains why it exits the mainline.
which also explains why it exits the mainline. 13:47
m: for 7, 42, 1 { say "iterate"; when 1 { say "one" }; when 42 { say "forty-two" }; when * { say "ANYTHING!!!" } }
camelia rakudo-moar 8b3e8c: OUTPUT«iterate␤ANYTHING!!!␤iterate␤forty-two␤iterate␤one␤»
laben masak: i understand that's because "when" succeeds, what i'm saying is that it should do that inside a given, not everywhere
masak here you also see the succeed-means-exit-topicalizer-block semantics, but without the `given` block. 13:48
laben to me a lone "when" is like sugar over if smartmatch
13:48 FROGGS joined
masak laben: well, I I'm arguing that the fundamental problem is that `when` is allowed outside of a topicalizer block, not that it has that semantics there. 13:49
to me a lone `when` is an error. preferably a compile-time error.
note: only the block form. the statement modifier form is fine and works like you prescribe.
laben ofc you could disallow when outside of given, after all if smartmatch works fine 13:50
masak I'm suggesting that what I want is already spec. except that the spec doesn't go all the way and say `hen` is forbidden outside of a topicalizer block.
laben masak: is there another way outside of given to have a topicalizer block?
masak laben: `given`, `for`, pointy block with $_ parameter, routine with $_ parameter, `CATCH`, `CONTROL` 13:51
those are the ones that come to mind :) 13:52
laben mmh, but inside a for it doesn't exit the for which to me seems the surrounding block... 13:53
blegh, need some coffee
masak it exits the block, which cuts short the iteration 13:54
then the next iteration begins :)
peteretep or else it gets the hose again
laben basically like calling next
13:54 Hor|zon joined
masak yes, `succeed` and `next` are synonymous with for/when. 13:54
laben mh, i'm not sure i don't want when inside a for 13:55
PerlJam good morning #perl6
laben o/ PerlJam 13:56
13:56 FROGGS left
PerlJam masak: that when-in-mainline behavior is surprising to me. 13:58
masak laben: that's OK -- I'm sure I want it :) 13:59
PerlJam: a number of bugs have been discovered in which the mainline acts a bit too much like a block that can be returned from.
m: return 42; say "hello?" 14:00
camelia ( no output )
PerlJam seems like that could be a feature though (if we nailed it down sufficiently) 14:01
laben masak: right, and you're advocating to remove the possibility of having a when block outside a topicalizer since it brings that strange behaviour. if i understand your position correctly, i completely agree 14:02
14:02 leont joined 14:03 kaleem joined 14:06 ptc_p6 left
PerlJam m: return 42; say "hello?"; sub MAIN { say "hi" } 14:07
camelia ( no output )
Mouq m: say "hello?"; sub MAIN { say "hi" }; return 42
yoleaux 13 Oct 2014 16:27Z <moritz> Mouq: I finally installed your ssh key for perl6/doc - sorry that it took me so long
camelia rakudo-moar 8b3e8c: OUTPUT«hello?␤»
Mouq moritz: Thanks!
PerlJam m: $_ = 42; when 42 { say "hello?"; } sub MAIN { say "hi" } 14:08
camelia rakudo-moar 8b3e8c: OUTPUT«===SORRY!=== Error while compiling /tmp/S9h0hAVs85␤Missing semicolon.␤at /tmp/S9h0hAVs85:1␤------> $_ = 42; when 42 { say "hello?"; } ⏏sub MAIN { say "hi" }␤ expecting any of:␤ postfix␤ …»
PerlJam m: $_ = 42; when 42 { say "hello?"; }; sub MAIN { say "hi" }
camelia rakudo-moar 8b3e8c: OUTPUT«hello?␤»
masak m: say "one"; sub MAIN { say "two" } 14:09
camelia rakudo-moar 8b3e8c: OUTPUT«one␤two␤»
14:10 laben left
gtodd would bugs where mainline acts like returning from a block go away with a little more strictless and less "sugary" 'when'? 14:11
strictness
PerlJam gtodd: it would go away if the mainline didn't act like a topicalizer and if we did some static analysis during compilation that forbade "when" outside of a topicalizer 14:12
masak I don't think the mainline is acting like a topicalizer, even.
it's just that we don't check.
