»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:04 xinming left 00:05 xinming joined 00:09 BenGoldberg joined
BenGoldberg r: my @a := 1, { ($^a+1) / $^a } ... *; .say for @a[92] 00:10
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«1.618033988749894848205␤»
BenGoldberg r: my @a := 1, { ($^a+1) / $^a } ... *; .say for @a[93]
camelia rakudo-jvm 106959: OUTPUT«1.618033988749895␤»
..rakudo-parrot 106959: OUTPUT«1.6180339887499␤»
00:11 gtxcomm joined
webchat_masak my post: gist.github.com/masak/7966845 00:15
(advent post)
again, as stated, I will publish it tomorrow. submitting it for review now, and going to bed.
'night, #perl6
00:15 pmurias left
raydiak www.cl.cam.ac.uk/teaching/1999/AGra...ml#foot238 00:23
^ anyone feel like factoring that for me real quick? :)
00:25 rurban1 left, webchat_masak left
lue raydiak: perhaps, but I either have to discover how tell Maxima to interpret implied multiplication or insert the * symbols. :) 00:28
jnthn raydiak: heh, I think I mighta taken that course...or maybe only the second year one by the same lecturer... :) 00:29
Sadly, I think I forgot it all...
colomon lue: sounds like you might need… a grammar!
BenGoldberg r: sub infix:<oof> ($a, $b) { $b, $a }; say [oof] 1..9; # good 00:33
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«9 8 7 6 5 4 3 2 1␤»
BenGoldberg r: sub infix:<oof> (*@a) { @a.reverse }; say [oof] 1..9 # confused
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«9 7 5 3 1 2 4 6 8␤»
BenGoldberg Why does the .reverse version produce that odd order? 00:35
(Or, alternativly, if that's *
if that's right, then why does the $b, $a produces such a normal looking order?)
jnthn r: sub infix:<oof> (*@a) is assoc('list') { @a.reverse }; say [oof] 1..9 00:36
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«9 8 7 6 5 4 3 2 1␤»
lue r: sub infix:<oof> (*@a) { @a.reverse }; say 1 oof 2 oof 3 oof 4 oof 5 oof 6 oof 7 oof 8 oof 9;
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«9 7 5 3 1 2 4 6 8␤»
00:37 tmetro left
colomon jnthn++ 00:37
lue heh, it's fun trying to visualize what happens there, without list assoc :)
colomon r: sub infix:<oof> (*@a) { say :@a.perl; @a.reverse }; say [oof] 1..9 # confused
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«"a" => [1, 2]␤"a" => [2, 1, 3]␤"a" => [3, 1, 2, 4]␤"a" => [4, 2, 1, 3, 5]␤"a" => [5, 3, 1, 2, 4, 6]␤"a" => [6, 4, 2, 1, 3, 5, 7]␤"a" => [7, 5, 3, 1, 2, 4, 6, 8]␤"a" => [8, 6, 4, 2, 1, 3, 5, 7, 9]␤9 7 5 3 1 2 4 6 8␤»
BenGoldberg Ahh, I see! :) 00:38
r: sub infix:<oof> (*@a) is assoc('list') { say :@a.perl @a.reverse }; say [oof] 1..9 00:39
camelia rakudo-jvm 106959: OUTPUT«===SORRY!=== Error while compiling /tmp/2eidhXO_FS␤Two terms in a row␤at /tmp/2eidhXO_FS:1␤------> > (*@a) is assoc('list') { say :@a.perl ⏏@a.reverse }; say [oof] 1..9␤ expecting any of:␤ method ar…»
..rakudo-parrot 106959: OUTPUT«===SORRY!=== Error while compiling /tmp/zpumwE9aGC␤Two terms in a row␤at /tmp/zpumwE9aGC:1␤------> > (*@a) is assoc('list') { say :@a.perl ⏏@a.reverse }; say [oof] 1..9␤ expecting any of:␤ method…»
colomon BenGoldberg: if it's list assoc, the entire list passed to [oof] is sent all at once
(and only once) 00:40
BenGoldberg Got it. Is , (comma) list assoc?
r: sub infix:<oof> (*@a) is assoc('list') { say :@a; @a.reverse }; say [oof] 1..9 00:41
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«␤9 8 7 6 5 4 3 2 1␤»
BenGoldberg r: sub infix:<oof> (*@a) is assoc('list') { say :@a.perl; @a.reverse }; say [oof] 1..9
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«"a" => [1, 2, 3, 4, 5, 6, 7, 8, 9]␤9 8 7 6 5 4 3 2 1␤»
colomon forgot the .perl
raydiak jnthn: you have a leg up on me, then...I didn't graduate high school :)
colomon BenGoldberg: I dunno about comma operator. I think of it as list assoc, but the spec doesn't seem to mention it. (Compare perlcabal.org/syn/S03.html#Comma_op...precedence to perlcabal.org/syn/S03.html#List_infix_precedence ) 00:42
actually, looking again, I don't think the comma operator can be. 00:43
lue preflex: tell tadzik I'm having trouble with panda: gist.github.com/lue/01c2f45754f97f98460a I don't know why this is :( 00:46
preflex Consider it noted.
