»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:01 y3llow left, y3llow joined 00:03 y3llow left 00:04 cognominal_ joined, y3llow joined 00:06 y3llow left, cognominal left, y3llow joined 00:08 y3llow left 00:09 y3llow joined 00:11 y3llow left, y3llow joined 00:13 whiteknight joined, y3llow left, y3llow joined 00:15 y3llow left 00:16 y3llow joined 00:17 raiph joined 00:18 wolfman2000 joined, y3llow left, y3llow joined 00:20 y3llow left 00:21 y3llow joined 00:23 y3llow left, y3llow joined 00:25 y3llow left, y3llow joined 00:27 y3llow left 00:28 y3llow joined 00:29 silug left, japhb_ joined 00:30 y3llow left, y3llow joined
japhb_ phenny, tell jnthn In the if block at github.com/rakudo/rakudo/commit/4d...6b#L0R1217 , it looks like 'if @name' is a redundant test in the second if. 00:31
phenny japhb_: I'll pass that on when jnthn is around.
00:32 japhb_ left, y3llow left 00:33 y3llow joined 00:34 y3llow left 00:35 y3llow joined 00:37 y3llow left, y3llow joined 00:39 y3llow left, y3llow joined 00:41 y3llow left 00:42 y3llow joined 00:43 y3llow left 00:44 y3llow joined 00:46 y3llow left, y3llow joined 00:48 y3llow left 00:49 y3llow joined, bluescreen10 left 00:50 y3llow left 00:51 y3llow joined, silug joined 00:53 y3llow left, y3llow joined 00:55 y3llow left, y3llow joined 00:57 clkao_ is now known as clkao, y3llow left 00:58 y3llow joined 01:00 y3llow left, y3llow joined 01:02 y3llow left, y3llow joined 01:04 y3llow left 01:05 y3llow joined 01:07 y3llow left, y3llow joined 01:09 y3llow left, y3llow joined 01:11 y3llow left, y3llow joined 01:13 y3llow left 01:14 y3llow joined 01:16 y3llow left, y3llow joined 01:19 y3llow left, y3llow joined 01:22 y3llow left, y3llow joined 01:25 y3llow left, y3llow joined 01:28 y3llow left, y3llow joined 01:31 y3llow left, y3llow joined 01:34 y3llow left, y3llow joined 01:37 y3llow left, y3llow joined 01:40 y3llow left, y3llow joined 01:43 y3llow left 01:44 y3llow joined 01:46 y3llow left 01:47 y3llow joined
timotimo o_O 01:48
01:49 y3llow left 01:50 y3llow joined 01:51 _jaldhar left 01:52 y3llow left 01:53 y3llow joined 01:54 Chillance left 01:55 _jaldhar joined, whiteknight left, y3llow left 01:56 y3llow joined 01:58 y3llow left 01:59 y3llow joined 02:02 y3llow left, y3llow joined 02:05 y3llow left, y3llow joined 02:08 raiph left, y3llow left, y3llow joined 02:11 y3llow left, y3llow joined 02:16 orafu left 02:18 orafu joined 02:20 fgomez left 02:28 xinming joined 02:30 tokuhirom joined 02:31 xinming_ left 02:33 tyatpi joined 02:38 raiph joined, fgomez joined 02:40 cognominal___ joined 02:44 cognominal_ left
raiph phenny tell lichtkind FYI: open source git-backed-wiki written by/for github. Supports Markdown | ReST | etc. More details at github.com/features/projects/wikis. 02:47
phenny raiph: 08 Apr 13:18Z <lichtkind> tell raiph the topics of tablets 2 can be made in the same wash :)
raiph phenny: tell lichtkind FYI: open source git-backed-wiki written by/for github. Supports Markdown | ReST | etc. More details at github.com/features/projects/wikis. 02:50
phenny raiph: I'll pass that on when lichtkind is around.
03:03 tyatpi left
mucker p6: my $long…list = 5; 03:05
p6eval rakudo 3bd91f: OUTPUT«===SORRY!===␤Confused␤at /tmp/JzHA9q6vzb:1␤»
..pugs: OUTPUT«*** ␤ Unexpected "\8230"␤ expecting word character, "?", "!", trait, "=", infix assignment, term postfix or operator␤ at /tmp/6JbKzV1R3B line 1, column 9␤»
..niecza v15-6-gefda208: OUTPUT«===SORRY!===␤␤Confused at /tmp/9Os0KqFOZb line 1:␤------> my $long⏏…list = 5;␤␤Parse failed␤␤»
mucker p6: say "∴ foo" if 1; 03:08
p6eval pugs, rakudo 3bd91f, niecza v15-6-gefda208: OUTPUT«∴ foo␤»
sorear mucker: you can only use alpanumeric characters in names, with a small number of exceptions 03:11
mucker p6: my $αβ = "5";
p6eval niecza v15-6-gefda208: OUTPUT«Potential difficulties:␤ $αβ is declared but not used at /tmp/fUGmtIuDEb line 1:␤------> my ⏏$αβ = "5";␤␤»
..pugs, rakudo 3bd91f: ( no output )
mucker by alphanumeric do you mean alphabet in any language and numeric in any language ? 03:12
or just [A-Za-z0-0]
geekosaur as defined by the Unicode spec 03:13
(your Greek example worked, you got a defined-but-not-used warning) 03:14
mucker gr8 .. just wanted to confirm with the blead versions
is niecza an experiment ? or is it a serious attempt at perl6 on clr ? 03:16
geekosaur "yes" (I think it started as the former and is now the latter but don't quote me) 03:17
sorear that's about where I'd put it too
if you told me when I started that I'd be working on this thing for two years I'd have laughed
mucker awesome :) two competing implementations
geekosaur ...ohai
sorear ? 03:18
geekosaur three if you count pugs, then there's perlito as well.
benabik Two and two halves?
geekosaur that's a major part of the intent of perl6, that there isn't a single implmentation
03:19 PacoAir left
mucker reading the unicode spec (and I thought only lawyers would wade through 1000 pages of jargon) 03:23
03:23 overrosy_ joined
benabik mucker: If we don't hear from you, we'll send out the search dogs. ;-) 03:23
03:23 cxreg2 joined
sorear Better 1000 pages of jargon than the alternative... 03:24
03:24 yeltzooo7 joined
sorear most language specs are that long or longer; there's a reason we call people "language lawyers" 03:24
mucker tried reading common lisp spec in one sitting. had hallucinations the next day. 03:25
03:26 yeltzooo left, BinGOs left, felipe left, overrosy left, cxreg left, pernatiy_ left
sorear You should have read R5RS instead 03:26
03:26 pernatiy joined
sorear It's the odd one, only around 50 pages 03:26
03:27 buubot_backup left
mucker I Plan on reading the r7. Should be out soon me hopes. 03:27
sorear I hear they abandoned the shortness that made r5 so awesome 03:28
benabik Wasn't R6RS not worthwhile?
mucker r6 was huuge. they splan on making r7 in the spirit of r5 03:29
But no matter what the spec, every fucking language is a FAIL at building a CPAN.
03:30 FACEFOX left 03:31 FACEFOX joined, IngisKahn left 03:32 buubot_backup joined
mucker the horror ! the horror ! 03:38
03:40 raiph left 03:43 BinGOs joined 03:47 kranius left 03:59 thelazydeveloper left, preflex left 04:01 preflex joined 04:31 birdwind1pbird joined 04:44 mucker left 04:45 birdwind1pbird left 04:59 IngisKahn joined 05:04 araujo left, raiph joined
raiph phenny: tell lichtkind for example, something along the lines of github.com/raiph/tablets/wiki 05:05
phenny raiph: I'll pass that on when lichtkind is around.
moritz \o 05:07
05:15 spaceships left
raiph moritz: hi and goodnight 05:19
05:20 raiph left 05:30 zhutingting joined 05:51 cognominal___ left 05:52 cognominal joined 06:00 oiig_ left, oiig_ joined 06:04 bacek left 06:13 fgomez left, sisar joined 06:17 bacek joined 06:30 zhutingting left 06:39 tokuhirom left, tokuhirom joined 06:42 lestrrat left 06:43 lestrrat joined 06:44 tokuhirom left 07:01 rsimoes left, rsimoes joined 07:13 orafu left, orafu joined 07:16 BooK__ is now known as BooK 08:01 mucker joined 08:09 answer_42 joined
tadzik o/ 08:13
08:19 answer_42 left 08:23 thou left 08:24 grondilu joined
mucker \o 08:26
grondilu perl6: my (Int $a, Int $b); 08:34
p6eval pugs, rakudo 3bd91f: ( no output )
..niecza v15-6-gefda208: OUTPUT«Potential difficulties:␤ $a is declared but not used at /tmp/9PTOqdZrh2 line 1:␤------> my ⏏(Int $a, Int $b);␤ $b is declared but not used at /tmp/9PTOqdZrh2 line 1:␤------> my ⏏(Int $a, Int $b);␤␤»
grondilu How can I define a bunch of variables of the same type, without repeating the type?
mucker use an array ? 08:36
sorear p6: my Int ($a, $b) #OK
p6eval pugs, rakudo 3bd91f, niecza v15-6-gefda208: ( no output )
sorear looks like that is at least valid syntax
tadzik p6: my Int ($a, $b); $b = "foo"; 08:37
p6eval pugs: ( no output )
..rakudo 3bd91f: OUTPUT«Type check failed in assignment to '$b'; expected 'Int' but got 'Str'␤ in block <anon> at /tmp/mQXHoO3AJ9:1␤␤»
..niecza v15-6-gefda208: OUTPUT«Potential difficulties:␤ $a is declared but not used at /tmp/r5gr4kRIPR line 1:␤------> my Int ⏏($a, $b); $b = "foo";␤␤»
08:38 alvis``` joined 08:39 alvis`` left
grondilu sorear: indeed. Silly of me not to have tried that. 08:54
08:59 birdwindupbird joined 09:05 zhutingting joined 09:11 fsergot joined
fsergot Hi #perl6 o/ 09:11
tadzik hello hello 09:14
sorear o/ 09:17
09:40 odoacre left 09:41 odoacre joined
fsergot nom: say (my $s = "abc"); 09:56
p6eval rakudo 3bd91f: OUTPUT«abc␤»
10:10 grondilu left 10:11 Patterner left 10:14 jaffa4 joined, sisar left, overrosy_ left 10:15 Psyche^ joined, Psyche^ is now known as Patterner, overrosy joined 10:26 zhutingting left 10:27 zhutingting joined 10:30 araujo joined, araujo left, araujo joined 10:31 whiteknight joined 10:33 cognominal left 10:35 cognominal joined 10:36 baest_ is now known as baest 11:00 JimmyZ joined 11:25 bluescreen10 joined 11:32 NamelessTee joined 11:36 JimmyZ left 11:52 PacoAir joined
mucker is there a way to to have non-alphnumeric character in a variable name by some escape mechanism like using a \ 12:01
before the uni character ?