PerlJam right. 14:13
I was trying to be explicit in that we don't *want* the mainline to act as a topicalizer
masak *nod* 14:14
gtodd I didn't realize it was doing that
masak there's nothing in the spec to indicate it does, or should.
14:16 FROGGS joined
PerlJam in a bifurcated universe, we could have the mainline act as topicalizer and support when-in-mainline as a feature. Not sure if there's enough upside to even think about it though. 14:16
Mouq m: given 42 { { say "test"; when 42 { }; say "test2" }; say "test3" }
gtodd and if there's no million line program being used in some giant bank that depends on that ... 14:17
camelia rakudo-moar 8b3e8c: OUTPUT«test␤test3␤»
masak PerlJam: pretty sure there isn't.
:q
hugme hugs masak, good vi(m) user!
masak :)
gtodd hehe
Mouq I think that should only output "test"?
gtodd :w
masak , just like Perl 6, tries to quit the wrong thing
Mouq m: given 42 { ->{ say "test"; when 42 { }; say "test2" }(); say "test3" } # should *definitely* only be "test" 14:18
camelia rakudo-moar 8b3e8c: OUTPUT«(timeout)»
PerlJam Mouq: I think *all* blocks get a $_ parameter and this is part of the problem 14:19
(maybe)
Mouq m: {;}.signature.say; ->{}.signature.say 14:20
camelia rakudo-moar 8b3e8c: OUTPUT«(timeout)»
Mouq :($_? is parcel)
:()
14:21 Hor|zon left
PerlJam huh. 14:21
14:22 xfix is now known as notfix 14:24 abraxxa left 14:27 Hor|zon joined 14:28 chenryn left, sjn_ is now known as sjn 14:29 flaviusb joined 14:35 Mouq left 14:40 notfix is now known as xfix 14:41 treehug88 joined 14:42 yeahnoob left 14:48 kaleem left 14:51 _slade_ joined
tadzik Today is World Standards Day. Except in the US, which celebrates it on the 23rd 14:52
FROGGS :D
14:52 chenryn joined
masak ^ autopun 14:53
tadzik it's also true
en.wikipedia.org/wiki/World_Standards_Day
FROGGS that's sad 14:54
14:55 rurban joined 14:57 pepl left
carlin I would like to wish everyone a happy World Standards Day 15:01
but I can't find the RFC that specifies the protocol for doing so
FROGGS :S 15:04
15:05 jepeway joined
gtodd 15:06
15:09 FROGGS left 15:10 molaf joined 15:11 rurban left
gtodd looks on rosettacode for examples of "continuations" 15:12
... a continuation sandwich in perl6
15:20 araujo left, pmurias joined, denis_boyun_ left
masak m: macro M() {}; say &M 15:20
camelia rakudo-moar 8b3e8c: OUTPUT«macro M () { #`(Macro|74368744) ... }␤» 15:21
15:22 rurban joined, rurban1 joined, araujo joined 15:23 rurban left
[Coke] even more parrot failures: S06-signature/slurpy-params.rakudo.parrot aborted 57 test(s) 15:26
dalek rl6-roast-data: cfa87b8 | coke++ | / (4 files):
today (automated commit)
[Coke] this on top of the two failures we've been getting for some time
15:28 rurban1 left
masak lol, I blogged! strangelyconsistent.org/blog/macros...ing-macros 15:33
I probably shouldn't expect the same awesome feedback as yesterday, when lots of people were collected at a #hackathon :)
15:33 FROGGS[mobile] joined
masak but, as always, feedback appreciated. 15:33
15:36 azawawi joined 15:39 zakharyas left 15:40 azawawi left
gtodd masak: keep up the good work ... oh wait maybe you wanted useful feedback ... anyway your blogs are like a short grad course in CS so ++ 15:43
:)
15:43 kaare_ joined
colomon ye gads, masak's done three blog posts already this week!?! 15:44
masak gtodd: as someone who never really did a CS degree, I am very flattered.
gtodd heh me neither 15:45
colomon did a CS degree, but it was back in the dark ages when they thought it was a good idea to make IBM mainframe assembly language a required course.
