»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.
00:03 sivoais left 00:05 Psyche^ joined 00:06 sivoais joined 00:09 Patterner left, Psyche^ is now known as Patterner 00:10 hash_table joined 00:15 hash_table left, cognominal joined 00:20 bowtie left 00:29 larks left 00:30 larks joined 00:47 atrodo left 00:49 cognominal left 01:03 anuby joined 01:04 grondilu left 01:16 PerlPilot joined 01:17 gfldex_ joined, LoRe_ joined 01:18 drbean left, PerlJam left 01:19 drbean joined 01:20 _ilbot joined 01:24 cognominal joined, hypolin joined 01:28 ggoebel_ joined 01:36 s1n left 01:38 cognominal_ joined, cognominal left 01:45 cognominal_ left
pmichaud noteworthy date: Dec 18 will be Perl's 25th birthday. Might be worth an article/mention on p6advent. :-) 01:49
timotimo pmichaud: can i ask you for a brief answer on the last paragraph of this comment here? rt.perl.org/rt3/Ticket/Display.htm...xn-1177716 01:51
01:52 cognominal_ joined 02:03 whiteknight left
pmichaud timotimo: looking 02:06
02:09 ggoebel_ left
pmichaud timotimo: initially it looks good to me, although I'm not up to date on the latest parser changes. Definitely jnthn++ should be the one to approve it for now 02:09
if it follows the w\ay that STD.pm6 handles the error, then it's very likely correct; if it doesn't follow STD.pm6, then perhaps it should :) 02:10
timotimo oh, i was just wondering about adding a test case 02:11
pmichaud test cases are good but it's sometimes hard to trap compile-time errors and test the message
timotimo yes, my first attempts to get the error message as a string or something have failed
pmichaud you might be able to test things with an eval 02:12
timotimo say, what could be the cause of getting "Missing or wrong version of dependency 'src/Perl6/Actions.pm" when starting rakudo?
pmichaud Missing or wrong dependency almost always means there's an out of date .pbc somewhere
timotimo oh, that's interesting. i wonder how i made that come up
02:13 FROGGS_ joined
timotimo i'm trying to make rakudo spit out "did you mean @arr?" when the user writes $arr[1] instead of only "$arr not declared" 02:13
(when the user writes that and there exists a @arr)
02:13 hash_table joined
pmichaud that would be cool. It kind of means that the error handler itself needs to look ahead (continue parsing) to see if there's a .[] or .{} after the unrecognized variable 02:14
timotimo i think as a first approximation it would be enough to just see if the other var exists 02:15
because i have no idea how to work the parser :D
pmichaud ...variables, since it could also be $arr{'foo}
er, $arr{'foo'}
02:16 FROGGS left
timotimo by changing a bit of code in the Actions.pm i get the "missing or wrong dependency". i wonder if the build silently went wrong? 02:17
don't see any errors though
kind of don't want to run a make clean now. not on this laptop. 02:18
i think i'll go to bed instead of continuing to try. 02:20
popl Don't do it in your sleep.
timotimo why not? git allows reverts easily ;) 02:21
it might just impact the quality of sleep negatively
pmichaud I'm afk again 02:23
timotimo thank you for your help :)
02:23 s1n joined 02:31 ggoebel_ joined
timotimo t.h8.lv/0001-WIP-on-arr-1-error.patch - anyone who would like to can jump in and add the finishing touches and take as much blame/praise as they'd like 02:35
02:44 dayangkun joined 02:48 dayangkun left 02:52 Util left 02:53 Util joined 02:54 tadzik joined 02:59 topo joined 03:00 gootle left 03:12 hypolin left 03:13 tadzik left, topo left, topo joined, hypolin joined, tadzik joined 03:15 jaldhar_ joined 03:16 cognominal_ left 03:22 cognominal joined 03:24 s1n left 03:27 fgomez left 03:30 jaldhar_ left 03:32 jaldhar_ joined 03:37 fgomez joined 03:38 s1n joined 03:48 hash_table left 03:49 am0c joined 03:56 noggle_ joined 04:00 noggle left
colomon is almost done adding operators to Rakudo for the Day 13 Advent calendar post. 04:22
04:31 wamba joined
colomon are people getting rakudo test fails in is-prime.t and dash-e.t? 04:44
dalek ast: 2085636 | (Solomon Foster)++ | S03-operators/bag.t:
Fudge for Rakudo.
 04:45
rl6-roast-data: 1bbebbf | coke++ | / (3 files):
today (automated commit)
colomon nr: (1..100).grep(*.is-prime) 04:46
p6eval rakudo 78b53a, niecza v24-5-g599cbcb: ( no output )
colomon nr: say (1..100).grep(*.is-prime)
p6eval rakudo 78b53a, niecza v24-5-g599cbcb: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97␤»
dalek kudo/nom: 798dbdd | (Solomon Foster)++ | src/ (3 files):
Add Texas versions of the Set and Bag operators.
 04:48
colomon forgive me if that introduces failures into is-prime.t -- I can't see how it could, and I need to get to bed.
05:06 PerlPilot left 05:08 PerlJam joined 05:12 s1n left 05:13 dalek left 05:17 topo left, Util left 05:19 dalek joined, ChanServ sets mode: +v dalek 05:20 fgomez left 05:21 rummik left, topo joined, Util joined, rummik joined 05:22 telex left 05:23 skids left, telex joined 05:24 sftp left 05:25 fgomez joined 05:29 dalek left 05:32 telex left 05:35 dalek joined, ChanServ sets mode: +v dalek, telex joined 05:42 PerlJam left 05:43 PerlJam joined 05:50 Juerd left 05:52 SamuraiJack joined, Util left 05:54 dalek left 05:56 tadzik left, Juerd joined, Util joined 05:58 topo left, dalek joined, ChanServ sets mode: +v dalek 06:00 telex left, topo joined 06:02 telex joined 06:09 kurahaupo left 06:10 tadzik joined 06:12 topo left 06:13 PerlJam left 06:14 wamba left, Util left 06:20 Util joined, PerlJam joined, topo joined 06:25 Util left 06:27 tadzik left 06:28 Util joined, tadzik joined 06:29 kaleem joined 06:34 Bzek joined 06:36 Util left 06:38 Util joined 06:41 ggoebel_ left 06:45 dalek left 06:49 hugme left, Juerd left, sergot left 06:50 tadzik left, PerlJam left, topo left, pmichaud left, masak left, Util left 06:54 quester joined 07:02 quester left 07:37 Guest31156 is now known as ponbiki 07:38 tadzik joined 07:40 hugme joined, ChanServ sets mode: +v hugme
tadzik what happened? /o\ 07:40
07:40 Juerd joined
sorear tadzik: looks like a classic case of "feather dropping off the internet" 07:43
Juerd That's an accurate description of what happened 07:44
More specifically, the underlying physical machine had a kernel oops. Cause unknown.
sorear just an oops, not a panic?