12:01 answer_42 joined 12:04 answer_42 left 12:09 Chillance joined 12:11 leprevost joined 12:12 baest left 12:23 spider-mario joined 12:25 bluescreen10 left 12:28 thelazydeveloper joined 12:33 _jaldhar left
masak_ mucker: I think you can always quote variable names and thus put anything you want in them. 12:35
(hi, #perl6)
arnsholt mucker: Don't think so. But identifiers can contain more than [a-zA-Z_0-9]
masak_ mucker: then there's the apostrophe-dash exception.
r: say my $foo'bar-akbar = 42 12:36
p6eval rakudo 3bd91f: OUTPUT«42␤»
12:36 mdupont joined
arnsholt r: use NativeCall; class Foo is repr('CStruct') { has Mu $.obj; }; say('alive'); 12:39
p6eval rakudo 3bd91f: OUTPUT«===SORRY!===␤Could not find NativeCall in any of: lib, /home/p6eval/.perl6/lib, /home/p6eval/nom-inst2/lib/parrot/4.2.0-devel/languages/perl6/lib, .␤»
masak_ std: my Int ($a, $b, Str ($c, $d)) #OK 12:40
p6eval std 1ad3292: OUTPUT«ok 00:00 42m␤»
fglock I think it possible to add random unicode using string interpolation in variable names with the $::("xyz") syntax 12:41
masak_ there's an interesting tension between "wow, declaration syntax and signature syntax are unified. that's pretty neat" and "g'ah, this doesn't apply at all! what were we thinking?"
fglock: yeah, that's what I was groping for above. 12:42
12:45 masak_ is now known as masak
arnsholt Well, that was odd 12:49
Compiler internals hacking is always interesting =) 12:50
13:00 birdwindupbird left
arnsholt jnthn: *prod*? 13:04
13:06 JimmyZ joined
jnthn arnsholt: Did you mean "star:" ? 13:13
phenny jnthn: 00:31Z <japhb_> tell jnthn In the if block at github.com/rakudo/rakudo/commit/4d...6b#L0R1217 , it looks like 'if @name' is a redundant test in the second if.
arnsholt jnthn: Indeed 13:15
But I've come across something else
13:15 Shozan left 13:16 Shozan joined
arnsholt Should REPR_name_to_id be available from a REPR? 13:16
jnthn arnsholt: May be challening to call from a dynamic REPR.
arnsholt: You have the name available in REPR(obj)->name though, iirc. 13:17
arnsholt It is. I get a run-time error from the linker when the code tries to call it =)
jnthn Yeah, not surprised
13:17 Shozan is now known as SHODAN
jnthn That's why certain things are passed in explicitly to dynamic representations. 13:18
arnsholt Right
jnthn What do you need the ID for, ooc?
arnsholt Checking for the REPR of a member in the CStruct REPR
To handle CPointer/CArray appropriately
jnthn You could always compare the name. 13:19
arnsholt Troo
masak arnsholt: what're you building? 13:20
jnthn Zavolaj improvements, it seems :D
arnsholt I'd like to be able to embed CPointer/CArray in CStruct 13:21
jnthn Embedding a pointer to a CStruct in a CStruct shouldn't be hard either if you're doing those bits :)
.oO( I heard you liked CStrcuts... )
13:21 baest joined
arnsholt Oooh, that's an idea as well 13:22
Possibly the best place to start, come to think of it
jnthn They're relatively similar.
Pointers are the easiest.
arnsholt 'k 13:23
jnthn Since you need never keep any "shadow references" around.
To be clearer
my $x = SomeCStruct.new();
my $y = SomeCStruct.new();
$x.lemonade = $y;
Here we have a choice 13:24
The memory in this case is managed.
arnsholt Right
jnthn That is, if we don't have the GC trace the reference to $y through $x, we'll end up with the underlying struct being freed
arnsholt Yup 13:25
jnthn But the O(1) marshalling means that what we store inside the CStruct of $x (in the bit of the memory we pass off to C land) has to be the pointer to *inside* of $y
So we have to keep $y around too
arnsholt Definitely
jnthn Pointers lack this complexity
Because they just...point somewhere 13:26
There's no memory being managed.
arnsholt So, $y has to go in child_objs, and $y.actual_c_struct into $x's struct layout, no?
jnthn Right 13:27
Something like that.
arnsholt And similarly for CArray and CPointer
jnthn CArray, certainly
CPointer - good question.
I guess if we want to be able to write stuff like 13:28
my $x = SomeCStruct.new();
my $p = pointer-to($x);
Then yeah, we need to consider it having a child_obj too
I've only really handled CPointer in the case it's retunred from a library and thus we aren't managing the memory of the thing it points to. 13:29
arnsholt++ # digging in to this fun :)
arnsholt Yeah, this is a layer or two of abstractions below what I usually do, so it's quite enlightening =) 13:30
(Interestingly, this can be tied in with the webby trend of what people want to do for the Oslo hackathon)
Since I'd like to talk to Mongrel2, which works over ZMQ 13:32
jnthn ooh :) 13:33
arnsholt The idea was to abstract away the whole deal of talking HTTP yourself 13:34
So that a Perl 6 webapp can focus on logic rather than HTTP and such
But that requires being able to talk to Mongrel2 in the first place 13:35
Thus, Zavolaj hacking
jnthn Happy yak is happy. 13:36
.oO( I'm naming this yak "Shawn" )
arnsholt Hmm. Maybe I'll rename my local branch shawn =D 13:37
masak .oO( And this over here I'll dub "Shorn" ) 13:38
felher moritz: ICMP ECHO REQUEST 13:39
tadzik hugme: add sergot to ecosystem 13:42
hugme tadzik: sorry, I don't know anything about project 'ecosystem'
tadzik hrm
fsergot What is the ecosystem? :) 13:43
tadzik github.com/perl6/ecosystem
list of our modules lives here
masak popularized the term
tadzik I like it 13:44
masak the idea being that yes, we find bugs directly in the compilers, but many bug reports actually flow in from the set of community modules.
jnthn masak things about everything in terms of bug reports :P
*thinks
masak the more we can *harness* that and do things like regularly smoke the ecosystem, the sooner we'll fix those bugs.
moritz hugme: add sergot to perl6 13:45
hugme hugs sergot. Welcome to the perl6 github organization
moritz felher: ICMP ECHO RESPONSE
masak jnthn: bug reports are merely the digital reification of community love :) 13:46
fsergot \o/ !
13:47 zhutingting left
felher moritz: gist.github.com/2384508 | if you don't have any objections, you use/apply the Exception.Bool patch. I can push the spectest-patch myself i think, since i have a perl6 commit bit :) 13:48
moritz: i did do make spectest after the patch and the only things it breaks are the things that i patched in the spectest-patch 13:50
dalek kudo/nom: 1fa8fb5 | (Felix Herrmann)++ | src/core/Exception.pm:
Make Exception.Bool return True

Signed-off-by: Moritz Lenz [email@hidden.address]
moritz felher: no objections 13:51
felher \o/ :)
dalek ast: 0eba93a | (Felix Herrmann)++ | S (4 files):
rewrite tests that rely on Exception to return False on .Bool
13:53
13:58 felipe joined 13:59 tokuhirom joined
masak the part of dalek that finds the lowest common denominator between changed paths shouldn't split things in the middle of a path fragment. 13:59
moritz 14:00
masak Ronja, that you? 14:01
tadzik :) 14:02
moritz it was her, yes 14:05
masak funny, it's what I imagine a young child would say, too. 14:06
".ä!"
moritz .u � 14:07
phenny U+FFFD REPLACEMENT CHARACTER (�)
jnthn If anybody wants to spectest the name-cleanup branch and confirm it looks clean, I may as well merge it. I can do the changes beyond that fairly undisruptively in nom.
dalek ast: 4d5e023 | jnthn++ | S02-names-vars/variables-and-packages.t:
Remove test for removed SUPER and fix tests that check you can't declare the pseudo-packages.
