»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.
ggoebel jnthn: regarding looking up the unicode codepoint for a character name... you may want to google: apache fop getcodenameforglyph 00:04
source code and binaries aren't much different in size from icu though... 00:05
00:09 alec__ joined 00:14 AndChat|98100 joined, AndChat|98100 left 00:15 AndChat|98100 joined 00:17 AndChat|98100 left 00:28 mh joined 00:29 thou joined
ggoebel phenny tell jnthn: or perhaps just process www.unicode.org/Public/6.2.0/ucd/NamesList.txt 00:36
dalek rl6-most-wanted: 504e1fa | (Geoffrey Broadwell)++ | most-wanted/modules.md:
Continuing through task-belike-*
 00:37
00:37 mh left 00:38 AndroUser2 joined, alec__ left 00:43 AndroUser2 left, alec__ joined
[Coke] phenny: ask curtispoe if github.com/Ovid/Config-Tiny is gone or moved. 00:48
phenny [Coke]: I'll pass that on when curtispoe is around.
[Coke] japhb: ^^ looks like someone started on an ini configurator 00:51
japhb [Coke], commits welcome (easier now that it's in the perl6 org) :-) 00:54
Or I will eventually get to it if it's listed in the Perl 6 module list
[Coke] I found a blog post that said he started it, but it's gone.
00:55 alec__ left
japhb :-( 00:55
01:01 b1rkh0ff left, anuby joined 01:06 PhilipS__ joined 01:07 PhilipS__ left 01:16 hypolin joined, drbean joined 01:17 anuby left 01:21 alec__ joined 01:23 dayangkun left
dalek ecs: 703df65 | lue++ | S02-bits.pod:
[S02] Fixed double angle bracketed formatting code.
 01:26
01:27 alec__ left 01:29 Chillance left 01:31 alec__ joined 01:38 shinobicl joined
shinobicl hi... are there any plans for 'improving' META.info in perl6 packages? Like the thing that Maven's "pom.xml" does in java?. 01:38
01:41 Exodist left 01:52 Exodist joined 01:55 skids joined
dalek rl6-most-wanted: a792408 | (Geoffrey Broadwell)++ | most-wanted/modules.md:
More task-belike-* results
 02:00
02:03 sevin joined
colomon n: say 4.bin 02:10
p6eval niecza v24-23-g0520c7c: OUTPUT«Unhandled exception: Unable to resolve method bin in type Int␤ at /tmp/wUqKivx7va line 1 (mainline @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4233 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 4234 (module-CORE @ 580) ␤ at /home/p6…
02:14 dayangkun joined 02:16 MikeFair left 02:17 MikeFair joined 02:24 alec__ left 02:30 alec__ joined 02:31 s1n joined 02:33 sevin left, lustlife joined
dalek rl6-most-wanted: 284cccf | (Geoffrey Broadwell)++ | most-wanted/modules.md:
Finish going through current set of task-belike-*
 02:34
02:35 s1n left
colomon forgive me, I have used a side-effect in a grep block. 02:41
diakopter karma will introduce wrongness into your primality tests 02:48
benabik retroactively. 02:49
02:52 FROGGS_ joined 02:56 FROGGS left
dalek rl6-most-wanted: 0fa6f10 | (Geoffrey Broadwell)++ | tools/prettify-json:
Sort prettified JSON
 02:56
colomon I'm distressingly pleased by the code: gist.github.com/colomon/4974879 03:00
benabik colomon: It's definitely clever. By which I mean I don't understand it. 03:02
colomon benabik: I had to do research.
benabik colomon: You might want to add comments so people don't have to research it to figure it out. ;-)
colomon and it's probably overkill.
or wrong. 03:03
I've actually twisted my brain in knots trying to figure out how to gracefully test it.
The grep selects things from @items based on the bits of $gray. 03:05
$gray is the Gray number corresponding to $i. Though now that I'm thinking about it I'm not sure why it's useful here.
dalek rl6-most-wanted: 808d96f | (Geoffrey Broadwell)++ | most-wanted/modules.md:
Start marking up existing WIP modules from the Perl 6 modules list
 03:08
colomon yeah, the Gary number thing is totally useless in my implementation. 03:13
at least according to my tests.
03:13 s1n joined
colomon oh, my routine is very inefficient in this case. 03:16
03:18 MikeFair left 03:21 s1n left
colomon so probably I need to figure a better way to use the Gray number thing. 03:21
03:29 orafu left, orafu joined 03:32 shinobicl left 04:22 dayangkun left 04:31 dayangkun joined 04:33 odoacre joined
skids colomon: Are you trying to figure out a useful ordering for the combos? en.wikipedia.org/wiki/Combinatorial...ber_system might help. 04:40
04:43 kaleem joined 04:48 preflex_ joined, preflex left 04:49 preflex_ is now known as preflex 04:59 ElDiabolo joined
bonsaikitten hrm. can anyone help me figure out a parrot issue? it's not detecting perldoc, which is installed and as far as I can tell functional 05:03
I don't quite understand the Configure.pl and the machinery behind it yet
moritz bonsaikitten: maybe try #parrot? 05:05
(on irc.perl.org, that is)
dalek rl6-most-wanted: 106229f | (Geoffrey Broadwell)++ | most-wanted/bindings.md:
WIP bindings from the Perl 6 modules list
rl6-most-wanted: 9cc09f6 | (Geoffrey Broadwell)++ | most-wanted/modules.md:
WIP modules from the Perl 6 modules list
japhb OK all -- that's a reasonable first take on perl6-most-wanted, but it's far from perfect. 05:06
bonsaikitten moritz: hrm ok
japhb I'm burning out a bit, so if there is anyone else interested in keeping perl6-most-wanted, please do make commits at will 05:07
*keeping up 05:08
I guess before I completely let go, I should note the current location on the most-wanted wiki page ...
Done. 05:12
*whew*
moritz japhb++ # cleanup before letting go
japhb Feels good to have it off my mind. 05:13
I knew it needed to be done, but it was a ton of pretty much slogging drudgery work to get it to a stage useful to the community at large. 05:14
(And hopefully it is, at this point. :-)
05:21 sevin joined 05:43 bluescreen10 left, ElDiabolo left 05:57 ElDiabolo joined 06:02 ElDiabolo left 06:28 ElDiabolo joined
ElDiabolo when was operator overloading added to rakudo ? 06:31
06:32 quester joined
TimToady it's pretty much been part of the design from the beginning, though not always implemented quite right 06:35
and we don't really talk about "operator overloading", because it falls out from multiple dispatch and the fact that operators are just funny names for functions 06:36
ElDiabolo TimToady, I am getting a "Cannot add tokens of category 'operator'" from my debian sid rakudo and wanted to check if this is just not implemented there. It is however a 2012-10 thingy, so thats not the problem. 06:39
TimToady there is no category "operator" 06:41
you must be more specific: "prefix", "postfix", "infix" etc.
ElDiabolo TimToady, Uh, thanks :) 06:42
benabik r: multi postfix:<!>(1) { 1 }; multi postfix:<!>($i) { $i * ($i - 1)! }; say 5! 06:45
p6eval rakudo b9ee89: OUTPUT«120␤»
06:51 FROGGS_ left, rindolf joined 06:57 lue left 07:07 rindolf left 07:08 rindolf joined 07:10 lue joined, sizz left, sizz joined 07:15 bapa left, ElDiabolo left 07:17 PacoAir joined, PacoAir left 07:18 sizz_ joined 07:19 sizz left 07:21 bapa joined 07:22 sftp left 07:26 rindolf left, rindolf joined 07:27 ElDiabolo joined
ElDiabolo r: map {say "::$_";} <== (1, 2, 3); 07:27
p6eval rakudo b9ee89: OUTPUT«::1␤::2␤::3␤»
07:33 sevin_ joined 07:34 sevin left 07:38 rindolf left, rindolf joined 07:40 daxim_ joined, daxim left 07:43 kaleem left, kaleem joined 07:45 SamuraiJack joined, daxim_ left, proller joined, daxim_ joined, daxim_ is now known as daxim 07:47 arlinius left 07:50 FROGGS joined
FROGGS morning 07:50
07:53 domidumont joined 07:57 tadzik1 joined, tadzik left 07:58 tadzik1 is now known as tadzik
ElDiabolo r: sub prefix:<select>(@table) is looser(&infix:<+>) {} 07:59
p6eval rakudo b9ee89: ( no output )
ElDiabolo r: sub prefix:<select>(@table) is looser(&infix:<+>) is export{} 08:01
p6eval rakudo b9ee89: OUTPUT«===SORRY!===␤Missing block␤at /tmp/LkYh9EmZB9:1␤------> able) is looser(&infix:<+>) is export{}⏏<EOL>␤ expecting any of:␤ statement list␤»
moritz ElDiabolo: add a space before the { 08:02
ElDiabolo: otherwise it's parsed as a postcircumfix
like in %hash{...}
08:02 rindolf left 08:03 quester left, rindolf joined
ElDiabolo moritz, Thx :) 08:06
r: sub prefix:<select>(@table) is looser(&infix:<+>) is export {} 08:07
p6eval rakudo b9ee89: ( no output )
FROGGS ElDiabolo: why a select-prefix and not just a function? 08:11
08:11 am0c joined
ElDiabolo FROGGS, Thoughtlessness 08:12
08:14 dayangkun left 08:17 domidumont left
ElDiabolo Thx. 08:17
08:17 arlinius joined
ElDiabolo r: map {say "::$_";} <== (1, 2, 3); 08:18
p6eval rakudo b9ee89: OUTPUT«::1␤::2␤::3␤»
ElDiabolo r: map {say "::$_";}, (1, 2, 3);
p6eval rakudo b9ee89: OUTPUT«::1␤::2␤::3␤»
08:18 domidumont joined
ElDiabolo The above does nothing in my test script unless I do my @dumy = map ... . Any hints on why ? 08:20
moritz ElDiabolo: map itself is lazy 08:21
ElDiabolo: and it is only evaluated when there's some indication that the elements are needed 08:22
assignment to an array is such an indication
as well as sink (void) context
ElDiabolo Good. Thx.
