»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, std:, or /msg camelia p6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by masak on 12 May 2015.
[Coke] m: dir | say 00:00
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5===␤Argument to "say" seems to be malformed␤at /tmp/1AH9y6GQg4:1␤------> 3dir | say7⏏5<EOL>␤Other potential difficulties:␤ Unsupported use of bare "say"; in Perl 6 please use .say if you meant $_, or use an explicit invoca…»
[Coke] m: dir | .say
camelia rakudo-moar 1b537b: OUTPUT«(Any)␤Unhandled exception: Failed to get the directory contents of '/home/camelia/True': chdir failed: Unknown system error␤ at <unknown>:1 (/home/camelia/rakudo-inst-2/share/perl6/runtime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE…»
00:00 AlexDani` joined
psch m: dir | say() 00:00
camelia rakudo-moar 1b537b: OUTPUT«␤Unhandled exception: Failed to get the directory contents of '/home/camelia/True': chdir failed: Unknown system error␤ at <unknown>:1 (/home/camelia/rakudo-inst-2/share/perl6/runtime/CORE.setting.moarvm:throw:4294967295)␤ from src/gen/m-CORE.sett…»
psch m: dir() | say() 00:01
camelia rakudo-moar 1b537b: OUTPUT«␤»
psch m: say | say()
camelia rakudo-moar 1b537b: OUTPUT«␤True␤»
00:01 AlexDaniel left
[Coke] r: my $x = 'ab'; say ('ab' ~~ / a | b | $x /).Str 00:02
camelia rakudo-{moar,jvm} 1b537b: OUTPUT«a␤»
psch m: say + | say()
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/2dk_vm0u6_␤Variable '&prefix:<|>' is not declared␤at /tmp/2dk_vm0u6_:1␤------> 3say + 7⏏5| say()␤»
[Coke] r: constant $x = 'ab'; say ('ab' ~~ / a | b | $x /).Str
psch m: say +() | say()
camelia rakudo-{moar,jvm} 1b537b: OUTPUT«a␤»
rakudo-moar 1b537b: OUTPUT«␤any(0, True)␤»
labster m: &infix:<|>.() 00:03
camelia ( no output )
labster m: say &infix:<|>.() 00:04
camelia rakudo-moar 1b537b: OUTPUT«any()␤»
psch m: say [|]
camelia rakudo-moar 1b537b: OUTPUT«any()␤»
psch m: say &[|] [|], [|]
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/QSEB6FQy98␤Preceding context expects a term, but found infix , instead␤at /tmp/QSEB6FQy98:1␤------> 3say &[|] [|],7⏏5 [|]␤»
psch m: say [&[|]] [|], [|]
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/U76prd4Q3C␤Preceding context expects a term, but found infix , instead␤at /tmp/U76prd4Q3C:1␤------> 3say [&[|]] [|],7⏏5 [|]␤»
psch ehh, nevermind :)
00:05 cognominal left 00:06 pecastro joined
dalek ast: cfc0897 | coke++ | S05-metasyntax/longest-alternative.t:
Add tests for RT #113884
 00:09
00:10 Ben_Goldberg joined
psch implicit returns are broken on jvm :/ 00:10
00:10 Ben_Goldberg left
psch well, in one test file at least 00:11
r: sub f { sub g { 1 } }; my $a = f(); say $a()
camelia rakudo-jvm 1b537b: OUTPUT«java.lang.NullPointerException␤ in block <unit> at /tmp/tmpfile:1␤␤»
..rakudo-moar 1b537b: OUTPUT«1␤»
psch r: sub f { return sub g { 1 } }; my $a = f(); say $a()
camelia rakudo-{moar,jvm} 1b537b: OUTPUT«1␤»
psch actually, i guess it's the sub declaration as implicit return 00:12
00:12 BenGoldberg left
psch and jdb backtraces really don't help hah 00:13
00:14 BenGoldberg joined 00:15 Ben_Goldberg joined
psch m: sub f { $_ }; my $a = f(1); say $a 00:16
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Pek_L_ZtF5␤Calling f(int) will never work with declared signature ()␤at /tmp/Pek_L_ZtF5:1␤------> 3sub f { $_ }; my $a = 7⏏5f(1); say $a␤»
psch m: sub f { @_[0] }; my $a = f(1); say $a
camelia rakudo-moar 1b537b: OUTPUT«1␤»
psch m: sub f { @_[0] }; my $a = f(sub g { 1 }); say $a
camelia rakudo-moar 1b537b: OUTPUT«sub g () { #`(Sub|57502528) ... }␤»
psch m: sub f { @_[0] }; my $a = f(sub g { 1 }); say $a()
camelia rakudo-moar 1b537b: OUTPUT«1␤»
psch m: sub f { sub { @_[0] } }; my $a = f(1); say $a() 00:17
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/3aglJckKZM␤Calling f(int) will never work with declared signature ()␤at /tmp/3aglJckKZM:1␤------> 3sub f { sub { @_[0] } }; my $a = 7⏏5f(1); say $a()␤»
psch m: sub f($a) { sub { $a } }; my $a = f(1); say $a()
camelia rakudo-moar 1b537b: OUTPUT«1␤»
psch m: sub f($a) { sub g { $a } }; my $b = f(1); say $b() 00:18
camelia rakudo-moar 1b537b: OUTPUT«1␤»
00:18 gfldex left
psch ok, i'm baffled 00:18
next test failure...
actually 00:19
habitual m:
j: sub f($a) { sub g { $a } }; my $b = f(1); say $b()
camelia rakudo-jvm 1b537b: OUTPUT«java.lang.NullPointerException␤ in block <unit> at /tmp/lPIFkZleLR:1␤␤»
psch j: sub f { sub { @_[0] } }; my $a = f(1); say $a()
camelia rakudo-jvm 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/FZ4wXMzTu7␤Calling f(int) will never work with declared signature ()␤at /tmp/FZ4wXMzTu7:1␤------> 3sub f { sub { @_[0] } }; my $a = 7⏏5f(1); say $a()␤»
psch j: sub f { @_[0] }; my $a = f(sub g { 1 }); say $a 00:20
camelia rakudo-jvm 1b537b: OUTPUT«sub g () { #`(Sub|1241353615) ... }␤»
psch j: sub f { @_[0] }; my $a = f(sub g { 1 }); say $a()
camelia rakudo-jvm 1b537b: OUTPUT«1␤»
psch j: sub f { @_[0]; { @_[0] } }; my $a = f(sub g { 1 }); say $a()
camelia rakudo-jvm 1b537b: OUTPUT«Method 'Any' not found for invocant of class 'Parcel'␤ in block <unit> at /tmp/58Qw6DAJL3:1␤␤»
psch j: sub f { @_[0]; anon sub { @_[0] } }; my $a = f(sub g { 1 }); say $a()
camelia rakudo-jvm 1b537b: OUTPUT«(Any)␤»
psch j: sub f($a) { anon sub { $a } }; my $a = f(sub g { 1 }); say $a()
camelia rakudo-jvm 1b537b: OUTPUT«java.lang.NullPointerException␤ in block <unit> at /tmp/WlK79q_M0Y:1␤␤»
psch m: my int $a is default(42); 00:22
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/T29ipfnEQo␤Can't use trait 'is default' on a native.␤at /tmp/T29ipfnEQo:1␤------> 3my int $a is default(42)7⏏5;␤ expecting any of:␤ constraint␤»
psch j: my int $a is default(42);
camelia ( no output )
00:26 dustinm` left
[Coke] m: my \\a = any set <1 2 3>; \n say 1 ~~ a 00:29
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/SLdnZehN8u␤Malformed my␤at /tmp/SLdnZehN8u:1␤------> 3my7⏏5 \\a = any set <1 2 3>; \n say 1 ~~ a␤»
[Coke] m: my \a = any set <1 2 3>; ␤ say 1 ~~ a 00:30
camelia rakudo-moar 1b537b: OUTPUT«False␤»
[Coke] j: my \a = any set <1 2 3>; ␤ say 1 ~~ a
camelia rakudo-jvm 1b537b: OUTPUT«False␤»
skids m: my \a = any set <1 2 3>; a.perl.say 00:31
camelia rakudo-moar 1b537b: OUTPUT«any(set("1","3","2"))␤»
00:32 telex left
[Coke] does someone have a java verison built locally? 00:32
skids kinda maybe I'll check. 00:33
[Coke] just realized that RT# 115270 is talkinga bout the reply.
00:34 telex joined
psch seems to work here 00:35
gist.github.com/peschwa/11b1f3fd9921b4b7700a
i don't know if i would call "0" as return value of ~~ happy, though, but that's on moar and i don't have that built atm 00:36
skids Yeah here too.
psch $ echo -ne "start { while true { 1 } }\n" | ./perl6-j --ll-exception # that's a line that still does something wonky 00:39
i don't remember if there was an RT
00:39 leont left
psch r: start { while true { 1 } }\n 00:39
camelia rakudo-{moar,jvm} 1b537b: OUTPUT«5===SORRY!5===␤Function true needs parens to avoid gobbling block␤at /tmp/tmpfile:1␤------> 3start { while true { 1 }7⏏5 }\n␤Missing block (apparently taken by 'true')␤at /tmp/tmpfile:1␤------> 3start { while true { 1 } 7⏏5…»
psch ehh, spurious \n
well, the error is still the expected one
the echo line gives a NPE 00:40
utterly weird :P
labster m: say 2 > 1 ~~ ( More or Less ) 00:44
camelia rakudo-moar 1b537b: OUTPUT«True␤»
labster m: say(i before e unless i after "c") 00:46
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/RyLBJuDfPk␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/RyLBJuDfPk:1␤------> 3say(i before e unless 7⏏5i after "c")␤»
labster m: say i before e unless i after "c"; 00:47
camelia rakudo-moar 1b537b: OUTPUT«True␤»
psch labster++
00:48 rindolf joined 00:57 vendethiel left 00:58 rindolf left 01:00 Ben_Goldberg left 01:03 vendethiel joined 01:06 rindolf joined 01:15 rindolf left 01:17 rmgk_ joined, rmgk is now known as Guest99707, rmgk_ is now known as rmgk, Guest99707 left 01:19 atroxaper joined 01:24 atroxaper left 01:30 tinyblak_ joined 01:33 tinyblak left
b2gills .tell cognominal `say True.'!'()` calls a method by its name at runtime ( think `$a."$b"()` parens are required ). Currently to call a prefix operator as a method: `say True.&prefix:<!>` ( any code reference would work `$b = &say; $a.$b` ) I think? S03 should be changed to match 01:37
yoleaux b2gills: I'll pass your message to cognominal.