14:09
arnsholt Grumblegrumble
jnthn: NQP and Rakudo are too thorough in making sure the versions match up =) 14:10
mucker probably a silly thing, but can I send eval to a perl6 bot in private ?
masak yes, you can.
jnthn mucker: /msg p6eval r: say "yay"
arnsholt: ;) 14:11
mucker (:
jnthn Actually testing the branch will be a pain because we'll get confused with the exceptions fixes that just went in, whihc it doesn't have.
moritz jnthn: I'll first run a spectest with felher++'s changes, then with name-cleanup
jnthn I'll just merge it and hope I didn't miss anything :)
moritz (I planned to merge the branch locally before spectesting
jnthn moritz: Am running that build/test here 14:13
moritz: I just reviewed the commits...there really should be nothing platform specific in there. 14:14
14:16 lichtkind joined
lichtkind moritz: ok gitit failed it broke syntax almost like socialtext, i will go on with html 14:17
phenny lichtkind: 02:50Z <raiph> tell lichtkind FYI: open source git-backed-wiki written by/for github. Supports Markdown | ReST | etc. More details at github.com/features/projects/wikis.
lichtkind: 05:05Z <raiph> tell lichtkind for example, something along the lines of github.com/raiph/tablets/wiki
lichtkind k will test gollum too 14:18
mucker p6: @::("long…list") = 1,2,3; print @::("long…list"); #following fglocks advice ! n works fine
p6eval rakudo 3bd91f: OUTPUT«Symbol '@::long…list' not found␤ in method <anon> at src/gen/CORE.setting:8229␤ in <anon> at src/gen/Metamodel.pm:2256␤ in any find_method_fallback at src/gen/Metamodel.pm:2254␤ in any find_method at src/gen/Metamodel.pm:828␤ in block <anon> at /tmp/VqeLeBgU7a… 14:19
..niecza v15-6-gefda208: OUTPUT«123»
..pugs: OUTPUT«*** Undeclared variable: ("@long\226\128\166list",MkPad (padToList [("$_",PELexical {pe_type = (mkType "Scalar"), pe_proto = <Scalar:0xf6ed9505>, pe_flags = MkEntryFlags {ef_isContext = True}, pe_store = <ref:0xf6ede2a5>}),("@_",PELexical {pe_type = (mkType "Array"), pe_pro…
mucker … is unicode for ...
mucker also going to install n on my sys 14:20
masak I think it's just NYI in Rakudo.
interesting quip in an interesting thread. news.ycombinator.com/item?id=3840145 -- discuss. 14:25
moritz if Eliezer succeeds, then yes :-) 14:26
jnthn wonders why it works on n.
I mean, looks like an undeclared variable to me. 14:27
masak I sure hope that Elezier won't have to do *all* the heavy lifting himself :)
moritz r: my $a::('!!'); say $a::('!!)
p6eval rakudo 3bd91f: OUTPUT«===SORRY!===␤Cannot declare a variable by indirect name (use a hash instead?)␤at /tmp/IZAN79oGtz:1␤»
tadzik p6: my %a = any('a', 'b') => 3; say %a<b>; say %a<a> 14:28
p6eval niecza v15-6-gefda208: OUTPUT«Unhandled exception: Unmatched key in Hash.LISTSTORE␤ at /tmp/fQcWzhuQiS line 0 (mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3842 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3843 (module-CORE @ 65) ␤ at /home/p6eval/niecza…
..pugs: OUTPUT«␤␤»
..rakudo 3bd91f: OUTPUT«3␤Any()␤»
moritz anyway, that can't work in the general case, only when the expression is constant-foldable
tadzik nah, almost :)
14:28 xinming_ joined
moritz because lexpads are immutable at run time 14:29
mucker n: my $a::('!!'); say $a::('!!')
p6eval niecza v15-6-gefda208: OUTPUT«Use of uninitialized value in string context␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1222 (warn @ 3) ␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 230 (Mu.Str @ 10) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/src/Ni…
mucker n: $a::('!!'); say $a::('!!')
p6eval niecza v15-6-gefda208: OUTPUT«Any()␤»
mucker n: $a::('!!')=5; say $a::('!!')
p6eval niecza v15-6-gefda208: OUTPUT«5␤»
mucker "Use of uninitialized value in string context␤" what does this error mean ?? 14:30
i get this even when I init the value
n: $my a::('!!')=5; say $a::('!!'); #but only when I use my at the beg 14:31
p6eval niecza v15-6-gefda208: OUTPUT«===SORRY!===␤␤Variable $my is not predeclared at /tmp/RFG9yCErdc line 1:␤------> <BOL>⏏$my a::('!!')=5; say $a::('!!'); #but on␤␤Two terms in a row at /tmp/RFG9yCErdc line 1:␤------> $my ⏏a::('!!')=5; say …
mucker n: my a::('!!')=5; say $a::('!!'); #but only when I use my at the beg
p6eval niecza v15-6-gefda208: OUTPUT«===SORRY!===␤␤GLOBAL::a does not name any package at /tmp/Hh8TnOqoKe line 1:␤------> my a::('!!')⏏=5; say $a::('!!'); #but only when I use␤␤A type must be provided at /tmp/Hh8TnOqoKe line 1:␤------> my a::('!!')[33…
mucker n: my $a::('!!')=5; say $a::('!!');
p6eval niecza v15-6-gefda208: OUTPUT«Use of uninitialized value in string context␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1222 (warn @ 3) ␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 230 (Mu.Str @ 10) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /home/p6eval/niecza/src/Ni…
moritz it might be related to niecza implemting 'no strict', and assuming that it's a package variable or so
mucker oh ok. 14:32
14:32 xinming left
mucker sorry for clobbering the screen 14:32
mucker still a noob with irc
JimmyZ good evening, #perl6 14:33
masak JimmyZ! \o/ 问候! 14:36
JimmyZ 麦高: 下午好!
masak 下午好,下午好. 14:37
JimmyZ: 你住在哪里?广州?我忘了。 14:38
JimmyZ masak: 我在深圳
masak 啊,对。
14:38 kranius joined
masak 我想起了全省正确 :) 14:39
14:55 JimmyZ_ joined 14:57 JimmyZ left 14:58 JimmyZ_ is now known as JimmyZ
dalek Heuristic branch merge: pushed 16 commits to rakudo/nom by jnthn 15:01
jnthn moritz: In a recent patch: ".HOW.WHAT is the same as just .HOW. Simplify" 15:03
moritz: That's not actaully the case.
moritz: In this case you get away with it though.
moritz: But the meta-object will usually be an instance. 15:04
felher spectests that merge ooc.
dalek kudo/nom: ac083c5 | jnthn++ | docs/ChangeLog:
Fill out ChangeLog a little more.
15:09
jnthn Mostly, the branch started cleaning up name handling 15:10
Of note, not re-parsing longname in a bunch of places
15:14 lestrrat left 15:15 lestrrat joined
felher jnthn/moritz: i got two fails: "t/spec/S32-exceptions/misc.rakudo ......................... Failed 2/282 subtests", namely "not ok 216 - right exception type (X::Augment::NoSuchType)" and "not ok 67 - right exception type (X::Redeclaration)" 15:27
15:41 vlixes joined
TimToady n: my \term:<☻> = 42; say ☻ 15:44
p6eval niecza v15-6-gefda208: OUTPUT«42␤»
TimToady mucker: ^^^
mucker (: 15:46
15:47 alvis``` left
jnthn TimToady: In S02: 15:48
MY # Symbols in the current lexical scope (aka $?SCOPE)
15:48 Psyche^ joined
jnthn TimToady: Just to clarify, does this mean if you do MY::<$x> it means anything *visible* from that scope? 15:48
TimToady: As opposed to anything declared precisely in that scope? 15:49
On a similar note, what about MY::.keys?
15:50 tokuhirom left
TimToady std: my $good; say MY::<$good>; my $bad; { say MY::<$bad> } 15:50
p6eval std 1ad3292: OUTPUT«===SORRY!===␤Variable MY::<$bad> is not predeclared at /tmp/Wi8jY37TKR line 1:␤------> $good; say MY::<$good>; my $bad; { say ⏏MY::<$bad> }␤Check failed␤FAILED 00:00 42m␤»
jnthn I assume that OUR:: really *is* just $?PACKAGE and no more magical than that?
TimToady ^^
jnthn n: my $good; say MY::<$good>; my $bad; { say MY::<$bad> } 15:51
p6eval niecza v15-6-gefda208: OUTPUT«Potential difficulties:␤ $good is declared but not used at /tmp/HXuYtn5iOc line 1:␤------> my ⏏$good; say MY::<$good>; my $bad; { say M␤ $bad is declared but not used at /tmp/HXuYtn5iOc line 1:␤------> my $good; say MY::<$good>; my…
jnthn n: my $good; say MY::<$good>; my $bad; { say MY::<$bad> } #OK
p6eval niecza v15-6-gefda208: OUTPUT«Any()␤Any()␤»
jnthn n: my $good = 1; say MY::<$good>; my $bad = 2; { say MY::<$bad> } #OK
p6eval niecza v15-6-gefda208: OUTPUT«1␤2␤»
jnthn So, it finding 2 is wrong here?
TimToady interesting divergence from STD there...
15:51 Patterner left, Psyche^ is now known as Patterner
jnthn TimToady: Yeah, I'm just reviewing pseudo.t and the spec before implementing various of the pseudo-packages. 15:52
TimToady: My question applies to OUTER and CALLER too
Well, and UNIT and SETTING :)
I know DYNAMIC has to be different and go scanning. As does ::<$x>
15:53 fgomez joined
jnthn std: my $good = 1; say MY::<$good>; { say OUTER::<$good> } 15:53
p6eval std 1ad3292: OUTPUT«ok 00:00 42m␤»
jnthn std: my $bad; { my $good; say MY::<$good>; { say OUTER::<$good>; say OUTER::<$bad>; } 15:54
p6eval std 1ad3292: OUTPUT«===SORRY!===␤Variable OUTER::<$bad> is not predeclared at /tmp/TsZLX9aMEf line 1:␤------> MY::<$good>; { say OUTER::<$good>; say ⏏OUTER::<$bad>; }␤Unable to parse block at /tmp/TsZLX9aMEf line 1:␤------> my $bad; ⏏…
jnthn Guess that's my answer.
For OUTER at least.