Juerd I mean panic.
The thing where numlock and capslock blink :)
It's configured to reboot on panic, but it didn't do so 07:45
Anyhow, I'm going back to bed.
'nite
sorear g'nite Juerd 07:47
am0c pity p6eval.. 07:54
07:55 cibs joined, p6eval joined, ChanServ sets mode: +v p6eval 08:02 Kharec joined 08:06 pseudo1 joined 08:08 benabik left 08:10 pseudo1 left
cedrvint pmichaud: OK, I will mention this 25th anniversary :) 08:16
08:25 benabik joined
cedrvint colomon: why your article isn't published yet? It looks good (colomon++) to me. 08:38
08:47 wamba joined 08:56 cognominal left 08:57 sahadev1 left, yves_ left
cedrvint Street Fighter, Zelda and Megaman turned 25 this year too! It can't be a coincidence :) 09:00
moritz ENODALEK 09:01
09:01 yves_ joined
moritz timotimo: I've pushed your patch from the backlog, with one small change: I've changed '.' to <sym>, as all the other operators do 09:02
and I've added a small test
09:02 sahadev joined 09:03 cognominal joined 09:07 kurahaupo joined 09:23 popl left 09:24 cognominal left
felher Good morning, folks. 09:34
09:36 fhelmberger joined
moritz good morning, felher 09:38
09:41 dakkar joined 09:42 wamba left, wamba joined
cedrvint r: sub term:<fourty-two> { 42 } 09:44
p6eval rakudo 58233d: OUTPUT«===SORRY!===␤Cannot add tokens of category 'term'␤at /tmp/Rm5DrwGbof:1␤------> sub term:<fourty-two> ⏏{ 42 }␤ expecting any of:␤ colon pair␤ quote words␤»
cedrvint NYI?
n: sub term:<fourty-two> { 42 } 09:46
p6eval niecza v24-5-g599cbcb: OUTPUT«Potential difficulties:␤ &term:<fourty-two> is declared but not used at /tmp/C5BA4p_1zq line 1:␤------> sub term:<fourty-two> ⏏{ 42 }␤␤»
cedrvint n: sub term:<fourty-two> { 42 }; fourty-two.say
p6eval niecza v24-5-g599cbcb: OUTPUT«42␤»
09:47 rindolf joined
cedrvint is there a way to add such terms with Rakudo? 09:47
moritz r: constant fourty-two = 42; fourty-two.say 09:49
p6eval rakudo 58233d: OUTPUT«42␤»
moritz r: macro fourty-two { quasy { 42 } }; say fourty-two.say 09:50
p6eval rakudo 58233d: OUTPUT«===SORRY!===␤Could not find sub &quasy␤at /tmp/wm_7iEMfyC:1␤------> ␤»
moritz r: macro fourty-two { quasi { 42 } }; say fourty-two.say
p6eval rakudo 58233d: OUTPUT«42␤True␤»
cedrvint I meant a term like "rand", not constant. 09:51
moritz r: sub term:sym<fourty-two> { 42 }; say fourty-two
p6eval rakudo 58233d: OUTPUT«===SORRY!===␤Cannot add tokens of category 'term'␤at /tmp/j5zT0e5uJx:1␤------> sub term:sym<fourty-two> ⏏{ 42 }; say fourty-two␤ expecting any of:␤ pair value␤ quote words␤»
moritz no, doesn't work
cedrvint ok, thanks. 09:52
moritz might be added to src/Perl6/Grammar.pm around line 3214 09:53
09:56 wamba left, wamba joined 09:58 Kharec left 10:00 rindolf left, rindolf joined 10:01 hypolin left 10:03 Kharec joined 10:05 anuby left 10:11 wamba left, wamba1 joined 10:22 bowtie joined 10:51 grondilu joined
grondilu rn: print -1 mod 10 10:51
p6eval niecza v24-5-g599cbcb: OUTPUT«9»
..rakudo 58233d: OUTPUT«-1»
grondilu ? 10:52
eval: print -1 mod 10
buubot_backup grondilu: ERROR: syntax error at (eval 20) line 1, near "1 mod "
grondilu eval: print -1 % 10
buubot_backup grondilu: 91
grondilu eval: -1 % 10
buubot_backup grondilu: 9
moritz bug in rakudo
11:03 spacebat joined 11:05 birdwindupbird joined, Su-Shee_ joined 11:07 kurahaupo left 11:09 Su-Shee left, Su-Shee_ is now known as Su-Shee 11:48 amkrankruleuen joined
colomon cedrvint: because it's only half-finished. 12:04
cedrvint: maybe not quite even that.
cedrvint: but thanks! 12:05
moritz still 9 hours until the 13th ends in my time zone
speaking of which 12:06
tomorrow is PerlJam day, but he hasn't shown up for quite some time
anybody want to claim tomorrow instead?
colomon moritz: and even more in my time zone!