b2gills TimToady: Is there a (good) reason that `$*OUT.'say': 5` doesn't parse? 01:39
m: $*OUT.'say': 5 01:40
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/q9yB9rpbTT␤Quoted method name requires parenthesized arguments. If you meant to concatenate two strings, use '~'.␤at /tmp/q9yB9rpbTT:1␤------> 3$*OUT.'say'7⏏5: 5␤»
b2gills m: $*OUT.'say'( 5 )
camelia rakudo-moar 1b537b: OUTPUT«5␤»
01:46 vendethiel left, aborazmeh joined, aborazmeh left, aborazmeh joined 01:51 colomon left 01:52 colomon joined, vendethiel joined 01:54 khw joined 01:57 colomon left 02:03 colomon joined 02:11 dayangkun joined 02:15 vendethiel left 02:17 noganex joined 02:19 noganex_ left
kanl is there a way to determin the type of a Range? .WHAT of bounds may be inconsistent. 02:32
m: ( -Inf .. 1 ).bounds.map: *.WHAT.say 02:34
camelia rakudo-moar 1b537b: OUTPUT«(Whatever)␤Unhandled exception: Method 'count' not found for invocant of class 'Bool'␤ at <unknown>:1 (/home/camelia/rakudo-inst-2/share/nqp/lib/Perl6/BOOTSTRAP.moarvm::4294967295)␤ from src/gen/m-CORE.setting:10417 (/home/camelia/rakudo-inst-2/s…»
kanl m: ( -Inf .. 1 ).bounds.map: { $_.WHAT.say }; 02:36
camelia rakudo-moar 1b537b: OUTPUT«(Num)␤(Int)␤»
kanl m: ( 'a' .. * ).bounds.map: { $_.WHAT.say };
camelia rakudo-moar 1b537b: OUTPUT«(Str)␤(Num)␤»
kanl m: ( 'a' .. 5 ).bounds.map: { $_.WHAT.say }; 02:37
camelia rakudo-moar 1b537b: OUTPUT«(Str)␤(Int)␤»
kanl i suppose assuming .min.WHAT would be correct for most of the cases. and for the case of * being one of the bounds, just assume the .WHAT of another? 02:47
02:53 colomon left, colomon joined 02:57 baest_ is now known as baest 03:03 shinobi-cl joined 03:08 [Sno] left 03:13 shinobi-cl left 03:16 nys left 03:20 aborazmeh left 03:36 atroxaper joined 03:37 khw left 03:41 dustinm` joined 03:45 vendethiel joined 03:52 davido__ joined 03:54 davido___ left 03:59 [Sno] joined 04:02 BenGoldberg left 04:07 vendethiel left 04:23 ggoebel2 joined 04:24 ggoebel left 04:25 vendethiel joined 04:42 xinming left 04:48 vendethiel left 04:56 vendethiel joined 05:01 skids left 05:07 kaare_ joined 05:13 rurban joined 05:28 rurban left 05:29 TimToady left 05:40 vendethiel left 05:42 nightfrog left 05:43 nightfrog joined 05:45 vendethiel joined 05:47 gfldex joined 05:53 diana_olhovik joined 06:01 kurahaupo left 06:04 aborazmeh joined, aborazmeh left, aborazmeh joined 06:07 vendethiel left 06:12 vendethiel joined 06:14 aborazmeh left 06:16 atroxaper left 06:17 atroxaper joined 06:21 TimToady joined
TimToady b2gills: the reason that doesn't parse is because we have to be careful when interpolating into the end of a sentence in quotes that ends with a period: "The answer is $foo." 06:23
yoleaux 9 Jul 2015 23:43Z <[Coke]> TimToady: if we can get a ruling on RT#92742 - should enum <a b c>; c++ give c, a, or die?
06:24 rurban joined
b2gills Fair enough 06:25
TimToady m: enum <a b c>; say c.succ
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/JM6GHLIco0␤Undeclared routine:␤ c used at line 1␤␤»
TimToady m: enum ABC <a b c>; say c.succ 06:26
camelia rakudo-moar 1b537b: OUTPUT«3␤»
TimToady m: enum ABC <a b c>; say b.succ
camelia rakudo-moar 1b537b: OUTPUT«2␤»
06:26 FROGGS joined
TimToady .tell [Coke] c++ is an error because c is a constant, but c.succ gives 3 because c is really just a name for an integer 06:26
yoleaux TimToady: I'll pass your message to [Coke].
06:28 AlexDani` left, rurban1 joined
b2gills m: class C is Int {}; C.new(4).succ.^name 06:31
camelia ( no output )
b2gills m: class C is Int {}; say C.new(4).succ.^name
camelia rakudo-moar 1b537b: OUTPUT«Int␤»
06:33 vendethiel left 06:35 RabidGravy joined, vendethiel joined 06:48 domidumont joined 06:52 domidumont left 06:53 domidumont joined
[Tux] test 50000 44.875 44.787 06:53
test-t 50000 46.206 46.117
on retry even worse
test 50000 47.233 47.140
test-t 50000 47.031 46.939
the system is not busy
07:01 aborazmeh joined, aborazmeh left, aborazmeh joined
RabidGravy marning! 07:02
07:02 atroxaper left 07:04 atroxaper joined 07:13 brrt joined 07:18 darutoko joined 07:19 vendethiel left 07:21 leont joined 07:40 vendethiel joined
masak morning, o #perl6 07:42
07:45 amurf left
masak Perl 6 day! \o/ 07:46
m: class Year { has $.year; has Date $.start-date handles <is-leap-year> .= new(:$!year, :month(1), :day(1)); has Date $.end-date .= new(:$!year, :month(12), :day(31)) }; say .end-date and say .is-leap-year given Year.new(:year(2012)) 07:49
camelia rakudo-moar 1b537b: OUTPUT«2012-12-31␤True␤»
masak Perl 6 is awesome...
nine masak: indeed it is 07:56
moritz m: class Year { has $.year; has Date $.start-date handles <is-leap-year> .= new(:$!year, :month(1), :day(1)); has Date $.end-date = self.start-date.later(:1year).earlier(:1day) 07:59
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/VDRndcoU77␤Missing block␤at /tmp/VDRndcoU77:1␤------> 3.start-date.later(:1year).earlier(:1day)7⏏5<EOL>␤ expecting any of:␤ postfix␤ statement end␤ statement modifier…»
moritz m: class Year { has $.year; has Date $.start-date handles <is-leap-year> .= new(:$!year, :month(1), :day(1)); has Date $.end-date = self.start-date.later(:1year).earlier(:1day) }; say Year.new(year => 2015).end-date 08:00
camelia rakudo-moar 1b537b: OUTPUT«2015-12-31␤»
08:01 rindolf joined 08:02 vendethiel left 08:06 vendethiel joined, atroxaper left
masak m: class C { has $.x = $.y + 1; has $.y = $.x + 1 }; say 1; C.new; say 2 08:10
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/VHor7xneh7␤Virtual call $.y may not be used on partially constructed objects␤at /tmp/VHor7xneh7:1␤------> 3class C { has $.x = $.y7⏏5 + 1; has $.y = $.x + 1 }; say 1; C.new;␤ expecting any of:…»
masak m: class C { has $.x = $!y + 1; has $.y = $!x + 1 }; say 1; C.new; say 2
camelia rakudo-moar 1b537b: OUTPUT«1␤Use of uninitialized value of type Any in numeric context in method at /tmp/70rhlVR3zt:1␤2␤»
masak oh, ok :)
08:12 zakharyas joined
masak every time I change a `say` to a `print`, I get a little bit uneasy, because I'm not always sure whether I'm also implicitly changing a .gist to .Str somewhere. 08:17
08:22 sivoais left 08:23 sivoais joined
masak is once again reminded of the importance of finding the right method name 08:26
08:26 atroxaper joined
masak just used `.allot` in some code that lays things out on a page :) 08:26
not to be confused with hyperboleandahalf.blogspot.se/2010/...thing.html 08:27
DrForr Or .shallot. 08:34
brrt what is .allot 08:35
masak brrt: similar to 'allocate', but it felt a little bit more right when talking about making room for things on a canvas
brrt aha :-) 08:36
moritz
.oO( allotace )
 08:42
08:42 huf left 08:44 huf joined 08:47 leont left 08:51 vendethiel left 08:52 atroxaper left 08:54 atroxaper joined, araujo left
brrt perl6 grammars are still dark voodoo magic to me ^^ 08:55
masak brrt: it gets better.
brrt i hope
masak enjoys the new constant + heredoc niceness for the first time in actual code
DrForr github.com/drforr/perl6-ANTLR4 # may offer some inspiration.
Woodi
.oO( light voodoo magic ? )
 08:56
hi #perl6 today :)
08:57 vendethiel joined 08:59 araujo joined, araujo left
DrForr Oh, masak - Was reading strangelyconsistent.org/blog/counti...igurations - There's a simple argument for exact cover. There are 38 hexes on the board. 38 isn't divisible by 3, so there there must be at least one length-2 piece. You can't remove a single length-3 piece because that leaves a length-1 piece, ergo they must be removed in pairs. 09:00
09:01 araujo joined 09:03 rindolf left 09:04 espadrine joined
DrForr Alternatively, 2x+3y=38 only has solutions for even values of y. 09:08
Woodi jnthn: arrays 10x10x10 bytes in 1000 bytes blob cannot be more perfecter :) but there is something like en.wikipedia.org/wiki/Curse_of_dimensionality - data in arrays with high number of dimensions are sparse. but things are similiar when just one index grows in 2D or 3D - blobs can be very empty. would be nice to have other representation of native arrays, linked list or something... then blobs cou
ld be converted to linked list or ...