15:46 chenryn left, ghostlines left, chenryn joined
masak colomon: expect blogging to slow down from now on. I've been a bit lucky these past two days with $work not interfering too much with my Perl 6 time. 15:47
colomon ou' 15:48
you're not going to be average 1.5 blog posts a day forever? Damn.
masak :) 15:49
dalek line-Python: a64f55a | nine++ | / (3 files):
Todo failing tests for now and add some more.
 15:50
line-Python: 252bf41 | nine++ | / (3 files):
Support passing Python dictionaries back to Perl 6.
 15:59
tadzik "The {{{ }}} syntax being universally hated" -- that should now be your signature masak :P
16:03 treehug8_ joined
dalek line-Python: 5852ce2 | nine++ | t/py_to_p6.t:
Fix UTF-8 test and un-todo it.
 16:04
16:05 treehug88 left
dalek line-Python: cd96b59 | nine++ | t/py_to_p6.t:
Add one more test for passing non-Unicode strings back to Perl 6
 16:06
16:06 anaeem___ left 16:08 Hor|zon left
dalek rlito: 89bf744 | (Flavio S. Glock)++ | misc/typeglob/import.pl:
Perlito5 - typeglob vs. import behaviour investigation
 16:09
16:13 Hor|zon joined 16:14 kurahaupo joined 16:15 Mouq joined
pmurias masak: my thought on the blog: gist.github.com/pmurias/e341037f5e8495aaf457 16:17
dalek rlito: 9b70863 | (Flavio S. Glock)++ | / (2 files):
Perlito5 - create $Perlito5::PHASE
 16:18
16:20 vendethiel joined
vendethiel Home! whew. 16:20
masak: my Text::Intended is complete
16:21 anaeem1 joined
vendethiel ...I was more complete during that train ride than during those 2 hackaton days :-) 16:21
I guess getting rid of timo gave me an energy boost :P
16:21 anaeem1 left
dalek line-Python: a65b7df | nine++ | / (3 files):
Support passing Bufs to Python
 16:22
vendethiel nine_++
I'll push it. I'm pretty happy with it, considering it's my 2nd grammar (the first one was when trying to explain actions to nine) 16:23
16:24 fhelmberger left 16:25 anaeem1 joined 16:29 anaeem1 left
dalek line-Python: 351e2b9 | nine++ | / (3 files):
Support passing Rats to Python
 16:30
16:33 anaeem1 joined, chenryn left
dalek line-Python: da0ed60 | nine++ | lib/Inline/Python.pm6:
Remove unused method unpack_return_values
 16:33
16:34 anaeem1 left 16:37 flaviusb left
gtodd vendethiel: is it Text::Intended or Text::Indented .... :-) 16:39
vendethiel gtodd: It's Intended to be Indented :)
16:40 Mso150 joined
gtodd I personally would like a Text::Intended ... indented of course ... that would write things as I intended 16:40
instead of DWIM a sort of "do what I meant to do" ... it parses my applications and groks the overall intent and suggests fixes ... PBP and perlcritic but with a kind of neural network enhanced intelligence 16:43
vendethiel that seems terrible :-)
gtodd heheh
vendethiel there's a reason we don't write in natural languages 16:44
16:44 immortal left 16:46 erkan joined
dalek line-Python: 95b6488 | nine++ | / (3 files):
Support passing floats from Python to Perl
 16:47
gtodd :-) users will perhaps need to choose a voice for perl6 in a future version ... at install time
dalek line-Python: 8849d3f | nine++ | / (2 files):
Support passing Nums from Perl to Python
 16:48
16:51 rindolf joined
gtodd n: say "are you dead?" 16:52
camelia niecza : OUTPUT«Cannot open assembly './run/Niecza.exe': No such file or directory.␤»
gtodd k
16:55 pmurias left
[Coke] niecza doesn't build on mono 3 16:55
gtodd rats 16:57
nine_ Is there some simple way to iterate over a list and get the index as well? 16:58
gtodd since it wasn't a rakudo backend I only ever had it running once or twice
colomon nine_: .kv
m: my @a = 132..138; say @a.kv