colomon is still debugging the next stage of his post. 12:08
and the perl6-debugger isn't working for me atm. :( 12:09
moritz fails at implementing terms 12:15
oh, actually it works, but only for term:sym<foo>, not for term:foo 12:16
n: sub term:sym<foo>() { 42 }; say foo 12:19
p6eval niecza v24-5-g599cbcb: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'foo' used at line 1␤␤Potential difficulties:␤ &term:sym<foo> is declared but not used at /tmp/d0o33tw7Lv line 1:␤------> sub term:sym<foo>⏏() { 42 }; say foo␤␤Unhandled exception: Che…
moritz n: sub term:foo { 42 }; say foo
p6eval niecza v24-5-g599cbcb: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'foo' used at line 1␤␤Potential difficulties:␤ &term:foo is declared but not used at /tmp/P69Rb4nBfs line 1:␤------> sub term:foo ⏏{ 42 }; say foo␤␤Unhandled exception: Check failed␤␤ …
moritz n: sub term:<foo> { 42 }; say foo
p6eval niecza v24-5-g599cbcb: OUTPUT«42␤»
moritz oh, also works for term:<foo> 12:20
just not for term:foo
colomon rn: my $bag = bag a = > 10, b => 2; my $bag2 = bag a => 3, c => 3; say $bag (-) $bag2 12:25
p6eval niecza v24-5-g599cbcb: OUTPUT«===SORRY!===␤␤Preceding context expects a term, but found infix = instead at /tmp/WXnqLzycec line 1:␤------> my $bag = bag a ⏏= > 10, b => 2; my $bag2 = bag a => 3, c␤␤Parse failed␤␤»
..rakudo 58233d: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/NmF5HbIbn_:1␤------> my $bag = bag a =⏏ > 10, b => 2; my $bag2 = bag a => 3, c ␤»
colomon rn: my $bag = bag a = > 10, b => 2; my $bag2 = bag a => 3, c => 3; 12:26
p6eval niecza v24-5-g599cbcb: OUTPUT«===SORRY!===␤␤Preceding context expects a term, but found infix = instead at /tmp/lIe6psHNW9 line 1:␤------> my $bag = bag a ⏏= > 10, b => 2; my $bag2 = bag a => 3, c␤␤Parse failed␤␤»
..rakudo 58233d: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix = instead␤at /tmp/TP0TKA3UsW:1␤------> my $bag = bag a =⏏ > 10, b => 2; my $bag2 = bag a => 3, c ␤»
colomon rn: my $bag = bag a => 10, b => 2; my $bag2 = bag a => 3, c => 3;
p6eval rakudo 58233d: OUTPUT«2 unexpected named parameters passed (a, b)␤ in sub bag at src/gen/CORE.setting:11886␤ in block at /tmp/7Dd621p6v8:1␤␤»
..niecza v24-5-g599cbcb: OUTPUT«Potential difficulties:␤ $bag is declared but not used at /tmp/bRW28pkkZw line 1:␤------> my ⏏$bag = bag a => 10, b => 2; my $bag2 = b␤ $bag2 is declared but not used at /tmp/bRW28pkkZw line 1:␤------> my $bag = bag a => 10, b => 2…
colomon rn: my $bag = bag a => 10, b => 2; my $bag2 = bag a => 3, c => 3; say $bag (-) $bag2
p6eval niecza v24-5-g599cbcb: OUTPUT«Unhandled exception: Excess arguments to bag, unused named a, b␤ at /home/p6eval/niecza/lib/CORE.setting line 0 (bag @ 1) ␤ at /tmp/v1VZqx5eVE line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4208 (ANON @ 3) ␤ at /home/p6eval/niecza/…
..rakudo 58233d: OUTPUT«2 unexpected named parameters passed (a, b)␤ in sub bag at src/gen/CORE.setting:11886␤ in block at /tmp/FuLesyBuU6:1␤␤»
colomon rn: my $bag = bag a, b, a; my $bag2 = bag a, c; say $bag (-) $bag2 12:27
p6eval niecza v24-5-g599cbcb: OUTPUT«===SORRY!===␤␤Undeclared routines:␤ 'a' used at line 1,1,1␤ 'b' used at line 1␤ 'c' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1443 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6…
..rakudo 58233d: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&a' called (lines 1, 1, 1)␤Undefined routine '&b' called (line 1)␤Undefined routine '&c' called (line 1)␤»
colomon rn: my $bag = bag <a, b, a>; my $bag2 = bag <a, c>; say $bag (-) $bag2
p6eval rakudo 58233d, niecza v24-5-g599cbcb: OUTPUT«set(b,, a)␤»
colomon yeah, that's my bug all right 12:28
rn: my $bag = bag <a, b, a>; my $bag2 = bag <a, c>; say set($bag); say set($bag2); say $bag (-) $bag2
p6eval rakudo 58233d, niecza v24-5-g599cbcb: OUTPUT«set(a,, b,, a)␤set(a,, c)␤set(b,, a)␤»
moritz are the commas inside the <...> intentional?
colomon rn: my $bag = bag <a b a>; my $bag2 = bag <a c>; say set($bag); say set($bag2); say $bag (-) $bag2
p6eval rakudo 58233d, niecza v24-5-g599cbcb: OUTPUT«set(a, b)␤set(a, c)␤set(b)␤» 12:29
colomon okay, not my bug at all
rn: my $bag = bag <a b a>; my $bag2 = bag <a c>; say set($bag); say set($bag2); my $diff = $bag (-) $bag2; say $diff 12:32
p6eval rakudo 58233d, niecza v24-5-g599cbcb: OUTPUT«set(a, b)␤set(a, c)␤bag("a" => 2, "b" => 1)␤»
colomon that's my bug
moritz what should the result be? 12:33
bag("a" => 1, "b" => 1) ? 12:34
colomon no, the problem here is that it's actually doing (my $diff = $bag) (-) $bag2 12:35
rather than my $diff = ($bag (-) $bag2)
TimToady: ^^
moritz r: rn: my $bag = bag <a b a>; my $bag2 = bag <a c>; say ($bag (-) $bag2) 12:37
p6eval rakudo 58233d: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/1d8SHyipwo:1␤------> rn⏏: my $bag = bag <a b a>; my $bag2 = bag ␤ expecting any of:␤ argument list␤ postfix␤ infix or meta-infix␤ infix stopper␤ state…
moritz rn: my $bag = bag <a b a>; my $bag2 = bag <a c>; say ($bag (-) $bag2)
p6eval rakudo 58233d, niecza v24-5-g599cbcb: OUTPUT«set(b)␤»
moritz rn: my $bag = bag <a b a>; my $bag2 = bag <a c>; my $diff = $bag (-) $bag2; say $diff 12:38
p6eval rakudo 58233d, niecza v24-5-g599cbcb: OUTPUT«bag("a" => 2, "b" => 1)␤»
12:41 benabik left 12:47 LoRe_ left 12:50 LoRe joined
colomon moritz: (-) is purely a set operation, there is no bag equivalent, so it converts its arguments to Sets and returns a Set. 12:55
so the say ($bag (-) $bag2) version is correct.