moritz ElDiabolo: though it might be interesting who you managed to not put it into sink context (or maybe sink context detection is imperfect) 08:23
s/who/how/
ElDiabolo moritz, Sink context is wrong in my rakudo, I think. 08:29
08:29 Ovidius joined
ElDiabolo moritz, map {say "::$_";}, @table; as one line in a script fails. 08:30
fails to say anything, to be precise.
08:32 fhelmberger joined
FROGGS I would might do: say map {"::$_";}, @table; 08:32
Ovidius Can anyone point me to recent benchmarks for Rakudo? Someone's taking issue with my performance comments on my recent blog post.
08:35 jerome left
ElDiabolo FROGGS, Yeah, I'll work around, just wondering if a bug report makes sense. 08:35
FROGGS ya, thats the question, bug or feature? 08:36
08:37 sevin_ left
moritz ElDiabolo: what rakudo version are you using? 08:40
ElDiabolo OK, now I can do say select {"--$_";} from @table;
This means I can tackle the first (as masak has pointed out) problem. Table aliases. 08:41
08:41 dayangkun joined
ElDiabolo What I would like to do is select {"--$t";} from @table $t; 08:42
rindolf ElDiabolo: what does select do here? 08:43
ElDiabolo rindolf Nothing by now. 08:44
rindolf ElDiabolo: Nothing yet?
ElDiabolo moritz, 2012.10-1 (as debian sid calls it). 08:45
08:47 cosimo left
ElDiabolo rindolf, I am currently learning p6 by implementing a LINQuish SQL lookalike. 08:47
rindolf ElDiabolo: ah, OK.
moritz ElDiabolo: that explains it. sink context was implemented in 2013.01 08:48
(well, that was the first release containing it)
ElDiabolo The problem with the table alias $t is 1. that the statement declares a local variable and I don't know how to implement this and 2. that said $t is first used and then declared . 08:49
arnsholt To implement aliases you're probably going to need macros 08:50
ElDiabolo moritz, Ah, thx. Hope the debian package maintainer stays active :)
08:51 dayangkun left
arnsholt masak++ has implemented a fair chunk of macros, but I don't think creating variable bindings is supported yet 08:51
ElDiabolo rindolf, I would prefer to stay as close as possible to SQL.
rindolf ElDiabolo: OK. 08:52
08:53 cosimo joined
moritz ElDiabolo: what do you want the syntax to look like? 08:56
ElDiabolo moritz, SQL. 09:00
moritz ElDiabolo: well, SQL doesn't use sigils
ElDiabolo moritz, hehe. Embedded SQL to data. 09:01
09:03 dayangkun joined
moritz ElDiabolo: you might be able to pull of some tricks with self-declared formal parameters 09:04
for example if you look at perlgeek.de/blog-en/perl-6/physical...lling.html
ElDiabolo moritz, In the long run an SQL to the database could be implemented this wax, too.
09:04 labster left
moritz y_acceleration => { $:force / $:mass }, 09:04
here $:force and $:mass are named parameters
which can be discovered from the outside by introspecting the signature 09:05
which is how Math::Model tracks interdependencies between the various expresions
ElDiabolo s/wax/way/
jnthn morning o/
moritz so, you might be able to get something running along the lines of 09:06
09:06 thou left
moritz { select $:alias.name from table as $:alias } 09:06
\o jnthn 09:07
ElDiabolo moritz, Oh, cool. I thought that I may have to get away from using variable declaration for aliases but did not know how yet. 09:08
FROGGS moritz: is it possible to declare variables like $<id> for use in a while loop when doing: while $sth.fetch { ... } ? 09:13
so you dont need to explicitly bind them? 09:14
moritz FROGGS: $<id> just desugars to $/{ 'id' }
FROGGS hmmm
moritz but I'd do it like this:
for $sth.rows { say $:id, $:name }
tadzik $:id? 09:15
std: $:id
p6eval std 7551b8f: OUTPUT«===SORRY!===␤Placeholder variable $:id may not be used outside of a block at /tmp/ZOWbTMREwf line 1:␤------> <BOL>⏏$:id␤Check failed␤FAILED 00:00 41m␤»
tadzik hm
moritz r: { say $:id }.(id => 42)
p6eval rakudo b9ee89: OUTPUT«42␤»
moritz tadzik: like $^id, only named
tadzik today I learned
nice 09:16
FROGGS so, $sth.rows returns hash and it will just work like that?
r: my @rows = { id => 42 }, { id => 43 }; for @rows { say $:id } 09:18
p6eval rakudo b9ee89: OUTPUT«Required named parameter 'id' not passed␤current instr.: '' pc 311 ((file unknown):146) (/tmp/MUwLLQ2ri_:1)␤called from Sub '' pc 256919 (src/gen/CORE.setting.pir:116553) (src/gen/CORE.setting:5590)␤called from Sub 'reify' pc 255934 (src/gen/CORE.setting.pir:116188…
moritz erm, no :(
I didn't consider that
you'd have to do differently after all
$sth.rows: { $:id, $:name }
$sth.rows: { say $:id, $:name }
jnthn r: my @rows = { id => 42 }, { id => 43 }; for @rows { say .<id> }
p6eval rakudo b9ee89: OUTPUT«42␤43␤»
moritz then $sth.rows can introspect the closure, and decide which arguments to pass into it 09:19
or like jnthn++ said :-)
jnthn r: my @rows = { id => 42 }, { id => 43 }; for @rows -> (:$id) { say $id }
p6eval rakudo b9ee89: OUTPUT«42␤43␤»
jnthn r: my @rows = { id => 42 }, { id => 43 }; for @rows -> $/ { say $<id> }
p6eval rakudo b9ee89: OUTPUT«42␤43␤»
jnthn :)
moritz my @rows = { id => 42 }, { id => 43, name => "foo" }; for @rows -> (:$id) { say $id }
r: my @rows = { id => 42 }, { id => 43, name => "foo" }; for @rows -> (:$id) { say $id } 09:20
p6eval rakudo b9ee89: OUTPUT«42␤Unexpected named parameter 'name' passed in sub-signature␤current instr.: '' pc 383 ((file unknown):210529660) (/tmp/fDATWtoqwK:1)␤called from Sub '' pc 256919 (src/gen/CORE.setting.pir:116553) (src/gen/CORE.setting:5590)␤called from Sub 'reify' pc 255934 (src/g…
moritz that's the disadvantage if you go that route: you have to use all the elements
jnthn Or throw in a *%
But yes
FROGGS okay, maybe .<id> might be the sanest 09:21
r: my @rows = { id => 42 }, { id => 43 }; for @rows { say "$.<id>" }
p6eval rakudo b9ee89: OUTPUT«===SORRY!===␤Non-variable $ must be backslashed␤at /tmp/N9LdLJTryB:1␤------> > 42 }, { id => 43 }; for @rows { say "$⏏.<id>" }␤ expecting any of:␤ argument list␤ prefix or term␤ prefix or meta-prefix␤ …
FROGGS r: my @rows = { id => 42 }, { id => 43 }; for @rows { say "&.<id>" } 09:22
p6eval rakudo b9ee89: OUTPUT«&.<id>␤&.<id>␤»
FROGGS hmmm
r: my @rows = { id => 42 }, { id => 43 }; for @rows { say "$(.<id>)" }
p6eval rakudo b9ee89: OUTPUT«42␤43␤»
FROGGS ahh, right
09:26 ObseLeTe joined 09:27 spider-mario joined
dvj perl6: my $a = "#12"; my $c = $a.match(/^\# (\d+) $/); 09:33
p6eval rakudo b9ee89: ( no output )
..niecza v24-23-g0520c7c: OUTPUT«===SORRY!===␤␤No unspace allowed in regex; if you meant to match the literal character, please enclose in single quotes ('#') or use a backslashed form like \x23 at /tmp/e6pQHSS86S line 1:␤------> my $a = "#12"; my $c = $a.match(/^\…
dvj perl6: my $a = "#12"; my $c = $a.match(/^\# (\d+) $/); say $c;
p6eval niecza v24-23-g0520c7c: OUTPUT«===SORRY!===␤␤No unspace allowed in regex; if you meant to match the literal character, please enclose in single quotes ('#') or use a backslashed form like \x23 at /tmp/mmdva6p85p line 1:␤------> my $a = "#12"; my $c = $a.match(/^\…