... even user could predeclare native array as specific representation.
but probably this is for post-6.0... 09:09
09:11 ab6tact joined
masak by the way, I have a resolution to make. 09:12
it's kind of a long-term one, with a horizon of 2020 or so.
ab6tact good *, #perl6
masak ab6tact: \o
ab6tact masak: seems I have arrived just in time for an exciting announcement :)
masak I want to focus some of my Perl 6 work on making gorgeous graphics.
like, I enjoyed making the graphics for strangelyconsistent.org/blog/youre-...-all-alike 09:13
but it's much to cumbersome to start from scratch every time, writing scripts to generate SVG and then saving them in the right format and uploading them.
ab6tact masak: I've had a distant goal of 'write a wrapper for Processing for perl6-j' 09:14
masak I would much prefer it if psyde (the static webpage thing I'm using) could handle making such diagrams inline in the Markdown
09:14 brrt left
masak and at most I'd have to create a small plugin or something to handle mazes 09:14
ab6tact oo, that does sound helpful :)
masak so, anyway, that's my goal. now you know.
ab6tact m: Date.new(:1999year).prev.say 09:15
camelia rakudo-moar 1b537b: OUTPUT«Method 'prev' not found for invocant of class 'Date'␤ in block <unit> at /tmp/oHEZhEkSmF:1␤␤»
ab6tact masak: is that expected? It seems weird that Date would have that behavior (can't get .prev from the first of a year) and makes getting the date behavior in overlaps-with trickty 09:16
masak m: say Date.new(:year(1999))
camelia rakudo-moar 1b537b: OUTPUT«1999-01-01␤»
ab6tact m: Date.new(:1999year).succ.say
camelia rakudo-moar 1b537b: OUTPUT«1999-01-02␤»
masak ab6tact: could it be that .prev is simply missing on Date?
ab6tact m: Date.new(:1999year).succ.prev.say
camelia rakudo-moar 1b537b: OUTPUT«Method 'prev' not found for invocant of class 'Date'␤ in block <unit> at /tmp/cqBWiGWkUN:1␤␤»
ab6tact wha? that's weird!
ShimmerFairy m: Date.new(:1999year).pred.say 09:17
camelia rakudo-moar 1b537b: OUTPUT«1998-12-31␤»
ab6tact p6doc -f pre returns Range.prev :/
*prev
ShimmerFairy Isn't it usually .pred ?
ab6tact ahgads
masak ah, there ya go :)
we oughta have Levenshtein for that... :)
ab6tact that'll fix. thanks ShimmerFairy
masak: well, when I grow up I hope to write a 'use no-shortforms' schema, and banish all abbreviations to the compile time error bin ;) 09:18
ShimmerFairy pred/succ are shortforms? 09:19
ab6tact predecessor / successor
masak .oO( previous / succious )
ShimmerFairy Do we really have those as methods though O_o ?
ab6tact maybe 'no-abbreviations' is a better name for the schema 09:20
masak 'no-abbrs'
:P
ab6tact jesus, s/schema/pragma/
was up late last night :P
masak s/pragma/stigma/
s/stigma/zeugma/
...what is this, Friday? :P
ab6tact hehe
ShimmerFairy wonders if use no-* wouldn't be better written as no * :P
masak m: no zeugma; 09:21
camelia rakudo-moar 1b537b: OUTPUT«===SORRY!===␤Don't know how to 'no zeugma' just yet␤»
ab6tact :)
masak m: no everything;
camelia rakudo-moar 1b537b: OUTPUT«===SORRY!===␤Don't know how to 'no everything' just yet␤»
ab6tact but I imagined that to be a 'negatiing pragma' form.. the abbreviations are not brought in by 'use abbrevs;' ,so 'no abbrevs' would be weird 09:22
masak m: no your-mom;
camelia rakudo-moar 1b537b: OUTPUT«===SORRY!===␤Don't know how to 'no your-mom' just yet␤»
ab6tact m: Date.new('1999-01-02').prev
camelia rakudo-moar 1b537b: OUTPUT«Method 'prev' not found for invocant of class 'Date'␤ in block <unit> at /tmp/nnrLtlLXBE:1␤␤»
masak it's .pred 09:23
m: Date.new('1999-01-02').pred
camelia ( no output )
masak m: say Date.new('1999-01-02').pred
camelia rakudo-moar 1b537b: OUTPUT«1999-01-01␤»
ab6tact that is so bizarre.. I guess the docs are wrong, because p6doc -f pre returns info about "method prev(Date:D:) return Date:D"
ShimmerFairy That'd be a typo then, I'd imagine.
masak ab6tact: ooh! plz to fix!
++ab6tact
ab6tact ohhh... that's why I keep spelling 'prev' wrong.. DAMN YOU APPLE AUTOCOMPLETE
DrForr I'm just a girl who can't spell n.. n... DAMNIT. 09:24
09:24 telex left
ab6tact \o/ 09:25
overlaps-with now only needs special casing for the infinite forms 09:26
09:26 telex joined
ab6tact .tell kanl checking only one side of the bounds won't work for determining the "type" of the range 09:26
yoleaux ab6tact: I'll pass your message to kanl.
ab6tact I did something like
09:27 cognominal joined
ab6tact my @types = $range.bounds: .WHAT; 09:27
and then did some grepping and some such
.tell kanl checking only one side of the boundary _will_ work if !$range.infinite 09:28
yoleaux ab6tact: I'll pass your message to kanl.
09:34 amurf joined 09:35 virtualsue joined
dalek c: 3bbbb81 | ab5tract++ | lib/Type/Date.pod:
It is .pred, not .prev
 09:36
09:36 bin_005 joined
masak ab6tact++ 09:36
09:36 haroldwu left 09:37 haroldwu joined 09:38 rindolf joined 09:39 dayangkun left, amurf left 09:42 teodozjan joined
teodozjan p6: Array[Str].new("ABCD") Z Array[Str].new("||") 09:43
camelia ( no output )
dalek ast: d6fe98b | lizmat++ | S32-array/splice.t:
Add todo test for typecheck failure

Somehow somewhere in the throw-like internals, we get a different typecheck failure than the one we expect :-(		 09:44
c: cf9ad41 | (Steve Mynott)++ | lib/Type/Proc.pod:
fix minor typos
teodozjan p6: (Array[Str].new("ABCD") Z Array[Str].new("||")).join(".").perl 09:48
camelia ( no output )
teodozjan p6: say (Array[Str].new("ABCD") Z Array[Str].new("||")).join(".").perl
camelia rakudo-moar 1b537b: OUTPUT«"ABCD ||"␤»
RabidGravy supermarket! 09:53
itz__ beer section? 09:54
09:57 RabidGravy left 10:05 aborazmeh left 10:07 g4 joined, g4 left, g4 joined 10:08 teodozjan left 10:10 teodozjan joined 10:15 woolfy joined, woolfy left
masak m: class C { has $.x is required }; C.new 10:16
camelia rakudo-moar 1b537b: OUTPUT«The attribute '$!x' is required, but you did not provide a value for it.␤ in block <unit> at /tmp/ZgjCxliD7R:1␤␤»
masak by the way, why is that block called '<unit>'? could we call it something else, like '<mainline>'? 10:17
it feels odd to use '<unit>' when we haven't declared something with the `unit` keyword.
teodozjan hi, please someone review perl6/form pull requests i didn't know that one waits for 25days and uploaded similar solution to Paul 10:18
itz__ virtualsue: github.com/retupmoca/P6-UUID/ 10:19
moritz teodozjan: fwiw you can push directly to perl6/form (and most other repos in the perl6/ org) 10:20
dalek kudo/nom: ae32e12 | lizmat++ | src/core/List.pm:
Add one more Int coercer, just in case
masak getting a review of one's PR might still be nice, though.
moritz aye
just sayin'
masak *nod* 10:21
it's useful to know one has commit access :>
or push access, rather.
virtualsue itz__: thanks
teodozjan I ddin't know that i can push but not merge :) 10:22
10:22 cschwenz joined
cschwenz o/ #perl6 10:23
masak cschwenz: \o
cschwenz Where would I report a warning being generated in Panda::Builder?
masak here :)
cschwenz Use of Nil in numeric context in block at lib/Panda/Builder.pm:125
i'm seeing that a lot during a clean build of Task::Star 10:24
masak it's this one: github.com/tadzik/panda/blob/maste...er.pm#L125
teodozjan @masak could You please merge it for last time then?
masak ok. 10:25
done.
teodozjan thanks, now it will pass tests :-) 10:26
masak \o/
cdc Hello #perl 10:29
*#perl6
masak \o
masak .oO( nothing wrong with #perl... just ain't us ) :P 10:30
cdc m: Blob.new(0xff, 0xff).decode; CATCH { .WHAT.say }
camelia rakudo-moar 1b537b: OUTPUT«(X::AdHoc)␤Malformed UTF-8 at line 1 col 1␤ in block <unit> at /tmp/tn3MAgXTii:1␤␤»
cdc I was looking to get a X::Malformed::UTF-8 instead of X::AdHoc
but it appears it comes from MoarVM
and there's not a lot of typed exceptions emitted from MoarVM 10:31
so, is this a good idea or not?
dalek kudo/nom: fadca9d | lizmat++ | lib/Test.pm:
Don't let the feedback on a failed test explode
lizmat afk& 10:32
cdc (from MoarVM/src/strings/utf8.c)
hum, never mind, I'll use "$str = try Blob.new(...).decode" instead. 10:38
10:38 TEttinger left
cdc it's even better in my case since I don't really care why Blob.decode fails 10:39
itz__ cschwenz: are you on OS X? 10:44
cschwenz yep :-)
itz__ rt.perl.org/Ticket/Display.html?id=125516 10:46
cschwenz I also ran into MoarVM spinning its wheels when compiling SVG::Plot. ^C and re-running resulted in SVG::Plot compiling and Task::Star eventually finishing successfully. 10:48
kanl gurus, how should i define/overload postcircumfix:<[]> ?
yoleaux 09:26Z <ab6tact> kanl: checking only one side of the bounds won't work for determining the "type" of the range
09:28Z <ab6tact> kanl: checking only one side of the boundary _will_ work if !$range.infinite
kanl m: class Node { has ( $.Str, @.list ); submethod BUILD( :$node ) { @!list = ( $!Str = $node ).comb: / \D+ | \d+ / } method postcircumfix:<[]>( *@i ) { @!list[@i] } }; my $a = Node.new( :node<abc000> ); say $a[0];
camelia rakudo-moar 1b537b: OUTPUT«5===SORRY!5=== Error while compiling /tmp/ilehdBeM9I␤Strange text after block (missing semicolon or comma?)␤at /tmp/ilehdBeM9I:1␤------> 3 ( $!Str = $node ).comb: / \D+ | \d+ / }7⏏5 method postcircumfix:<[]>( *@i ) { @!li␤ expecting an…»
masak kanl: I think postcircumfix:<[]> is a sub these days. 10:49
kanl m: class Node { has ( $.Str, @.list ); submethod BUILD( :$node ) { @!list = ( $!Str = $node ).comb: / \D+ | \d+ / }; method postcircumfix:<[]>( *@i ) { @!list[@i] } }; my $a = Node.new( :node<abc000> ); say $a[0];
camelia rakudo-moar 1b537b: OUTPUT«Node.new(Str => "abc000", list => ["abc", "000"]<>)␤»
masak kanl: if I were you, I'd find some nice example in the spectest suite.
kanl oh. alright. do you happen to have a link handy? thanks~ 10:50
masak looking now. I might be missing something, but there don't seem to be tests for this :/ 10:52
I've found a few `method postcircumfix:<{ }>` ones.