TimToady well, we'll have to negoitiate MY with sorear++ I guess
*got
jnthn n: my $x; { my $y; say MY::.keys } # curious 15:55
p6eval niecza v15-6-gefda208: OUTPUT«Potential difficulties:␤ $y is declared but not used at /tmp/s2Rfs0SdX7 line 1:␤------> my $x; { my ⏏$y; say MY::.keys } # curious␤ $x is declared but not used at /tmp/s2Rfs0SdX7 line 1:␤------> my ⏏$x; { my $y; say MY::.…
jnthn n: my $x; { my $y; say MY::.keys } # curious #OK
p6eval niecza v15-6-gefda208: OUTPUT«0␤»
jnthn Hm.
I worry if we give it the Niecza meaning then .keys will be *huge* :)
TimToady nodnod 15:56
jnthn And useless for walking the pads.
mucker what is the p6 equivalent of *foo{NAME} (p5) ? 15:57
15:59 JimmyZ_ joined 16:00 JimmyZ left 16:01 JimmyZ joined, fgomez left
mucker TimToady : is \term an experimental extension in n ?? 16:01
jnthn mucker: It's spec, though a relatively recent addition. Rakudo didn't get around to adding it yet. 16:02
It's not hard.
Just nobody got a round tuit.
mucker It's a neat feature :) no other lang has it me thinks 16:03
16:04 JimmyZ_ left
TimToady the hard bit is generating a new lexer/parser at that point 16:04
re *foo{NAME} Perl 6 doesn't have typeglobs, so there's no equivalent 16:05
16:15 geistteufel joined
TimToady actually, we know how to generate a new language at that point; what we don't know yet is how best to avoid regenerating lexers that are trivially different from the previous when that will turn out to be unnecessary 16:16
or how to avoid regenerating lexers that are identical to the one we generated last time we compiled this program
either of those might speed up a Perl 6 parser greatly
cognominal geistteufel, hello to the devil 16:17
geistteufel hi from hell :) 16:18
TimToady programming in PHP, are you? :)
masak or maintaining a major VB.NET code base... 16:19
16:20 plobsing left
tadzik guys, guys, there are ladies in the room :/ 16:20
TimToady sorry
masak we don't talk about lexer generation in front of ladies now? 16:22
tadzik don't we?
masak well, what bit of the conversation above was it that you :/-ed on? 16:23
tadzik but, well, VB.NET? There are limits, y'know
masak oh, I see now.
yeah, sorry.
tadzik concrete joke I guess ;)
masak .oO( if the joke was concrete, that explains why I got it so slowly... ) 16:24
dalek osystem: 830a24d | tadzik++ | META.list:
Add Text::T9
TimToady
.oO(went over like a concrete balloon...)
TimToady wonders how big a concrete balloon would have to be before the volume effects overcome the surface effects... 16:25
I suppose it depends on how thinly you can pour concrete and still call it concrete 16:26
"yeah, it's concrete, but it's reinforced with fiberglass and epoxy"
masak that's legit, in my book. 16:27
because you usually do that.
TimToady well, in the abstract, anyway 16:28
most concrete is not reinforced that way concretely
masak :) 16:29
TimToady
.oO(masak's new hobby: trying to trick people into autopunishment)
16:29 xdbr left 16:30 xdbr joined
masak most funny things I've seen in the past few days have turned out to be non-auto puns on closer inspection. 16:30
this one, fr'ex: twitter.com/sharonGOONer/status/190...0570595328 16:31
TimToady that's why I hedged it with "trying to" :P
well, not unless the *facepsalm* is also a reaction to making a bad pun 16:33
which I can read it as
just because you can argue yourself out of an autopun doesn't make it right :) 16:34
16:38 spider-mario left
TimToady and anything as far into the realm of pragmatics as *facepalm* will be subject to more than just semantic analysis; it is not enough that you did a facepalm, but that you must be seen doing a facepalm by the reader, with the explicit intent of producing an almost whiplash reaction in the reader 16:38
and anything in the realm of pragmatics takes its meaning from wherever the hearer can find it in the context 16:40
16:40 cognominal left
TimToady but this is highly variable from hearer to hearer 16:40
swearwords like PHP and VB.NET work the same way 16:41
16:41 cognominal joined
arnsholt There. The basics of my code seems to be in place 16:42
tadzik my $perl = ':('; 16:44
Signature.pm is sad
jnthn heh :)
I think I wrote that line, then was like, "lol" 16:45
tadzik .WHY _must_ use this for subs
tadzik dives in
masak TimToady: what was interesting to me wasn't that tadzik++ implied that PHP and VB.NET were swearwords, which is easy to grok, but that he did it by invoking 'dudes, not in front of the ladies', which is an antiquated meme in Sweden. 16:49
TimToady it's antiquated here too 16:51
for some definition of "here" that may not include the midwest :)
masak heh.
masak imagines a torus-shaped US
I've also noted that the concentration of blonde jokes is slightly higher in Poland. 16:52
TimToady
.oO(Missouri, the Doughnut Hole State)
16:53
or maybe that's Iowa
16:54 JimmyZ_ joined
TimToady the usual joke here is that the midwest is "Flyover Country" 16:54
16:54 JimmyZ_ left
TimToady funny how far the midwest is from the mideast though... 16:55
arnsholt jnthn: What roughly is the purpose of initialize_slots?
jnthn arnsholt: Context?
arnsholt Durr, sorry. In CStruct.[ch] 16:56
16:56 JimmyZ left
arnsholt CStructREPRData has an int array called initialize_slots, whose purpose I'm not quite able to figure out 16:57
jnthn /* Slots holding flattened objects that need another REPR to initialize
* them; terminated with -1. */
INTVAL *initialize_slots;
That one
arnsholt: See the initiailize REPR function 16:58
arnsholt Yah
jnthn arnsholt: initialize is basically "set up this memory like it needs to be if we're making a fresh object"
16:59 thou joined
jnthn arnsholt: It's the way that we end up with integer slots defualting to 0, num slots defaulting to NaN, etc. 16:59
arnsholt: In P6opaque if you inline something like a bigint that needs some fancier initialization, it takes care of doing that.
arnsholt Right, right.
jnthn arnsholt: Basically, it's just delegation to flattened in REPRs.
It only applies to things you flatten in. But you are working with reference types.
arnsholt Cool. So I'll probably need another similar array to set up the child objects 17:00
jnthn No
Oh
Yeah.
:)
Sorry, misunderstood at first :)
Note that I already prepared some of the way though
/* GC-marked objects that our C structure points into. */ 17:01
PMC **child_objs;
arnsholt Yeah, my code allocates that memory now
jnthn But that is currenlty unused.
oh, cool :)
17:03 isBEKaml joined
masak isBEKaml! \o/ 17:04
isBEKaml hola, masak! Hi, #perl6! :)
masak good, we're all here. today's agenda: taking over the world. 17:05
moritz start by reading halfhalf.posterous.com/dont-work-be...ve-someone :-)
17:05 cxreg2 is now known as cxreg
isBEKaml masak: did something just happen? :D 17:06
moritz: oh, I was just reading a blog post from adzerk this afternoon, similar things said (the title was a bit of a bait, though) 17:08
arnsholt jnthn: Is where a given child should be placed what you intended to use the CSTRUCT_ATTR_* #defines for? 17:09
jnthn arnsholt: yeah 17:10
Well, if by "where it should be placed" you mean "where the shadow copy goes"
arnsholt: IN_STRUCT just means it's a value we return directly.
masak moritz: nice one. I got my current job from play -- ultimately from getting to know jnthn++ here on #perl6. 17:11
isBEKaml: you must have missed many meetings, sir. we've been trying to take over the world for years :P
isBEKaml masak: Yes, I hope I'm not too late to the party! 17:12
moritz: Yes, that was a good read (resonated with me as much as Job's SHSF talk). 17:13
geistteufel TimToady: well, I'm just playing with my nickname. I'm a perl5 developer 17:14
I let PHP behind several years behind
isBEKaml moritz: erm, SHSF -- Stay Hungry, Stay Foolish.
masak isBEKaml: you're not too late to the party. just find a passion and go with the flow. 17:16
TimToady I'd categorize perl5 programming as more like purgatory... :)
17:16 GlitchMr joined
isBEKaml TimToady: If I understand the word purgatory, that's weird coming from you. :) 17:17
masak not really.
TimToady has been biased towards Perl 6 for several years now. :)
TimToady purgatory: seems real nice when you first come from hell... 17:18
arnsholt jnthn: Quite, quite. And "return directly" is the flatteing we have ATM, right?
isBEKaml masak: I know, maybe I'm just surprised he used that word. :)
jnthn arnsholt: Yeah.
masak isBEKaml: "people, please! there are atheists around!" :P 17:19
isBEKaml *lol*
arnsholt That reminds me, I need to finish the last half of Paradiso =)
Maybe I'll get around to that roughly around the same time as Perl 6.0.0
That'd be an excellent conincidence ^_^ 17:20
TimToady we're trying to finish the last half of Paradisio too, as it were
or maybe it's the third 80% by now
cognominal masak: and ignostic (note the initial) too because atheism is a very intolerant sect 17:21
*initial i
cognominal is probably a gnostic ignostic. The two sides don't come from the same part of thebrain. 17:22
TimToady well, a lot of folks are antignostic: they don't wanna know :)
masak cognominal: ignosticism, not to be confused with apatheism :P 17:23
cognominal :)
isBEKaml masak: There's something about #perl6 that keeps bringing all that fun back... :)
masak isBEKaml: I know!
TimToady I care!
cognominal the feat is doing it without coming immediately to insults. 17:24
TimToady ah, the French, always worried about insults...
in what other language can you insult someone by saying "I insult you!"