..rakudo b9ee89: OUTPUT«｢#12｣␤ 0 => ｢12｣␤␤»
dvj perl6: my $a = "#12"; my $c = $a.match(/^'#' (\d+) $/); say $c; 09:35
p6eval rakudo b9ee89, niecza v24-23-g0520c7c: OUTPUT«｢#12｣␤ 0 => ｢12｣␤␤»
dvj My rakudo installation gives me: 0 => ｢12｣
Is the regex matching wrong? 09:36
arnsholt It matched the whole string, but you're only capturing the numbers 09:37
moritz dvj: I agree with what rakudo and niecza do here on the channel
dvj: if you just want the string, use .Str or prefix ~
dvj rakudo: my $a = "#12"; my $c = $a.match(/^'#' (\d+) $/); say ~$c; 09:38
p6eval rakudo b9ee89: OUTPUT«#12␤»
moritz the 0 => ｢12｣ means "the first positional capture (with index 0) matched the string '12')
in general, regex matches return Match objects, which contain much richer information than just the matched string 09:39
doc.perl6.org/type/Match
dvj rakudo: my $a = "#12"; my $c = $a.match(/^'#' (\d+) $/); say $c.Int;
p6eval rakudo b9ee89: OUTPUT«Cannot convert string to number: base-10 number must begin with valid digits or '.' in '⏏#12' (indicated by ⏏)␤ in method Int at src/gen/CORE.setting:10512␤ in method Int at src/gen/CORE.setting:2392␤ in block at /tmp/IEz8E8D2Qq:1␤␤»
moritz r: my $a = "#12"; my $c = $a.match(/^'#' (\d+) $/); say $c[0].Int 09:40
p6eval rakudo b9ee89: OUTPUT«12␤»
moritz r: '#12' =~ /'#' (\d+) $ / and say $0.Int
p6eval rakudo b9ee89: OUTPUT«===SORRY!===␤Unsupported use of =~ to do pattern matching; in Perl 6 please use ~~␤at /tmp/XBUFUM_rUh:1␤------> '#12' =~⏏ /'#' (\d+) $ / and say $0.Int␤»
moritz r: '#12' ~~ /'#' (\d+) $ / and say $0.Int
p6eval rakudo b9ee89: OUTPUT«12␤»
dvj ah, ok. I see. Takes some time getting used to stuff being wrapped into Match and Junctions coming from Perl5 :)
moritz dvj: yes. But once you've got used to Match objects, you won't ever want to go back 09:41
especially when you try to extract the value of quantified capturing groups in Perl 5
and I mean all values, not just the last one :-) 09:42
09:42 kresike joined
kresike hello all you happy perl6 people 09:42
09:43 ElDiabolo left
mathw happy people? where? 09:51
FROGGS me is somewhat happy
diakopter I'm happy
FROGGS hi kresike
09:52 proller left
kresike FROGGS, o/ 09:53
masak hello happy kresike 09:55
and good forenoon, #perl6
09:55 dakkar joined
mathw actually, I'm fairly happy this morning 09:55
although I have lots of dull work to do
kresike hello masak o/ 09:56
FROGGS mathw: well, "dull work" is the definition of $work, isnt it? 10:02
or better: $day_job
10:02 hypolin left
diakopter FROGGS: nah, some people are blessed with jobs they enjoy 10:03
mathw FROGGS: Not always! My other job's usually rather more interesting. However in two weeks I'm starting a new job to replace both of them so I don't know how interesting that one's going to be. I'm hoping for 'occasionally' or better.
FROGGS I wanna have an interesting job too ó.ò 10:04
mathw so does everyone else 10:05
however, the last year or so I've come to really value a stable job too
even if it's a bit less interesting
10:08 dayangkun left
masak FROGGS: my $dayjob is not dull. 10:09
mathw my saturday job certainly isn't dull :D 10:12
10:14 jerome joined
moritz 's job isn't dull, but being paid to work on Perl 6 would be even better :-) 10:14
mathw :)
masak I guess that's why we're here. 10:23
mmcleric [Coke], japhb: about play-perl and p6mw - play-perl.org/quest/512203d8e05a54a63a000022 - can you comment on that? 10:40
tadzik \o/ 10:41
mmcleric ohai 10:43
moritz \o mmcleric 10:48
10:54 mmcleric left
FROGGS r: say "foobar" ~~ / [f|oo||b||a||r]+ / # is it valid to mix | and || ? 11:04
p6eval rakudo b9ee89: OUTPUT«｢foobar｣␤␤»
11:18 b1rkh0ff joined 11:23 salv0 joined 11:36 ObseLeTe left 11:39 Targen left 11:42 rindolf left 11:46 xinming left
moritz FROGGS: they simply have different precedences 11:47
S05 has the details 11:48
11:49 xinming joined 11:57 hubba joined, hubba left 12:00 renormalist left, brrt joined
masak TimToady: why are "replication" (x xx) and "concatenation" (~) their own precedence levels, rather than "multiplicative" and "additive", respectively? 12:01
moritz nr: say "we have " ~ 1000 * rand " open tickets" 12:04
p6eval niecza v24-23-g0520c7c: OUTPUT«===SORRY!===␤␤Two terms in a row at /tmp/IoD0owAAoI line 1:␤------> say "we have " ~ 1000 * rand ⏏" open tickets"␤␤Parse failed␤␤»
..rakudo b9ee89: OUTPUT«===SORRY!===␤Two terms in a row␤at /tmp/0fHdHa4z2V:1␤------> say "we have " ~ 1000 * rand ⏏" open tickets"␤ expecting any of:␤ postfix␤ infix or meta-infix␤ infix stopper␤ statement end␤ sta…
moritz nr: say "we have " ~ 1000 * rand ~ " open tickets"
p6eval niecza v24-23-g0520c7c: OUTPUT«we have 965.92582015596611 open tickets␤»
..rakudo b9ee89: OUTPUT«we have 1.56008228393034 open tickets␤»
moritz nr: say "we have " ~ 20 + 100 * rand ~ " open tickets"
p6eval niecza v24-23-g0520c7c: OUTPUT«we have 102.22957611187806 open tickets␤»
..rakudo b9ee89: OUTPUT«we have 31.8332726971374 open tickets␤»
moritz ok, this one actually illustrates the point I wanted to make :-)
masak yes, I accept your point. that's the consequence of the separation of precedence levels -- we can leave off the parens there. 12:05
but... hm. the fact that we can there is still (mildly) surprising to me.
12:05 mberends joined
masak mberends! \o/ 12:05
moritz: I guess my expectations differ from the precedence table, and that's it. 12:06
mberends masak! moritz! 12:07
moritz mberends!
tadzik \o/ 12:10
masak .oO( tvtropes.org/pmwiki/pmwiki.php/Main...uousScream ) 12:11
mberends!
moritz: as in, I'd totally write that as "we have " ~ (20 + 100 * rand) ~ " open tickets" -- even *knowing* that the precedence table is that way. 12:12
tadzik I'll go for 12:14
(damnit, Return key) "foo bar {100 * 20 + rand} baz"
masak tadzik: then for the sake of the example let's make it $prefix ~ (20 + 100 * rand) ~ $suffix 12:16
tadzik well, I'd still go for "" with a block inside :) 12:17
masak I guess the situation can be summarized as "the separation of those levels in the precedence table allows for numeric calculation to be done inside a chain of ~s, without parenthesizing" 12:18
tadzik I guess so
moritz aye
masak I think my surprise comes from the fact that I consider infix:<x> to be a multiplication operator, and thus wouldn't be at all surprised if it were sorted with infix:<*> 12:19
...and indeed, that's how Perl 5 does it.
12:20 SmokeMachine joined 12:21 SmokeMachine left, SmokeMachine joined
moritz fwiw I don't feel strongly either way 12:22
masak I'm curious because (a) it's a difference from Perl 5, and (b) I know TimToady wouldn't frivolously add prec levels, so there has to be a correspondingly good reason for it. 12:23
my suspicion is that the reason comes from TimToady, not from the RFCs. :)
12:30 hoelzro|away is now known as hoelzro
moritz r: my $x = 42; $x; say $x 12:39
p6eval rakudo b9ee89: OUTPUT«42␤»
moritz huh, still doesn't work
have I forgotten again to push my commits?