lue preflex: tell masak I left some comments on your advent gist. 00:52
preflex Consider it noted.
lue masak++ for the post :)
01:04 jeff_s1 left 01:05 jeff_s1 joined
jnthn sleep & 01:09
dwarring r: grammar G {rule TOP {<ident> {say $<ident>}}}; G.parse("hi") 01:27
camelia rakudo-parrot 106959: OUTPUT«「hi」␤␤»
..rakudo-jvm 106959: OUTPUT«(Any)␤»
dwarring r: grammar G {rule TOP {<ident> {say $/}}}; G.parse("hi") 01:30
camelia rakudo-jvm 106959: OUTPUT«Nil␤»
..rakudo-parrot 106959: OUTPUT«「hi」␤ ident => 「hi」␤␤»
dwarring r: grammar G {rule TOP {(<ident>) {say $0}}}; G.parse("hi") 01:31
camelia rakudo-parrot 106959: OUTPUT«「hi」␤ ident => 「hi」␤␤»
..rakudo-jvm 106959: OUTPUT«(Any)␤»
dwarring is rakudo-parrot 'right' here? 01:33
01:33 Rotwang left
dwarring (in the above three examples) 01:34
[Coke] woolfy: I think lizmat picked that topic, actually. Just along for the ride, me. 01:42
dalek rl6-roast-data: 919435a | coke++ | / (6 files):
today (automated commit)
01:43
rl6-roast-data: d0b5ef6 | coke++ | bin/rakudo.moar.sh:
get ./perl6 so we can run tests
01:48
02:03 tgt left, beastd left 02:28 raiph joined 02:33 diakopter joined 02:41 ssutch joined 02:44 diakopter left
dwarring has looked at roast - thinks it may need some test cases for the above 02:47
02:50 diakopter joined 02:58 ssutch left 03:01 xenoterracide left 03:07 _ilbot left 03:08 _ilbot joined 03:24 ssutch joined 03:28 xinming_ joined 03:29 daxim joined, xinming left 03:31 gdey joined
BenGoldberg r: sub infix:<L> (*@a) is assoc('list') { @a }; say [L] 1..10 03:39
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
BenGoldberg r: sub infix:<L> (*@a) is assoc('list') { @a }; say [RL] 1..10
camelia rakudo-parrot 106959: OUTPUT«Too many positional parameters passed; got 10 but expected 2␤ in block at gen/parrot/CORE.setting:16523␤ in sub at gen/parrot/CORE.setting:16630␤ in block at /tmp/dv2sL_OiUl:1␤ in any at /tmp/dv2sL_OiUl:1␤ in any at gen/parrot/stage2/…»
..rakudo-jvm 106959: OUTPUT«Too many positional parameters passed; got 10 but expected 2␤ in block at /tmp/War5E7NecT:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1086␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1292␤ in any command_eval at gen/jvm/stage2/NQPHLL.nqp:119…»
BenGoldberg Apparently you can't do [R<foo>] where <foo> is assoc('list')... 03:40
03:46 ssutch left 03:47 ssutch joined 03:48 ssutch left 03:51 hummeleB1 left, gfldex left 03:54 gfldex joined 03:55 rurban1 joined, jeff_s1 left 03:56 cognominal left, jeff_s1 joined, cognominal joined 04:00 atroxaper joined 04:06 preflex_ joined, ChanServ sets mode: +v preflex_, preflex left, preflex_ is now known as preflex 04:17 gdey left
lue r: sub infix:<☃> (*@a) is assoc('list') { @a }; say [R☃] 1..10 04:32
camelia rakudo-jvm 106959: OUTPUT«Too many positional parameters passed; got 10 but expected 2␤ in block at /tmp/Q_e_9gLmzV:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1086␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1292␤ in any command_eval at gen/jvm/stage2/NQPHLL.nqp:119…»
..rakudo-parrot 106959: OUTPUT«Too many positional parameters passed; got 10 but expected 2␤ in block at gen/parrot/CORE.setting:16523␤ in sub at gen/parrot/CORE.setting:16630␤ in block at /tmp/wuSLkaZwul:1␤ in any at /tmp/wuSLkaZwul:1␤ in any at gen/parrot/stage2/…»
04:34 atroxaper left 04:35 atroxaper joined 04:50 thou joined 04:59 Mouq left 05:21 MikeFair__ left, MikeFair joined 05:23 wooden left 05:25 raiph left 05:27 JimmyZ joined 05:28 MikeFair left 05:29 wooden joined 05:31 MikeFair joined 05:35 BenGoldberg left 05:46 fridim_ left 06:09 Mouq joined 06:12 kivutar left 06:13 Mouq left 06:26 atroxaper left 06:30 rurban1 left
