»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.
00:02 xenoterracide joined 00:10 hummeleB1 joined
timotimo i'm going to bed now. if you find any more typos, feel free to directly fix them 00:14
lue
.oO(Help, I'm trapped in a Parrot source code factory!)
 00:16
00:19 Ben_Goldberg joined 00:20 btyler left, BenGoldberg left
lue I think the reason why I only got that "Error executing" message when using qx// is because nqp uses a separate but similar fork/child process for shell, while qx uses parrot's. Parrot's has the error message. 00:26
From what I can tell, parrot's child process thing checks for the issue, while nqp (seemingly) doesn't. 00:35
what troubles me is that the only commonality I can see is the use of the system child process-related functions :/ 00:42
00:46 beastd left
lue (though parrot catches the error before using any of those functions, IIRC) 00:48
00:50 ajr_ left
lue would like to know what Proc::Status.status means/does, if only someone spec'd the class. 00:59
01:21 perigrin_ left, colomon left 01:27 raiph joined
lue Ah, so something in the process (fork() perhaps?) is throwing an ENOMEM for me :/ 01:34
01:35 hummeleB1 left
lue Which tells me this is a problem with my system rather than parrot or nqp or rakudo or panda. This should be fun :/ 01:36
lue &
01:37 colomon joined 01:48 rurban1 joined 02:03 tgt left 02:15 rurban1 left, jnap left 02:17 raiph left 02:33 perigrin joined 02:46 raiph joined, colomon left 02:48 colomon joined 02:57 obra_ is now known as obra
japhb How can I refer to attributes of the outer class inside methods of an inner class, as in the following (broken) code? 03:03
r: class Foo { has %.items; my class Bar { method find-item($name) { %!items{$name} } }; method new-bar { $!items<baz> = "quux"; Bar.new() } }; Foo.new-bar.find-item("baz").say;
camelia rakudo-parrot 413650: OUTPUT«===SORRY!=== Error while compiling /tmp/j380BLimoF␤Attribute %!items not declared in class Bar␤at /tmp/j380BLimoF:1␤------> od find-item($name) { %!items{$name} } }⏏; method new-bar { $!items<baz> = "quux"␤ …»
..rakudo-jvm 413650: OUTPUT«===SORRY!=== Error while compiling /tmp/9LH70N75Kv␤Attribute %!items not declared in class Bar␤at /tmp/9LH70N75Kv:1␤------> od find-item($name) { %!items{$name} } }⏏; method new-bar { $!items<baz> = "quux"␤ …»
lue r: class Foo { has %.items; my class Bar { method find-item($name) { PARENT::items{$name} } }; method new-bar { PARENT::items<baz> = "quux"; Bar.new() } }; Foo.new-bar.find-item("baz").say; 03:16
camelia rakudo-parrot 413650, rakudo-jvm 413650: OUTPUT«quux␤»
japhb lue, thanks!
lue japhb: ^^^ Note that this is accessing the public Foo.items method.
japhb lue: Understood. Is there no way to do it using the actual attribute, rather than the accessor? 03:18
After all, the inner class is lexically inside the outer one ...
lue japhb: %!PARENT:: didn't work, lemme try the alternate form after I find it...
r: class Foo { has %.items; my class Bar { method find-item($name) { PARENT::<%!items>{$name} } }; method new-bar { PARENT::<%!items><baz> = "quux"; Bar.new() } }; Foo.new-bar.find-item("baz").say; 03:19
camelia rakudo-parrot 413650, rakudo-jvm 413650: OUTPUT«quux␤»
lue gets the sense that %!PARENT::items not working is a bug 03:20
03:20 MikeFair_ joined
lue japhb: ^^^ yes you can, it seems 03:20
japhb That PARENT::<%!items> syntax seems a tad syntax-abusey. But what the hey, I'll take it. 03:21
lue japhb: it's perfectly valid syntax :)
r: my $::<#$$###!#> = 5; say $::{'#$$###!#'} 03:22
camelia rakudo-parrot 413650: OUTPUT«===SORRY!===␤QRPA: index out of bounds␤»
..rakudo-jvm 413650: OUTPUT«===SORRY!===␤VMArray: Index out of bounds␤»
lue r: my $::<#$$###!#> = 5; say $::<#$$###!#>
camelia rakudo-parrot 413650: OUTPUT«===SORRY!===␤QRPA: index out of bounds␤»
..rakudo-jvm 413650: OUTPUT«===SORRY!===␤VMArray: Index out of bounds␤» 03:23
japhb lue: I didn't mean it wasn't to spec, I meant that in this case a package naming like syntax is use to reach something that is in a different namespace dimension.