12:39 mmcleric joined
dalek kudo/nom: 018b73b | moritz++ | src/Perl6/ (2 files):
warn about variables in sink context
 12:41
12:43 renormalist joined 12:47 kaleem left
masak moritz: bloody distributed version control systems. people commit and commit, but what good does it do? it's all local. :P 12:56
arnsholt =D
moritz masak: indeed. And it doesn't help that I program rakudo on four different hosts 12:57
arnsholt My boss (to the extent that I have on, being in academia) tried to sell me on his "industrial-strength" version control system (SVN) the other day
masak "I forgot to push" is the new "the dog ate my homework".
moritz and occasionally remotely from one on the other
arnsholt All in good humour of course, but I've still to needle him about the deficiencies of SVN =) 12:58
13:03 kaleem joined
dvj perl6: my $a = "asdf"; $a.substr(/\s+.*$/, ''); 13:09
p6eval niecza v24-23-g0520c7c: OUTPUT«Unhandled exception: System.InvalidCastException: Cannot cast from source type to destination type.␤ at Anon.1.MAIN.3e9c0f7f-de90-4af4-8b94-dbbd4ca5c7c3.C1ANON (Niecza.Frame ) [0x00000] in <filename unknown>:0 ␤ at Niecza.Kernel.RunCore (Niecza.Frame& cu…
..rakudo b9ee89: OUTPUT«No such method '!cursor_start_all' for invocant of type 'Int'␤ in regex at /tmp/OkAeclaTjF:1␤ in method substr at src/gen/CORE.setting:3904␤ in block at /tmp/OkAeclaTjF:1␤␤»
13:10 kaleem left, kaleem joined
dvj oh, I was using substr instead of subst 13:10
still, weird error message by rakudo 13:11
13:11 FROGGS left 13:20 spider-mario left
moritz I agree 13:20
r: so /\s+.*$/
p6eval rakudo 018b73: OUTPUT«No such method 'match' for invocant of type 'Any'␤ in method Bool at src/gen/CORE.setting:10772␤ in sub prefix:<so> at src/gen/CORE.setting:2587␤ in block at /tmp/Bcj0Dp8iyr:1␤␤»
jnthn Well, you can pass a *-1 there
So it thinks it's getting a whatever block, and invokes it with the number of elements 13:21
r: /\s+.*$/(1) 13:26
p6eval rakudo 018b73: OUTPUT«No such method '!cursor_start_all' for invocant of type 'Int'␤ in regex at /tmp/RCQPow3JFD:1␤ in block at /tmp/RCQPow3JFD:1␤␤»
moritz that's the old "violating the regex calling conventions should produce better error messages" problem 13:27
masak aye. 13:28
13:29 FROGGS joined, daxim_ joined
jnthn yes. 13:30
13:33 daxim left
FROGGS jnthn: I'm trying to match stuff in <?before ...> tokens, is it possible to pass these matches to the Actions.pm methods? 13:35
I have problems getting these values
moritz <?before> doesn't capture
FROGGS damn it
13:35 xenoterracide left
moritz I wonder if if you can say <alias=?before ...? 13:36
erm. <alias=?before ...>
FROGGS will try
moritz: github.com/rakudo/rakudo/commit/93...4a6247a97d 13:37
I cen use $<twigil> in that grammar, but I am unable to bind it, say, to $*SOMETHING
but I can do: :my $*SOMETHING := "1"
I do get the "1" within the actions 13:38
moritz well, you have to declare $*SOMETHING in an outer rule
jnthn r: say '1234' ~~ /<before 2> \d+/ 13:39
p6eval rakudo 018b73: OUTPUT«｢234｣␤ before => ｢｣␤␤»
jnthn oh, yeah, it'z zerowidth... :)
FROGGS r: say '1234' ~~ /<before $<thing>=[2]> \d+/
p6eval rakudo 018b73: OUTPUT«｢234｣␤ before => ｢｣␤␤»
moritz otherwise it's gone by the time the action method for this rule is called
jnthn r: say '1234' ~~ /<before (2)> \d+/
p6eval rakudo 018b73: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter ( (must be quoted to match literally)␤at /tmp/jMZD9COhbS:1␤------> say '1234' ~~ /<before (2)> \d+⏏/␤Regex not terminated␤at /tmp/jMZD9COhbS:1␤------> say '1234' ~~ /<before (2)…
jnthn Yeah 13:40
FROGGS r: my $*THING; say '1234' ~~ /<before $<thing>=[2]> \d+ { $*THING := $<thing> } /; say $*THING
p6eval rakudo 018b73: OUTPUT«===SORRY!===␤Cannot use bind operator with this left-hand side␤at /tmp/VKQOlWVzZI:1␤------> $<thing>=[2]> \d+ { $*THING := $<thing> ⏏} /; say $*THING␤ expecting any of:␤ postfix␤ infix or meta-infix␤ infi…
jnthn Try assignment
moritz FROGGS: try assigning (in rakudo at least)
FROGGS nqp: my $*THING; say('1234' ~~ /<before $<thing>=[2]> \d+ { $*THING := $<thing> } /); say($*THING)
p6eval nqp: OUTPUT«234␤␤»
13:40 xenoterracide joined
jnthn That's an option, yes :) 13:41
FROGGS nqp: my $*THING; say('1234' ~~ /<before $<thing>=[2]> \d+ { $*THING := $<thing>.Str } /); say($*THING)
p6eval nqp: OUTPUT«234␤␤»
FROGGS I though I've seen that somewhere...
moritz r: '1234' ~~ /<before $<thing>=2>234/; say $<thing>
p6eval rakudo 018b73: OUTPUT«Any()␤»
moritz heh, would have been too easy
FROGGS r: say('1234' ~~ /$<zw>=<before $<thing>=[2]> \d+ /);
p6eval rakudo 018b73: OUTPUT«｢234｣␤ zw => ｢｣␤ before => ｢｣␤␤»
FROGGS ya, zerowidth 13:42
r: say('1234' ~~ /$<zw>=<?[2]> \d+ /);
p6eval rakudo 018b73: OUTPUT«｢234｣␤ zw => ｢｣␤␤»
FROGGS r: say('1234' ~~ /$<zw>=<?[2]>? \d+ /);
p6eval rakudo 018b73: OUTPUT«｢1234｣␤ zw => ｢｣␤␤»
moritz r: 'abc' ~~ /<before <alpha>> bc/; say $<before><alpha>
p6eval rakudo 018b73: OUTPUT«Any()␤»
FROGGS r: say('21234' ~~ /$<zw>=<?[2]>? \d+ /);
p6eval rakudo 018b73: OUTPUT«｢21234｣␤ zw => ｢｣␤␤»
13:42 skids left
FROGGS moritz: is it not <?before ...> ? 13:43
with an "?" ?
moritz well, usualy yes :-) 13:44
13:44 hoelzro is now known as hoelzro|away
arnsholt Oh, how I wish I could have sub MAIN in Perl 5 13:45
13:45 sevin joined
FROGGS okay, so if I cant capture it, then I might do that at runtime... like in Cursor.INTERPOLATE 13:45
but then I might run into problem when / || @a / is optimized to / @a /, right? 13:46
s/problem/problems/
moritz erm, what problems?
neither || @a nor @a captures
FROGGS no, that's not what I'm trying to do 13:47
@a = <a b c>; / @a / is equal to /[a|b|c]/
but / || @a / should switch to sequential 13:48
like /[a||b||c]/
it is tricky to make these things right 13:51
13:51 benabik left
masak arnsholt: you mean there isn't a CPAN module for sub MAIN yet? :) surprising. 13:51
arnsholt: OTOH, it's not really nice unless you also have multis. 13:52
moritz www.digi.no/911787/opera-sendte-90-paa-doren # now I'm glad I didn't join Opera last year
masak "Perl 6 isn't a menu of features you can pick and choose from."
moritz first you need proper subroutine signatures
and preferrably a type system
masak right. and make sure not to break prototypes in the process! 13:53
or the existing light-weight types system.
arnsholt moritz: Wow. That's the department where I interned
And yeah, I've actually toyed with the idea of making a MAIN module for Perl 5, but decided the tools aren't really there 13:54
13:57 zamolxes_ left 14:00 dayangkun joined, dayangkun left 14:01 PacoAir joined 14:12 lustlife left, lustlife1 joined 14:21 Psyche^ joined 14:24 Patterner left, Psyche^ is now known as Patterner 14:32 kaleem left 14:33 bluescreen10 joined 14:37 bluescreen10 left 14:41 hoelzro|away is now known as hoelzro 14:42 bluescreen10 joined, bluescreen100 joined 14:43 pmurias joined
pmurias hi 14:44
tadzik hi 14:45
14:46 PacoAir left
colomon n: sub postfix:<!>($n) { [*] 1..$n; }; say 26! / (2! * 24!) 14:46
p6eval niecza v24-23-g0520c7c: OUTPUT«325␤»
14:46 bluescreen10 left
nwc10 masak: who was it who said "Perl 6 isn't a menu of features you can pick and choose from." ? 14:46
I've forgotten
14:46 PacoAir joined
daxim_ it's got layers. 14:47
14:49 am0c left
moritz I think it was TimToady, not sure though 14:49
14:49 phenny left 14:50 phenny joined
nwc10 I'm not sure if anyone said *that* directly :-) 14:51
FROGGS nwc10: looks like it's about that: www.gossamer-threads.com/lists/engi...ply;so=ASC 14:52
nwc10 that's me :-)
masak nwc10: it was someone on p5p.
heh, yeah seems it was nwc10 :)
nwc10 "Message-ID: [email@hidden.address] is in my sent mail folder. :-)
FROGGS nwc10: I guessed that :o) 14:53
masak 'Larry's overriding design decision for Perl 6 seems to be "I can't fix this in Perl 5".' -- I like that one too.