17:24 fgomez joined
isBEKaml smooth insults? o_O 17:25
masak oh, I just realize that that old joke is an autopun: "Is it ignorance or apathy? I don't know, and I don't care."
realized*
it's something as unusual as a *tandem* autopun! 17:26
arnsholt idly wonders why there's an i in ignorance
TimToady it's more ignoble that way
masak arnsholt: er, it negates "gnosis"?
as in "not knowing".
TimToady probably derives linguistically from the in- negator 17:27
isBEKaml masak: shouldn't that be agnorance? I don't care about "gnosis"!
17:27 awwaiid left
arnsholt Well, yeah. But the greek negative prefix is a(n) 17:27
cognominal masak: "er", as german inchative?
TimToady inconceivable!
isBEKaml ergnorance!
masak isBEKaml: shouldn't be "irrogance"? I don't know about "rogance"!
arnsholt But yeah, probably what TimToady suggests. Greek root with Latin negation 17:28
masak er, shouldn't that be "ipathy"? I don't know about "pathy"!
isBEKaml masak: Bah! It's all greek and latin to me!
benabik Reminds me of a T-shirt: "Polyamory is wrong! It's either multiamory or or polyphilia. But mixing Greek and Latin roots? Wrong!"
masak hah!
well, "hexadecimal" is just as wrong.
17:28 bluescreen10 joined 17:29 wolfman2000 left
isBEKaml so is saying "runny nose" and "smelly feet"! 17:29
back to engreesh, forget g and l.
TimToady and don't be "hypersensitive" 17:30
masak language is fundamentally broken. it's causes side effects in people's brains. not FP at all.
ideally, we should do language in a monad.
arnsholt Oxford Latin Dictionary to the rescue. It is indeed in-gnotus, from Greek agnostos
benabik OED++ 17:31
isBEKaml OLD++
arnsholt benabik: I need to get that shirt! =D
TimToady actually, I have no problem with mixing roots; we borrow productive language bits, not fossils, and there's no rule that says the new language cannot mix and match
benabik isBEKaml: Maybe it should be Oxford++
moritz Oxfnord :-) 17:32
arnsholt TimToady: Yeah. This is where my prescriptivist and descriptivist brains start to disagree =)
TimToady OED is hyperlikeable
not to mention superlikeable
isBEKaml or supralikeable 17:33
arnsholt Hehe
masak moritz: Oxfnord -- that's why I keep forgetting it exists! 17:34
tadzik r: my $a = method (Str $a) { ... }; say $a.perl
p6eval rakudo 3bd91f: OUTPUT«method <anon>(Mu , Str $a, Mu %_!) { ... }␤» 17:35
17:35 awwaiid joined
jnthn tadzik: huh, the $_! is wrong 17:35
tadzik the first and the last parameter are surely correct, yet not really friendly
jnthn tadzik: Should be *%_
tadzik yeah, should be... exactly
moritz and Mu :
jnthn Right.
isBEKaml jnthn: I vaguely recall you saying that back when nom was in the works. That's still in?
TimToady this fossilish view of language works elsewhere too; I predict that other languages will borrow the syntax of Perl 6's meta-operators without borrowing the idea of meta-operators
tadzik I wonder if Pod::To::Text should just use Method.perl for pretty-printing them for docs 17:36
that seems like the most obvious way, but I suppose the output can be shorter and more clear
moritz tadzik: in time, that will include the source code :/
isBEKaml And, I don't see why we need the Mu before Str in the signature..
tadzik oh, then it's a no-go anyway
isBEKaml: the invocant
moritz isBEKaml: that's the invocant marker gone wrong
isBEKaml ah 17:37
tadzik I wonder if I should skip the invocant when printing the method in --doc 17:38
or maybe just look into it and add a 'Class method' or 'Instance method' marker to it
17:46 NamelessTee left
tadzik gist.github.com/2386191 -- this deliberately skips the invocant marker for the sake of readability and no confusion. Thoughts? 17:47
dalek p/cstruct-work: 51a3250 | (Arne Skjærholt)++ | src/6model/reprs/CStruct.c:
Start letting CStruct have CArray members.
17:48
p/cstruct-work: 6700db0 | (Arne Skjærholt)++ | .gitignore:
Ignore OS X bundles.
moritz tadzik: +1
tadzik +1 from my mentor; settled then :) 17:49
arnsholt jnthn: github.com/perl6/nqp/commit/51a325...b940c76aa1 # First commit in branch cstruct-work
benabik method outside of a class is a warning, not an error?
tadzik warning
moritz and only if it's named 17:50
tadzik you can my $a = method(); and then A.^add_method
moritz r: my $x = method () { };
p6eval rakudo 3bd91f: ( no output )
arnsholt Oh, right. dalek++ prints all the commits here as well. I forgot =)
moritz r: my $x = method foo() { };
p6eval rakudo 3bd91f: OUTPUT«Useless declaration of a has-scoped method in mainline␤»
jnthn arnsholt: "The initialize_slots member in the REPRData needs to be more complex" - not sure it does... 17:51
arnsholt: At least, not for the CArray case stuff you're doing
TimToady it's interesting to see 下午 "down noon" for afternoon, when I just learned yesterday that 昃 means both "afternoon" and "decline". Are these both references to the sun starting down from the zenith?
jnthn arnsholt: Since that isn't "flattened in"
TimToady I suppose I should ask that when it isn't in the middle of the night in China 17:52
dalek kudo/nom: 73ce76a | tadzik++ | lib/Pod/To/Text.pm:
[Pod::To::Text] Print subs and methods prettier

This delibarately skips invocant and optional named params in method signature for the sake of readability and the least amount of confusion.
We're not using simply Sub.perl and Method.perl since they may contain the source code one day, and we almost certainly don't need that.
17:53
tadzik this is cool. I can now write "#= get a list of words from @words matching $input" and it formats nicely
moritz \o/
arnsholt jnthn: Right. In that case, initialize() needs to be more clever, since right now it passes the sub-initializer a pointer directly into the cstruct blob
I just couldn't think of a way to do that without a bit more information 17:54
jnthn arnsholt: But a pointer to an array is a reference type, so there's no need to call a sub-initializer. 17:57
arnsholt: It just starts life NULL.
arnsholt: And if it's requested, we know the type object.
arnsholt Oh, right. Good point =) 17:58
masak TimToady: no, there's just a general space/time synesthesia meme in China, saying that time moves downwards, like in a schedule app. 18:06
TimToady: there is interesting research proving that such a synesthesia exists for exactly those groups that have such metaphors in their language. 18:07
s/proving/strongly indicating/
sorear good * #perl6 18:09
jnthn: It succeeds in n because n doesn't enforce use strict for runtime lookups - it goes to GLOBAL 18:10
TimToady though 昃 breaks down into "sun" and "slant", and we can say "the slanting sun"
and I do believe in "Metaphors We Live By", I kinda doubt your synesthesia thesis in this particular case, since time flows uphill for 上午 :) 18:12
*and while
dalek p/cstruct-work: 3bde5a3 | (Arne Skjærholt)++ | src/6model/reprs/CStruct.c:
Remove some TODOs since they weren't actually necessary. jnthn++
sorear masak: My impression of Eliezer is that he is far more concerned with stopping other people from getting there first than actually doing anything himself
masak TimToady: nono, that's to be read as "above noon". 18:13
sorear: really? I hadn't gotten that impression so far. :/
moritz jnthn: fwiw I got the same spectest results as felher. One is probably due to a test stupidity, the other one might be a regression that an error previously thrown with a dedicated type is now just a string 18:14
jnthn moritz: I knew there was something with misc so I ignored it, but I may well have caused a regression.
moritz jnthn: no, it's an actual regression
./perl6 -e 'use MONKEY_TYPING; augment class NoSuchClass { }'
===SORRY!===
get_pmc_keyed_str() not implemented in class 'LongName'
jnthn Oh 18:15
moritz jnthn: I fixed the failure in misc.t that we had for a few weeks
jnthn yeah, I pass soemthing that was a string before that is now an object
moritz d0568017e25387bfe78831eb9e2f5b54e507c391 from yesterday
sorear .go SIAI scary idea
huh, I thought phenny had that feature. 18:16
moritz jnthn: if you fix that one, I'm +1 to merging, and will handle the other test failure in misc.t
though the wrong other test that fails hides a real error
class A { }; subset A of Any;
doesn't die, but should
18:18 eoh| joined, eoh| left, eoh left, eoh joined, eoh left
sorear masak: not 100%, but relevant: multiverseaccordingtoben.blogspot.c...a-and.html 18:18
jnthn moritz: Will take care of it. 18:23
cognominal jnthn, have you looked at gist.github.com/2343861? 18:24
jnthn no 18:25
(looking)
lichtkind jnthn: do you have any clue why slightest change even innocent like my $var = 3; in rakudo made it uncompileable? 18:26
jnthn cognominal: Interesting to see. Unfortunately I suspect it'll slow parsing down a lot if we apply it right now.
cognominal That's could be a good comment
jnthn cognominal: But we want that stuff written in NQP in the end.
cognominal my point is not that it could replace existing code today but that it should compile. 18:27
jnthn cognominal: Well, the QAST work and the NQP::Optimizer that'll come after it should help :)
18:28 lichtkind left
jnthn moritz: fixed locally 18:29
moritz jnthn: \o/
jnthn: I'm working on improved tests too
jnthn cognominal: OK, I don't know why the NQP compiler would be unhappy at first glance.