*is used
It feels just a little YTTCN 03:24
lue PARENT is specifically designed to access the parent package.
r: my $a = 5; { say $PARENT::a } 03:25
camelia rakudo-parrot 413650, rakudo-jvm 413650: OUTPUT«(Any)␤»
lue r: my $a = 5; package Foo { say $PARENT::a }
camelia rakudo-parrot 413650, rakudo-jvm 413650: OUTPUT«(Any)␤»
03:34 raiph left 03:35 raiph joined 03:38 hugme left 03:39 hugme joined, ChanServ sets mode: +v hugme 03:40 BAMbanda joined
BAMbanda I'm really confused. I started reading Modern Perl and now I find out that Perl 6 and 5 are two different languages, but compatible? 03:40
where do I start to get acquainted with the culture, philosophy, and of course proficient with coding using Perl? 03:41
diakopter BAMbanda: not really compatible 03:42
perigrin "which perl"
diakopter: they hang out and drink together.
diakopter in a bathtub of koolaid 03:43
BAMbanda dangg, that speaks volumes
:)
diakopter well
03:43 cognominal__ left
perigrin well ... koolaid was added to the liquids in teh bathtub ... let's just go with that. 03:43
.oO(a Camel is really a harry buffalo?)
 03:44
BAMbanda hmm, so Perl6 is for the cool hipsters
Perl5 is for the mature people settled in their ways?
diakopter I'm not sure you'll find a generalization that simple.. it's really complex.
03:44 cognominal joined
perigrin hipsters are into retro man, they like the polyphonic sounds of perl5 ... the perl6 people are the cutting edge 03:44
03:44 cognominal left
perigrin edges so sharp ocassionally you may have to implement them ... 03:45
BAMbanda so we should reconcile with a healthy mix of koolaid and other bathtub fluid? 03:46
03:46 cognominal joined, cognominal left
BAMbanda is perlmonks 5 only? 03:46
perigrin no.
diakopter mostly
perigrin again it's not that simple :) 03:47
BAMbanda damn you all
:)
diakopter feels damned
perigrin feels all.
03:47 cognominal joined, cognominal left
BAMbanda georg hegel needs to put his tidbit on this matter 03:47
diakopter is felt by perigrin
03:48 cognominal joined
BAMbanda are concepts descriped in "High Order Perl" available in p6? 03:48
perigrin felts diakopter
03:48 cognominal left
perigrin yes 03:49
mostly
BAMbanda alright, rakudo it is. koolaid is always refreshing
03:49 cognominal joined
diakopter BAMbanda: yes, recursively. Also, iteratively. 03:50
03:50 cognominal left
perigrin and hopefully soon concurrently 03:50
BAMbanda sweet, jvm sounds fun
clojure and perl6, what beautiful shoes for the mind 03:51
03:51 cognominal joined, cognominal left 03:52 cognominal joined, cognominal left, gtxcomm left 03:53 gtxcomm joined, cognominal joined 03:54 cognominal left 03:55 cognominal joined, cognominal left 03:56 cognominal joined, cognominal left
lue wonders which ulimit settings to fiddle with to discern the cause of an ENOMEM 03:56
03:58 cognominal joined, cognominal left
lue
.oO(Or maybe I can replace forks with vforks...)