I have that one as "You can't get here from there."
nwc10 which is actually more serious than simonkidd.wordpress.com/2010/08/12/...from-here/ 14:55
"I can't fix this in Perl 5" is heresy in some circles. 14:58
masak yes, of course.
nwc10 excellent topic to rant about over beer :-) 15:01
but, one's two biggies for starters
Perl 5 (the current implementation) is never going to have decent threads
Perl 5 (the current implementation) is never going to have better than O(n) Unicode [for code points] and O(n**2) Unicode [for graphemes] 15:02
daxim_ lwn.net/Articles/537621/ has this article been already discussed? the log search didn't find anything, so I don't know whether it's because I'm using it wrong or not
moritz daxim_: the fulltext search just plain sucks 15:03
masak daxim_: ISTR it was discussed a few days ago. 15:04
moritz anyway, just seems like chromatic's usual Perl 6 FUD and ranting
brrt daxim_, imho (haven't seen it yet), its a distraction
lets keep trucking on
masak wow, only now I see that the title of that one is "Goodnight, Perl 6" (when chromatic's article was "Goodnight, Parrot")
FROGGS nwc10: the link "Taking YAPC::EU 2003 in Paris seriously" on your page is borken
nwc10 FROGGS: I should ask Elaine where that image lives now. 15:05
moritz oh sorry, this one wasn't by chromatic. I misread the meta data. 15:06
daxim_ I didn't notice that this article is a reply/comment, either
brrt anyway, a lot of people don't believe in perl6, and luckily, some do
masak oh, it appears to be two distinct lwn posts. one the child of the other. 15:07
15:07 stopbit joined
huf why are all these people personally insulted that p6 isnt ready by their concepts of ready? 15:08
masak it appears to be a kind of cargo-culted anger.
15:09 sevin left
masak there's not much thought about it. it reaches ridiculous proportions when people write "I waited five years for Perl 6, and then I used X instead", as if all they did for five years was to sit on a stool outside of a closed door waiting for something to happen. 15:10
I bet no-one ever did that. :)
moritz well, they would have starved first
evolution takes care of us, so to speak
masak hugs evolution
brrt :-) 15:11
tadzik that'd be natural selection, actually
evolution is when they mutate, no?
masak tadzik: last I looked, evolution and natural selection were, at the very least, good buddies.
tadzik oh they are 15:12
tadzik hums De-evolving by Jonathan Coulton
huf one drives the other
masak random genotype changes add entroy. natural selection removes it.
daxim_ bribe me to port Algorithm-Combinatorics. that would be a low-level yak for me in order to do more interesting stuff in the language. 15:13
tadzik daxim_: I'll teach you an optimal way to tie a shoe
masak :P
tadzik serious offer
masak can you do it with one hand?
15:13 spider-mario joined
tadzik nah 15:13
daxim_ tadzik, I already know that! plus.google.com/108247278178090416...d69rD4pHTk 15:14
tadzik oh, that's merely about, er, Węzeł Płaski and Węzeł Babski, however you call that in English :) 15:15
huf hmm?
is this some nursery rhyme?
tadzik decomutees 15:16
daxim_ okay, if you have something better to offer than this, then I accept your proposal! deal? 15:17
colomon daxim_: For what it's worth, I just added combinations to List::Utils.
daxim_ oh, good 15:18
masak huf: seems to be related to en.wikipedia.org/wiki/File:Granny_knot.svg
the terms seem to mean "flat knot" and "woman's knot". 15:19
colomon r: my $a = [].shift; say $a
p6eval rakudo 018b73: OUTPUT«Element shifted from empty list␤ in method gist at src/gen/CORE.setting:10516␤ in sub say at src/gen/CORE.setting:7601␤ in block at /tmp/Ml5j36QVfu:1␤␤» 15:20
colomon r: my $a = [].shift; say $b; say $a
p6eval rakudo 018b73: OUTPUT«===SORRY!===␤Variable '$b' is not declared␤at /tmp/yQKgiaQWKG:1␤------> my $a = [].shift; say $b⏏; say $a␤ expecting any of:␤ postfix␤»
colomon r: my $a = [].shift; say 42; say $a
p6eval rakudo 018b73: OUTPUT«42␤Element shifted from empty list␤ in method gist at src/gen/CORE.setting:10516␤ in sub say at src/gen/CORE.setting:7601␤ in block at /tmp/fY0o3SYuJ5:1␤␤»
huf masak: how do you call one half of that?
masak: the most basic building block of knots
masak huf: no idea. :)
huf cause yeah, they taught us this to tie our sails on the boom 15:21
colomon r: my $a = [].shift; say 42; say $a ~~ Failure
p6eval rakudo 018b73: OUTPUT«42␤True␤»
colomon r: my $a = [].shift; say 42; say $a !~~ Failure 15:22
p6eval rakudo 018b73: OUTPUT«42␤False␤»
nwc10 I had thought that the anger about lateness is sort of www.corsinet.com/braincandy/hlife.html
but I'm not so sure
Ovidius I don't see it as anger so much as being tired of waiting and giving up. 15:23
nwc10 there's this feeling that "something should be done to improve Perl 5" "Preferably yesterday" "without changing the stuff that I like"
without, I feel, most people who want it (quite legitimately, I agree)
arnsholt To me, chromatic seems inordinately concerned with what other people spend their time on, as well 15:24
nwc10 really having a good feeling for how hard it is to do.
Do too little, and you ship, but are as popular as Python 3
do too much, and you don't ship (yet)
arnsholt Although, to be fair, I'm not really familiar with the history (of which there seems to be a fair amount) of that particular story 15:25
nwc10 There are a lot of things I'm not going to say
but I will say that chromatic gave a lot of his own time fixing hard, nonfun bugs in Parrot which no-one else wanted to get their hands dirty with.
which has defineately helped Rakudo get where it is today 15:26
arnsholt That's good at least
nwc10 and, that I don't think that there is anyone now doing that unfun job that he used to do
masak indeed. 15:27
colomon r: my $a = [].shift; say 42; say $a.defined
arnsholt Probably not, yeah
p6eval rakudo 018b73: OUTPUT«42␤False␤»
PerlJam good $localtime all 15:32
masak PerlJam! \o/
PerlJam masak!
:)
pmurias nwc10: "re do little" there were a couple implementations compiling Perl 6 to Perl 5 (sacrificing some of the harder features), that were relatively usable but there was little interest in them
15:34 sftp joined
colomon moritz: when you get a chance, can you look at the test failures in List::Utils? The code works fine in Niecza, I believe it used to work fine in Rakudo, but something weird is happening with exceptions (?) with the current Rakudo. 15:36
r: my $a = [].iterator.list.shift; say 42; say $a.defined 15:37
p6eval rakudo 018b73: OUTPUT«42␤False␤»
15:37 tokuhiro_ joined, sftp left 15:39 crab2313 joined, thou joined 15:41 sftp joined 15:42 rindolf joined
moritz n: my $a = [].iterator.list.shift; say 42; say $a.defined 15:44
p6eval niecza v24-23-g0520c7c: OUTPUT«42␤False␤»
moritz colomon: I'll try
colomon moritz: danke
moritz though tuits are stretched thin, and will be for quite a while :(
dalek kudo-js: 8854e88 | pawel++ | runtime.js:
Bring our implementation of 6model closer to nqp-jvm.
 15:45
kudo-js: 4f7e3dd | (Pawel Murias)++ | runtime.js:
Remove *loads* of dead code from before the serialization refactor.