10:54 atroxaper left, xnrand joined 10:55 user3 joined
kanl that can help me monkeying after too, i suppose. what i don't understand with the example i posted is that the syntax seems ok ( compiler didn't complain ) but clearly not DWIM .. 10:55
masak kanl: try this instead: `Node does Positional`, and then define the method `AT-POS` 10:56
kanl: I think that's better than directly overloading .[]
10:56 user3 left
kanl masak: thanks, will give it a try 10:57
10:57 virtualsue_ joined
kanl that aside, would the thing i posted be indicating a bug, or just a NYI feature? 10:58
masak doesn't look wrong to me at first blush.
but these things are not NYI either.
jnthn If you implement the .AT-POS method, you get the various postcircumfix:<[ ]> forms "for free", as they all call your .AT-POS
yoleaux 9 Jul 2015 22:31Z <RabidGravy> jnthn: that it seems that the rt.perl.org/Ticket/Display.html?id=125408 has gone away at 2015.06-226-g01a60df built on MoarVM version 2015.06-88-g647df11
9 Jul 2015 22:56Z <TimToady> jnthn: "to implementing through implementing"?
9 Jul 2015 23:55Z <[Coke]> jnthn: I bet RT #107204 is closable with tests after your nfg work.
10:59 virtualsue left
kanl jnthn: will do. thanks. 10:59
10:59 virtualsue_ is now known as virtualsue
masak are we likely to get rid of Rakudo's Perl 5 dependencies before Christmas? (guessing "no".) 11:00
jnthn masak: As in, Configure.pl? 11:01
masak yes, mainly.
and for nqp too, I guess.
jnthn I'd say "no"
masak yeah, thought so.
maybe a reasonable goal for somewhat later, though?
jnthn If nothing else, because we should try to do things that will create the best experience for those trying out Perl 6 11:02
And I imagine we could spend the time that'd take on more worthy things.
Oh, eventually for sure.
11:02 Begi joined
Begi Hey ! How can I get the most important value of an @array ? Is there a special function ? Thanks 11:04
moritz Begi: what's the most important value?
masak haha
moritz "it's called .[]; you just have to feed it your definition of 'most important'" 11:05
jnthn suspects a translate-o :)
masak Begi: largest? smallest? largest absolute value? funniest?
jnthn m: my @a = 1,5,3,10,7; say @a.max # if largest
camelia rakudo-moar fadca9: OUTPUT«10␤»
masak jnthn++ beat me to it 11:06
Begi Yes, largest
jnthn .tell TimToady fixed, thanks!
yoleaux jnthn: I'll pass your message to TimToady.
moritz m: sub importance($x) { abs($x - 5) }; say (1, 2, 4, 8, 10).max(&importance)
camelia rakudo-moar fadca9: OUTPUT«10␤»
moritz m: sub importance($x) { abs($x - 5) }; say (1, 2, 4, 8, 10).min(&importance) 11:07
camelia rakudo-moar fadca9: OUTPUT«4␤»
Begi It works, thanks ! Do this function is in the documentation ? 11:08
moritz doc.perl6.org/routine/max 11:13
huh, only for Range
11:14 bin_005 left 11:15 rindolf left
Begi I didn't found it... 11:20
11:24 gcole joined 11:25 cschwenz left
dalek p: f847e7b | jnthn++ | src/vm/jvm/QAST/Compiler.nqp:
Map multi-dim ops for JVM backend.

None of them are yet implemented, but this is enough to get the test file to compile.		 11:35
11:38 rurban left 11:42 brrt joined 11:43 rurban joined
dalek p: 4b6c325 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/SixModelObject.java:
Extend JVM REPR API with multi-dim functions.
 11:45
masak I'm pretty happy with github.com/perl6/specs/commit/5f132abb41 -- jnthn++ -- especially the way it's based on actual usage. 11:49
colomon +1 11:50
jnthn :)
masak but (a) I have *never* set $_ manually and then done `when` stuff, so I'm pretty surprised about that usage :)
and (b) I have *never* nested `when`s, so I'm pretty surprised about that usage :)
(it does make sense, and I'll even consider (b) in the future)
jnthn masak: On (a), nor had I, and I'm not sure it's a pattern I'd use, but in a sense there's nothing wrong with it.
colomon Yeah, (b) seems like a great idea
jnthn And on (b), I think it can clean things up a good bit in some cases. 11:51
masak yeah.
I don't know of any other language that allows that.
DrForr m::2<_01>+1
m: :2<_01>+1
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Wverya0d6i␤Malformed radix number␤at /tmp/Wverya0d6i:1␤------> 3:2<7⏏5_01>+1␤ expecting any of:␤ number in radix notation␤»
11:51 abraxxa joined
DrForr ^^^ bug? Inasmuch as :2<0_1> is allowed. 0b_01 works in p5. 11:51
masak I think the rule is that _ must occur between digits. 11:52
m: say 100__000
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5===␤Only isolated underscores are allowed inside numbers␤at /tmp/EtvHVgMJ1w:1␤------> 3say 1007⏏5__000␤Confused␤at /tmp/EtvHVgMJ1w:1␤------> 3say 1007⏏5__000␤␤»
11:52 virtualsue left
jnthn m: say _100_000 # interpreted as an identifier 11:52
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5=== Error while compiling /tmp/HADQEGBoPT␤Undeclared name:␤ _100_000 used at line 1␤␤»
11:52 brrt left
DrForr That's acceptable, I was wondering if it was an oversight. 11:52
masak yeah, I think parsing _01 in there would be inconsistent. 11:53
and there's no good use for that initial _ anyway
11:54 virtualsue joined
DrForr Agreed, there's no real use for it but they're allowed in p5. 11:54
m: :2<0_>
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5=== Error while compiling /tmp/Sr4RkyyTup␤Malformed radix number␤at /tmp/Sr4RkyyTup:1␤------> 3:2<0_7⏏5>␤ expecting any of:␤ number in radix notation␤»
DrForr Right.
masak t/spec/S17-lowlevel/lock.t failed test 7 and then aborted here during the spectest run.
not reproducible with `make t/spec/S17-lowlevel/lock.t` or `perl6 t/spec/S17-lowlevel/lock.t` 11:55
11:55 abraxxa left, atroxaper joined
masak er, `t/spec/S17-lowlevel/lock.rakudo.moar` 11:55
11:55 abraxxa joined
FROGGS masak: that's a test file that usually gets skipped for releases fwiw 11:56
DrForr m: :3<0213> 11:57
camelia rakudo-moar fadca9: OUTPUT«===SORRY!===␤Invalid character '3' in number literal␤»
DrForr Nice.
Now for balanced ternary :)
kanl gurus, i'm experiencing some weirdness with eqv, and i can't quite golf it down. could someone please help me have a look at this code snipet? i'm at my wit's end with this :( 11:58
gist.github.com/anonymous/571b2a3c97a6176a6ecf
11:59 atroxaper left
kanl i even tried to print every element out, it's like staring straight at madness 11:59
i'm fine with just using 'ne', but i still like to know why '!eqv' didn't work at expected.. 12:00
^as 12:01
moritz kanl: eqv tests types as well (both from the container type and from the element types) 12:05
kanl: so you should compare .^name from both sides too, if you want to know why eqv doesn't work 12:06
kanl ah, i see. so it's actually comparing Int Int and Str Str and getting True for both?
12:07 teodozjan left
kanl err, not that. i see. container. 12:08
moritz: thanks very much.
12:09 AlexDani` joined 12:10 brrt joined
kanl m: my @a = 'abc000'.comb: / \d+ | \D+ /; my @b = 'abc001'.comb: / \d+ | \D+ /; my @i = @a.keys.grep: { @a[$_] !eqv @b[$_] }; @i.say; 12:11
camelia rakudo-moar fadca9: OUTPUT«1␤»
kanl hmm. but shouldn't this golfed down form work the same way?
this is behaving as expect. 12:12
what had really made it different in the gist i posted though?
now i'm really confused :/ 12:15
12:15 kaare_ left 12:16 kaare_ joined
timotimo o/ 12:19
masak \o
12:20 virtualsue left
masak kanl: your code in the gist is too complicated for me to understand and debug. 12:21
kanl: the fact that you golfed it down and *got rid of your problem* is an indication that it's too complicated for you to understand and debug, too :P
kanl: if you are unable to golf the problem down to something simpler, consider building your full solution out of simpler parts instead. 12:22
ShimmerFairy If you're using ne for string comparison, then you're looking for !eq , and not !eqv (since 'eq' is for string comparison like 'ne', 'eqv' is something different)
masak "Debugging is twice as hard as writing the code in the first place. Therefore, if you write the code as cleverly as possible, you are, by definition, not smart enough to debug it." -- Brian Kernighan 12:23
ShimmerFairy: that might be it.
ShimmerFairy++
kanl ShimmerFairy: well, there are chunks that are not Str, but Int, and i just thought eqv is a generic comparison that works for both Str and Int. but you're right, and like i said, i'd be fine with 'ne'. it has the same effect i wanted. 12:24
it still doesn't explain the thing i just run in channel. 12:25
12:25 zakharyas1 joined
moritz what doesn't it explain? 12:26
masak it kinda does explain it.
masak .oO( the burden of evidence that the bug reporter is not confused falls on the bug reporter )
moritz m: my @a = 'abc', '001'; my @b = 'abc', '000'; my @i = @a.keys.grep: { @a[$_] !eqv @b[$_] }; @i.say; 12:27
camelia rakudo-moar fadca9: OUTPUT«1␤»
ShimmerFairy Also, something like / \d+ | \D+ / would be more succinctly written as / .+ / , at least in the code sample you ran in the channel
moritz ShimmerFairy: no
ShimmerFairy ah, yeah, misunderstood the regex :)
moritz I don't see the problem in that snippet
m: my @a = 'abc', '001'; my @b = 'abc', '000'; my @i = @a.keys.grep: { @a[$_] ne @b[$_] }; @i.say;
camelia rakudo-moar fadca9: OUTPUT«1␤»
moritz same result
masak you would've been right 'f't'were / [\d | \d]+ /
12:28 zakharyas left
[Coke] TimToady: the original issue was that it int'd it at all; the request seems to be that given 'enum ABC <a b c>' that 'a++' should give 'b', not an int at all. if you disagree with that premise, then the current code is probably fine. 12:29
yoleaux 06:26Z <TimToady> [Coke]: c++ is an error because c is a constant, but c.succ gives 3 because c is really just a name for an integer
masak [Coke]: I disagree with that premise. 12:30
[Coke]: why would `a++` make sense if `a` is an enum?
m: True++ 12:31
camelia rakudo-moar fadca9: OUTPUT«Cannot modify an immutable Bool␤ in block <unit> at /tmp/XlE16Z_7xd:1␤␤»
moritz a.pred, not a++
masak not in what [Coke] said above.