 03:59
03:59 cognominal joined
perigrin or sporks. 03:59
03:59 cognominal left 04:00 cognominal joined, cognominal left 04:01 cognominal joined 04:02 cognominal left, rurban1 joined 04:03 cognominal joined, cognominal left, preflex left, preflex_ joined, ChanServ sets mode: +v preflex_ 04:04 cognominal joined, preflex_ is now known as preflex, cognominal left 04:07 cognominal joined, cognominal left 04:08 cognominal joined, cognominal left 04:09 grep0r left, cognominal joined, cognominal left 04:11 cognominal joined, cognominal left 04:12 cognominal joined, cognominal left 04:13 cognominal joined, cognominal left 04:14 cognominal joined 04:15 cognominal left 04:16 cognominal joined, cognominal left 04:17 grep0r joined
lue dangerously replaces occurrences of fork with vfork without other changes... 04:18
04:18 cognominal joined 04:19 cognominal left 04:20 cognominal joined, cognominal left 04:21 cognominal joined, cognominal left 04:22 cognominal joined, cognominal left 04:24 cognominal joined 04:29 cognominal left 04:56 perigrin left, perigrin joined 05:01 preflex_ joined, ChanServ sets mode: +v preflex_ 05:02 preflex left, preflex_ is now known as preflex 05:05 xenoterracide left 05:19 MikeFair_ left 05:21 raiph left 05:25 [Sno] left 05:36 Ben_Goldberg left
lue tadzik: on panda, I found the issue to be fork() in nqp and parrot sometimes failing with an ENOMEM. Sadly I've yet to find out why or how to fix it :( 05:39
(tadzik: not helping is JSON/Tiny.pm all of a sudden demanding I precompile previous files :/ That's probably my fault though.)
05:43 dayangkun left
lue tadzik, jnthn, other interested parties: I apparently have panda's bootstrap working. AFAIK the only change left sitting around is changing the fork to vfork in nqp/src/vm/parrot/ops/nqp.ops and parrot/src/platform/generic/exec.c 05:59
I'll test it out in more detail tomorrow, see if that's really what fixed it.
06:04 rurban1 left 06:07 geekosaur left, flussence left, geekosaur joined 06:08 flussence joined
lue to clarify, I'm talking about the forks in the non-WIN32 Run_OS_Command function in the NQP file and the Parrot_proc_exec function in the Parrot file. 06:08
BAMbanda sorry for the newb question, how can i place my /install/bin into my $PATH variable? 06:09
please don't redirect me to #ubuntu :(
the perl6 REPL is beautiful, i just wanna start hacking 06:10
or should i say, parrot REPL?
lue BAMbanda: "export PATH=$PATH:new/path/to/search/bin" in my ~/.bashrc seems to do the trick. (then you have to either logout/login or IIRC type `source ~/.bashrc` in each terminal you want the changes to take hold.) 06:14
BAMbanda lue, thanks 06:17
06:17 hugme left 06:18 hugme joined, ChanServ sets mode: +v hugme 06:19 fridim_ left
tadzik lue: interesting. Did it use to be vfork and got changed a while ago? 06:22
lue tadzik: no clue. I'll go find whatever git util lets me find out :) 06:23
tadzik okay :) 06:24
lue for parrot at least, the entire function has been in its current state since 2011, with just a couple of not-touching-fork changes a month later and also 2012. 06:26
06:26 [Sno] joined 06:27 dwarring_ joined
lue tadzik: parrot's fork has been around since 2009 at least, I don't think a move from vfork caused it :) . 06:29
tadzik weird 06:31
lue nqp's use of fork is more recent (2013); though I suspect the function is heavily inspired by similar older functions in parrot. 06:32
06:32 ivan`` left
lue tadzik: I should note I'm not entirely sure that vfork was the solution. I had also cleared panda of any local changes before testing the vfork addition. 06:32
tadzik oh, plenty of people have problems without having local changes 06:33
and I have no idea what to do about it. Extreme tuit shortage may be one of the reasons