18:29 lichtkind joined
jnthn ooh, dinner time...I should find something to eat :) 18:30
jnthn has some work towards pseudopackage stuff locally...hopefully land some of those later tonight :)
dalek ast: c9e3611 | moritz++ | S32-exceptions/misc.t:
more robust name clash tests
18:33 leprevost left
geistteufel I wonder perl6 is now in development since several years, and seems completly a new language. this language will be ready for production someday ? in years ? month ? is they a cpan for perl6 or perl6 include already so much stuff that cpan is not really a need actually ? 18:36
masak geistteufel: it's a gradual thing. 18:37
moritz geistteufel: "ready for production" greatly depends on the production. It's already useful for some kinds of production environment
masak geistteufel: I'm using Perl 6 for things that I would consider to be "in production". it's ready for some things, not for others. 18:38
moritz geistteufel: and we're slowly evoliving a module system. module.perl6.org give a short overview
geistteufel I see, but they is not yet interpretor with good speed, right ?
masak geistteufel: the progress is not defined so much by "we need to put this feature in place" but by "we need to make things smaller, faster, and more polished".
geistteufel: try Niecza, it's got surprisingly good runtime speed. try Rakudo, it's got surprisingly good startup time :) 18:39
18:39 birdwindupbird joined
geistteufel actually we can use it with command line tools only. Or solve stuff with math. But not really open for the web. They is a Plack equivalent, CGI ? FCGI ? 18:41
sorear the main thing is that all known Perl 6 implementations have agonizingly slow parsers
We have a rudimentary Plack equivalent
geistteufel and connection to db, like MySQL, PostgreSQL, MongoDB? 18:42
sorear rudiments of those too 18:43
masak yeah. we have stuff, but it's not always up-to-date, or very powerful.
really what we need is more hands to use stuff, and fix things up rather than get frustrated about their current shoddiness :) 18:44
sorear I think a lot of the library stuff hasn't matured because nobody wants to build on such slow interpreters... but this is probably a self-serving delusion, since my research focus is speeding up the parsers ;)
masak some of it is bitrot.
bitrot has a slightly demotivational effect on ecosystem people. :)
sorear wonders if masak is interested in the Goertzel link
geistteufel I have take a look of the coding. It's really impressive. 18:45
The good point with perl5 is that they is 1 interpreter, so every energy is focus on this one. 18:46
masak sorear: I am. haven't read it yet. will react to it when I have.
geistteufel have multiple interpreter spare the effort, so it is slower to get something fast
tadzik 9 women can't birth a child in a month
benabik geistteufel: The downside is that perl5 is defined by the interpreter, so future versions have to be bug-compatable.
moritz geistteufel: that's a common fallacy 18:47
geistteufel: you can only try a single design with a single compiler
benabik (Or perhaps bug-compatable isn't quite it. "Implementation quirk" compatible, perhaps.)
geistteufel but if an interpreter can do stuff, and not the over, it will be like having multiple browser. IE / Mozilla / Opera / Chrome ... a really complain of implementation of the language for developer ?
moritz that's why we have a test suite, and bugreports for missing features 18:48
TimToady every implementation has blind spots, and if there is only one implementation, these blinds spots get set in concrete
to mix a metaphor or two
geistteufel ok so any interpreter will have to follow the last implementation?
moritz no
there's a spec
geistteufel if all interpreter do the same stuff the same way, what are the point to have many one ?
tadzik what's the point of clang if there's gcc? 18:49
moritz geistteufel: doing something isn't the same as doing it well
TimToady there are many ways to do the same thing
geistteufel That"s the modo of perl language :)
cognominal grosso motto 18:50
sorear *motto
geistteufel that's funny. It's really a new way to do. Write spec, let over people to implement the spec. It is more open
moritz well, it's not quite so unidirectional as you make it sound 18:51
TimToady that's more of a waterfall descriptiong than a whirlpool description
nobody is smart enough to spec Perl 6 in advance 18:52
geistteufel TimToady: I'm not english native (like you should have notice), so it's hard for me to understand the metaphore
I see.
TimToady so there has to be feedback from implementations back to spec too
and there is a lot of useful information going side-to-side between the different implementations too 18:53
18:53 fsergot left
geekosaur multiple implementations is key to this as well: it helps distinguish "hard in this implementation" from "hard in general" 18:54
geistteufel ok. I see.
moritz oh, and we are at the point where the implementation can exchange code
TimToady especially the bits written in Perl 6 :) 18:55
moritz because both are partially written in Perl 6
geistteufel So interpreter say "perl6 - spec 20120425 compatible" or something like that
TimToady none of them are yet spec compatible, or even test-suite compatible
geistteufel and so they is compiler; can we compile our perl6 code ?
I create a pm, and it build a pmc or something ? 18:56
TimToady so far we just say "works in niecza 2012-04-14" and such
but we'll converge on the test suite eventually 18:57
geistteufel some of them implement already some part of the spec, over one over part. if we need feature of part impleted in 2 separate compiler, we are bloody annoying ?
moritz the compiler writers know that the spec writer are annoying :-)
TimToady especially that snotty guy that wrote the STD parser :) 18:58
sorear tadzik++ "9 women cannot birth a child in a month"
geistteufel I'm just curious, having multiple compiler, multiple implentation, sound like 'HTML & CSS' interpreter. And It's sometime hell to solve stuff for all interpreter version. That is why I wonder how can that work for perl6. 18:59
sorear geistteufel: "waterfall" means en.wikipedia.org/wiki/Waterfall_model
masak sorear: Brooks takes that further. he says there are O(N**2) costs in making a team bigger.
geistteufel sorear: we can't parallelize everything :)
tadzik sorear: oh, "birth" was gramatically correct in this context?
cognominal TimToady, is your secret projet is to get a language capable to translate the bible in any language?
*project 19:00
masak tadzik: maybe "gestate" covers it better. but I'm no native speaker either :)
sorear tadzik: I think so? I'm not fretting it
tadzik ok then. I wasn't sure :)
TimToady cognominal: no, my not-so-secret project is to translate theological concepts into the society of hackers 19:01
sorear geistteufel: "whirlpool" is a metaphorical contrast to a waterfall, where water moves in every direction, not just one
geistteufel I get it; the waterfall model (I use this at work) is not always nice to apply, and it take time to release something concistant. 19:02
cognominal TimToady, that's ambitious indeed.
TimToady geistteufel: there's only one person smart enough to design things in advance, and He's theological... :) 19:03
tadzik eek 19:04
TimToady so all we can do is approximate that with time travel
geistteufel so the language progress with experience.
dalek osystem: 3b739ab | tadzik++ | META.list:
Add Grammar::Debugger
TimToady geistteufel: yes
cognominal geistteufel, whirlpool is about feedback 19:05
TimToady fortunately, the future is unevenly distrubuted, so we can steal ideas from the future to help the other parts catch up :)
cognominal the pi factor in waterfall projects is due to lack of appropriate feedback.
19:06 u2012 joined
geistteufel I have to read a book on it 19:06
TimToady it's difficult to come up with a good strange attactor in real life because of inertia
geistteufel and try it a little, just play around
the issue is to find something interesting I can build with it.
TimToady but to get good feedback at all scales requires that our "particles" stay light-on-their-feet so they can turn quickly 19:07
19:07 fsergot joined
geistteufel I love create API, working with REST or Plack. And I don't know if I can create this kind of stuff with perl6 19:07
TimToady you can, eventually :)
moritz you can now
masak right now, the answer is, "you can, but it might not be 100% smooth" 19:08
moritz though maybe not as easily as with p5 right now
TimToady but we have to help redistribute that future to you :)
geistteufel with rakudo ?
moritz yes
masak for example.
geistteufel the guy of rakudo working hard to implement fully the spec, or to speedup their interpreter right now ? 19:09
tadzik huh. Does Grammar::Tracer work for you?
TimToady geistteufel: it's not either/or, but both/and
19:09 NamelessTee joined
geistteufel I thing if they is a really fast interpreter, they will be a lot more active developer on the language 19:09
19:10 u2012 left
masak geistteufel: a few months back, startup time got significantly faster in Rakudo. and yes, that does make a huge difference, just as you say. 19:10
moritz masak: where "a few" is "less than two"
19:11 uvtc joined
geistteufel they is something like Test::Simple for perl6 ? 19:11
TimToady but Perl 6 has the opposite problem from Perl 5. Perl 5 is a big glob of premature optimizations, while Perl 6 is a big glob of postmature optimizations."
moritz use Test;
masak but it's more like Test::More :) 19:12
uvtc Is there a shortcut for referring to a particular message on this channel in the irclog? 19:13
geistteufel yeah I have to read a book, I think it is like learning a new language. with perl5 based. The hard stuff is to switch back afterwhile :)
moritz uvtc: the timestamp on the left is a link to that particular line
TimToady uvtc, look at the timestamp in the log
uvtc Right, but I mean, instead of copying/pasting that link into #perl6, I thought there might be a special syntax that gets recognized. The way S02 gets recognized. 19:14
For example.
moritz uvtc: nope
uvtc moritz, k, thanks.
moritz hm, it's possible 19:16
we could make an irclog:<id here> syntax
and since it's the irclog that renders it, it can look up channel and day too 19:17
tadzik is there anything known to be broken with nom grammars? Like, regressions against b?
uvtc moritz, All you'd need is some ⍟'s.
;)
moritz but that would only be useful if there was an easy way to get at the ID
tadzik Text::CSV is broken, Config::INI too, and I just figured that Text::CSV grammar behaves correctly on niecza
masak geistteufel: I regularly switch between Perl 5 and Perl 6. sometimes I slip, but it's fairly benign.
moritz tadzik: yes
tadzik moritz: what are those? 19:18
masak geistteufel: I did a Perl 5 course last week. kept writing `perl -e 'say ...'` :)
TimToady -E works better there
moritz tadzik: match object construction is sometimes incomplete... lemme find the ticket
masak TimToady: exactly.