camelia rakudo-moar 8b3e8c: OUTPUT«0 132 1 133 2 134 3 135 4 136 5 137 6 138␤»
nine_ colomon: excellent! thanks 16:59
timotimo vendethiel: i'm sorry i didn't jump off the train earlier 17:00
17:00 slavik joined
dalek line-Python: 46c420b | nine++ | / (3 files):
Support passing arrays to Python
 17:01
vendethiel timotimo: don't be :)
timotimo on the train i was sort of spitballing something like this: 17:02
17:02 ajr_ left 17:03 rurban joined
timotimo proto method foobar is proxy { * }; method foobar() { $!frobfrob * 2 }; method foobar($val) { $!frobfrob = $val div 2 } 17:03
17:03 guru joined 17:04 guru is now known as Guest55848, Guest55848 left
timotimo it seems like that's already doable with regular traits and a bit of wrap 17:05
17:06 dakkar left
gtodd .kv is so cool :-) it's one of those "little things" that is very nice about perl6 ... of course you can do keys and values on an array in perl5 too (say "$_ $a[$_]" for keys @a ; #errm or something) but 17:06
vendethiel timotimo: maybe =) 17:07
17:07 ajr joined, Hor|zon left, virtualsue left
timotimo gtodd: now here's a mind-boggler for you: 17:07
m: say "hello".kv
camelia rakudo-moar 8b3e8c: OUTPUT«0 hello␤»
17:08 ajr left
dalek line-Python: efc8caa | nine++ | / (3 files):
Support passing hashes to Python
 17:08
17:08 gfldex joined
gtodd heheh 17:09
17:10 Mouq left
gtodd m: say "hello person".kv 17:10
camelia rakudo-moar 8b3e8c: OUTPUT«0 hello person␤»
gtodd doh
m: say <hello person persons>.kv 17:11
camelia rakudo-moar 8b3e8c: OUTPUT«0 hello 1 person 2 persons␤»
17:11 kjs_ left 17:12 leont left
dalek line-Python: caf7b40 | nine++ | / (3 files):
Support passing Any to Python (ends up as None)
 17:12
line-Python: a0c7035 | nine++ | t/ (3 files):
Add test for precompilation
 17:15
17:25 isBEKaml joined 17:29 ptc_p6 joined, anaeem1 joined, anaeem1 left
gtodd if you try to reverse a hash in p6 %hash.reverse nothing happens and you have to remember .inverse ... should there be an error message? 17:30
colomon warning, anyway, I'd say
gtodd I say this because the autocreating of arrays (so dup keys don't clash) part of .inverse is just *awesome* 17:31
vendethiel can't really do anything. This is in the bag of "everything on any"
gtodd this will make people want inverse in perl5 or ... add "inverse" to their growing list of perl6 lovely things 17:32
colomon: oops yeah I meant a SORRY! :-)
17:33 grettis joined, Hor|zon joined
gtodd I could add an issue somewhere ... just to watch it get added so I can learn about warnings :-) 17:34
or perhaps the plan is to eventually merge slip/reverse/invert into a some kind of overloaded data type aware version of reverse ? 17:37
colomon hashes aren't ordered, reverse never makes sense on them. 17:39
vendethiel m: my $a = 1; say $a.push(3).perl # gtodd
camelia rakudo-moar 8b3e8c: OUTPUT«Cannot call 'push'; none of these signatures match:␤:(Any:U \SELF: *@values, *%_)␤ in block <unit> at /tmp/tXY5UZk0DY:1␤␤»
vendethiel meh
17:39 Hor|zon left 17:41 CRM114 joined
gtodd ok good 17:41
17:41 jercos_ is now known as jercos
isBEKaml heh. reverse on hashes. :-) 17:43
yeah, what colomon said - hashes don't have any ordered nature like lists/arrays do. So reverse doesn't make any sense. 17:44
17:44 camelia left
isBEKaml where's camelia? 17:45
dalek line-Perl5: 2f51f44 | nine++ | / (3 files):
Support import list in $p5.use
isBEKaml m: say 42 17:46
gtodd isBEKaml: agreed ... reverse becomes flip/reverse/invert ( of course I meant "flip" above ) and the invert makes more sense (I think people even talk about inverting hashes with reverse in perl5) 17:47