lue (as a further note, IIRC shell() ends up in the NQP file, while qx// takes its merry time in Parrotland. (in fact, qx// uses nqp::open, which is mapped straight to parrot's function.)) 06:34
tadzik: I still want to be more scientific about it before being definitive though :)
06:35 thou left
lue I also want to make sure my perception that panda's mem usage dropped from 40% on fork to ≈17% on vfork is correct. I just don't feel like recompiling parrot and nqp and rakudo tonight :P 06:36
06:37 ivan`` joined
lue Even if vfork is what fixes it, I can't say for sure if it won't cause different issues. I was kinda worried that vfork wouldn't even work in those functions. 06:40
moritz_ \o 06:41
06:41 moritz_ is now known as moritz, rurban1 joined 06:45 darutoko joined 06:58 rurban1 left 07:06 SamuraiJack_ joined, benabik left 07:12 kaleem joined 07:16 SamuraiJack_ left 07:19 SamuraiJack joined
timotimo o/ 07:21
07:23 ssutch left 07:24 geekosaur left 07:34 Guest40081 is now known as ponbiki 07:40 bbkr joined, ribasushi left 07:50 denisboyun joined
jnthn morining/ #operl6 07:53
*morning, *#perl6 :)
timotimo mhh, new language? objective perlscal6? 07:54
07:57 hugme left
moritz objectivist Perl 6 07:57
07:58 BAMbanda left 07:59 hugme joined 08:00 ChanServ sets mode: +v hugme 08:03 FROGGS joined
tadzik hugme: hug me 08:03
hugme hugs tadzik
tadzik :)
moritz also hugs tadzik 08:04
timotimo tadzik (>^_^)> 08:06
tadzik (: 08:07
I usually don't like when people laugh at php on perl channels, but I just _have_ to show you this: news.php.net/php.internals/70691 08:13
it's related because it's about compilers, sort of
08:14 zakharyas joined
timotimo fails to have wordpress publish his article on his g+ profile :\ 08:14
moritz tadzik: :-)
08:14 nnunley left 08:16 dmol joined
timotimo i think i could make a habit out of writing Perl 6 with a non-breaking space in the future :P 08:17
moritz non-breaking habit!
timotimo is now going to spend the rest of the day obsessively checking the view count on his article 08:19
08:19 pecastro left
timotimo i noticed i forgot to mention what file of perl6/roast to put the test i suggested in. i hope my readers will regard that as a puzzle/challenge instead of a roadblock 08:19
moritz they can just ask in here if they feel lost 08:20
tadzik WOW 50% OFF FIRST COMMITS :) 08:25
08:26 denisboyun left 08:28 kivutar left
FROGGS \o/ 08:33
timotimo \o/ 08:36
08:36 frettled_ is now known as frettled
timotimo tadzik, hoelzro, you may get a kick out of "creeper worlds 3", which is on steam greenlight at the moment 08:36
tadzik looks 08:37
timotimo it also appears to have its own scripting language, which looks a bit strange
tadzik I don't see Linux as a supported platform
roman_from_gladiator_showing_a_thumb_down.png 08:38
frettled tadzik++ pretty cool link
tadzik oh, at least they promise one
timotimo yes, it's promised 08:40
tadzik timotimo: did you see Race the Sun? It has like the best promotional video I've seen
flippfly.com/racethesun/
it's already steam-available too
timotimo i played the demo on wine, which worked fine except if i maximized the window in which case the UI became super super sluggish
haha, cute 08:42
ah, cool, it has user-created content 08:43
that is very good to have
hm. i can't get it through the humble widget and unlock it on steam at the same time? 08:45
and no official linux support on the steam page :\ 08:46
tadzik yeah 08:47
according to them, they had a last-minute bug with the linux release on steam
so they delayed it
moritz timotimo++ # advent calendar post 08:49
timotimo EVERY CUSTOMER GETS A STEAM KEY 08:50
tadzik timotimo: waitwhat. Buying on the widget doesn't give you a steam key?