"...that 'a++' should give 'b'..."
moritz yes
ShimmerFairy a.succ you mean?
moritz erm, yes 12:32
just a sign error :-)
dalek pan style="color: #395be5">perl6-examples: 4123799 | paultcochrane++ | .travis.yml:
Use community-supported perl6 for Travis
ShimmerFairy m: say 42++
camelia rakudo-moar fadca9: OUTPUT«Cannot assign to an immutable value␤ in block <unit> at /tmp/USdw6Drs4h:1␤␤»
masak physicist :P
ShimmerFairy ^ that's why a++ shouldn't work
moritz masak: *bows*
masak ShimmerFairy: well, [Coke] (correctly) pointed out that it's *not* an Int, it's an enum.
moritz it's both, no? 12:33
ShimmerFairy masak: but why would you essentially do a += 1 ?
moritz m: enum A <a b c>; say a ~~ Int
camelia rakudo-moar fadca9: OUTPUT«True␤»
moritz it's an Int.
m: enum A <a b c>; say a.succ
camelia rakudo-moar fadca9: OUTPUT«1␤»
FROGGS m: enum A <a b c>; say a.succ; a
camelia rakudo-moar fadca9: OUTPUT«1␤»
FROGGS m: enum A <a b c>; say a.succ; say a 12:34
camelia rakudo-moar fadca9: OUTPUT«1␤a␤»
FROGGS m: enum A <a b c>; say a.succ; say +a
camelia rakudo-moar fadca9: OUTPUT«1␤0␤»
12:34 muraiki joined
FROGGS why would one try to modify the value of an enum? 12:34
enums are pretty constant to me
12:34 Begi left
ShimmerFairy precisely, if you mean a.succ then don't say a++ :P 12:34
FROGGS nods
[Coke] masak,TimToady: original ticket is RT #92742
masak oh, aha. 12:35
rt.perl.org/Ticket/Display.html?id=92742
[Coke] the ticket is about .succ, not ++ - I took a shortcut.
masak and it has nothing to do with... right.
[Coke] which masak asked for. :P
masak it's not a shortcut that preserves semantics :)
[Coke] this is why I mentioned the ticket whenever I brought it up. :P 12:36
masak I'm sorry, I missed that.
kanl moritz: i see where the problem is now. it's { method AT-POS( *@i ) { @!list[@i] } }
i'd changed it to AT-POS( $i ) { ... } 12:37
masak [Coke]: the ticket expects a string back. I don't anymore. I expect an enum back.
[Coke] down to 1,025 rakudo tickets.
masak: I updated the subject to reflect that yesterday.
masak ah, sugoi.
ShimmerFairy masak: I think c.succ depends on what kind of mathematical model you want, whether you want it to loop (c.succ -> a), treat the bounds of the enum as though they were infinity or something (c.succ -> c), or something else/not-math-related (c.succ -> ERROR)
kanl moritz: so you're right. it's the container type, i.e. list being compared. thanks!
moritz++
12:37 vendethiel left
masak ShimmerFairy: it's spec that c.succ -> c 12:38
ShimmerFairy: if you want other semantics, define your own sub :)
ShimmerFairy ah, ok then :) I was unaware
[Coke] Please add a pointer to the design docs in the ticket to make me happy.
kanl thanks everybody for helping me debug!
[Coke] r: say "no hugmebot?" 12:39
camelia rakudo-{moar,jvm} fadca9: OUTPUT«no hugmebot?␤»
masak [Coke]: hum. I don't see any mention of .succ in S12...
but there's prior art :)
m: say False.pred; say True.succ 12:40
camelia rakudo-moar fadca9: OUTPUT«False␤True␤»
ShimmerFairy masak: Admittedly, the only number ring (if it's even the correct terminology) I know of off-hand is the one that makes 1/0 = ∞ , at the expense of nice everyday things (like 1 < 2) :)
[Coke] m: say False.succ;
camelia rakudo-moar fadca9: OUTPUT«True␤»
masak ShimmerFairy: it's not the correct terminology :)
[Coke] m: say Bool.WHAT
camelia rakudo-moar fadca9: OUTPUT«(Bool)␤»
masak ShimmerFairy: you might be looking for a concept such as en.wikipedia.org/wiki/Cyclic_group
[Coke] m: say Bool ~~ Enum
camelia rakudo-moar fadca9: OUTPUT«False␤»
masak Enum is not what you think it is... 12:41
m: enum A <b c>; say A ~~ Enum
camelia rakudo-moar fadca9: OUTPUT«False␤»
masak m: enum A <b c>; say b ~~ Enum
camelia rakudo-moar fadca9: OUTPUT«False␤»
masak o.O
[Coke] m: say Bool ~~ enum
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5===␤Whitespace required after keyword 'enum'␤at /tmp/Ax08Hqkszk:1␤------> 3say Bool ~~ enum7⏏5<EOL>␤Undeclared routine:␤ enum used at line 1␤␤␤»
masak bloody Perl 6.
what the heck is Enum, then?
m: enum A <b c>; say b.WHAT
camelia rakudo-moar fadca9: OUTPUT«(A)␤»
[Coke] Anyway, I probably enough info to try to make the ticket work. Thanks TimToady, masak.
m: say Enum.WHAT 12:42
camelia rakudo-moar fadca9: OUTPUT«(Enum)␤»
masak no, thank *you* :)
[Coke] ^^ :)
masak TimToady: one of us should spec .pred and .succ on enums.
12:42 breinbaas joined
ShimmerFairy masak: I *believe* this is what I was thinking of, though it's been a while since I read up on the weird variations on real numbers :) en.wikipedia.org/wiki/Real_projective_line 12:43
masak oh, you were thinking of that one. yeah, that's more towards the direction of topology.
DrForr As long as it forms an algebraic field :) 12:44
masak yeah, that's what I thought ShimmerFairy was talking about :) 12:45
12:45 uncleyear left
ShimmerFairy I used "ring" in the sense that it looks like one in the graphic :P 12:46
masak heh.
12:46 vendethiel joined, uncleyear joined
masak ShimmerFairy: in algebra, a ring is something with addition and multiplication, plus some simple laws that regulate those. 12:47
ShimmerFairy Yeah, I've seen that ring has a specific mathematical meaning before, so I had a sneaking suspicion I wasn't using it in the formally correct sense :) 12:48
DrForr You are correct, sir :)
jnthn
.oO( algebreic topology taught me that two rings would make a donut, but the last thing I tried with two rings ended up with me married :P )
 12:52
12:52 RabidGravy joined
moritz jnthn: were your wedding rings edible, at least? :-) 12:52
12:53 breinbaa1 left, abraxxa left
timotimo or perhaps eligible for trade-in at a Dunkin' Donuts? 12:55
jnthn moritz: I just tried, and wife gave me a funny look :P 12:56
masak jnthn: oh, in the sense of direct product, yes. 12:57
jnthn: but please call them "circles" :) 12:58
RabidGravy warm
jnthn But "if you liked it shoulda put a circle on it" breaks meter... :P
masak :P
the product space of two circles is a torus. you could think of the two original circles as helping to form the "inner tube" and the "hole in the middle", respectively. 12:59
RabidGravy adjustable by changing to other words though
two
13:03 espadrine left, espadrine_ joined 13:08 vendethiel left 13:09 vendethiel joined, lucasb joined
muraiki jnthn++ # two rings joke 13:11
13:12 amurf joined
hoelzro good morning, #perl6 13:12
masak hoelzro! \o/
vendethiel o/ #perl6, \o hoelzro
masak vendethiel! \o/ 13:13
hoelzro greetings masak, vendethiel
13:16 amurf left 13:19 AlexDani` left
[Coke] gah, someone opened a ticket! it's going in the wrong direction! 13:19
FROGGS /o\ 13:20
nwc10 [Coke]++ # fighting to reduce the ticket mountain 13:23
[Coke] jnthn++ and bartolin++ get all the credit these days. 13:24
can someone give me a oneliner to see if a string is valid utf8? 13:25
moritz in Perl 6? 13:26
try $buffer.decode('UTF-8'); say not $! 13:27
if something is inside a Str, it's already been decoded, so not much point in checking
masak oh, say it ain't $!
[Coke] so, echo "possibly invalid" | perl6 -pe ".say" should squawk? 13:29
13:30 ab6tact left
masak oh, this is a ticket, isn't it? 13:30
was this Zefram's ticket?
moritz [Coke]: yes
hoelzro [Coke]: sorry, that was me =/ 13:32
at least it's related to another ticket =P
cdc [Coke]: echo $'\xff\xff' | perl6 -e '.say' 13:38
(Any)
however:
timotimo you need to .get.say or something
cdc m: Blob.new(\xff\xff).decode
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5=== Error while compiling /tmp/p6a47tx7qA␤Unable to parse expression in argument list; couldn't find final ')' ␤at /tmp/p6a47tx7qA:1␤------> 3Blob.new(\xff\7⏏5xff).decode␤»
timotimo er 13:39
13:39 vytas joined
moritz m: Blob.new(0xff, 0xff).decode 13:39
camelia rakudo-moar fadca9: OUTPUT«Malformed UTF-8 at line 1 col 1␤ in block <unit> at /tmp/dOU8uRmg4a:1␤␤»
cdc m: Blob.new(0xff, 0xff).decode
camelia rakudo-moar fadca9: OUTPUT«Malformed UTF-8 at line 1 col 1␤ in block <unit> at /tmp/w2oNp0XuzF:1␤␤»
13:39 vytas left
timotimo i meant perl6 -e 'get.say' 13:39
cdc moritz: you were faster :)
moritz cdc: I've been probably doing Perl 6 a bit longer than you did, too :-)
hoelzro I don't think Regex.gist/Regex.perl has any tests in roast; should it?
cdc moritz: for sure! :) 13:40
moritz hoelzro: the problem with both is that they have lots of variability in their output
cdc timotimo: $ echo $'\xff\xff' | perl6 -e get.say
Malformed UTF-8
hoelzro moritz: true 13:41
I'm just thinking of doing RT #125383 real quick, and am wondering if I should add tests
kanl m: my $m = 0 but False; say ?$m; $m += 1; say ?$m; +$m += 1; 13:42
camelia rakudo-moar fadca9: OUTPUT«False␤True␤Cannot modify an immutable Int␤ in block <unit> at /tmp/lj_miRI_Hf:1␤␤»
FROGGS kanl: +$m results in a constant 13:43
I mean, there is no container you can assign to 13:44
kanl oh. i.c but does += get applied to both 0 and False?