moritz timotimo: gist.github.com/moritz/2375e30d8c3bf35bcf04 06:36
06:48 ponbiki left 06:56 SevenWolf joined 07:03 laidback_01 left, laidback_01 joined 07:31 rurban1 joined 07:35 rurban1 left 07:36 atroxaper joined 07:52 kaare_ joined 07:58 Mouq joined 08:00 darutoko joined 08:02 Mouq left 08:07 atroxaper left
woolfy Coke: could well be that lizmat picket the topic. You wrote the thing. Nice work. Useful. Thanks! 08:11
08:14 nnunley_ left 08:15 ssutch joined 08:19 ssutch left 08:20 SamuraiJack_ joined
dwarring .tell pmurias just your your question about PDF::Grammar. In it's present form its a a grab-bag of parsing rules for different PDF elements. Doesn't handle compression, which happens with most real world PDFs 08:32
dalek ast: 31bff3d | (David Warring [email@hidden.address] | S05-grammar/action-stubs.t:
test for availability of $/ etc within tokens (fudged on jvm)
08:33
dwarring ^ thats for the stuff I was trying out earlier 08:34
r: grammar G {rule TOP {<ident> {say $<ident>}}}; G.parse("hi") 08:35
camelia rakudo-parrot 106959: OUTPUT«「hi」␤␤»
..rakudo-jvm 106959: OUTPUT«(Any)␤»
dwarring I've looked the the spec again and thinking that it's a nyi in jvm
FROGGS yeah, that needs fixing 08:36
woolfy masak++ for nice advent post... I've sent you a couple of comments, I found some typos 08:45
masak woolfy: nice, thank you. 08:54
preflex masak: you have 1 new message. '/msg preflex messages' to read it.
masak preflex: messages
preflex lue said 8 hours, 2 minutes and 6 seconds ago: I left some comments on your advent gist.
masak \o/
diakopter masak: wb :) 08:56
masak lue: all fixed; thank you.
raiph: all fixed; thank you.
woolfy masak++ 08:58
masak woolfy: all fixed; thank you.
need to breakfast over here -- after that I'll do the conversion into Wordpress, and publish le post. 08:59
09:01 xinming joined 09:04 xinming_ left 09:13 SamuraiJack_ left
dwarring looks at the advent brainstorming list "perl resources". there's a lot of good stuff out there that help me get going: specs, roast, perl6-examples, irc (live and logs), rossetta code, rakudo source (mostly written in Perl 6), doc.perl6.org, perl6 modules. and of-course the advent calendar 09:16
09:21 xinming left
diakopter is there a comprehensive list of all Perl 6 blog posts? with ones that are bitrotteng/out-of-date clearly marked? that would be very helpful to someone trying to learn from blog posts 09:22
09:23 xinming joined
dwarring good question. didn't mention 6guts.wordpress.com/ for a start 09:24
dwarring also uses reddit some times 09:29
09:34 xinming left 09:35 xinming joined 09:46 Mouq joined 09:50 Mouq left 09:54 denis_boyun joined 10:06 SevenWolf left 10:12 atroxaper joined 10:16 atroxaper left 10:18 Rotwang joined 10:37 denis_boyun left 10:39 integral left 10:43 denis_boyun joined
masak ok, back. 10:46
conversion to Wordpress begins.
diakopter masak: howdy :)
10:49 thou left 10:50 denis_boyun left
masak diakopter: ahoj. :) 10:50
10:51 spider-mario joined
masak (I categorized days 10, 11, 12, and 14 as "2013" on Wordpress) 10:53
post is live: perl6advent.wordpress.com/2013/12/1...ting-them/ 10:57
sorry about the delay :)
for future reference, here is how you can easily convert from (Github gist) Markdown into (Wordpress <p>-less) HTML:
$ pandoc -f markdown -t html foo.md > foo.html 10:58
$ perl6 -p -e 's/^\<p\>//; s[\<\/p\>] = "\n"; s[\<pre\>\<code\>] = "<pre>"; s[\<\/code\>\<\/pre\>] = "</pre>\n"' foo.html > foo.wordpress
lue++ # 'adverbly' post 11:00
I'm reading the part about adverbs and precedence, and I'm thinking that adverbs are probably just complicated enough for the average programmer not to grok them. :) 11:01
a bit like variant sigils in Perl 5.