moritz colomon: why did you give it %list_infix precedence? 12:57
then it's no wonder it's looser than item assignment
colomon moritz: I just did it as TimToady had it.
but yes, now I'm thinking that's wrong.
moritz it seems those ops aren't in S03 12:58
I wonder if the precedence should be the same as junctive and/junctive or 12:59
12:59 [Coke] joined
bbkr should custom exception be defined in lib/X/ directory or is it better to declare package like lib/JSON/RPC/X which provides X:: exceptions? the later method looks more friendly for package management but may cause unexpected namespace conflicts when multiple modules are loaded. 12:59
moritz bbkr: so far the recommendation is X::JSON::RPC::WhatGoesWrong 13:00
bbkr in lib/X/JSON/RPC ? 13:01
moritz the module -> path mapping doens't change just because there's an X in the name 13:02
so yes
bbkr moritz: thanks
colomon btw, the script I've now got working can give you the ten most frequently used words that are in Hamlet and not in Much Ado About Nothing -- takes Rakudo 1m35s. 13:08
moritz what about niecza?
colomon now running 13:09
8.5 seconds
had not expected such a big difference! 13:10
moritz r: say 95 / 8.5
p6eval rakudo 58233d: OUTPUT«11.176471␤»
13:13 amkrankruleuen left 13:14 MayDaniel joined 13:22 Util joined 13:30 GlitchMr joined, kaleem left 13:44 benabik joined 13:53 cognominal joined
cedrvint is there a documentation on how to add a new metaoperator? 14:00
14:00 atrodo joined
timotimo hey moritz, i've made another small change later that i forgot to make public: - <sym> {} <?before \s | ',' | <terminator> > + [ <sym> <!before \w | '('> | <sym> {} <?before \s | ',' | <terminator> > ] 14:03
i wasn't sure what exactly that other part did that i just removed willy-nilly, so i added it back as an alternative 14:04
oh, that was for something completely different i think 14:10
14:11 bluescreen10 joined
timotimo it was for making $. give a better error, that's it! 14:13
14:16 PacoAir joined, sahadev left
cedrvint read "it was for making $" ;) 14:20
timotimo i'd love to write a patch like that one day :D
GlitchMr read "it was for making $" ;) as $" variable should have better error. 14:21
std: $"
p6eval std a8bc48f: OUTPUT«===SORRY!===␤Unsupported use of $" variable; in Perl 6 please use .join() method at /tmp/M02Syhr8_W line 1:␤------> $"⏏<EOL>␤Parse failed␤FAILED 00:00 41m␤»
GlitchMr Except this error is good error 14:22
timotimo well, i was talking mostly about:
[Coke] r: use Test; is("\c[LINE FEED (LF)]", "\c10", '\c[LINE FEED (LF)] works');
p6eval rakudo 9df474: OUTPUT«ok 1 - \c[LINE FEED (LF)] works␤»
timotimo r: $.
p6eval rakudo 9df474: OUTPUT«===SORRY!===␤Non-declarative sigil is missing its name␤at /tmp/KmSv8HnS2W:1␤------> $⏏.␤ expecting any of:␤ statement list␤ prefix or term␤ prefix or meta-prefix␤»
[Coke] that dies locally with "Unrecognized character name LINE FEED (LF)" 14:23
GlitchMr std: $.
p6eval std a8bc48f: OUTPUT«===SORRY!===␤Unsupported use of $. variable; in Perl 6 please use the filehandle's .line method at /tmp/pUiMczYiQh line 1:␤------> $.⏏<EOL>␤Parse failed␤FAILED 00:00 41m␤»
GlitchMr STD can already give good errors
cedrvint r: /{ make '$' }/
p6eval rakudo 9df474: ( no output )
cedrvint it makes '$' :)
diakopter std: std: say: ; 14:24
p6eval std a8bc48f: OUTPUT«ok 00:00 40m␤»
timotimo it may make $, but no output? seems fishy
cedrvint :)
timotimo r: sub get-rich-quick() { make '$' }; say get-rich-quick(); 14:25
p6eval rakudo 9df474: OUTPUT«Cannot bind attributes in a type object␤ in sub make at src/gen/CORE.setting:10305␤ in sub get-rich-quick at /tmp/8okUQM9d_c:1␤ in block at /tmp/8okUQM9d_c:1␤␤»
timotimo so, i added a bit of code to Grammar.pl that made r: $. work correctly, but the original bug report complained about r: / $. / which my code doesn't catch (and i don't know why or how to figure it all out) 14:27
14:29 PacoAir left, PacoAir joined 14:30 sahadev joined 14:32 cognominal left 14:35 GlitchMr left 14:38 cognominal joined
colomon rn: my $bag = KeyBag.new(<a a a b>); say $bag; $bag{b}--; say $bag; 14:45
p6eval rakudo 9df474: OUTPUT«===SORRY!===␤CHECK FAILED:␤Undefined routine '&b' called (line 1)␤»
..niecza v24-5-g599cbcb: OUTPUT«===SORRY!===␤␤Undeclared routine:␤ 'b' used at line 1␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1443 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 1147 (P6.comp_unit @ 37) ␤ at /home/p6…
colomon rn: my $bag = KeyBag.new(<a a a b>); say $bag; $bag<b>--; say $bag;
p6eval rakudo 9df474, niecza v24-5-g599cbcb: OUTPUT«keybag("a" => 3, "b" => 1)␤keybag("a" => 3)␤»
timotimo colomon: did you mean {b} and not <b>? 14:46
oh, i see yo ucorrected it
14:47 GlitchMr joined 14:51 cognominal left, cognominal_ joined 14:53 stopbit joined 15:00 cognominal_ left, stopbit left
bbkr is there a way to run prove on Perl 5 and Perl 6 tests at the same time? prove seems to not recognize Perl6 shebang. and -e flag changes executable globally for all tests. 15:02
15:03 cognominal_ joined, benabik left 15:05 kaare_ joined 15:08 kivutar joined
timotimo wasn't perl5 supposed to launch perl6 if it encounters use v6; at the beginning of a file? 15:08
seems like that's not implemented :( 15:09
[Coke] $ panda install SDL 15:10
connect failed: Connection refused
(after a rakudo * 2012-11 install)
ah, golfed it to just "panda". :( 15:11
15:11 benabik joined 15:13 cognominal_ left
timotimo installs panda to see what's up 15:13
15:13 stopbit joined
timotimo i thought panda used github for its information? interesting 15:15
15:17 cognominal joined
timotimo ah, it's feather.perl6.nl:3000 15:18
and that's down, so ... not perl6s fault i guess
GlitchMr The problem with shebangs is that those won't work in Windows. 15:19
[Coke] ah. be nice if the error said what connect was refused. 15:20
GlitchMr I would say to use some program that does redirects when it doesn't recognize shebang. The problem is that the only such program I know is perl.