13:44 vytas joined
FROGGS I dunno 13:44
kanl it appears so. 13:45
just want to know if it's the expected behavior 13:46
dalek kudo/nom: bd0401b | hoelzro++ | src/Perl6/Actions.nqp:
Include (regex|token|rule) if present in Regex.perl

Fixes RT #125383		 13:47
13:48 vytas left 13:50 vytas joined
masak m: my $m = 0 but False; say $m.Bool 13:51
camelia rakudo-moar fadca9: OUTPUT«False␤»
moritz kanl: looks fne to me
*fine
masak m: my $m = 0 but False; $m++; say $m.Bool
camelia rakudo-moar fadca9: OUTPUT«True␤»
masak m: my $m = -1 but False; $m++; say $m.Bool
camelia rakudo-moar fadca9: OUTPUT«False␤»
masak seems $m simply loses its `but`-ness.
just a normal Int after incrementing it. 13:52
kanl thanks for the clarification
13:53 espadrine_ left
masak m: class C { has $.x = 42; submethod BUILD {} }; say C.new.x 13:54
camelia rakudo-moar fadca9: OUTPUT«42␤»
masak oh, so attributes keep their default values, even in the presence of a submethod BUILD? 13:55
m: class C { has $.x = 42; submethod BUILD(:$.x) {} }; say C.new.x
camelia rakudo-moar fadca9: OUTPUT«Cannot find method 'STORE'␤ in submethod BUILD at /tmp/M2zHHBCwKr:1␤ in block <unit> at /tmp/M2zHHBCwKr:1␤␤»
masak m: class C { has $.x = 42; submethod BUILD(:$!x) {} }; say C.new.x 13:56
camelia rakudo-moar fadca9: OUTPUT«(Any)␤»
masak m: class C { has $.x = 42; submethod BUILD(:$!x = 5) {} }; say C.new.x
camelia rakudo-moar fadca9: OUTPUT«5␤»
masak m: class C { has $.x = 42; submethod BUILD(:$!x = 5) { $.x } }; say C.new.x
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5=== Error while compiling /tmp/zQ0mW0_cGX␤Virtual call $.x may not be used on partially constructed objects␤at /tmp/zQ0mW0_cGX:1␤------> 3.x = 42; submethod BUILD(:$!x = 5) { $.x7⏏5 } }; say C.new.x␤ expecting any of:␤ …»
masak that "Virtual call" error message should probably happen for :$.x as well...
masak submits rakudobug
vendethiel masak: for :$.x? you have :$!x 13:57
13:57 FROGGS left
masak further up. 13:57
hoelzro timotimo: do you approve of my fix for rt.perl.org/Ticket/Display.html?id=125383?
13:59 espadrine_ joined
timotimo oh RT, come on! "Could not load ticket 125383?" %) 14:00
yeah, i do approve :) 14:01
hoelzro \o/ 14:02
dalek kudo/nom: 26bc4d5 | hoelzro++ | src/Perl6/Actions.nqp:
Trim trailing whitespace from Regex source
 14:05
masak m: class C { has $.x; has $.y = 42; submethod BUILD(:$!x) {} }; (.say for .x, .y) given C.new(:x(5)) 14:07
camelia rakudo-moar fadca9: OUTPUT«5␤42␤»
masak \o/
14:07 kanl left 14:08 skids joined
lucasb m: do { say 1 } if True 14:15
camelia rakudo-moar fadca9: OUTPUT«1␤»
lucasb m: do { say 1 } given True
camelia rakudo-moar fadca9: OUTPUT«5===SORRY!5=== Error while compiling /tmp/fCJpGZs8zM␤Unsupported use of do...given; in Perl 6 please use repeat...while or repeat...until␤at /tmp/fCJpGZs8zM:1␤------> 3do { say 1 } given True7⏏5<EOL>␤»
lucasb I do see anything wrong if 'do {} given' or 'do {} for', but the error message tells me to use repeat for both cases 14:16
*with
*I don't see anything wrong with 14:17
masak it's because it considers `given` to be a loop. 14:18
14:18 brrt left
masak I agree with you, that error message shouldn't trigger on do..given 14:18
lucasb: feel free to submit a rakudobug.
14:19 cognominal left
lucasb masak: ok, thanks. what's your opinion on 'do {} for ...' ? 14:19
14:21 diana_olhovik left 14:22 diana_olhovik joined 14:27 pullphinger joined 14:29 Ven joined
lucasb "for @a { s/a/b/ }" works ok, but when I do "for $fh.lines { s/a/b/ }", I get "Cannot assign to an immutable value". It works if I add 'is copy', but not with 'is rw'. 14:40
14:40 VinceDee joined
timotimo right, the value you get passed from lines can't accept changes because of what the underlying storage is 14:41
14:42 virtualsue joined
lucasb timotimo: so, .lines return literal strings that are not stored in a container, right? ok, thanks. 14:43
timotimo at least that's my assumption
dalek p: 9a94d3e | jnthn++ | src/vm/jvm/QAST/Compiler.nqp:
Add missing :tc argument.
 14:44
p: c1b189a | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/reprs/VMArrayInstance (11 files):
Implement dimensions/set_dimensions in VMArray.
p: 79a13b0 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
Implement [num|set|<?>]dimensions ops on JVM.
jnthn Yes, timotimo++ is correct.
timotimo BBIAB 14:45
jnthn Well, that's 14/188 tests...
timotimo if it's an S-shaped curve again, that's promising :)
b2gills `do { say 1 } given True` doesn't make sense to me unless you mean `{ say 1 }() given True` 14:48
14:51 khw joined, khw_ joined, cschwenz joined
b2gills ... or you want to change it to parse like `my $a = ({ say 1} given True);` instead of `(my $a = { say 1 }) given True` 14:52
m: sub f { { say 1 } given True }; say f 14:53
camelia rakudo-moar 26bc4d: OUTPUT«1␤True␤»
14:53 khw_ left
lucasb I used 'do {}' because I was playing with writing a 'for' statement modifier, followed by a 'given' statement modifier 14:54
cschwenz When testing, is there a version of is-approx() for which I can specify the tolerance? (So, for example, if I wanted to verify a timing is within a microsecond [1e-6] of what I expect.) 14:55
lucasb I think I like chained/cascading statement modifiers, but I understand they are not supported
cschwenz Or do I have to roll my own?
dalek kudo/nom: 4bdbe83 | lizmat++ | src/core/List.pm:
Some cosmetics for easier testing
kudo/nom: 406aae9 | lizmat++ | src/core/native_array.pm:
Give array[int|num].splice same as List.splice
ast: c3d8169 | lizmat++ | S32-array/splice.t:
Added support for array[int|num].splice testing

Alas, there's still a lot of breakage, so still work in progress		 14:57
14:59 khw left 15:00 khw joined 15:01 g4 left
lizmat afk again& 15:02
15:06 diana_olhovik left 15:07 VinceDee left 15:11 FROGGS joined, zakharyas1 left
ugexe cschwenz: multi sub is-approx(Numeric $got, Numeric $expected, Numeric $tol, $desc = '') is export { 15:13
cschwenz ah, thanks ugexe! :-)
ugexe there is a 5 argument version as well that takes these in place of Numeric $tol: 15:14
Numeric :$rel_tol = 1e-6, Numeric :$abs_tol = 0
dalek p: 2b4221c | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/SixModelObject.java:
Multi-dim REPR API correctoins.
 15:15
p: ef20fbf | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
Implement generic multi-dim read/write ops.
p: 404e842 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/reprs/VMArrayInstanceBase.java:
Implement multi-dim access funcs in VMArray.
jnthn *groan* 15:16
15:16 cognominal joined
jnthn Of all the words to typo... 15:16
muraiki haha 15:25
ugexe better revert 15:26
15:27 virtualsue left, virtualsue joined 15:30 nys joined 15:34 spider-mario joined, cschwenz left 15:36 ShimmerFairy left
dalek p: a21ba08 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/ (3 files):
Stub in MultiDimArray REPR on JVM.
 15:38
itz__ is there any way of introspecting Int to show Int.DEFINITE? 15:44
TimToady m: say Int.DEFINITE 15:48
yoleaux 11:06Z <jnthn> TimToady: fixed, thanks!
camelia rakudo-moar 406aae: OUTPUT«False␤»
15:48 aborazmeh joined, aborazmeh left, aborazmeh joined
itz__ I was thing of something like Int.^methods displaying DEFINITE 15:49
errr s/thing/thinking
PerlJam itz__: since DEFINITE is macroish, probably not. 15:51
lucasb m: $_ = 'a1'; s/(\d+)/<$0>/; .say 15:52
camelia rakudo-moar 406aae: OUTPUT«a<1>␤»
lucasb m: my @a = <a1 b2 c3>; for @a { s/(\d+)/<$0>/ }; say @a
camelia rakudo-moar 406aae: OUTPUT«Use of Nil in string context in code at /tmp/dZX6Wl36PP:1␤a<> b<1> c<2>␤»
lucasb ^^ I think this is related to the bug mentioned a few days ago about the example in S05 "s:g[\$ <( \d+ )>] *= 2" not working with 'perl -pe'
In a for loop, it seems like $/ is only updated in the next iteration. Any confirm my interpretation?
*anyone
15:53 aborazmeh left 15:56 virtualsue left 15:58 virtualsue joined, ShimmerFairy joined 16:03 domidumont left 16:05 Alina-malina left
moritz confirms 16:07
16:08 Begi joined
Begi Hey ! I'm still trying to improve my Perl6. I've code a converter °C -> °F & °F -> °C. Can you take a look at it and tell me what can I improve ? Thansk for advance ! 16:10
[Coke] url?
Begi The code is here : github.com/Emeric54/perl6/blob/mas...tisseur.p6
[Coke] while 1 -> loop
Begi Oh yes, I always forget :/ 16:11
hoelzro sub mode1/mode2 could easily be a multi
dalek p: 032f577 | jnthn++ | src/vm/jvm/runtime/org/perl6/nqp/sixmodel/reprs/MultiDimArray (2 files):
Fill out MultiDimArray base class.