timotimo i was wrong 08:51
normally it says explicitly that you get a steam key from the widget, this one doesn't
tadzik *relief*
timotimo but their blog does say you do get it
gog.com sells it for 5 dollars at the moment
and you get a steam key as well
hint hint :)
tadzik from gog?
that's news 08:52
timotimo www.gog.com/game/race_the_sun
tadzik "Now with 30 Day Money Back Guarantee"
is that ever true, or: does anyone ever try? :)
timotimo hehe. 08:53
tadzik but really, steam key from gog?
timotimo yes.
tadzik also, I don't think gog gives you a linux version, does it?
timotimo "every customer gets a steam key"
tadzik where does it say that?
timotimo one second
flippfly.com/news/race-the-sun-comi...ember-9th/ 08:54
i shot them a tweet to find out if gog will get me access to the linux version as well 08:55
or rather: buying it on gog
tadzik yeah, if that's possible, I'd buy it right away 08:56
timotimo well, it's 3am where they are located
the question is, is that offer for race the sun limited to the next 5 hours? 08:57
oh well. worst case: 30 day money back guarantee
tadzik hehe, let's see if that works as advertised :) 08:58
timotimo 7. GRAAAAH! I'm angry! Can I get my money back right away?
We always aim to provide our users with working games, so first we'll try to get your purchase working on your system. If we can't get it working, we'll get your money back. No worries. :)
but also: If your game doesn't work because you misread our system specs, all we can really offer is that we're sorry for you. :( 08:59
tadzik yeah, gog doesn't list linux compatibility 09:00
timotimo i'll shoot gog's support team a question 09:05
masak mornin', #perl6 09:07
tadzik hey hey masak 09:08
masak I just wanted to report a semi-visual autopun I saw.
diakopter BL INK
timotimo tadzik: let's start #perl6-gaming so we don't annoy non-gamers in here :P
tadzik timotimo: yeah, maybe we should :)
masak a t-shirt with "I ♥ REALISM". but the heart is a picture of an actual anatomical heart.
dwarring_ masak: thanks for the pandoc instructions 09:20
timotimo wow. a pacemaker that gets inserted into any vein that autonomically finds its way to the heart and hooks up to it by itself 09:24
tadzik I wonder how long until someone invents something that travels to the heart and explodes 09:25
timotimo well, if you get the plans from that company, you can easily develop that, but what would be the point? 09:26
tadzik kickstarter? :P
timotimo if you can already put a catheter into a vein of someone
... yeah right
tadzik kidding
but yeah, that's pretty awesome
it sounds like something from a sf/horror movie, hopefully the military is not into these sort of things these days 09:27
dwarring_ I'll tackle wordpress after some zzz's - gist.github.com/dwarring/7848868 09:28
will need to arrange access 09:29
moritz dwarring_: /msg me your email address
dwarring_: and I'll send you an invitation
dwarring_ moritiz: did you get that 09:31
jnthn Probably not if you send it to moritiz :P
moritz invitation sent :-)
dwarring_ arrgh getting late
moritz dwarring_: use tab completion for nicks :-) 09:32
timotimo tadzik: i think the military has no reason to bother with artery-traveling devices if they can just let a bullet travel through the arteries by aiming roughly in the right direction and holding down the trigger for a few seconds
tadzik I guess so
dwarring_ moritz++ 09:33
dwarring_ got invite zzzzz&
09:41 nnunley joined, ribasushi joined
timotimo i may do a lightning talk at the congress this year; but i don't know what about! :( 09:44
at least i'll be reachable under the vanity number PRL6 through the internal GSM & DECT network :D 09:45
09:45 nnunley left 09:46 ribasushi left
arnsholt timotimo: If we wrap up the ZMQ stuff, ipython and Perl 6 perhaps? =) 09:50
09:50 perigrin left, perigrin joined 09:52 fhelmberger joined 09:53 xinming left 09:57 xinming joined
timotimo arnsholt: you do realize the congress is in 1.5 weeks? ;) 09:58
10:00 pecastro joined 10:02 dakkar joined
timotimo i'm not quite sure what exact commit caused the stage0 to not be appropriate any more 10:34
i'm also surprised. i thought i had been able to build nqp-j recently 10:36
but now i really can't
FROGGS (git bisect)++ 10:37
arnsholt timotimo: Oh. That's a tad close, I agree =D 10:38
11:03 xinming_ joined, xinming left 11:06 sqirrel joined 11:07 kaare_ joined
timotimo is going to supply a java bootstrap based on newest nqp/master 11:13
11:24 xinming_ left, xinming joined 11:25 rindolf joined 11:29 xinming left 11:30 xinming joined
timotimo there we go 11:37
dalek p: 468bd92 | (Timo Paulssen)++ | src/vm/jvm/stage0/ (10 files):
get a newer JVM bootstrap in place.