11:02 integral joined
jnthn Thankfully, the only place the average programmer is likely to actually run into them is on a subscript, which I guess can be explained in a grokkable way. 11:08
moritz masak++ # advent post 11:21
11:22 denisboyun joined 11:25 daniel-s_ joined 11:29 denisboyun left 12:01 sqirrel joined 12:27 xinming left 12:28 rindolf joined, xinming joined
masak jnthn: that's a good point. 12:31
jnthn: fwiw, I believe they're actually fairly sane -- unless you think about the implementation too much ;)
jnthn Do Not Mention The Adverbs Implementation :P 12:50
FROGGS :P, indeed :o) 13:04
:o() even
13:09 xenoterracide joined 13:17 gfldex left 13:21 xenoterracide left, xenoterracide joined
moritz :P⊆ :NP 13:24
13:27 tgt joined 13:30 rindolf left 13:49 daniel-s_ left 13:53 sqirrel left 14:01 Maddingu1 is now known as Maddingue 14:06 timotimo left 14:15 spider-mario left 14:25 pecastro left, zakharyas joined
masak waves from a train o/ 14:28
14:30 pecastro joined
nwc10 going anywhere interesting? 14:31
masak I'm going up to .se's city #2 to teach. 14:33
nwc10 it is left as an excercise to the reader, as to whether this is exciting? :-) 14:34
14:35 rindolf joined
jnthn Is rain exciting? :) 14:38
tadzik lucky it's not city #17 ;) 14:48
preflex tadzik: you have 1 new message. '/msg preflex messages' to read it.
tadzik what did I do :o
14:48 beastd joined
tadzik oh, lue, I don't know why it is either :( 14:49
it happenened some time after shell/run started returning objects instead of numbers, now sometimes the process succeeds but rakudo gets a false value
I tried debugging it, no luck 14:50
masak nwc10: teaching is always a little exciting. but frankly I'm more excited by the prospect of doing my last five days of $work for the year, and going back home to @family. 14:51
rjbs Hm. It's good train-riding weather here. 14:56
Sadly, America. So no trains.
14:58 census joined
census hi masak! hi labster! hi FROGGS! long time no speak 14:58
masak census! \o/ 15:05
rjbs: waitwait, the US has no trains?
rjbs: I'm sure I saw one in Back to the Future III.
census the US has trains masak unless i'm jumpng into the converstaion at a bad point 15:06
masak census: I like your sense of humility, and bravery in jumping into a discussion ;) 15:07
census: and no, it isn't a bad point. 15:08
census thanks masak :)
masak (and yes, I know you have trains) :P
census masak i figured you did :)
masak :D
census: how're you doing? made any population-wide statistical analyses lately? 15:09
census i'm about to!
i'm lke moments away from it
thanks for askign!
i was workign on one project and it is goign horribly
so i've tabled it for the moment and am doing another one
15:09 xenoterracide left
census all the same data set 15:09
masak nice. I think. 15:10
census this other one is working out much better so i am hoping to just finish it up quickly
and then i'll just go back to the other one and finish it even though my results on it were not making sense
masak general notice: please everyone who wants to help, join #help-census. it's a joy, but I can't do it all alone :) 15:11
15:11 jferrero joined
census hahahaha 15:13
yes census is a handful
masak it's a labor of love :)
15:17 atroxaper joined
raydiak g'morning #perl6 15:17
masak raydiak! \o/ 15:18
raydiak masak: either it's not 7 AM in your timezone, or you drink way too much coffee :) 15:19
15:20 rurban1 joined
masak raydiak: definitely the former. 15:22
raydiak: I'm one of 'em Ur-opeans.
TimToady masak: you have a comment 15:25
FROGGS hi census
census long time no speak FROGGS!Q 15:26
FROGGS :o) 15:27
raydiak masak: Ah, that explains it. Local time for me is behind nearly the entire rest of the world. Which country?
15:28 jferrero left
masak TimToady: looking 15:28
15:28 jferrero joined
masak raydiak: .se 15:28
raydiak: IKEA, herring, Pirate Bay
jnthn ABBA :P 15:29
15:30 rurban1 left
raydiak masak: What is it like in Sweden? 15:31
masak raydiak: generally? very nice and quiet and civilized. right now? dark and wet and spells of saffron. 15:32
smells*
census codl 15:34
i imagine it is cold
raydiak masak: sounds like what I've heard elsewhere...I ought to visit some day; I'd love to experience that 15:35
masak: as for where I am, the nature is beautiful, but major cities where I've lived are not peaceful, and "civilized" is stretching it, IMHO :) 15:37
masak TimToady: perl6advent.wordpress.com/2013/12/1...mment-3482
15:38 jferrero left
masak census: it's cold if you're used to Saudi Arabia. not so cold if you're used to England. 15:38
[Coke] looks up from shoveling snow. What?
masak raydiak: if you're ever in the region, stop by and we'll have a beverage of your choice.
15:39 jferrero joined
jnthn [Coke]: All our snow melted already :( 15:40
15:41 census left
raydiak masak: That's very kind, thanks for the offer! I'll take you up on it if I ever manage to escape the US 15:41
15:42 jferrero left 15:43 timotimo joined
raydiak heh, google says it's 5 F degrees warmer in stockholm than here atm 15:43
timotimo moritz: i'm not sure if these failures from regex_optimizer are all legit :\ 15:44
but i'll look into it, thanks!