vendethiel m: 1.push(3).perl.say; #gtodd 17:48
:( camelia
17:49 SamuraiJack left 17:50 camelia joined 17:51 ChanServ sets mode: +v camelia 17:52 Mouq joined
gtodd the term invert is consistent, lovely and worthy of being highlighted when someone tries to reverse a hash :) 17:52
isBEKaml gtodd: :-)
vendethiel m: 1.push(3).perl.say; #gtodd 17:53
camelia rakudo-moar 8b3e8c: OUTPUT«Cannot call 'push'; none of these signatures match:␤:(Any:U \SELF: *@values, *%_)␤ in block <unit> at /tmp/w4EikOzYpf:1␤␤»
vendethiel what was it again :P
isBEKaml gtodd: I somehow confuse you with tobyink on perlmonks. No idea why :-)
PerlJam can't stand much perlmonks these days ... too much NodeReaper and other such nonsense 17:55
Mouq m: my @a = 132..138; say @a[@a.keys.pick(3)]:kv # gtodd 17:56
camelia rakudo-moar 8b3e8c: OUTPUT«3 135 4 136 5 137␤»
17:58 nwc10 left 18:03 isBEKaml left, dwarring joined 18:04 ggherdov left 18:06 Rounin left 18:07 Rounin joined, ggherdov joined, isacloud left 18:08 isacloud______ joined 18:11 kjs_ joined 18:16 SamuraiJack joined
jnthn is home safely 18:16
vendethiel nice:) ! 18:19
gtodd m: "hair".flip # pay stylist 18:20
camelia ( no output )
18:20 Mso150 left
gtodd m: "hair".flip.say # pay stylist 18:20
camelia rakudo-moar 8b3e8c: OUTPUT«riah␤»
jnthn The headwind that gave me a tight connection on the way out was a tailwind on the way home, so we were 25 mins early landing at CPH :) 18:21
18:21 SamuraiJack left
gtodd m: "hair".comb.reverse.join.say # do it yourself 18:21
camelia rakudo-moar 8b3e8c: OUTPUT«riah␤»
18:22 lucs_ is now known as lucs
tony-o m: 'hair'.comb.perl.say; 18:27
camelia rakudo-moar 8b3e8c: OUTPUT«("h", "a", "i", "r").list␤»
18:30 kjs_ left, Mso150 joined 18:34 anaeem1_ joined 18:36 Mso150 left 18:37 Mso150 joined, kjs_ joined 18:38 Mso150 left 18:39 Mso150 joined, itz_ joined 18:41 grondilu left 18:43 grondilu joined 18:45 ptc_p6 left
bartolin I want to use something like the "std: <CODE>" locally. I checked out and built perl6/std from github, but I don't get an executable 'tools/std' 18:47
is the following mostly equivalent?
$ viv -c -e '<CODE>'
or how would I do that? 18:48
18:48 kjs_ left
masak hm, I guess the question is "how does camelia do it?" 18:48
bartolin yes, that's what I mean ;-) 18:49
masak github.com/perl6/evalbot/blob/mast...ot.pl#L178
bartolin aha, thanks! 18:50
bartolin looks
18:50 _slade_ left 18:52 Hor|zon joined
moritz (camelia never passes actual code on the command line, because of quoting trouble 18:54
)
vendethiel kwotes 18:55
18:59 xfix is now known as notfix 19:15 Hor|zon left 19:19 Mso150 left 19:20 Mso150 joined 19:22 darutoko left 19:29 virtualsue joined
masak pmurias: re gist.github.com/pmurias/e341037f5e8495aaf457 -- yep. 19:31
pmurias: from the blog post: "I think any more complicated than this and it'd be a slang. In fact, I don't mind if there's a nice, sliding scale, so that you can essentially evolve a cluster of macros of this type into a slang if you want." 19:32
19:40 molaf left 19:42 kjs_ joined
gtodd m: my @powers-of-two := 1, 2, 4 ... *; 1; @powers-of-two[^10] # from perl6advent.wordpress.com/2010/12/0...-operator/ 19:43
camelia rakudo-moar 8b3e8c: OUTPUT«WARNINGS:␤Useless use of constant integer 1 in sink context (line 1)␤»
gtodd does the syntax from perl6advent.wordpress.com/2010/12/0...-operator/ need updating in a comment?
colomon m: my @powers-of-two := 1, 2, 4 ... *; say @powers-of-two[^10] 19:46
camelia rakudo-moar 8b3e8c: OUTPUT«1 2 4 8 16 32 64 128 256 512␤»
colomon the sequences are fine, I think
the 1l was there to make the REPL happy.