[Coke] Thanks for tracking that down.
GlitchMr And Perl only redirects when shebang doesn't contain "perl".
But "perl6" contains "perl".
timotimo [Coke]: um, the error did say "connection refused" :)
it didn't say "connection resuf^H^H^Hfused to feather.perl6.pl, though 15:21
GlitchMr Feather works fine.
timotimo then the server that does the ecosystem on feather has crashed?
hm, brb 15:22
[Coke] feather was recently restarted. 15:23
GlitchMr glitchmr@feather ~> uptime
16:23:42 up
Recently, you say?
Well, it was restarted 16 hours ago. 15:24
15:33 skids joined 15:34 rindolf left 15:38 fgomez left 15:47 bluescreen10 left 15:52 birdwindupbird left
colomon Advent post up: perl6advent.wordpress.com/2012/12/1...-and-sets/ 15:53
please let me know if you find any typos! though I'm going to be at a preschool Christmas part for the next hour or so...
15:54 spider-mario joined 15:55 cognominal left
diakopter colomon++ lgtm 15:56
colomon *party 16:02
16:04 cognominal joined 16:07 PerlJam joined 16:17 erkan left 16:23 benabik left 16:29 benabik joined
GlitchMr colomon: I didn't knew you can simply call bag() 16:30
Well, looks I'll have to update that page second time 16:31
rosettacode.org/mw/index.php?title=...did=148507
16:34 daniel-s__ left 16:41 ingy left, ingy joined 16:45 cognominal left, cognominal joined
TimToady colomon: "compute the set of works" ~~ s/works/words/ 16:48
jnthn evening o/ 16:52
diakopter o/ 16:53
www.webupd8.org/2012/01/install-ora...u-via.html - the bestest 16:54
TimToady colomon: perhaps ⊍ is good for calculating degrees of freedom for when you add independent dimensions to the "space" of each named muldimensional switch 16:59
*multi
timotimo colomon is a preschooler and does perl6 development? amazing! 17:00
TimToady moritz: p5ism in your post: if ($!) { 17:05
17:15 cognominal left
TimToady moritz: "interesting varibles" 17:20
17:21 Chillance joined
TimToady and an agreement problem in "write tests that checks" 17:22
17:22 cognominal joined
TimToady dunno whether you meant "a test that checks" or "tests that check" 17:22
(or both :) 17:23
17:29 popl joined
japhb
.oO( Chancellor of the ExChecker )
 17:32
17:33 benabik left 17:38 sftp joined 17:40 cognominal left 17:44 benabik joined
colomon TimToady++ 17:46
TimToady: thoughts on the list infixish of the set/bag union/intersection operators? 17:47
17:47 dakkar left 17:48 kurahaupo joined 17:59 kaleem joined
moritz \o 17:59
btw colomon++'s patch to add the set infix operators nearly tripled parse time of the setting for me 18:00
colomon oh! you mean it's not always so ungodly slow?
some sort of LTM side effect because the operators all start with ( ? 18:02
jnthn Unicode? 18:05
18:06 kaleem left
jnthn yeah 18:07
Just 'cus the lines are commented out doesn't mean they don't slow the parse down a load.
japhb Well, now that they're committed, you can remove them in a separate commit, and then at some later date bring them back by reverting the removal commit. :-) 18:11
jnthn Aye 18:12
moritz r: sub postfix:['!']($x) { [*] 1..$x }; say 5! 18:13
p6eval rakudo 9df474: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/4EL3Se__hO:1␤------> postfix:['!']($x) { [*] 1..$x }; say 5!⏏<EOL>␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤»
moritz r: sub postfix:('!')($x) { [*] 1..$x }; say 5!
p6eval rakudo 9df474: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/Z5D9K6AWuu:1␤------> postfix:('!')($x) { [*] 1..$x }; say 5!⏏<EOL>␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤»
jnthn sub postfix:<!>($x) { [*] 1..$x }; say 5! 18:14
r: sub postfix:<!>($x) { [*] 1..$x }; say 5!