15:45 sqirrel joined
FROGGS p: multi P5Numeric('' \SELF) is export { '' }; say P5Numeric("") # I love that mmd! it makes it easy to implement absurdity :o) 15:45
camelia rakudo-parrot 106959: OUTPUT«␤»
FROGGS TimToady++ and jnthn++ and (whoever who is involved)++ 15:46
15:48 jferrero joined 15:52 jferrero left 15:53 jferrero joined
timotimo moritz: in fact, the roll.t test succeeds locally; must be a bad assumption, since it's about random seeds giving different values: ok 40 - different results due to random random-number seed 15:53
15:54 jferrero left
timotimo oh, of course! 15:54
my local regex_optimizer was recently rebased, the remote one wasn't
15:54 ajr joined 15:55 ajr is now known as Guest72225, Guest72225 is now known as ajr_
jnthn train & 15:55
masak .oO( because we work with training ) 15:57
FROGGS >.< 15:58
15:58 Psyche^_ joined 16:02 Psyche^ left
timotimo moritz: all the failures that are extra in your gist are non-failures locally for me, but i can't tell if there are no new failures on jvm when rebasing regex_optimizer. i'll push the new version of that branch; can you re-run the tests? 16:03
16:04 jeff_s1 left 16:05 Rotwang left, jeff_s1 joined
masak timotimo: usually when I rebase locally, I don't push to the same branch remotely, as that would require a push --force. 16:06
timotimo: (don't know if I got your use case exactly, but sounds like that one.)
timotimo right. i'll push to r-regex-optimizer :) 16:07
masak timotimo: so I tend to rename it both locally and remotely from e.g. macros to macros2
that works too.
timotimo there we go
dalek kudo/r-regex_optimizer: a1f1ec5 | (Timo Paulssen)++ | src/Perl6/Optimizer.nqp:
run the QRegex optimizer from rakudo's optimizer.
timotimo even rebased it further into new nom 16:08
16:15 raiph joined
woolfy Ooh, nice, I know about the use of i in Perl 6, did not realise that Perl 6 therefore has a Complex variable type 16:26
( perl6advent.wordpress.com/2013/12/1...mment-3482 )
timotimo well, it kind of has to, in order to represent it properly :)
woolfy So Inf is a constant in Perl 6. But Googol isnt? 16:27
16:27 SamuraiJack_ joined
woolfy (I saw in masak's advent blog define $googol but you could make Googol a constant just like Inf. Would be cool too. 16:27
)
Oh, lizmat just tells me you had a discussion about that recently and that I should read the backlog... sorry for re-opening fresh wounds... :-) 16:29
timotimo yeah, you could
r: constant googol := eval '1' ~ '0' x 100; say googol;
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«10000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000␤»
woolfy timotimo++
16:29 sqirrel left
raydiak you could spell it as "1e100", too 16:35
16:35 dwarring left 16:37 SamuraiJack_ left
timotimo that wouldn't be an integer then 16:37
or would it?
r: say (1e100).WHAT
raydiak ah, didn't think of that
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«(Num)␤»
timotimo r: constant googol := eval '1' ~ '0' x 100; say googol.WHAT;
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«(Int)␤» 16:38
raydiak r: constant googol := 1e100.Int; say googol.WHAT;
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«(Int)␤»
16:40 xinming left
raydiak r: say (1.Int / 2.Int) 16:46
camelia rakudo-parrot 106959, rakudo-jvm 106959: OUTPUT«0.5␤»
raydiak should numeric constants perhaps automatically be integers if they don't have a fractional part? 16:48
16:49 tgt left 16:52 tgt joined, hummeleB1 joined, atroxaper left 16:55 Piers_ joined 17:00 Rotwang joined, Mouq joined 17:09 zakharyas left 17:15 sqirrel joined, BenGoldberg joined
moritz no: We have very clear rules of which literals produce what types 17:18
and since for example Num and Int arithmetic produce different results, it must be possible to write an integral Num 17:19
timotimo right
moritz: ack/nak on running the jvm spectests on r-regex_optimizer?
moritz timotimo: I'll start another spectest run 17:20
timotimo thank you! :)
BenGoldberg Speaking of regex optimization ... will there be something like perl5's 'study' operator, or is that for the distanct future? 17:26
timotimo please explain what that does to me? i never did any perl5
BenGoldberg It did some analysis of a string, so that future regex matches on that string ran faster. 17:28
timotimo ah, ok
sounds like a fun project
but probably for much later :)
BenGoldberg distant future, then :) 17:31
timotimo well, it really depends; rakudo development has a few people who are not focusing their efforts optimally (me, for example) 17:32
optimally, we would always be doing the most pressing bottleneck and the most interesting feature, but I tend to go around looking for fun optimizations to do and lacking features that bother me directly
moritz timotimo: now I get "nqp-j is too old" 17:34
timotimo er, whoops. i think you should be able to just rebase the nqp branch on current master
moritz I can also locally merge the nqp branch into master
timotimo sure
moritz does, and recompiles NQP 17:36
timotimo: should I do the same for the rakudo branch?