gtodd in the REPL I get
@powers-of-two[^5]
Cannot call 'postcircumfix:<[ ]>'; none of these signatures match:
:(Any \SELF, int $pos)
... 19:47
PerlJam gtodd: did you put "my @powers-of-two := ..." and "@powers-of-two[^5]" on separate lines in the REPL? 19:49
gtodd let me look
PerlJam I think there's some bug there, but I'm not sure.
yeah ... 19:50
> my @x := 5,8 19:51
5 8
> say @x;
(Mu)
gtodd hmm yes and I got WARNINGS: Useless use of constant integer 1 in sink context (line 1) ... then I ran it again on separate lines :-|
on one line it works
so is it easier to comment on the advent blog or fix any perl6 bug :) 19:52
later is probably more useful ... wasn't sure if it was an already reported bug 19:54
PerlJam RT#12294 btw
gtodd ok thanks
PerlJam er, #122914
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=122914
gtodd relatively new ... or newly reported ... 19:58
20:01 guru joined, guru is now known as Guest71001 20:04 Guest71001 is now known as ajr_ 20:05 virtualsue left 20:06 rindolf left 20:09 rindolf joined 20:11 leont joined 20:12 Sqirrel left 20:15 new_listener joined, new_listener left 20:20 Mso150 left
pmichaud good afternoon, #perl6 20:20
jepeway good pm, pm. 20:21
20:21 kaare_ left
jepeway so...got a q re .WHAT: i don't know how to append it to a string. 20:23
m: say 1.WHAT
camelia rakudo-moar 8b3e8c: OUTPUT«(Int)␤»
jepeway m: say 1 ~ "eek"
camelia rakudo-moar 8b3e8c: OUTPUT«1eek␤»
dalek line-Python: 8d6a813 | nine++ | / (3 files):
Support passing Python objects through Perl space and calling methods
line-Python: 8b44f1c | nine++ | / (3 files):
Allow passing Any to Python methods
jepeway m: say 1.WHAT ~ "eek"
camelia rakudo-moar 8b3e8c: OUTPUT«use of uninitialized value of type Int in string context in block <unit> at /tmp/TQgFvxvwGk:1␤␤eek␤»
jepeway hrm?
moritz m: say 1.^name ~ "eek" 20:25
camelia rakudo-moar 8b3e8c: OUTPUT«Inteek␤» 20:26
moritz jepeway: .WHAT returns the actual type object. say() knows how to print it, but it's the empty string (+ warning) in string context
jepeway aha.
moritz to get the class name, use the 'name' method of the metaclass 20:27
and .^method calls a method on the meta class object
jepeway aha-er - say doesn't impose string context.
moritz no, it calls .gist
20:28 rurban left, zakharyas joined
pmichaud So, I have a blog article drafted about the GLR, but before I post it.... anyone else besides me have any qualms (or answers) to the potential code breakage we're about to inflict on existing Perl 6 users...? 20:28
jepeway m: say 1.WHAT.gist ~ "eek" 20:29
camelia rakudo-moar 8b3e8c: OUTPUT«(Int)eek␤»
jepeway ok. ty, moritz++
20:31 azawawi joined
azawawi hi #perl6 20:31
masak pmichaud: oh, I'm never looking forward to this kind of overarching changes... but I do believe we're in a better position than ever before to detect and handle the ecosystem upheaval. 20:32
pmichaud: and I do believe that it will be "worth it" for the consistency and performance benefits.
colomon pmichaud: as an existing p6 user, I have some qualms. 20:33
[Coke] I would recommend any changes for the GLR be done in a branch.
colomon pmichaud: but if we do it all at once, it shouldn't be too bad.
[Coke] then when we're closing to done, we can worry about bringing it back into nom.
*close to
20:33 Hor|zon joined
dalek line-Python: 3f06038 | nine++ | t/call.t:
Fix and enable test for passing hashes to Python methods
 20:33
colomon [Coke]: yes, absolutely. 20:34
and I can easily hack the module smoke tester to work in a branch, so we can check out the changes against the ecosystem before merging them into nom.