 11:38
tadzik doom dooom doom
11:39 atroxaper joined 11:51 kivutar joined 12:05 rurban1 joined 12:10 ribasushi joined 12:11 fridim_ joined, tgt joined 12:20 atroxaper left 12:21 atroxaper joined 12:22 denis_boyun joined
FROGGS perl -E 'my %h = b => 3; say scalar %h' 12:26
1/8
wth is that?
12:26 denis_boyun left
FROGGS ahh, number of used slots of allocated slots 12:27
timotimo FROGGS: "one out of 8 buclkets"
yeah
12:27 fridim_ left 12:29 tgt left 12:35 xinming left 12:36 xinming joined 12:39 ivanshmakov joined
hoelzro afternoon 6ers 12:44
timotimo: that game looks pretty sweet
timotimo hoelzro: visit us in #perl6-gaming :)
12:47 cognominal joined 12:57 kaleem left 13:04 breinbaa1 left 13:05 breinbaas joined 13:07 geekosaur joined 13:08 woosley left 13:16 woosley joined 13:21 denis_boyun joined 13:24 hummeleB1 joined 13:36 kivutar left 13:41 rindolf left 13:43 kivutar joined 13:46 denis_boyun left 13:48 rurban1 left 13:50 raiph joined, atroxaper left, atroxaper joined 13:51 lue left 13:54 ajr joined, ajr is now known as Guest43855, Guest43855 is now known as ajr_ 13:55 rurban1 joined 13:57 kivutar left, xenoterracide joined 14:05 lue joined 14:11 kivutar joined 14:16 eternaleye left, PacoAir joined 14:25 Bucciarati_ is now known as Bucciarati 14:27 eternaleye joined 14:33 kaleem joined 14:34 denis_boyun joined 14:35 raiph left 14:36 araujo left, araujo joined 14:49 btyler joined 14:51 prammer joined 14:54 rurban1 left 14:57 xinming left 14:58 xinming joined 15:02 nnunley joined 15:04 jnap joined, rindolf joined 15:07 benabik joined 15:08 xinming left 15:09 xinming joined 15:11 johnmilton joined 15:16 thou joined 15:17 johnmilton left 15:18 johnmilton joined 15:19 dmol left 15:20 johnmilton left 15:21 johnmilton joined, dmol joined 15:23 kaleem left 15:25 bluescreen10 joined 15:26 tgt joined 15:28 [Sno] left 15:30 sqirrel left 15:33 zakharyas left 15:34 woolfy left 15:38 lizmat_ left 15:39 eternaleye left 15:50 kivutar left 15:53 takadonet joined 15:58 Psyche^_ joined 16:02 Psyche^ left 16:05 fhelmberger_ joined 16:08 fhelmberger left 16:10 fhelmberger_ left, xinming_ joined 16:11 thou left 16:12 sqirrel joined
PerlJam builds rakudo-jvm for the first time in a long while 16:12
(heck it's been a while since I built any rakudo) 16:13
16:13 xinming left 16:23 jeffreykegler joined, xinming_ left 16:27 raiph joined 16:29 benabik left 16:30 xinming joined 16:35 jdv79 joined
jdv79 is planetsix.perl.org experiencing difficulties? 16:35
timotimo the difficulty of being taken down, i imagine 16:38
16:41 FROGGS left, estrabd_ left
moritz jdv79: use planeteria.org/perl6/ instead 16:41
timotimo oh cool, thanks! 16:42
jdv79 that's where i got redir'd and that also seems to suffer from some ailment
time heals all wounds? 16:43
timotimo huh, weird, it's down for me, too
16:46 ajr_ left
jdv79 its at least dns related 16:47