15:49 skids joined 15:50 kaare_ joined 15:57 kivutar joined 16:02 tokuhiro_ left 16:11 census joined
PerlJam Queen has taught me a couple of things this morning ... fat bottom girls make the rockin' world go 'round ... Queen will rock you ... and ... we are the champions. 16:15
:-)
masak "Queen" will rock you, but "we" are the champions?
that sounds rather inconsistent :)
PerlJam masak: beware the hobgoblins of foolish consistency ;) 16:16
colomon n: sub postfix:<!>($n) { [*] 1..$n; }; say 10! / (4! * 6!) 16:17
p6eval niecza v24-23-g0520c7c: OUTPUT«210␤»
masak PerlJam: heh :)
PerlJam (also, I said "a couple of things", but I listed 3. It's poetic license or something ;)
masak that one is fine. the formal definiton of "a couple" is 2..3 16:18
colomon n: sub postfix:<!>($n) { [*] 1..$n; }; say 10! / (5! * 5!)
p6eval niecza v24-23-g0520c7c: OUTPUT«252␤»
jnthn
.oO( Beware the Hobgoblins? But they taste so good! )
 16:19
PerlJam masak: I've never seen anyone go "oh, what a cute couple!" when they see three people holding hands going down the street.
jnthn PerlJam: hah, I was thinking the same :) 16:20
moritz PerlJam: that's why German has the distinction between "paar" and "Paar"
:-)
colomon remembers having the "3 is a couple" debate with his father back in elementary school. 16:21
16:21 domidumont left
PerlJam masak: and how many is "a few" ? :) 16:22
FROGGS at least two, that's all I know 16:23
kresike bye folks 16:29
16:29 kresike left, kivutar left
PerlJam I think something is defective in my brain ... everytime I see "kresike", I think of West Side Story and Officer Krupke 16:30
huf i think of a very small crash 16:31
but i have inside information
a ... crashling? 16:32
16:32 brrt left
skids prefers the trollish counting system. It has only three numbers: one, two, and "yes, definitely two". 16:33
masak :) 16:36
pmurias jnthn: what's the difference between isnull and isnull_s? 16:43
colomon is a bit puzzled why his combinations routine is so painfully slow on, say 10 choose 5
masak colomon: maybe it's generating big lists of things eagerly. 16:44
jnthn pmurias: Object vs. native string
colomon masak: nope, it's lazy 16:45
jnthn pmurias: On the JVM they go down to the same op
colomon well, I guess my test for it is not lazy. hmm.
pmurias jnthn: the distinction makes sense on Parrot?
jnthn pmurias: And anywhere else we need to keep native strings and objects distinct. 16:46
We only get away with it on the JVM because it can't tell one null from amother :)
uh, another
masak .oO( a Freudian slip is meaning one thing but saying your mother ) 16:47
pmurias jnthn: is null a valid native string?
jnthn pmurias: I guess it's more the absence of a string :) 16:48
pmurias but a native string variable can store those?
jnthn Yes
pmurias on parrot we have both a null and string null? 16:49
jnthn Correct.
colomon masak: aha, appears a lot of the slow is in my tests to verify the combine returned is correct. :)
jnthn On the JVM they're both just a NULL.
uh, null 16:50
aconst_null
:)
colomon r: say +("a" .. "m")
p6eval rakudo 018b73: OUTPUT«13␤»
colomon n: sub postfix:<!>($n) { [*] 1..$n; }; say 13! / (5! * 8!) 16:51
p6eval niecza v24-23-g0520c7c: OUTPUT«1287␤»
colomon oh
r: say +("a" .. "h")
p6eval rakudo 018b73: OUTPUT«8␤»
pmurias jnthn: do we have docs on nqp ops somewhere?
jnthn nqp-opcode.txt is the closest to that 16:52
pmurias: They need documenting much better, it's just getting around to it... :) 16:53
FROGGS moritz / jnthn: can you help me with this? gist.github.com/FROGGS/d704ce02283dfdde53df 16:59
TimToady masak: nah, it just means you're getting father from the truth
FROGGS I dont know where the error comes from
FROGGS .oO( Truth, I am your father! ) 17:00
pmurias I'll start documenting the more tricky opcodes soon 17:01
jnthn <.unset_SEQ> 17:04
That doesn't return a Cursor
{ self.unseq_SEQ }
But why not just { $*SEQ := 0 } :)
FROGGS tries
jnthn But...how will that ever actually match? :) 17:05
It can't be before a || *and* before a sigil :)
I think you'd have much better look setting a contextual in termaltseq and termalt 17:06
*luck
Or maybe nibbler 17:07
masak nibbler! \o/
jnthn om nom nom nom nom regex!
17:08 pmurias left, b1rkh0ff left
pmichaud jnthn: (MARKED/MARKER) feel free to adjust them however you wish 17:13
ParseShared is indeed where I expected it to end up 17:14
jnthn pmichaud: OK, if it feels right to you too, I'm happy to go ahead and do it. 17:15
pmichaud yeah, this week is going to be a little hectic for me :-/
jnthn I'm gonna be fairly tied up tomorrow/Wednesday, but Thursday looks tuitful. 17:16
Maybe get some time this evening...depends how my slide writing godes :)
uh, goes
17:17 Chillance joined 17:20 b1rkh0ff joined 17:27 SmokeMachine left, hoelzro is now known as hoelzro|away 17:28 SmokeMachine joined 17:31 Chillance left, mberends left, fgomez left, avar left, sorear left, kshannon left 17:36 Chillance joined, mberends joined, fgomez joined, avar joined, sorear joined, kshannon joined
masak "tuitful", what a delightful neologism. 17:40
TimToady well, that's just using a productive suffix
round tuition is more like a neologism 17:41
not to mention a pun...
I suppose a real linguist might consider "round tuition" to be a portmanteau 17:42
TimToady tries to visualize an airportmanteau 17:43
huf ow
nwc10 milimetre backscatter? 17:44
at least, would that be the form that it takes?
TimToady backscattergun?
FROGGS jnthn: by "setting a contextual" you mean that I should add stuff to the AST that declares a var that I'll check in INTERPOLATE? 17:46
TimToady a new word from Indonesia might be a Borneologism? 17:47
jnthn FROGGS: I think you'd pass it as a parameter to interpolate
FROGGS I like INTERPOLITE though, always wanna type that instead...
jnthn FROGGS: But track the need to do so using the contextual 17:48
A little like we do with the case-insensitive flag
FROGGS so you mean like a global $*THING variable? then I must make sure to reset it right, in case of /[ 1 || @a] @b / 17:50
17:50 not_gerd joined
not_gerd o/ 17:50
jnthn Well, it's not really global, it's dynamic. The important thing is that you're looking for being directly lexically nested in such an alternation, and dynamics at compile time represent the lexical shape of the program. 17:51
not_gerd oO( s/expixit/explicit/ - the keys are like... right next to each other )
17:55 Targen joined
TimToady hmm, www.gossamer-threads.com/lists/perl...392#283392 seems just a bit revisionist to me; paraphrasing: "the initial four revisions of MOP weren't for Perl 6, they were for pugs." Right... 18:03
18:05 daxim_ left 18:06 dakkar left
rjbs That's weird. 18:06
sure is interesting times in perl, this month. 18:07
(although I see that's pretty old)
PerlJam 2013 is shaping up to be a watershed year for Perl (both 5 and 6) 18:08
rjbs watershed meaning that I'm shedding water from my eyes?
PerlJam I can't wait to see what happens in March and April! :)
rjbs Seems like the same old sound and fury to me.
Significand yet to be determined. 18:09
japhb mmcleric, re: irclog.perlgeek.de/perl6/2013-02-18#i_6466917 , I'd simply say two things: 1) What coke said -- the work lists would have to be shared, since (as predicted) they ended up being rather extensive, and 2) p6mw moved to github.com/perl6/perl6-most-wanted , where you can see the size and scope of the current lists 18:20
Also, while I accept the necessity of using github to work with the Perl 6 ecosystem, I chafe at having to use Twitter for ... well, anything, but in particular just to use Play Perl. 18:22
mmcleric auth with github is already in works, and auth with one-time email tokens is not far away too 18:23
I know that's a popular demand
those lists are huge
are they prioritized? do we have to import everything, or can we start with a most important subset?
japhb Yay, glad to hear about the new auth methods 18:24
mmcleric (I jus treplied to coke here: play-perl.org/quest/512203d8e05a54a63a000022)
japhb Yup, I told you the lists were big. ;-)
colomon woah, mmcleric is right. What's the point of having a most-wanted list if everything under the sun is on it?
cotto How much do people care about Parrot's current profiling tools?
keep them? Nuke them? Fix them?
PerlJam cotto: I typically want to profile at a level above Parrot, so ... 18:25
japhb colomon: Here's the problem -- we need both small and large things for people to work on; and clearly, a lot of people have been getting to work over the last couple years.
jnthn cotto: The sub-level profiler is useful
cotto jnthn: thanks 18:26
japhb But I think people were kidding themselves about how ready Star was for people to be using it heavily in the wild. There's just so much that people expect to do easily in Perl that isn't there yet.
jnthn perl6 --profile x.p6 2> foo.callgrind # I've used this kinda thing quite a bit. The output is often helpful.
japhb So I threw some effort into figuring out (and detailing) just how much was left to do.
cotto jnthn: is that the only one you personally care about? 18:27
18:27 SamuraiJack left
colomon japhb: if the list is just "things it would be nice if someone worked on sometime", isn't that just the "wanted" list? 18:27
jnthn cotto: I'm pretty sure --profile is doing the sub level profiling. I don't know that I've used (or found useful) anything else. In fact, I'm not aware of what else is out there ;) 18:28
colomon japhb: ie I was expecting "most wanted" to be, OMG, we need this ASAP.
18:28 SamuraiJack joined
jnthn colomon: I think that the "everything that's wanted" list would be a lot longer. :) 18:29
TimToady just use a level of indirection and call it "CPAN"
PerlJam TimToady: heh, I was just thinking that
cotto jnthn: thanks. That's a useful data point. 18:30
japhb colomon, Yes, I certainly see your point with that. However, I can't provide that prioritization by myself -- I basically took a collection of what different communities and people saw as important to their daily work, applied a modicum of sanity to it, and categorized it. But everything on there is important in some way. I'm not qualified without community input to judge *most* important.