20:34 CRM114 left 20:35 treehug88 joined
masak 'night, #perl6 20:35
colomon o/
20:35 treehug8_ left
[Coke] I can also update the daily spectest runs to cope. 20:37
dalek line-Python: d708019 | nine++ | / (2 files):
Support passing Python objects to Python methods from Perl 6
 20:38
20:38 Hor|zon left
[Coke] someone's being productive. 20:39
jnthn pmichaud: Guess you made it back home? :)
20:41 virtualsue joined
jnthn pmichaud: While I doubt the Parcel going away will be a big impact, I find it much harder to estimate how big an impact the flattening changes are going to have... 20:41
pmichaud: And I'm wondering how well we can know without "implement, run tests, run ecosystem code"
colomon jnthn: see what I was saying above -- do it in a branch and I'll hack the ecosystem smoking to run on the branch 20:42
guess we'll have to branch roast, too.
jnthn *nod* 20:43
vendethiel m&z²:1 20:44
ermm...
masak: my text::indented is here: github.com/nami-doc/nephrite
masak: sorry - github.com/nami-doc/nephrite.pl
20:54 bjz_ joined 20:55 bjz left 20:57 colomon left
vendethiel (I am very unhappy of how newlines are handled...) 20:58
20:58 virtualsue left
gtodd pmichaud: answers++ but if there is a comments section that might be useful for q/a as well 20:59
timotimo i'm glad pm and jnthn made it back safely 21:01
gtodd one question I will have will be does GLR changes effect the pretty amazing level of feature parity for the backends -m -p -j ... answer: probably no way to know until the switch is thrown :-D 21:03
jnthn gtodd: It will mostly be changes in CORE.setting, so in terms of feature parity I don't think there will be a big change here. 21:05
gtodd ++ 21:06
pmichaud sorry, got called away
yes, I made it home
yes, GLR changes will happen in a branch. I'm more concerned about massive breakage of stuff... such as we're changing the semantics of .map a bit so that it doesn't flatten its invocant.
That's the one that concerns me the most, actually.
timotimo gtodd: there's a reason why we have .flip, .reverse, .invert; it's the same reason why we have + and ~ instead of just +
pmichaud [Coke]: btw, I'm planning to claim the blog post as a status update for the grant. :) 21:08
jnthn pmichaud: Yes, that seems the most likely source of pain.
pmichaud: otoh, any .map on an Array (a common case) will likely be OK in a lot of cases, in so far as the assignment to the array is flattening. 21:09
pmichaud jnthn: oh, that's likely true. I'm thinking of things where we do .foo.bar.map
jnthn pmichaud: Indeed.
gtodd timotimo: I wondered about adding a SORRY warning for use of .reverse on a hash
pmichaud 20:32 <masak> pmichaud: and I do believe that it will be "worth it" for the consistency and performance benefits. 21:11
21:11 azawawi left
pmichaud oh, I don't doubt that consistency and performance will win in the long run, I'm just wondering if we'll get a lot of initial blowback about how we're often burning our existing userbase by introducing incompatible changes 21:11
leont still isn't sure he understands the GLR
yoleaux 11:10Z <FROGGS> leont: so, did I understand you right that I would need to look deeper into how IO::Uncompress::Gunzip achieves platformability in P5?
pmichaud but hearing you say it's probably worth it is helpful :)
leont: that's what my blog post will hopefully clarify. :) 21:12
and, fwiw, I was able to do an almost complete draft on the plane, as opposed to only select pieces.
bbiab
21:13 azawawi joined
timotimo i like the sound of that 21:14
azawawi what sound? :) 21:18
jnthn pmichaud: fwiw, I just went quickly over my own set of modules looking for uses of .map and .grep. I dunno it's an especially representative sample, but it is code I could look at quickly. I found only one place that I couldn't rule out needing an s/map/for/, but I think it probably would be fine without that anyway... 21:19
21:20 bartolin left
leont The only uses I had doubts about were range.map and split.map, though I suspect they'd both still work 21:20
jnthn I think they'd be unaffected. 21:21
21:21 jack_rabbit joined, colomon joined 21:22 zakharyas left
timotimo azawawi: your irc client joining #perl6 of course ;) 21:25
21:26 gfldex left 21:27 denis_boyun_ joined
azawawi is working on git diff/log/status support in farabi6 :) 21:28
timotimo oh wow
farabi6 is really learning a bunch of tricks
since you're looking to put farabi6 into star, would you be interested in giving farabi6 a conditionally active asynchronous ... thing? 21:29
21:29 rindolf left, leont left
timotimo if i understand correctly, running a long-running script will make farabi6's server-side features unavailable? 21:30
like, farabi6 would be blocked after executing some code would finish?