TimToady Exactly.
tadzik moritz: in this case it's overcomplete :)
masak :P
19:18 u2012 joined
moritz tadzik: could be the same underlying issue 19:19
jnthn Most of the issues boil down to the mark_commit fix that's needed
tadzik gist.github.com/2387188
masak TimToady: sometimes I think your brain runs on puns and autopuns...
tadzik: that looks like old '**' syntax, for one thing. 19:20
doesn't Niecza implement the new syntax?
tadzik niecza warns about that, yes. What's the new one?
just '%'? Just '%' doesn't work at all 19:21
masak tadzik: quantifier and '%'
tadzik oh
masak that's why the new syntax is better. see also, ingy ;)
tadzik okay, but Rakudo still fails here :)
moritz tadzik: rt.perl.org/rt3/Ticket/Display.html?id=107254 19:22
tadzik moritz: thanks
masak oh, it's that one :/
yeah, that one's basically blocking on pmichaud-shaped tuits, I think.
tadzik masak: github.com/masak/csv/pull/1 19:24
jnthn pmichaud shaped tuits or jnthn to drink enough... :)
Hopefully the pmichaud shaped tuits will show up at the upcoming hackathon though :)
tadzik if not, we'll just get you drunk ;P
jnthn The same fix will unblock a bunch of qbootstrap issues
tadzik: Is *Oslo*? :) 19:25
tadzik is there something about Oslo I need to know?
jnthn tadzik: *mumble*beer price*mumble*
s/Is/In/
tadzik aww 19:26
lichtkind am i late for exorcism?
geekosaur (jnthn-plus?) 19:27
TimToady no, just for the seance, but that's okay
we need someone late at the seance
uvtc lichtkind, re. irclog.perlgeek.de/perl6/2012-04-14#i_5442414 , can you tell where gitit failed? 19:28
19:28 brrt joined
masak tadzik: github.com/masak/csv/commit/9b731b...76e3163e2e 19:29
uvtc lichtkind, s/can you tell/can you tell me/
cognominal rt.perl.org/rt3//Public/Bug/Displa...?id=107254 seems releated to my pet bug rt.perl.org/rt3/Public/Bug/Display...?id=111288 19:30
dalek kudo/nom: 03edce1 | moritz++ | docs/running.pod:
document --doc
19:30 imarcusthis left
masak fondly remembers Ciechan Maciejowe... :) 19:33
lichtkind uvtc: yes i could not create with it a simple appendix a entry, my example was to create a bold triangle meta op within a marker that tells you to place here your operator which has the same syntax like a html tag , there seems to be no way to get this, i though how to use other brackets but i need all 3 kinds for various meanings
19:33 imarcusthis joined
moritz I find src/core/IO/ArgFiles.pm quite confusing 19:34
jnthn
.oO( it makes me say arg! )
19:35
moritz: In what way is it confusing? 19:36
jnthn has just glanced through it now
masak sometimes I wish we'd called .IO ".File"
$path.File
moritz jnthn: it keeps state in $!args, $!io and $!io.opened
uvtc lichtkind, sent you a private msg so as not to clog up this channel 19:39
moritz jnthn: how can I distinguish an exhausted IO::ArgFiles object from an empty one, ie one that is supposed to read only from $*IN? 19:42
19:42 mucker left
jnthn moritz: Hmmm. It just does: $!filename = $!args ?? $!args.shift !! '-'; 19:43
So, once it's eaten up all the args...it can't.
Is it meant to read from STDIN after exhausting the list of files? 19:44
moritz no
jnthn 'cus that's what it looks like it's doing today...
moritz it doesn't
jnthn $.args not containing a Positional could be done way
Though maybe that's a bit evil
moritz anyway, that's why I find it confusing 19:45
19:46 NamelessTee left 19:47 ggoebel left 19:48 ggoebel joined
masak sorear: "provably safe" is ridiculous, for the reasons TimToady touch upon in www.wall.org/~larry/pm.html 19:55
sorear: I can imagine a "hard takeoff", but, hm, www.scottaaronson.com/blog/?p=346
these days, my priors tell me we're more likely to have a disastrous global "event" caused by an energy crisis than we're to have it caused by awesome AI emerging out of a lab somewhere. 19:57
19:58 u2012 left
timotimo either energy crisis or any other resource. metals, say 19:58
kind of sad to think this might happen in my lifetime :| 19:59
noone's getting a break
masak "might" feels kinda weak here.
timotimo either you're old now and you've had a war in your life or you're young and you'll have energy crisis or any other resource shortage ruining a part of your life
moritz "(Also, if the Singularity ever does arrive, I expect it to be plagued by frequent outages and terrible customer service.)" 20:00
masak Scott++
sorear I'm sypathetic to the Hubbardian model of no hard runnings-out
20:00 ggoebel left
sorear we won't go from copper to no copper overnight - as nice mines are exhausted, the price will rise until it becomes economical to mine landfills, at which point it will be unpleasantly high but stationary 20:01
cognominal today we would need 1000 "human energy slaves" for each human being if we were to consume as much energy as today without machines to transform whatever in energy.
sorear cognominal: I couldn't parse that. 20:02
cognominal transitively probably more because one need infrastructure to care for these "slaves".
masak sorear: yes, but while increases are often steady and smooth, decreases seldom are.
20:02 GlitchMr left, ggoebel joined
dalek kudo/nom: e60c80c | moritz++ | src/core/IO/ArgFiles.pm:
implement ArgFiles.slurp
20:02
kudo/nom: 9d489dd | moritz++ | src/core/ (2 files):
rename ArgFiles to IO::ArgFiles
kudo/nom: d9753c1 | moritz++ | src/core/IO (2 files):
implement zero-arg slurp
masak sorear: for rough declines, see cluborlov.blogspot.se/2010/11/peak-...story.html 20:03
20:03 machine2 left
cognominal sorear: to produce the energy you use (thru manufactured objects and you daily life like transportation), you would need 1000 human on bicyles with dynamos, working 12 hours a day. 20:04
masak key idea "Planet Earth Can't Import Oil".
sorear I think that the human future, if there is one, will qualitatively resemble the distant past - equilibrium state with carrying capacity 20:05
moritz wonders if "Brave New World" is considered post-singularity, from a philosophical point of view
cognominal there is still slavey today but not that kind of slavery.
timotimo if only human civilization growth had been controlled so much sooner :|
masak sorear: it has to. exponential growths aren't sustainable.
sorear cognominal: that's just stupid. Every step you go up the food pyramid loses 90%. Do away with the slaves and just burn their food in a power-plant boiler. Better yet, avoid the 99% losses from croppery and build solar. 20:06
masak I find it interesting that previous generations worried about a nuclear threat. this generation worries about exponential curves hitting various roofs. population, global warming, oil consumption...
sorear cognominal: once you've eliminated the factor of 1000+ inefficiency of using humans as an energy source, it doesn't seem so bad 20:07
cognominal photovoltaic is build in China with coal so it has a big carbon footprint and it cannot provide enery when we need it and we don't know how to store energy except in hydroelectric barrage (pumping up in low cost hours). 20:08
sorear In the natural world, when a species is introduced into a new environment, it grows exponentially until it hits a roof, then stabilizes
timotimo sorear: i think the problem with humans is the kind of connectedness. with civilizations growing in developed countries, they are more likely to exploit third world countries even more than they already do 20:09
cognominal s/barrage/dam/
timoto: so true 20:10
masak sorear: yeah. there's some overdue "stabilization" up ahead, that's the main point. :/ 20:12
sorear there's pain ahead, but I don't see the need for nihilistic despair. 20:13
masak in that vein, anyone who hasn't seen www.youtube.com/watch?v=F-QA2rkpBSY probably should.
timotimo so we're heading towards a stabilization - do we have any idea how it'll pan out, given that we are kind of overstaying our welcome, so to speak?
masak sorear: no, despair won't help at all, and I'm not suggesting that. preparing helps, but the uncertainties makes it a bit vague what to do.
cognominal some cynic peiple contend that the recent slow down due to fincancial bubbles exploding has been beneficial because it has had a big impact on the energy comsumption.with the slowing of the global economy. 20:14
masak timotimo: let's just say there will be no more Apple stores.
cognominal: I doubt the effect has been that noticeable.
cognominal unemployed people don't drive cars to work. 20:15
masak I doubt the energy consumption has gone down by more than, say, 15%.
sorear masak: I have a student in a class of mine who rants about how the 2011 Tohoku earthquake and tsunami was a plot by environmentalists to forestall overpopulation. It's a bit depressing
masak cognominal: it's industries that consume the most energy, compared to individuals.
cognominal masak, probablt less than that, but noticable though
masak sorear: I hear that and think "only in the US..." 20:16
timotimo sorear: we have the power to make earthquakes? or does he mean the earthquake was used - via mismanagement - to get as much damage as possible done?
20:18 Araq joined
sorear masak: he's on visa from Kuwait; I don't know how much US culture has poisoned him 20:18
masak huh. 20:19
anyway, that's nuts.
moritz there are known cases where geothermal drilling has caused earthquakes, but on a much smaller scale 20:20
moritz wonders if availability of TV and birth rates are correlated 20:21
masak so it's claimed.
masak has no plans to get a TV :)
moritz if so, there's a much more effective (and equally cruel :) way to prevent overpopulation 20:22
timotimo wait, in what direction would they corellate?
corelate*
masak more TV, lower birth rates. 20:23
sorear meh, we all know affluence lowers birth rates
masak allegedly because they both occupy the "something fun to do in the evening" niche.
jnthn What on earth could be on TV that's so interesting?! 20:24
sorear (I know several people who argue that rising world standards of living will cause US population to tank before it reaches the stabilization point)
cognominal lower birth rates but dumb kids not informed about contraception
masak plans to even out the no TV with some affluence :)
cognominal: you get your contraception information through the tube in France? 20:25
we got it at school, as I recall. no TV required.
geistteufel is it normal that "make spectest" in rakudo doesn't pass all tests ?
lichtkind geistteufel: in nomeni patri et fili et spiritus sancti ... :)
masak geistteufel: it's worth mentioning here.
geistteufel: which test files? 20:26
geistteufel I copy the test result
in pastbin
masak thank you.