timotimo i'm trying it with a rebase, too 17:37
the rakudo branch is already rebased on newest nom, but you should just be able to merge nom into the old version of that branch
it's just a two-line change or something
17:37 BenGoldberg left
TimToady masak: you have another comment :) 17:40
17:41 BenGoldberg joined
masak TimToady: replying to it now :) 17:41
17:42 sqirrel left
TimToady remembers a former girlfriend almost taking his head off after he used the term "shanghai 17:42
so he doesn't use the verb form of that anymore... 17:43
masak strange. 17:44
TimToady: perl6advent.wordpress.com/2013/12/1...mment-3484
TimToady well, she was half Chinese :)
moritz masak++
masak by the way, "hurt the feelings of the Chinese people" is inherently amusing to me. but don't tell the Chinese people that. 17:45
see also languagelog.ldc.upenn.edu/nll/?p=3425 17:46
17:49 colomon left, colomon joined 17:50 rindolf left 17:53 Mouq left
TimToady "This facility contains chemicals known to the State of California to cause cancer." 17:54
17:54 Mouq joined 17:55 ajr_ left 17:56 ajr joined, ajr is now known as Guest55758 17:59 Guest55758 is now known as ajr_ 18:03 colomon left 18:04 sqirrel joined, colomon joined 18:08 Piers_ left 18:12 thou joined
ajr_ Everything known to the State of California causes cancer. 18:16
18:23 ajr_ left 18:26 sqirrel left 18:34 kaare_ left, kaare__ joined 18:39 darutoko left, fridim_ joined
moritz timotimo: gist.github.com/moritz/2375e30d8c3...imizer-log looks pretty good 18:54
timotimo i'm glad. do you think i should merge? 18:56
18:59 dwarring joined 19:00 colomon left 19:02 daxim left 19:04 slavik1 joined
timotimo should probably rather merge now and see if there's fallout than merge the day before the release again :) 19:04
moritz timotimo: +1 19:09
19:11 colomon joined
dalek p: e178ee3 | (Timo Paulssen)++ | src/vm/ (2 files):
teach parrot and jvm about zerowidth literals
19:13
nqp: 681a9b8 | (Timo Paulssen)++ | src/QRegex/P6Regex/Actions.nqp:
nqp: first draft of a before-assertion simplifier.
19:13 dalek left 19:14 dalek joined, ChanServ sets mode: +v dalek, nnunley joined
dalek kudo/nom: a1f1ec5 | (Timo Paulssen)++ | src/Perl6/Optimizer.nqp:
run the QRegex optimizer from rakudo's optimizer.
19:15
kudo/nom: e542a0c | (Timo Paulssen)++ | tools/build/NQP_REVISION:
bump NQP revision to get the regex optimizer.
19:16 daxim joined
Mouq timotimo++ 19:20
timotimo :) 19:28
19:29 WABQYYH joined 19:30 colomon left 19:38 daxim left 19:50 johnmilton joined 19:51 denis_boyun joined 19:53 BenGoldberg left
masak nwc10++ # doing cool stuff with the moar GC 19:53
19:57 colomon joined, denis_boyun_ joined, ajr joined, ajr is now known as Guest11405 19:58 denis_boyun left, Guest11405 is now known as ajr_ 20:05 rindolf joined 20:07 Mouq left 20:08 denis_boyun joined
jnthn Yay, there's now a way to get to Gothenburg that doesn't suck. 20:09
20:09 denis_boyun_ left
moritz Helicopter? :-) 20:10
FROGGS ó.ò
jnthn No, a new train service that's got comfortable seats, wifi, coffee and does the journey in half an hour less time. 20:11
20:13 Mouq joined 20:15 thou left
masak I can attest to its non-suckiness. 20:18
20:20 jeff_s1 left 20:21 jeff_s1 joined 20:24 denis_boyun_ joined 20:25 zakharyas joined, denis_boyun left 20:26 rindolf left, rindolf joined
lue hello world o/ 20:29
tadzik: really? If shell/run is the issue, perhaps I have a solution... /me tries something
(unfortunately it requires being platform specific, so I'm testing the version that'll work for me :P) 20:31
20:34 Mouq left
masak lue++ # S15 work 20:35
lue masak: still quite a bit to do, but I feel like Pod6-ifying our specs at the moment :) 20:36
20:36 rindolf left
lue perl6++ making ?qx'echo $?' do the right thing. 20:37
eh... wait. nevermind :P
FROGGS hehe
lue tadzik: my idea is that maybe, just maybe, using qqx// will fix the issue. 20:39
moritz runs ./perl6-m -e 'say "Hello World"' for the first time
lue The "fails when works fine" thing for shell/run may seem random, but its occurrences are... strangely consistent.