p6eval rakudo 9df474: OUTPUT«120␤»
jnthn Guess just a case of unhandled forms. I think you may be able to draw on some existing code to fix that. 18:15
18:16 Kharec left
moritz jnthn: I think so too. It was just that if it worked right now, I could rewritten the unicode ops as infix:["\c[ascii name of Unicode symbol]"] or so 18:19
diakopter rn: eval 'sub infix:<()>($a, $b) { $a+$b }; say '~('() 'x 500) ~ '3' 18:25
p6eval rakudo 9df474, niecza v24-5-g599cbcb: OUTPUT«3␤» 18:26
diakopter rn: eval 'sub infix:<()>($a, $b) { say $a.WHAT; $a+$b }; say '~('() 'x 4) ~ '3' 18:30
p6eval rakudo 9df474, niecza v24-5-g599cbcb: OUTPUT«Parcel()␤Int()␤3␤»
diakopter rn: eval 'sub infix:<()>($a, $b) { say $a.WHAT; $a+$b }; say '~('() 'x 6) ~ '3'
p6eval rakudo 9df474, niecza v24-5-g599cbcb: OUTPUT«Parcel()␤Int()␤Int()␤3␤»
diakopter rn: eval 'sub infix:<()>($a, $b) { say $a.WHAT; $a+$b }; say '~('() 'x 7) ~ '3'
p6eval rakudo 9df474: OUTPUT«===SORRY!===␤Two terms in a row␤at eval_0:1␤------> WHAT; $a+$b }; say () () () () () () () ⏏3␤ expecting any of:␤ postfix␤ infix or meta-infix␤ infix stopper␤ statement end␤ statement modif…
..niecza v24-5-g599cbcb: OUTPUT«Unhandled exception: Two terms in a row at (eval) line 1:␤------> WHAT; $a+$b }; say () () () () () () () ⏏3␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 1443 (die @ 5) ␤ at /home/p6eval/niecza/src/STD.pm6 line 5597 (STD.panic @ …
FROGGS_ wow, Stage parse : 245.012 18:33
moritz: exactly three times the time as before 18:34
diakopter std: gist.github.com/4278551 18:36
p6eval std a8bc48f: OUTPUT«ok 00:08 83m␤»
colomon working on patch...
GlitchMr .u ⊎ 18:37
phenny U+228E MULTISET UNION (⊎)
GlitchMr I've added "Day 13 – Bags and Sets" to 2012 category 18:38
18:40 kivutar left
colomon GlitchMr++ 18:41
moritz, jnthn: is it enough to delete the offending lines, or does some more substantial change need to be done to the nature of the file?
moritz colomon: deleting the offending lines should be enough 18:42
GlitchMr KeyBag actually reminds me collections.Counter in Python.
>>> from collections import Counter
>>> Counter("the world isn't perfect")
Counter({' ': 3, 'e': 3, 't': 3, 'r': 2, 'c': 1, 'd': 1, "'": 1, 'f': 1, 'i': 1, 'h': 1, 'l': 1, 'o': 1, 'n': 1, 'p': 1, 's': 1, 'w': 1})
colomon moritz: ah, looks like I missed one.
18:43 wamba1 left
colomon there we go! 18:44
TimToady
.oO(a file with fewer than 128 non-ASCII chars would fit into an nfg8)
 18:46
.o((albeit by cheating, insofar as it'd treat precomposed chars as non-composable)) 18:48
jnthn ETOOCHEATY :P 18:49
18:49 noggle_ left 18:50 noggle joined 18:51 wamba joined 18:54 werwerwer joined
colomon patch sent, and I turned on S03-operators/set.t and /bag.t, too. 18:57
GlitchMr Just wondering, isn't is-prime function NP-complete problem?
colomon GlitchMr: no (he said in a slightly uncertain voice) 18:59
moritz I think it's not known whether prime factoring is in NP 19:00
and prime testing is potentially easier than prime factoring 19:01
GlitchMr www.wolfram.com/technology/guide/N...nTracking/
Non-Mathematica should be "Neither Mathematica or Perl 6 ;-)". 19:02
19:02 fgomez joined
werwerwer if prime testing isn't NP then prime factoring isn't NP neither 19:02
moritz werwerwer: do you have any literature on that? 19:05
werwerwer moritz: nope, but i think that complexity_of_prime_factoring <= complexity_of_prime_testing * N^2, so <here goes my previous statement> 19:08
moritz werwerwer: if that were true, RSA wouldn't work 19:10
we already know that the public key isn't a prime, so factoring a 1024bit key would only be in the order 1mio iterations 19:11
(at least if there's a known algorithm, not just a proof of existence)
werwerwer mortiz: i suppose that prime testing is NP complete 19:12
grondilu Where in the specs are the Set operators described in the #13 advent calendar?
19:14 SamuraiJack left 19:16 SamuraiJack joined
moritz en.wikipedia.org/wiki/Primality_test#Complexity 19:16
"The existence of the AKS primality test finally settled this long-standing question and placed PRIMES in P. However, PRIMES is not known to be P-complete, and it is not known whether it lies in classes lying inside P such as NC or L."
19:17 SamuraiJack left
werwerwer RSA is saved :) wikipedia explains it all 19:18
19:26 fhelmberger left 19:51 kurahaupo left 19:52 bbkr_ joined 20:02 masak joined
masak evenin', #perl6 20:02
diakopter masak!
FROGGS_ hi @all
masak feather felt like rebooting, eh?
FROGGS_ or better: hi @*ALL ?
flussence I just read the day 11 post. That's a pretty clever solution, passing code instead of data between threads... 20:03
20:04 Bzek left 20:09 supernovus joined
supernovus So, how do we get panda working properly again now? 20:09
moritz it's broken? 20:11
supernovus Well, apparently when you install a package using it, it's not found, it can't even install other packages depending on the first package. I'm deleting my rakudo and panda source trees, and my ~/.panda and ~/.perl6 folders and recompiling from scratch, see if that fixes it. 20:12
moritz tries 20:14
supernovus For what it's worth, I install a lot of my own packages using a local folder rather than downloading them, but I tried both ways, and neither worked.