(phew) :)
20:27 simcop2387 left
masak lichtkind: he's an *un*holy ghost, judging by the nick... ;) 20:27
geistteufel pastebin.com/HedmJBTy 20:28
masak looks
ah, lots of "dubious" ones.
jnthn geistteufel: Windows?
geistteufel I working a lot in test since the last hackaton at Paris, last week. :)
masak Linux?
geistteufel MacOS Lion
masak ah.
jnthn Oh, interesting.
20:28 simcop2387 joined
geekosaur also notice breeding lots of kids because (1) high neonatal death rates (2) more help around the farm 20:28
masak they all seem to be of that kind. 20:29
jnthn Those tests are all the ones that use Test::Util.
masak geekosaur: right, both of which are obsoleted by rising affluence.
geekosaur bothof which tend to go away with the level of affluence (and free time) that enables buying/watching a TV
masak what geekosaur said ;)
moritz geekosaur: can you please run ./perl6 t/spec/S32-scalar/undef.rakudo and nopaste the output from that too?
masak s/geekosaur/geistteufel/ ? :) 20:30
20:30 brrt left
geistteufel no 20:30
moritz yes, sorry
geistteufel I'm not a geekosaur :)
moritz geistteufel: can you pleae do it anyway? :-) 20:31
masak geistteufel++
geistteufel yep
sorry :)
ok
I do it
masak as soon as he's done being offended that you mis-tabbed him :)
geekosaur :p 20:32
should we both be offended?
geistteufel pastebin.com/qZm1WPjc
masak geekosaur: you should be offended in equal and opposite directions :P
geistteufel: there's your error, right there. 20:33
question is why it runs the ICU tests when you don't have ICU installed.
moritz because it's not marked as # icu 20:34
geistteufel um, what's that ?
unicode
masak oh!
geistteufel ok, and should I install something ?
moritz yes, icu (and its headers)
and then reconfigure and recompile everything
though we should really warn about missing ICU much earlier 20:35
geistteufel yep
20:35 birdwindupbird left
masak maybe it's time to make ICU a hard requirement. 20:36
wasn't that discussed recently on parrot ML?
geistteufel I start install it. I need to rebuild perl6 from scratch ?
jnthn masak: Or just implement NFG :) 20:37
moritz geistteufel: you need to even rebuild parrot
rm -rf parrot
perl Configure --gen-parrot && make # should do it
erm
sorry
rm -rf install/
deleting parrot doesn't help, but doesn't hurt too much either
dalek kudo/nom: 2676549 | jnthn++ | / (2 files):
Start to sketch out a PseudoStash that will handle many of the pseudo-packages. Not yet wired up, let alone complete.
20:38
kudo/nom: 1c862bc | jnthn++ | src/Perl6/ (2 files):
Start to clean up indirect lookup handling a little. Also fix a bug in the compile time handling of indirect lookups.
kudo/nom: 68df8f5 | jnthn++ | src/ (3 files):
Switch indirect name lookups to use the new root PseudoStash approach. Still much missing in PseudoStash to make that work yet, mind. Also a little more elimination of duplicate logic.
kudo/nom: 7f28736 | jnthn++ | src/Perl6/Grammar.pm:
Unbust a typed exception.
kudo/nom: 5a31737 | jnthn++ | src/core/PseudoStash.pm:
Get things like ::('MY::$a') and $::('OUTER::b') working.
kudo/nom: c59411c | jnthn++ | src/ (2 files):
Flesh out pseudo-package handling a bit more. Now all of the interpolated lookup tests that now take a similar code path all pass again (and we should handle a bunch more cases now).
geistteufel icu is a hudge program
downloading ... 20:39
20:40 mdupont_ joined
masak Unicode is a huge standard. 20:41
masak almost said "Unicode is a huge WOMAN WITH BUNNY EARS"
huf and then you did 20:43
20:44 vlixes left
TimToady if meta-saying something is the same as saying it, we're all gonna be in deep trouble 20:44
huf i know, i know, use-mention error
masak I was almost going to say that.
TimToady you can almost say that again
masak that almost goes without saying. 20:45
geekosaur didn't Gödel have something to say about that too?
huf but perhaps we're all in deep trouble.
TimToady +1 to NFG anyway
masak geekosaur: Gödel showed that all you need in order to make "say"/"meta-say" collapse, is integer arithmetic. 20:46
TimToady after which great hilarity ensues
masak well, something like that but more formally.
TimToady "you can either be stupid or look stupid--pick two" 20:48
20:48 uvtc left
geekosaur except they don't quite collapse. you can encode omega but you still can't write it (that is in fact the point) 20:48
20:55 alvis joined
masak right. 20:58
the system becomes powerful enough to encompass theorems with a self-reference (albeit through the encoding mechamism), and that makes it either incomplete or inconsistent. 20:59
21:00 oiig_ left, NamelessTee joined 21:02 fgomez left 21:08 fgomez joined
geistteufel I have install icu4c on mac, then add export LDFLAGS='-L/usr/local/Cellar/icu4c/4.8.1.1/lib' CPPFLAGS='-I/usr/local/Cellar/icu4c/4.8.1.1/include' 21:12
but still the same issue 21:13
moritz what is icu4c?
geistteufel I have do "make distclean", Configure, make, make spectest
icu4c is the package for icu
icu4c 4.8.1.1 21:14
moritz geistteufel: did you remove the install/ dir?
geistteufel site.icu-project.org/
make distclean ?
21:14 Araq left
moritz distclean doesn't remove the install directory 21:14
masak make realclean, more like.
or the git clean -dfx thing.
moritz neither
yes, git clean -xdf would work 21:15
21:15 noam joined
masak and realclean wouldn't? huh. 21:15
moritz distclean = realclean + deleting Makefile
geistteufel ok I do both now
moritz note that the install location can be /usr/local/ too
geistteufel compiling agaib :)
again 21:16
moritz you don't want any make target to rm -rf /usr/local/
or /usr/ or / for tha matter :-)
moritz -> sleep
jnthn 'night, moritz
moritz \o
21:16 NamelessTee left
geistteufel why Configure compile stuff ? 21:17
masak because it's building parrot and nqp. 21:20
jnthn It's not Configure that does so itself; if you give it the right flags it just triggers the building of the dependencies for you.
Wow. Stopping term:sym<name> from re-parsing the name etc. seems to knock a little off the spectest time... 21:22
masak oh, and I believe people here might appreciate www.philosophersnet.com/games/identity.php 21:24
cognominal geistteufel, rakudo pulls to other github distribution, one for parrot and one for nqp, that it compiles.
parrot is the vm, nqp is used to bootstrap rakudo. Some nqp libraries are used by rakudo too. 21:25
s/pulls to/pulls from/
geistteufel ok 21:26
well, it still doesn't work 21:31
same issue
can't solve the ICU stuff
dalek kudo/nom: b82d905 | jnthn++ | src/Perl6/ (3 files):
Further clean up term:sym<name>, so we don't do any of the reparsing and don't rebuild the LongName object. Since we encounter this rule a fair amount, this seems to improve compilation performance a little.
21:32
tadzik surived :) 21:33
21:33 vlixes joined 21:34 lichtkind_ joined 21:36 NamelessTee joined, lichtkind left, lichtkind_ is now known as lichtkind
geistteufel some mac user just disactivate the ICU stuff. 21:39
the make test work
I will try to find out later 21:42
21:55 lestrrat left 21:57 lestrrat joined 21:59 bluescreen10 left 22:01 spadger joined 22:02 spadger left
fsergot 'night! o/ 22:13
sorear o/
22:13 fsergot left
tadzik 'night 22:13
22:14 fgomez left 22:15 fgomez joined 22:16 jaffa4 left
dalek kudo/nom: 84f4fd4 | jnthn++ | src/Perl6/ (2 files):
Implement ::<$x> style accesses, and make OUTER::<$x>, CALLER::<$x> and MY::<$x> work.
22:21
kudo/nom: 9b4f785 | jnthn++ | src/core/PseudoStash.pm:
Fix DYNAMIC, so it now basically works.
22:23 localhost left 22:25 localhost joined
cognominal jnthn, what the diff between nqp::clone and pir::repr_clone? 22:25
jnthn cognominal: First currently maps to the Parrot clone v-table 22:26
Latter maps to a 6model primitive. 22:27
Expect some unification (in favor of the latter) down the line :)
cognominal thx
masak 'night, #perl6 22:29
tadzik 'night
good knight
22:34 jferrero left, jferrero joined
lichtkind uvtc++ 22:40
he really showed that appendix aA is possible with markup
but maybe its not worth the effort since the effort to write such markup is almost harder than html which some proper defined css styles 22:41
masak: hai
dalek kudo/nom: c384ab7 | jnthn++ | src/core/PseudoStash.pm:
First crack at OUR.
22:42
22:55 uvtc joined
dalek ast: 5021c60 | jnthn++ | S02-names-vars/names.t:
Todo a slightly dubious test; unfudge a couple that we now pass.
22:57
ast: 2c98b03 | jnthn++ | S0 (2 files):
A few more OUTER:: unfudges for Rakudo.
ast: 4b79581 | jnthn++ | S02-names/symbolic-deref.t:
One more unfudge.
23:04
23:05 PacoAir left
jnthn More tomorrow...night o/ 23:10
sorear ooops, missed jnthn 23:14
23:25 pochi_ left 23:45 pochi joined
lichtkind it reall looks promising 23:48
already the second who want to help with docs