moritz: \o/ now I want to try compiling perl6-m :) 20:40
masak me too!
moritz compiling Test.pm6 still fails, but who needs tests anyway? :-)
Unhandled exception: Cannot invoke this object (REPR: Uninstantiable, cs = 0) 20:42
why does that remind me of "I will not buy this record, it is scratched"? :-)
FROGGS moritz++ 20:47
lue gaah, using qqx seemed to have made it worse! :(
moritz with qqx it'll try to interpolate the $? or some parts of it in perl code 20:49
lue moritz: I'm retrying with "&& echo \$?" after the build command, see how it goes.
21:00 colomon left, denis_boyun_ left
lue uhh... "Error executing process: Cannot allocate memory" 21:02
tadzik: ^^^ this comes from replacing the shell command with ?qqx"perl6 --target=pir --output=blib/lib/Panda/Project.pir lib/Panda/Project.pm && echo \$?" 21:03
tadzik :o 21:04
did our memory usage strangely grow recently?
lue tadzik: I will say that the => fetching panda step takes much longer than the other => fetching stages, from what I've perceived. 21:05
tadzik 'cos panda is much bigger :P 21:07
21:07 jeff_s2 joined 21:08 jeff_s1 left
lue tadzik: apparently, replacing that line in my modifications to the qqx// thing I just pasted turned all my previous "False" results with "Error executing process: Cannot allocate memory". Which might be helpful. 21:08
(somehow, rakudo(-parrot?) is having issues with allocating memory.) 21:09
21:09 denis_boyun joined
lue (the backtrace only mentions bin/panda (which catches the error) and rakudo's source files, btw.) 21:10
jnthn lue: Easiest thing is to grep for that message 21:11
lue jnthn: hey, I just typed ack -u in the panda repo :)
O.O $ ack -u 'Cannot allocate' ../rakudo 21:12
$
jnthn ack -u too :P
Try in nqp
Also for "Error executing" 21:13
lue Um... 'Cannot allocate' isn't in panda/ or rakudo/ or nqp/ or parrot/ . Lemme try searching other parts of the error. 21:14
21:14 denis_boyun left
lue would use -i to find more results, except he doesn't think error message should be transforming their casings :) 21:14
$ ack -u 'Error executing' ../parrot 21:15
../parrot/src/platform/generic/exec.c
289: "Error executing process: %s", strerror(errno));
I wonder if shell/run is experiencing the same issue, or if this is a result of using qqx// ... 21:16
FROGGS lue: shell/run is implemented in nqp in contrast to qx/qqx
lue FROGGS: unless you're saying qx is implemented in parrot, they could still share the same cause. 21:17
FROGGS lue: see rakudo/src/core/control.pm and look for QX 21:18
21:19 colomon joined 21:24 denis_boyun joined
lue still thinks there could be a parrot underpinning to this mess. 21:24
timotimo underpinning for the fjords? 21:25
i think i'll write a short bit on the regex optimizer branch into the changelog 21:29
dalek kudo/nom: 67fcb5f | (Timo Paulssen)++ | docs/ChangeLog:
put regex optimizer changes in the changelog
21:32
21:34 denis_boyun_ joined, denis_boyun left
lue I really, really can't wait for perl6-m to get in working order. I'm getting tired of parrot. 21:35
21:35 daxim joined
dalek kudo/moar-support: 524d2f5 | jonathan++ | src/Perl6/Actions.nqp:
Scatter declaration_static for Moar's sake.

It already uses the closure model that we need to move to on the JVM in the future.
21:36
21:37 kaare__ left 21:39 zakharyas left 21:41 cognominal left, ponbiki joined 21:42 cognominal joined
dwarring has added actions counter-example to advent draft - gist.github.com/dwarring/7848868 21:43
dwarring modest addition 21:44
raydiak r: say ( Int(1)/Int(2) === 1/2 ) 21:49
camelia rakudo-parrot e542a0, rakudo-jvm e542a0: OUTPUT«True␤»
raydiak when does Int math produce different results than Num math?
21:49 m_m joined
lue r: say 1/3; say 1e0/3e0; 21:50
camelia rakudo-jvm e542a0: OUTPUT«0.333333␤0.3333333333333333␤»
..rakudo-parrot e542a0: OUTPUT«0.333333␤0.333333333333333␤»
raydiak r: say(1/3).WHAT; say (1e0/3e0).WHAT;
camelia rakudo-parrot e542a0, rakudo-jvm e542a0: OUTPUT«0.333333␤(Num)␤»
raydiak r: say (1/3).WHAT; say (1e0/3e0).WHAT; 21:51
camelia rakudo-parrot e542a0, rakudo-jvm e542a0: OUTPUT«(Rat)␤(Num)␤»
lue r: say 1/3; say 1 div 3
camelia rakudo-parrot e542a0, rakudo-jvm e542a0: OUTPUT«0.333333␤0␤»
dalek kudo/moar-support: 864bec6 | jonathan++ | src/vm/moar/ops/perl6_ops.c:
Fix p6captureouters to do the right thing.