Sometimes cleaning out everything and performing a fresh install fixes everything, so I'll let you know what happens :-) 20:16
moritz supernovus: note that stuff is now installed into %*CUSTOM_LIB<site> by default, not ~/.perl6/lib anymore 20:17
Meh, I get No object at index 164 20:18
supernovus moritz: It may have something to do with the change. It wasn't finding the module regardless of where it's being installed now. I installed HTTP::Status and it showed that it was successfully installed, but when I tried to install a library that depends on it, it couldn't find it... again, if nobody else is reporting issues, a fresh install may fix things. 20:20
FROGGS_ can somebody please read the highlighted block please? gist.github.com/4223775#file-insta...xt-L60-L71 20:21
supernovus Hmm, well, I think I may have an answer to my problem... panda has moved repositories... 20:23
moritz hm?
supernovus hmm, never mind, I tried to clone panda and got an error "fatal: remote error: Repository not found." 20:24
colomon grondilu: they're not in the spec.
grondilu: they're in the spec tests, on the other hand, and in both Rakudo and Niecza. 20:25
20:25 kurahaupo joined, bbkr__ joined
supernovus guh, just a typo. 20:25
moritz moritz@jacq:~>panda install Math::RungeKutta 20:28
connect failed: Connection refused
maybe tadzik needs to restart some web service after the last feather reboot? 20:29
20:29 Tene_ joined, Tene_ left, Tene_ joined 20:30 smash_ joined, quietfan1tic joined, thou joined
tadzik yeah 20:31
ok it's up 20:32
20:33 chee2 joined 20:34 crazed- joined, bbkr_ left, wamba left
moritz tadzik: what command did you use? 20:35
(just so instructions are in the logs)
20:35 crazed- is now known as crazedpsyc, szbalint joined
tadzik moritz: perl ~/modules/bin/app.pl 20:35
I added it to crontab @reboot, but doesn't seem to work :/
and I didn't get any email from cron
moritz which feather is that? 20:36
1?
tadzik yes
20:37 spider-mario left
moritz ok, thanks 20:38
20:39 sirrobert left, sorear joined, bruges_ joined 20:40 aindilis` joined 20:41 _jaldhar_ joined 20:43 _jfried joined, cosimo joined 20:44 marloshouse_ joined, wamba joined 20:46 BooK_ joined
supernovus Crossing fingers, I have my PATH updated with the new versions. Installing my base set of libraries now. 20:47
20:47 p6eval left, jaldhar_ left, drbean left, Bucciarati left, jfried left, bruges left, marloshouse left, aindilis left, cosimo_ left, marloshouse_ is now known as marloshouse 20:48 drbean joined 20:49 p6eval joined, ChanServ sets mode: +v p6eval
moritz panda install Math::RungeKutta # succeeded for me 20:50
20:53 benabik left
felher colomon++ # blog post about sets :) 20:53
masak colomon++
20:53 Bucciarati joined
masak 'night, #perl6 20:53
supernovus panda installed everything successfully for me this time 20:54
tadzik: is the source to the module app available? 20:55
felher 'night, masak 20:56
supernovus have a good one masak!
tadzik supernovus: if not, it should be :)
20:57 doy left 20:58 _jfried is now known as jfried
supernovus tadzik: It's not in the 'ecosystem' or 'modules.perl6.org' or 'panda' repositories, which is why I'd asked. 21:01
tadzik yeah, it's not published
I'll put it somewhere, in the ecosystem maybe? 21:02
probably won't bother anyone there :)
I'll just clean it up a bit
21:02 kivutar joined 21:03 benabik joined
supernovus ecosystem seems like a sensible location, maybe in a 'app' or 'service' subdirectory? 21:03
moritz or 'server' maybe :-)
supernovus That works too :-)
Does the app auto-generate the projects.json file, or depend on the 'update.pl' already in ecosystem? Or maybe I should just wait and read the code. Impatient as I am ;-) 21:04
tadzik it needs a cronjob that'll run updatelist.pl 21:05
# DO NOT EVER USE THAT FEATURE IN PRODUCTION 21:06
# OR TINY KITTENS SHALL DIE WITH LOTS OF SUFFERING 21:07
yeah, uh, I guess I should've read that earlier :P
supernovus lovely warning... oO
21:09 GlitchMr left
tadzik there we are 21:11
dale<TAB><TAB> 21:12
21:12 tadzik is now known as da1ek
da1ek OMG GUYS, A COMMIT TO perl6/ecosystem! github.com/perl6/ecosystem/commit/26bf37641 21:12
21:12 da1ek is now known as tadzik
supernovus lol 21:13
21:13 doy joined
FROGGS_ hehe 21:14
tadzik++ # humor :o)
supernovus Nice simple app. 21:18
tadzik feel free to improve/clean it up 21:19
FROGGS_ tadzik: do you have time to read that? gist.github.com/4223775#file-insta...xt-L60-L71 21:21
popl tiny kittens are overrated.
FROGGS_ I'd like to have some feedback
popl: what about huge kittens? mine is bitte than my almost-three-year-old son when it stands upright on its hind paws 21:22
s/bitte/taller/ # I think 21:23
no idea why I wrote "bitte"
21:24 kivutar left
tadzik FROGGS_: hm. I guess MANIFEST could also keep the current "best" module version, and panda activate would then update that 21:25
by default it'd be the newest one or so 21:26
FROGGS_ but you will have four MANIFEST files (home, site, vendor, perl)
tadzik oh, yes
FROGGS_ every file has a "best" module version 21:27
supernovus I'd say if a user doesn't specify a desired version/auth in their use statement, we shouldn't need to load the MANIFEST files either, just use the current logic, and look in home, site, vendor and perl (in that order) for a matching module. It will use the currently activated module. 21:29
popl FROGGS_: Cats in general.
FROGGS_ supernovus: if you do "use Foo::Bar", you will maybe find home/Foo/Bar.pm, site/Foo/Bar.pm, and so on 21:30
supernovus: which one is the right one?
supernovus FROGGS_: If it exists in "home", that is the right one, otherwise look in "site", and so on. The manifest is only required if a specific version is requested. Also, how does @*INC interact with %*CUSTOM_LIB? Do we search @*INC first, before 21:31
FROGGS_ and what if the module is called Föö::Bär? whats the filename we should look for on a filesystem that can't handle ö and ä?
diakopter then the module writer obviously didn't want the module to be used on that platform 21:32
FROGGS_ supernovus: according to S11 the module with the highest version is the best, not the first we find
diakopter: but it should be possible to use unicode names...
diakopter FROGGS_: sure, but you're just limiting folks' usage of your software 21:33
supernovus Hmm, in which case we'll have to read the MANIFEST files, every time. I'd suggested that once before and was shot down as being excessive overhead, which is why I'd thought that maybe the manifest should only be required if we ask for a specific version/auth. 21:34
FROGGS_ diakopter: no, since there is a file that tells you what the right filename for a given module name is
supernovus: well, I know that it might be expensive to read four files... I guess asking TimToady won't hurt 21:36
supernovus Of course if we cache the MANIFEST information in a binary format that doesn't require significant parsing overhead (still having a plain text version that we can rebuild from if necessary) then that could mitigate some of the objections to having to read the manifest each time.