Though I guess that's the reasoning behind trying to port it to Play Perl.
Because then people can actually vote it around and bubble the real key stuff to the top.
mmcleric exactly, PP can help you to crowdsourse prioritization
[Coke] which is going to give us much better answers than #perl6 is. 18:31
s/is/can/
japhb heh
PerlJam mmcleric: so ... does PP have quests for others to take on?
mmcleric PerlJam: play-perl.org/quest/512203d8e05a54a63a000022 - read this and links from this 18:32
tl;dr: not yet, but it's my priority this week
PerlJam mmcleric++ :)
mmcleric note that my vacation is over, so I'll be a bit slower from now on than last week… but I still think it'll be ready in a few days 18:33
TimToady we need an abbreviation for tl;dr 18:34
mmcleric TimToady: new unicode character? 18:35
japhb
.oO( A little square box with tl at the top and dr at the bottom )
TimToady or just use ⏏
your program was too long, and I didn't read it...
std: your program was too long, and I didn't read it... 18:37
p6eval std 7551b8f: OUTPUT«===SORRY!===␤Bogus term at /tmp/Ascza9dlyL line 1 (EOF):␤------> am was too long, and I didn't read it...⏏<EOL>␤Undeclared routines:␤ 'it' used at line 1␤ 'long' used at line 1␤Parse failed␤FAILED 00:00 43m␤»
TimToady hmm, in this case, it did read it all
I guess it "didn't read" the <EOL> 18:38
colomon mmcleric++ japhb++ 18:47
18:49 not_gerd left 18:55 SamuraiJack left 18:58 fhelmberger left 19:10 jac50 joined 19:17 alester joined
Harzilein what's are example for crazy internal dsls that work with niecza and rakudo right now? 19:31
what* 19:32
examples*
19:32 PacoAir left
TimToady that would depend on how you define "crazy", "internal", and "dsl" 19:33
PerlJam Harzilein: Perl 6 ;)
TimToady but quotes and regex are the main ones
Harzilein i mean "programmer-defined" 19:34
19:34 PacoAir joined
PerlJam Harzilein: I believe ABC is a dsl that works on both R and N 19:34
TimToady nr: say Q:s/ { my pid == $*PID } /
p6eval rakudo 018b73: OUTPUT« { my pid == 2919 } ␤»
..niecza v24-23-g0520c7c: OUTPUT«Use of uninitialized value in string context␤ at /home/p6eval/niecza/lib/CORE.setting line 1290 (warn @ 5) ␤ at /home/p6eval/niecza/lib/CORE.setting line 266 (Mu.Str @ 15) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /tmp/kMgbioCqLW line 1 (mainline @ …
masak Harzilein: you can get pretty far by defining subs which use dynamic variables. 19:35
TimToady Q:s is a user-defined quote
masak Harzilein: the really crazy stuff will need macros and/or slangs.
PerlJam Harzilein: github.com/colomon/ABC (but ask colomon about it)
TimToady to me, user-defined != internal
PerlJam Harzilein: maybe git://github.com/tadzik/Bailador.git fits your criteria too? 19:36
TimToady or do you mean "defined by using Perl 6"?
masak right, the "internal" in "internal DSL" means "using the programming language itself to do the DSL".
ABC would be very much an external DSL. 19:37
that's Fowler's terminology, by the way.
Harzilein masak: i meant that with internal, right.
masak then I agree that Bailador is the best we have.
Harzilein: you might also want to check out my advent post on the topic.
dvj I wrote a module to strip HTML tags using grammar. It turned out to be *very* slow. I'm new to Perl6, so could anyone give me pointers on what I'm doing wrong? 19:38
masak but the *really* crazy stuff will come when we have sufficiently advanced macros and slangs.
PerlJam Harzilein: why do you ask anyway? Looking for something to work on? or just curious?
dvj module: github.com/dagurval/HTML-Strip/blo.../Strip.pm6
masak Harzilein: here: perl6advent.wordpress.com/2012/12/2...-y-things/
PerlJam dvj: what makes you think it's *you* that's done something wrong? ;) 19:40
masak dvj: looks good to me -- I don't see any obvious slowdowns.
TimToady you might try it with niecza, which might run it faster than rakudo, if it runs 19:41
masak dvj: of course, if all you want to do is strip HTML tags, then probably you could get away with .split(&html_tag_regex).join() :)
dvj good to hear :) Documents were taking up to 15min to parse so I thought it must be me 19:42
PerlJam dvj: Maybe it's IO that's causing your grief? How big are the HTML files you're throwing at it?
dvj masak: Then I wouldn't learn grammar ;)
FROGGS 15minutes? wow 19:43
PerlJam dvj: you can try github.com/perlpilot/Grammar-Profiler-Simple to see if there are any bottle-necks in your grammar.
TimToady dvj: currently grammars are designed for maximum flexibility, not maximum speed; the speed will come later (we hope) when we figure out how to optimize our grammars better 19:44
dvj PerlJam: doc.perl6.org/ <-- takes several minutes
TimToady: that sounds good to me. I'm just glad I'm using them right hehe
TimToady for instance, right now all method calls in our grammars are virtual 19:45
dvj PerlJam: I'll try out that module
TimToady but de-virtualizing method calls will tend to take whole-program analysis
and a certain amount of type inferencing
but also, pretty much everything runs slow on rakudo/parrot so far. hoping this changes some when we get rakudo/JVM, but we'll see 19:47
colomon Harzilein: masak++ is right, ABC is about processing external ABC files, not making your p6 code look like ABC.
alester masak: I haven't forogtten about our article. 19:52
masak colomon: would be kinda cool to have it as an internal DSL, I guess :) 19:53
but Perl 6 does not stop at the limits that, say, Ruby does.
we mean to do internal DSL for realz. :) 19:54
alester: oh, cool. PerlJam++ has supplied some starting material for when you're ready.
colomon masak: it's kind of a cool idea as a toy, but since the entire point of the project is to processing external files... 19:55
dalek osystem: f773b23 | Dagur++ | META.list:
Added HTML-Strip
 19:58
osystem: 70539c4 | colomon++ | META.list:
Merge pull request #18 from dagurval/patch-1

Added HTML-Strip
alester masak: Can you drop it in mail to me please?
20:01 benabik joined
masak alester: will do. 20:06
colomon: ah, ok. perhaps not so worthwhile, then :) 20:07
alester: sent. 20:08
20:08 benabik_ joined, benabik left, benabik_ is now known as benabik 20:14 benabik left 20:23 brrt joined 20:24 benabik joined 20:26 zamolxes joined 20:29 colomon left
ingy jnthn: are you going to .tw this osdc? 20:31
jnthn: and/or do you know anyone else who is?
TimToady r: 1+2 20:32
p6eval rakudo 018b73: ( no output )
TimToady r: 1+2; 42
p6eval rakudo 018b73: OUTPUT«WARNINGS:␤Useless use of constant expression "1+2" in sink context (line 1)␤»
TimToady r: $_ + 2; 42
p6eval rakudo 018b73: OUTPUT«use of uninitialized variable $_ of type Any in numeric context in block at /tmp/98pvsVH2_P:1␤␤»
TimToady that should warn about useless use of + in sink context
r: my $x = 1, 2; 42 20:33
p6eval rakudo 018b73: ( no output )
TimToady and that should warn about useless use of , in sink context, unless it distributes, in which case it's useless use of 2
jnthn ingy: No, not this year I'm afraid.