Fixes loads of the "Cannot invoke..." errors.
raydiak r: say 1.WHAT 21:52
camelia rakudo-parrot e542a0, rakudo-jvm e542a0: OUTPUT«(Int)␤»
raydiak ah...so it's the e that makes an int not an Int any more 21:53
r: say 1e0.WHAT
camelia rakudo-parrot e542a0, rakudo-jvm e542a0: OUTPUT«(Num)␤»
masak no, the scientific-notation 'e' makes a Num. 21:54
21:56 ajr_ left
raydiak that's what I meant...an integer will produce an Int, unless you add the 'e' part...which IRL would still produce something considered an integer, but in perl 6 does not...correct? 21:58
(unless the part before or after the e was fractional, of course...then it wouldn't be an integer in either case) 21:59
lue also unless it's part of a Rat literal. 22:01
masak raydiak: no, the 'e' produces a Num, not an int. 22:02
lue masak: raydiak is switching freely between Perl 6's idea of integers an Mathematics' idea of integers. 22:03
22:03 colomon left
raydiak ah, right...I should clarify: I often learn something new by mentally building a logical mapping between the new thing and something I already know 22:04
masak well, there are lots of ways a scientific-notation number could lead to Not An int.
p: say 5e-2
camelia rakudo-parrot e542a0: OUTPUT«0.05␤»
masak p: say 3.14e0 22:05
camelia rakudo-parrot e542a0: OUTPUT«3.14␤»
lue raydiak: just remember that in Perl 6, how you write the number is more important than what the number is :) .
22:05 colomon joined
masak well, certainly that's true from a type perspective. 22:06
this, for example, works in Math, but not in Perl 6:
p: my Num $n = 4
camelia rakudo-parrot e542a0: OUTPUT«Type check failed in assignment to '$n'; expected 'Num' but got 'Int'␤ in block at /tmp/eNA2MqxXAQ:1␤ in any at /tmp/eNA2MqxXAQ:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ in any eva…»
masak 'night, #perl6 22:07
raydiak lue++: that's exactly what I was forgetting to focus on
\o g'night masak++ thanks for explaining
lue masak: I call won't-do-type-casting shenanigans on that one; Int ⊂ Num.
masak o/
22:08 dmol left 22:09 denis_boyun_ left, Mouq joined 22:10 dmol joined 22:11 dmol left 22:16 raiph left
lue :( make: *** [CORE.setting.moarvm] Segmentation fault 22:19
22:26 jeff_s1 joined 22:28 jeff_s2 left 22:31 Mouq left 22:32 Mouq joined 22:37 colomon left 22:55 jeff_s1 left 22:56 jeff_s1 joined 22:57 colomon joined
FROGGS I am out of ideas atm... perl6advent.wordpress.com/?p=2101&a...eview=true 22:58
suggestions how to extend it a very appreciated 22:59
I feel a bit sick atm but I'd be happy to add a more paragraphs tomorrow morning (in about 9 hours) 23:01
23:01 raiph joined 23:03 johnmilton left
dalek p: 565cce8 | (Tobias Leich)++ | src/QRegex/P5Regex/Grammar.nqp:
implement m:P5/\Q\E/
23:06
kudo/nom: dc5a1ff | (Tobias Leich)++ | / (3 files):
implement m:/\Q\E/
23:10
FROGGS ohh, adverb gone missing :/ 23:11
23:15 djanatyn left, dwarring left 23:16 Tene left
FROGGS humm 23:19
--output=gen/moar/stage2/NQPHLL.moarvm gen/moar/stage2/NQPHLL.nqp
Cannot find method 'optimize'
--module-path=gen/parrot/stage2 --setting-path=gen/parrot/stage2 gen/parrot/stage2/NQPHLL.nqp 23:21
Method 'optimize' not found for invocant of class 'NQPMu'
:o(
timotimo: do you have uncommitted changes?
23:22 cognominal left 23:40 raiph left 23:42 vtayur joined, vtayur left 23:46 pecastro left
woolfy FROGGS: I think it is quite a nice blog. Some typos (march -> March, rakudo -> Rakudo) and some sentences that need a second look from you. 23:47
23:47 meralt joined, pecastro joined, meralt left
woolfy FROGGS: and you switch from using the word 'slangs' to the word 'grammar', and after 'A slang was born' the word 'slang' is no longer used. Did you intend to write about slang or about grammar? 23:48
FROGGS: just re-read your blog tomorrow morning. You will find the words you need, I am sure. 23:49
23:56 beastd left 23:57 sqirrel joined