FROGGS_ supernovus: good point 21:38
supernovus I would also recommend using JSON as the plain text format for the MANIFEST rather than inventing yet another text-based data storage format. 21:39
FROGGS_ supernovus: btw, your question about @*INC and %*CUSTOM_LIB is answered somewhere on that page
supernovus: sure, is just an example
21:40 leont joined
FROGGS_ @*INC comes without MANIFEST file, and does pretty much the same as Perl 5, if we find something, we will live with that 21:41
21:43 leont left
supernovus Given the way @*INC is described in the document, I think this could work well. Start with textual MANIFESTS, and add a binary format later. For a significant metadata speedup, the binary versions could be calculated using the combined data from all custom lib above. So the binary cache in 'perl' would contain just the manifest from 'perl', but the cache in 'site' for instance would contain 'site', 'vendor', and 'perl', and so for 21:45
th. So when using a binary cache, we'd only have to load one file.
FROGGS_ k 21:46
21:46 benabik_ joined, benabik left, benabik_ is now known as benabik
supernovus Again, binary caches being an optional feature that any given implementation may or may not support. 21:46
21:49 sivoais left
supernovus I guess one thing about using JSON as the data format for plain text MANIFEST files, is that JSON itself, and likely the JSON::Tiny library would have to become a part of the implementation (and likely the specification.) 21:50
21:50 benabik left
diakopter supernovus: rakudo would probably use a binary packed form that can be directly mmap'd as data structures. 21:51
supernovus diakopter: Yeah, I was thinking the binary format may become the standard, and leave it up to the installer, such as panda to build both JSON and binary manifest files. Rakudo wouldn't know anything about the JSON version then. 21:53
21:58 benabik joined 22:07 skids left 22:09 kaare_ left 22:10 sahadev left
FROGGS_ gmight and thanks for your help 22:11
22:14 Chillance left
tadzik I can has a draft of my advent calendar post: gist.github.com/4280548 22:14
Criticism will be very welcome 22:15
22:15 Tene_ is now known as Tene 22:25 supernovus left
diakopter nice 22:25
"coroutine to be ran" -> "coroutine to be run" 22:26
22:29 benabik left, cognominal joined
tadzik thanks, fixed 22:30
diakopter consider having schedule return a wrapper Cororoutine object that sets a flag when it's done, so the caller can optionally block/wait until a particular scheduled thing is done 22:31
er, s/roro/ro/ 22:32
22:32 whiteknight joined
tadzik I guess async() could return an object you can wait() on 22:34
diakopter oh, yes, I meant async, not schedule 22:35
kurahaupo diakopter: (that sounded like a skit from a famous British sitcom)
diakopter tadzik: probably you should show an example of using Coroutines to do the same thing as the LWP example 22:36
tadzik oh, yes 22:39
implement AnyEvent::HTTP the good way :)
diakopter also, I'm not clear on how the mainline ever actually passes control to the things that are scheduled
ie. what's actually doing the event loop 22:40
something in parrot? 22:41
tadzik nah, it's implementing in Pure Perl
and it sucks, too ;)
diakopter for the IO thing?
s/thing/stuff,module/ 22:42
tadzik hm, I'm confused :) 22:43
github.com/tadzik/MuEvent/blob/mas...MuEvent.pm is what we're at
diakopter so only one connection can be active at once? 22:44
tadzik no 22:46
you can have any number you want
diakopter how would one do that
if one thing sleeps for 10 seconds, does an every-1-second timer still fire during those 10 seconds? 22:49
or, s/sleeps/counts to a trillion/ if you like 22:54
tadzik yeah, it's not preemptive
if something runs for 10 seconds it will block the execution for 10 seconds
diakopter I don't see how more than one connection can be active 22:55
isn't a connection handled from start to finish in one execution unit (without yielding)?
tadzik like a network connection? Npe 22:57
see ttjjss.wordpress.com/2012/04/02/asy...in-perl-6/ for example
23:00 sftp left 23:01 PacoAir left
diakopter tadzik: that doesn't answer my question, afaict 23:08
tadzik: socket.pmc talks about many "asynchronous versions" of various methods, but I can't find code for them. 23:09
they all look synchronous to me. 23:10
tadzik ah 23:11
I've not much idea about socket.pmc
diakopter reading INET.pm
accept looks blocking 23:13
tadzik it is 23:14
you have .poll() though
it works for one socket at a time though
diakopter that's what I was asking
23:15 chee2 left, chee2 joined 23:16 chee2 is now known as chee
diakopter I don't see your http_get 23:20
oh there it is
so, when you're polling the sockets, you can only run callback at once. 23:23
tadzik that's right
23:23 wamba left 23:24 wamba joined
diakopter so a user can add a callback to fetch a url, but say they don't want to close the socket when they're done; they want to issue another HTTP request 23:25
they can add another MuEvent::socket with their callback that doesn't include .close? 23:27
ok, I fully understand it now 23:28
the parrot docs are disingenuous
"there are these async versions... nope just kidding" 23:29
23:31 wamba left 23:33 skids joined
diakopter tadzik: how does one tell the MuEvent loop to stop 23:35
(from one of the callbacks, obviously)
tadzik sorry, I'm a bit sleepish :) 23:38
I'll try to anwer you tomorrow, now I'm having trouble computing sentences
good night o/ 23:39
diakopter o/
23:44 stopbit left 23:47 MayDaniel left 23:50 popl left 23:59 sivoais joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!