21:31 kjs_ left 21:32 broquaint left
[Coke] pmichaud: (blog post, grant) Thank you for posting a status update. 21:35
azawawi timotimo: add an issue? :)
masak pmichaud: well, I never was in the group of people who might potentially complain about burning our existing userbase by introducing incompatible changes in the first place :) 21:37
timotimo we shall unleash some righteous perlfire on those who believe the list stuff would be fine staying the way it is now ... 21:39
colomon has pmichaud actually posted his blog post? 21:42
timotimo didn't see it yet 21:45
pmichaud range.map will be unaffected, definitely. split.map should be unaffected. 21:47
I've not posted it yet.
It's drafted, but I need to attend to some formatting issues, and I drafted it while in a long metal tube next to what appeared to be a contagiously sick elderly russian woman, so I should make sure what I wrote makes grammatical and semantic sense. :) 21:48
21:50 ajr_ left
jnthn О нет! Hope you didn't catch anything... 21:50
pmichaud I don't feel like I've caught anything, but it's a little soon to tell. I'll know by this evening. I do know I loaded up on orange juice as a result. :) 21:51
arnsholt jnthn: Let's just hope he didn't catch that weird one that makes people forget how definiteness works =)
timotimo agreed
pmichaud ...definiteness? 21:52
masak neither being elderly nor being Russian tends to be contagious...
timotimo you think being elderly is not contagious?
masak timotimo: it's more of an inherited thing.
pmichaud I know hanging around youthful people is contagious. :)
timotimo it seems like it's extremely easy to contract, but often has a very long incubation time
arnsholt pmichaud: Overly specific language joke, sorry =)
pmichaud arnsholt: phew. I was afraid I might've forgotten it already. 21:53
masak I guess it's related to Russian not having an article.
er, I mean "Russian not having article".
jnthn masak: That's the good theory, yes. :)
21:53 BenGoldberg joined
masak .oO( or is it "the article"? ) 21:54
arnsholt That's good theory, yeah =)
21:54 anaeem1_ left
pmichaud reviews what he wrote on the plane. 21:54
arnsholt I had a Ukranian researcher on the neighbouring desk for a few weeks. She occasionally asked me for advice on definite/indefinite corner cases 21:55
masak .oO( all the missing articles from Russian are counterbalanced by people who have to feed the the cat )
arnsholt *whoosh* 21:56
Turnabout's fair play, I guess =)
BenGoldberg Even native English speakers will overlook the the, if one the is on one line and the other the is on the next.
jnthn It's a the the cat like the nyan cat? 21:57
masak arnsholt: hint: count the words in "feed the the cat"
arnsholt Oh. Derp =D
21:57 colomon left
masak :P 21:57
arnsholt i rid reel gud
masak it's frightfully easy to just slip over
mauke /\<\(\w\+\)\s\+\1\>
masak indeed
timotimo oh lord
mauke: why would you do this %) 21:58
mauke oops, make that \_s
arnsholt Interestingly, definite/indefinite isn't cross-linguistically completely fixed. There are some examples (that escape me ATM) where English and Scandinavian disagree
mauke timotimo: not a vim fan? :-)
masak arnsholt: an easy one: en:"Nature"/sv:"naturen"
timotimo mauke: in my vim i can have \v and \V 21:59
masak arnsholt: another one: en:"life"/sv:"livet"
arnsholt Yeah, those are good ones
mauke en: "life is short" vs. de: "das leben ist kurz"
jnthn
.oO( Well, that settles the nature/nurture debate... :P )
mauke also, "back to school"
german would use a definite article (or is that "the definite article"?) 22:00
masak "the the" :)
22:01 treehug88 left 22:10 Hor|zon joined 22:17 Hor|zon left 22:22 azawawi left, FROGGS[mobile] left 22:23 FROGGS[mobile] joined 22:35 kurahaupo left, FROGGS[mobile] left 22:40 grettis left 22:43 denis_boyun_ left 22:47 colomon joined 22:52 grettis joined 22:58 dayangkun joined 23:14 bjz_ left 23:29 kurahaupo joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!