Has the storage-type-independent bits.		 16:13
Begi homzro : a multi sub ?
jnthn multi sub mode('1') { ... }; multi sub mode('2') { ... } 16:14
But even putting that aside, then: 16:15
if $mode eq '1' { ... } elsif $mode eq '2' { ... } else { ... } would be prettier with given/when
given $mode { when '1' { ... } when '2' { ... } default { ... } } 16:16
itz__ and just allow temps to be entered as 23C or 70F rather than using a numerical mode 16:18
jnthn That'd also be nice :)
Begi Good idea :) I'll try to correct that, thanks :)
16:19 TEttinger joined
Begi In fact, what are the adavantages of given? 16:21
16:21 uncleyea1 joined
jnthn Begi: The most immediate one is avoiding mentioning the $mode variable multiple times 16:22
16:22 uncleyear left
jnthn Begi: And when you get things to a single use, you can sometimes eliminate them 16:22
16:23 Ven left
jnthn e.g. given selectMode() { when '1' { ...} ... } 16:23
16:24 petercommand left 16:25 petercommand joined 16:27 telex left
Begi Ok, thanks. I'll try to use it more often :) I hop it's better now : github.com/Emeric54/perl6/blob/mas...tisseur.p6 16:28
16:28 telex joined 16:29 domidumont joined
Begi If it's correct now, I'll try to do like itz said 16:29
[Coke] the say should be in a default {} block. 16:36
otherwise I'd expect it to say that after anything but 3 was picked.
Begi Coke : Where should I place the say ? 16:39
[Coke] given $mode { when '1' { ... } when '2' { ... } default { PUT SAY HERE } }
16:39 virtualsue_ joined
[Coke] it's the default - it's what you get when none of the when blocks match. 16:39
where's the GLR these days? 16:40
Begi you answered my question :) Thnaks for all !
[Coke] Begi: ~~ 16:41
16:42 virtualsue left, virtualsue_ is now known as virtualsue 16:45 spider-mario left
[Coke] Are the macros we have going to be left in for Christmas? 16:45
Or are they going to be removed and come back later(/better/stronger/whatever)
(we don't really have a way to tag something as a macro-related ticket. thinking of adding [MACRO] to all of them in RT for easier sorting. 16:46
16:52 Begi left 17:00 xinming joined 17:04 Hor|zon left 17:23 virtualsue left 17:24 virtualsue joined 17:27 tinyblak_ left 17:28 uncleyea1 left 17:33 vendethiel left 17:37 vendethiel joined 17:47 tinyblak joined 17:48 virtualsue_ joined, diana_olhovik joined 17:49 domidumont left, virtualsue left, virtualsue_ is now known as virtualsue 17:51 kaare_ left 17:54 Alina-malina joined 17:56 diana_olhovik left 17:57 diana_olhovik joined 18:01 vendethiel left 18:03 vendethiel joined
masak [Coke]: I have no horse in the race of whether macros should be left in. 18:14
[Coke]: I think they see little use in the ecosystem, but nevertheless use. so I'd be inclined to leave them in unless I hear a strong argument against. 18:15
(one of jnthn's modules uses them, for example)
[Coke]: feel free to add [MACRO] to all such RT tickets. I think the union of the searches for 'macro' and 'quasi' will find everything macro-related in RT. 18:16
moritz masak: news.ycombinator.com/item?id=9864824 your maze blog post from February is on the hn front page
masak huh :)
grr, gotta find and fix that string-handling bug. 18:18
I meant to look at that today... but I got distracted by a really successful greenfield Perl 6 thing.
18:28 brrt joined 18:35 virtualsue left 18:36 virtualsue joined 18:42 brrt left
b2gills .tell Begi This is how I might have started: `loop {␤prompt('> ') ~~ /^ \h* $<d> = ['-'?\d+] \h* $<s> = <[CF]> \h* $/;␤given $<s> {␤when 'C' { say "$<d> °C = { $<d> * (9/5) + 32 } °F" }␤when 'F' { say "$<d> °F = { ($<d> - 32) * (5/9) } °C" }␤default { last }␤}␤}␤` ( What you have is perfectly fine as well ) 18:45
yoleaux b2gills: I'll pass your message to Begi.
18:47 [Tux] left 18:49 [Tux] joined 18:50 domidumont joined 18:59 virtualsue left 19:00 [TuxCM] joined 19:01 Alina-malina left, [Tux] left 19:09 espadrine_ left 19:11 virtualsue joined 19:28 FROGGS left 19:34 tinyblak left
ugexe freebsd recently (9th or 10th) started failing the last test of t/spec/S17-supply/start.t on MoarVM only (expected: [1, 1] got: [1, 1]) 19:42
gist.github.com/ugexe/6464a5d63b1ecf9a1dd0
masak quick poll: has anyone ever had to deal with processing instructions inside of their XML, HTML, or SGML? ('deal with' means either parse them, add them, or manipulate them in the document.) 19:45
19:46 domidumont left
hobbs masak: occasionally. <?xml-stylesheet?> :) 19:48
masak oh, ok. yeah, that's the example Wikipedia had, too.
[TuxCM] masak, I used xslt for that 19:49
and later also LibXML
masak [TuxCM]: ...to parse PIs?
[TuxCM] then I guess I misunderstood
adding/deleting/changing attributes/namespaces/subsections 19:50
masak ah, yes. sounds like the non-PI parts of XML :)
[TuxCM] fwiw speed is really bad today: 45/44 19:51
masak I can't recall ever seeing a processing instruction in the wild. I hadn't heard of the <?xml-stylesheet?> one until today.
19:52 vendethiel left 19:57 vendethiel joined 20:00 bin_005 joined 20:01 MueThoS76 joined 20:02 MueThoS76 left 20:04 muraiki left
bartolin ugexe: I run daily spectests on freebsd (10.1) and don't see failures in S17-supply/start.t 20:04
dalek kudo/nom: 5f7ce71 | lizmat++ | src/core/native_array.pm:
Don't bother type checking, the guts will do it
 20:05
lizmat S17-supply/start.t is a flapper for me, it flaps more or fewer times depending on the build :-( 20:06
aka, an unrelated change can make it flap more or fewer times
geekosaur yep. I've seen it flap on OS X and Linux as well 20:07
20:08 bin_005_v joined
bartolin hmm. I don't run tests parallel, maybe that's the reaseon for clean tests of that file 20:09
20:10 bin_005 left, darutoko left 20:19 vendethiel left
ugexe i ran the test by itself and continue to get the same result 20:21
in windows/linux using moar and jvm (or freebsd jvm) i cant get it to fail 20:22
jnthn It's only a gut feeling and I've no basis for this at all, but it seems the OSX memory and threading libraries are somehow good at showing up issues more often than they show up on other platforms. 20:24
bartolin could the c compiler used explain the different behaviour (ugexe++ gets failures on freebsd, I do not)? 20:26
20:27 diana_olhovik left, kaare_ joined
dalek ast: e26327b | usev6++ | S06-signature/type-capture.t:
Add test for RT #114724
 20:28
geekosaur hm, actually, come to think of it I don;'t think I'd seen that test flap on OS X. (not actually on OS X at the moment, and recent builds have all been Linux) 20:29
SO X and fbsd use the same compiler (assuming recent fbsd). threading is fairly different though 20:30
20:32 vendethiel joined
jnthn sleep time & 20:36
lizmat good night, jnthn 20:47
20:48 TEttinger left
dalek kudo/nom: 97ebe47 | lizmat++ | src/core/ (5 files):
No word parcels for descriptive texts, skids++
 20:49
20:54 vendethiel left, skids left 20:55 dha joined
lizmat .tell skids Looks like github.com/rakudo/rakudo/blob/nom/...ck.pm#L307 will never be reached 20:57
yoleaux lizmat: I'll pass your message to skids.
20:58 kaare_ left
bartolin ugexe: I stand corrected: I was able to get a failure with S17-supply/start.t on freebsd (build from 24 hours ago) 21:05
[Coke] %ENV{$key} = nqp::p6box_s(nqp::iterval($envelem)); # causes things to die if env vars contain invalid UTF8.
bartolin it failed with "expected: [1, 1] got: [1]"
[Coke] RT #125420
^^ Should env be a hash of bytes instead of strings? 21:06
21:06 vendethiel joined, bin_005_v left 21:07 bin_005_v joined 21:09 ohb joined
masak [Coke]: that sounds both like a solution and like an unfortunate solution at the same time... :/ 21:11
[Coke]: when thinking about it, I find I would prefer some seemingly impossible "best of both worlds" solution.
geekosaur envars are bytestrings on POSIX 21:12
masak something like, if it's utf8-decodable, I will get a Str, no fuss. if it isn't, I can still get it as a Buf or something.
in particular, the program should not fail just because %*ENV contains something not utf8-decodable. 21:13
masak names this tendency "utf8 chauvinism"
[Coke] doable if we try to create a string, catch an exception, and then fallback to a blob. 21:14
m: say Int < 0;
camelia rakudo-moar 5f7ce7: OUTPUT«Invocant requires a 'Int' instance, but a type object was passed. Did you forget a .new?␤ in block <unit> at /tmp/FxDY0SnX9r:1␤␤»
masak yes, but...
[Coke] m: say "cat" gt Str;
camelia rakudo-moar 5f7ce7: OUTPUT«Use of uninitialized value of type Str in string context in block <unit> at /tmp/2imsHVEvga:1␤True␤»
hobbs making it always bytes is a pain, but less of a source of bugs
masak ...anytime we create such a mixed %*ENV hash, it becomes much harder to program against.
at the very least, people will assume it's always Str all over the place. 21:15
maybe this actually merits its own Str|Buf-like type, something that puts people through the motions of always checking.
masak .oO( the solution to typing-related problems: more types! )
[Coke] m: use Test; throws-like '"cat" gt Str', X::AdHoc 21:16
camelia rakudo-moar 5f7ce7: OUTPUT« 1..2␤Use of uninitialized value of type Str in string context in block <unit> at EVAL_0:1␤ not ok 1 - '"cat" gt Str' died␤ ␤# Failed test ''"cat" gt Str' died'␤# at /tmp/aE2piWwcKE line 1␤ ok 2 - # SKIP Code did not die, can not c…»
[Coke] m: use Test; throws-like '1 > Int', X::AdHoc
camelia rakudo-moar 5f7ce7: OUTPUT« 1..2␤ ok 1 - '1 > Int' died␤ ok 2 - right exception type (X::AdHoc)␤ok 1 - did we throws-like X::AdHoc?␤»
[Coke] ^^ That seems inconsistent.
masak this is a problem with filenames/paths as well.