20:33 colomon joined
TimToady moritz: ^^ 20:34
ingy jnthn: thx 20:35
TimToady (I'm inclined to say that the sink context distributes to its list members)
20:35 brrt left
TimToady otherwise we'll get warnings on loop(my $i = 1, my $j = 2; ...; ...) 20:36
which should be perfectly fine
jnthn Space before paren :P 20:37
TimToady except for that
it's missing a block too :P
std: loop(my $i = 1, my $j = 2; ...; ...) # curious 20:38
p6eval std 7551b8f: OUTPUT«===SORRY!===␤Undeclared routine:␤ 'loop' used at line 1␤Check failed␤FAILED 00:00 44m␤»
TimToady std: sub loop {...}; loop(my $i = 1, my $j = 2; ...; ...) # curious
p6eval std 7551b8f: OUTPUT«ok 00:00 45m␤»
TimToady (wasn't sure whether an arglist allowed ; or not
) 20:39
FROGGS for Undeclared routine: '<built-in>' it might suggest to add a space to avoid function call syntax
TimToady std: if(42) {...}
p6eval std 7551b8f: OUTPUT«===SORRY!===␤Word 'if' interpreted as 'if()' function call; please use whitespace instead of parens at /tmp/wdpUkhsxpG line 1:␤------> if⏏(42) {...}␤Unexpected block in infix position (two terms in a row) at /tmp/wdpUkhsxpG line 1…
TimToady does for that one
std: loop(my $i = 1, my $j = 2; ...; ...) {...} 20:40
p6eval std 7551b8f: OUTPUT«===SORRY!===␤Word 'loop' interpreted as 'loop()' function call; please use whitespace around the parens at /tmp/aqITgBAQXS line 1:␤------> loop⏏(my $i = 1, my $j = 2; ...; ...) {...}␤Unexpected block in infix position (two terms i…
TimToady does for that one too
but it needs the block to trigger it
it's really yet another ttiar error
FROGGS right, but that's perfectly fine I guess 20:41
(that it needs a block)
TimToady if there isn't a block, then they obviously wanted to call it as a function :)
FROGGS I did: "if( bla bla );␤{ more bla }" years and years ago in php and was searching at least 30minutes for the error -.- 20:42
and then I found the problem sitting on my chair 20:43
20:44 rindolf left, rindolf joined
moritz TimToady: I intentionally don't warn about stuff in sink context that warns or throws exceptions 20:50
because of things like ~$failure 20:51
which blows up, and might be intentional
FROGGS \o/ locally: 20:52
./perl6 -e 'my @a = <x xxxxxxx xxx>; say "xxxxxxxxx" ~~ /1 | @a/' ==> ｢xxxxxxx｣
./perl6 -e 'my @a = <x xxxxxxx xxx>; say "xxxxxxxxx" ~~ /1 || @a/' ==> ｢x｣
though it needs something before || and |
but I'll get around that
jnthn \o/
20:53 jac50 left 20:54 jac50 joined
TimToady moritz: the warning is still more useful; I'd suggest some other way to suppress the warning if that's what they intended 20:56
moritz TimToady: I can experiment with it 20:58
21:01 adu joined
lue std: say#`[]42; # I'm curious as to how it's I<this> syntax error 21:03
p6eval std 7551b8f: OUTPUT«===SORRY!===␤Unsupported use of bare 'say'; in Perl 6 please use .say if you meant $_, or use an explicit invocant or argument at /tmp/EBTn6Y1j2f line 1:␤------> say⏏#`[]42; # I'm curious as to how it's I<t␤Two terms in a row (lis…
21:09 rindolf left 21:12 alec__ left, nik joined 21:13 nik is now known as Guest29967, Guest29967 left 21:16 benabik left 21:17 benabik joined 21:19 alec__ joined 21:20 SmokeMachine left 21:22 spider-mario left 21:28 colomon left 21:29 domidumont joined 21:34 spider-mario joined 21:40 kaare_ left 21:42 alec__ left, benabik left 21:49 domidumont left 21:50 domidumont joined, PacoAir left 21:51 PacoAir joined 21:52 jac50 left 21:53 jac50 joined
dalek kudo/nom: 5101a54 | moritz++ | src/Perl6/Optimizer.pm:
add some type annotations to the Optimizer code
 21:54
22:00 donaldh joined 22:03 alec__ joined, lustlife1 left 22:06 domidumont left 22:11 skids left 22:12 crab2313 left 22:17 bluescreen100 left 22:19 PacoAir left 22:27 proller joined 22:40 Targen left
lue What is a [...] composer? (I'm sure I've seen it, but don't know what the name means) 22:42
(from the top of S02, btw) 22:43
masak r: say [1, 2, 3]
p6eval rakudo 5101a5: OUTPUT«1 2 3␤»
masak lue: it's when the [] create an Array object, like that.
r: say [+] 1, 2, 3
p6eval rakudo 5101a5: OUTPUT«6␤»
lue See, I knew I've already seen it :) 22:44
masak lue: the parser has to keep these two cases (array composer, reduce metaop) apart.
and that means doing a bit of lookahead.
dalek p/target-pbc: 6840a47 | (Gerhard R)++ | src/HLL/Compiler.pm:
Restore old semantics of HLL::Compiler.compile()

Returning the main sub would have worked, but is conceptionally wrong For now, just add a note about the proper solution		 22:48
22:54 pmurias joined 22:59 adu left
pmurias masak: re DSLs isn't that something to be afraid of. That is libraries being created as DSL sacrificing composability to gain cuteness? 22:59
masak I'm not sure why that should be feared more for DSLs than for any modification of the current language. (adding global subs, or types) 23:00
DSLs are just that, but... more.
rjbs wants a zero-qualification form of orelse. 23:05
.commit orelse;
As usual, I have a remedial question about a basic concept. 23:06
infex:<orelse> returns the left argument if defined, otherwise, the right. Great, I get it. 23:07
It says it is for "the first argument that evaluates successfully" (in S03). I understand "first" because it's showing a chain of orelse clauses.
23:07 labster joined
rjbs Is "evaluates successfully" related to the concept that I understand only vaguely, of returning failures that seem undef, but are fatal if touched just so? 23:08
labster lue: I was thinking about playing around with Lingua::EN::Numbers::Ordinal, can you accept GlitchMr++'s commit to make it panda compatible? 23:11
rjbs is now reading S04/Exceptions.
labster *pull request
lue ah, let me look it over.
23:12 adu joined
lue done. :) 23:13
masak rjbs: I believe so.
rjbs: I tend to read "evaluates successfully" as "returns something defined", though. not sure that's an approximation that holds up. 23:14
rjbs sees that orelse-ing a Failure marks it handling like any other defined() test.
lue
.oO(I need to finish the "17th" form in Ordinal, IIRC)
23:14 Chillance left
rjbs I think it has to hold up for this to be useful, most of the time. 23:14
perl6++ 23:15
23:18 pmurias left
labster thanks lue++. I have this strange idea that I'd go through the ecosystem, find things that don't work, and see if I can bring it up to spec. 23:18
masak labster++ 23:19
jnthn labster++ # that's certainly a good hing
*thing
23:20 geekosaur left
masak 'night, #perl6 23:20
labster night, masak 23:21
I just got tired of waiting for Perl 6 to just happen, so I figure I should just contribute some code. There needs to be some sort of working ecosystem to attract developers, and I don't want people to get frustrated when things don't work and go to a different language. 23:22
23:22 geekosaur joined
rjbs I often find it useful to throw "__END__" in a Perl 5 program to block off my pseudocode yet to be converted to real code. 23:22
Is there something to use for that in p6? 23:23
jnthn =END I think 23:24
23:25 colomon joined
rjbs S16 tells me to use q:p{filename} but :p is unrecognized adverb for Rakudo. Just drop it?" 23:25
geekosaur I think most of that stuff is unimplemented as yet 23:26
jnthn rjbs: Yes, though then you'll get a string, not a...path, whatever that should produce 23:27
rjbs produced a file of 0xF000 nuls, so that's a start!
23:30 spider-mario left
rjbs How do I get a one-byte buffer containing (say) 0x01? 23:32
jnthn my $b = Buf.new(0x01); 23:34
my $b = Buf.new(0x01); my $fh = open('foo', :w, :bin); $fh.write($b); $fh.close # gets me a 1 byte file
rjbs Thanks. I was about to use pack.
23:36 alec__ left
donaldh jnthn: are findcclass and findnotcclass needed for nqp-jvm ? 23:37
colomon > use List::Utils 23:38
> combinations("a".."d").perl
([], ["a"], ["b"], ["c"], ["d"], ["a", "b"], ["a", "c"], ["a", "d"], ["b", "c"], ["b", "d"], ["c", "d"], ["a", "b", "c"], ["a", "b", "d"], ["a", "c", "d"], ["b", "c", "d"], ["a", "b", "c", "d"]).list
jnthn donaldh: Yes, I did iscclass already so they should not be hard to add
rjbs I'm sorry, please tell me if you need me to shut up with simple questions, but next I'm looking for docs on how I can concisely make a buffer from a(n all-ASCII) Str.
jnthn donaldh: Once you add them, some code that needs them in nqp-src/QRegex.nqp can be uncommented.
23:38 proller left
donaldh jnthn: Okay, I have started. Need to build some tests. 23:39
rjbs Ah, Str.encode?
Great.
jnthn r: my $s = 'omgz'; say $s.encode('ascii')
p6eval rakudo 5101a5: OUTPUT«Buf:0x<6f 6d 67 7a>␤»
rjbs Yeah, got it... on to my next failure. :) 23:40
jnthn should sleep...gotta get up at sensible time tomorrow :)
23:41 skids joined
rjbs gist.github.com/rjbs/4981789 23:45
In this program, the first write-at succeeds. The second one fails, "seek failed: -1"
[Coke] so, I have a branch to kill pasm. NQP requires some generated pasm files (even in the setting); suggestions on how to deal with this? The best I have so far is that if someone asks for "foo.pasm" in that parrot branch, give them "foo.pir" instead. 23:46
rjbs Am I doing something wrong, or is seek broken, or?
rjbs just added a second file with a simpler demonstration. 23:48
[Coke] er, s/setting/stage0/
23:53 stopbit left
rjbs gives up for now, back to the old grind. 23:53
23:59 benabik joined
rjbs Aha! 23:59
method seek(IO:D: Int:D $whence, Int:D $offset) {

Powered by Cro and the Raku® Programming Language.

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

Thank you!