[Coke]: you have to put the expected-failing code in a block. 21:17
[Coke]: likely you are getting an X::AdHoc in the second case for the wrong reasons.
[Coke] S03-operators/misc.t ends with several tests where it is not a block. 21:18
I had just copied that.
masak hm, maybe there's a string version... :/
what's the advantage of having it as strings, I wonder... maybe if a parsefail is expected.
[Coke] m: use Test; throws-like {1 > Int}, X::AdHoc 21:19
camelia rakudo-moar 5f7ce7: OUTPUT« 1..2␤ ok 1 - code dies␤ ok 2 - right exception type (X::AdHoc)␤ok 1 - did we throws-like X::AdHoc?␤»
[Coke] m: use Test; throws-like {Str gt "cat"}, X::AdHoc
camelia rakudo-moar 5f7ce7: OUTPUT« 1..2␤Use of uninitialized value of type Str in string context in block at /tmp/d5GIbA40QU:1␤ not ok 1 - code dies␤ ␤# Failed test 'code dies'␤# at /tmp/d5GIbA40QU line 1␤ ok 2 - # SKIP Code did not die, can not check exception␤…»
[Coke] ^^ same
masak ok. 21:20
I agree that it's inconsistent.
dalek ast: 99a57ae | coke++ | S03-operators/misc.t:
Add tests for RT#93978
masak I don't know of a good reason why it should be inconsistent.
[Coke] ^^ that ticket.
21:29 vendethiel left
lizmat m: "cat" gt Str 21:29
camelia rakudo-moar 97ebe4: OUTPUT«Use of uninitialized value of type Str in string context in block <unit> at /tmp/RFf59fcpgW:1␤»
lizmat I don't see anything inconsistent ? 21:30
[Coke] Perhaps it's failing instead of throwing.
lizmat: the test function thinks it's different.
m: "cat" gt Str; 2+3;
camelia rakudo-moar 97ebe4: OUTPUT«WARNINGS:␤Useless use of "gt" in expression "\"cat\" gt Str" in sink context (line 1)␤Use of uninitialized value of type Str in string context in block <unit> at /tmp/3h9mXb6tWI:1␤»
lizmat but but: "cat" gt Str only warns ?? 21:32
it does not die
hence the throws-like fails ? 21:33
dha So. in the p5->p6 project, I've done perlvar, perlfunc, and perlop. 21:34
Any suggestions what I should make the next thing? Assuming I should just run away screaming, of course. 21:35
masak well, first off, dha++ 21:37
dha s/should/shouldn't/ 21:38
dha bows
[Coke] dha++, of course.
masak dha: "should" seems oddly apt, too :P 21:39
dha Well, yes. :-) 21:40
21:40 Hor|zon joined 21:42 vendethiel joined
masak dha: here are the ones I think would be interesting to see a translation of: perlrun, perllol, perlre*, perlootut 21:43
dha I was thinking of looking at perlsyn, but since things like while and for loops are not solidly documented for p6... ugh.
masak "solidly documented"? at least those things are quite stable at this point.
and you can always ask here.
dha True.
masak in a way, Perl 6 syntax is in a better position than Perl 5 syntax, because STD.pm6 is more readable than toke.c 21:44
21:45 Hor|zon left
masak m: class C { has $.x is required; submethod BUILD(:$!x) {} }; say C.new.x 21:48
camelia rakudo-moar 97ebe4: OUTPUT«(Any)␤»
dha I imagine that's true, but in reality, the concept of "you can just look at the source code!" isn't actually useful for most people trying to program stuff.
masak is this... expected behavior? does the :$!x in a BUILD parameter trump the 'is required' trait of the attribute? 21:49
dha: I meant that it might be useful for people trying to port perlsyn. 21:50
dha *nod*
21:56 atroxaper joined
RabidGravy I also think there should be some equivalent to perlipc but it's so different I don't think a straight port is a good idea 21:56
22:01 atroxaper left
RabidGravy but hey I can do MP3 encoding now, not quite sure that it's fast enough to do it "live" but I'll test tomorrow 22:03
dalek ast: 5e2ad04 | lizmat++ | S32-array/splice.t:
Some more splice testing, still in progress
 22:07
ast: 29a3ff9 | lizmat++ | S0 (2 files):
Merge branch 'master' of github.com/perl6/roast
22:07 vendethiel left
lizmat and with that I wish #perl6 good night! 22:09
masak 'night, lizmat 22:11
cognominal not sure how to use run to create a pipe. 22:20
yoleaux 01:37Z <b2gills> cognominal: `say True.'!'()` calls a method by its name at runtime ( think `$a."$b"()` parens are required ). Currently to call a prefix operator as a method: `say True.&prefix:<!>` ( any code reference would work `$b = &say; $a.$b` ) I think? S03 should be changed to match
22:20 pullphinger left
cognominal how would I do a ls| with a run? 22:20
m: say True.'!'() 22:22
camelia rakudo-moar 97ebe4: OUTPUT«Method '!' not found for invocant of class 'Bool'␤ in block <unit> at /tmp/R_mwsQfd6n:1␤␤»
lucasb m: .say for run('ls', :out).out.lines[^5]
camelia rakudo-moar 97ebe4: OUTPUT«run is disallowed in restricted setting␤ in sub restricted at src/RESTRICTED.setting:1␤ in sub run at src/RESTRICTED.setting:14␤ in block <unit> at /tmp/UaAmhc2Y5k:1␤␤»
lucasb oops
cognominal m: say True.&prefix:<!>
camelia rakudo-moar 97ebe4: OUTPUT«False␤»
cognominal m: say True.&prefix:<!>()
camelia rakudo-moar 97ebe4: OUTPUT«False␤»
lucasb cognominal: that's how I think it could be 22:23
^^
cognominal thx lucasb
runs fine 22:24
22:27 rurban left 22:33 rurban1 left
masak 'night, #perl6 22:48
22:51 dj_goku left, raiph joined 22:53 skids joined 22:56 Exodist joined 22:57 dj_goku joined, dj_goku left, dj_goku joined 22:59 muraiki joined
skids o/ 23:01
yoleaux 20:57Z <lizmat> skids: Looks like github.com/rakudo/rakudo/blob/nom/...ck.pm#L307 will never be reached
23:02 RabidGravy left
skids .tell lizmat proceed resumes matching when clauses. Also that code is temporary and we'll just throw those once .assuming has burned it wrt ecosystem falout. 23:02
yoleaux skids: I'll pass your message to lizmat.
23:10 vendethiel joined
tony-o m: say 5.^infix:<+>(5); 23:18
camelia rakudo-moar 97ebe4: OUTPUT«Method 'infix:<+>' not found for invocant of class 'Perl6::Metamodel::ClassHOW'␤ in block <unit> at /tmp/GbhOhkGEXi:1␤␤»
tony-o m: say 5.&infix:<+>(5);
camelia rakudo-moar 97ebe4: OUTPUT«10␤»
tony-o m: say 5.&infix:<+>(5,6,7);
camelia rakudo-moar 97ebe4: OUTPUT«Too many positionals passed; expected 0 to 2 arguments but got 4␤ in block <unit> at /tmp/cLybWXozUp:1␤␤»
23:29 Hor|zon joined
skids That reminded me... does anyone keep a dogpile of ecosystem+any other per6 code they find, such that it can all be grepped? I'd be interested to know where .assuming is already bein used. 23:29
23:33 vendethiel left 23:34 Hor|zon left 23:35 telex left 23:36 telex joined 23:37 dha left 23:38 lucasb left, geekosaur left 23:39 geekosaur joined 23:40 virtualsue left 23:44 atroxaper joined 23:45 leont joined 23:47 AlexDani` joined, AlexDani` is now known as AlexDaniel, TEttinger joined
AlexDaniel m: loop (my $i = 0; $i <= 5; $i++;) { say } 23:48
camelia rakudo-moar 97ebe4: OUTPUT«5===SORRY!5=== Error while compiling /tmp/YgalfVZJ4f␤Missing block␤at /tmp/YgalfVZJ4f:1␤------> 3loop (my $i = 0; $i <= 5; $i++7⏏5;) { say }␤»
AlexDaniel m: loop (my $i = 0; $i <= 5; $i++ {;) { say } 23:49
camelia rakudo-moar 97ebe4: OUTPUT«5===SORRY!5=== Error while compiling /tmp/8DFY0Mjfta␤Unexpected block in infix position (missing statement control word before the expression?)␤at /tmp/8DFY0Mjfta:1␤------> 3loop (my $i = 0; $i <= 5; $i++7⏏5 {;) { say }␤ expecting any …»
23:49 atroxaper left
AlexDaniel sure the code is wrong, but what does this error message mean 23:49
"missing block", then I give it a block and it says "unexpected block"
skids In the second one, it is ike you said: 23:50
m: my $i; $i++ { }
camelia rakudo-moar 97ebe4: OUTPUT«5===SORRY!5=== Error while compiling /tmp/somNcZTmOb␤Unexpected block in infix position (missing statement control word before the expression?)␤at /tmp/somNcZTmOb:1␤------> 3my $i; $i++7⏏5 { }␤ expecting any of:␤ infix␤ …»
TimToady std: loop (my $i = 0; $i <= 5; $i++;) { say }
camelia std 28329a7: OUTPUT«5===SORRY!5===␤Malformed loop spec at /tmp/7S0kI3Pi34 line 1:␤------> 3loop (my $i = 0; $i <= 5; $i++7⏏5;) { say }␤ expecting any of:␤ POST␤ postfix␤ postfix_prefix_meta_operator␤Parse failed␤FAILED 00:00 137m␤»
TimToady I like STD's message better
AlexDaniel TimToady: now that's an error message! 23:51
skids: ok, but if it wants a block, what is the right way to make it eat it?
I mean, sure it does not make sense, but if the compiler asks for it...
TimToady it doesn't really want a block there 23:52
it's just loosing track of its immediate goal and reporting on the outer goal
*losing
AlexDaniel well, since the message could be better I think I will report it. Although it is not a big issue at all 23:53
TimToady probably LHF for someone
AlexDaniel should I mark it as lhf? 23:54
TimToady sure
AlexDaniel great
23:55 vendethiel joined
AlexDaniel it is the only place in the world where I could come with a tiny issue like this and people will take it seriously :) 23:56
so great.
TimToady just means we're neglecting all the important things :) 23:57

Powered by Cro and the Raku® Programming Language.

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

Thank you!