16:50 denis_boyun__ joined 16:51 denis_boyun left 16:58 ajr joined 16:59 atroxaper left, ajr is now known as Guest98543, Guest98543 is now known as ajr_ 17:02 rurban1 joined 17:05 SamuraiJack left 17:11 denis_boyun__ left, denis_boyun joined
japhb trips over newly-failing tests in p6-pb functionality that hasn't changed in weeks, gets the chilling thought that perhaps it's Rakudo that changed ... time for a deep debug session. :-/ 17:15
timotimo japhb_: oh noes :( 17:17
japhb exactly. 17:21
We'll see. Could have been seemingly unrelated changes somewhere else, not sure yet. 17:22
17:23 spider-mario joined
timotimo oh man. i think in my dream you wrote something like "i have lots of cool stuff, but i won't commit it now" and i only just now realized that you didn't say that in reality 17:27
diakopter yeah; he's so non-committal
timotimo isn't that a weird thing to dream about? >_>
17:28 rurban1 left 17:32 pecastro left 17:36 eternaleye joined, [Sno] joined
japhb
.oO( noncommitphobic )
 17:38
17:38 rurban1 joined
japhb
.oO( commitmentphobe-phobic )
 17:39
arnsholt I get two failures in t/hll/06-sprintf.t for nqp-p. Is that a known failure? 17:40
17:41 ajr_ left
moritz doesn't "make test" too often in nqp 17:41
17:42 rurban1 left
arnsholt As long as you don't make many changes to NQP either, that's OK =) 17:42
moritz I don't :-)
17:43 rurban1 joined
japhb Grammar::Tracer running a non-trivial grammar against even a medium-sized file can take a while ... 17:44
jnthn Well, sure...it intercpets every single method call on the grammar ojbect 17:46
japhb Watching the red and green FAIL and MATCH markers scroll by in little sine waves is like watching holiday bunting
jnthn Actually I suspect having to do the output is a bigger burden than the method intercept... 17:47
17:47 rurban1 left
jnthn Di you try with --optimize=off by the way? 17:47
I *think* the regex optimizer landed recently... 17:48
that's the only recent thing I can recall that'd affect parsing.
japhb Oooh, I should do that.
Still fails. Hmmm. 17:49
Meeting prep &
jnthn timotimo will be relieved it's not that :P 17:50
arnsholt jnthn: Do you get sprintf failures with NQP on Parrot?
17:51 dwarring left 17:53 dwarring_ left
jnthn Dunno, not run those tests on Parrot for a long while 17:55
arnsholt 'k 17:56
timotimo jnthn: phew! 17:58
imagine *my* relief :)
i landed the new stage0 for jvm, jnthn
jnthn timotimo: nice 17:59
timotimo jvmnthn :3
jnthn :P
18:02 cooper left
timotimo i wonder who'll grab the last two free slots and then christmas 18:04
jnthn Not I... :) 18:05
I'll be moslty offline from Sat
18:11 thou joined, dakkar left 18:12 rurban1 joined
jnthn phew, my post is on the 19th, not tomorrow... :) 18:15
timotimo i wonder if "what exactly is slow in rakudo" would be a good post. alas, i'm not nearly close enough to knowing enough to write a post on that :| 18:19
18:21 sqirrel left 18:31 btyler_ joined, btyler left 18:32 darutoko left 18:40 rindolf left 18:42 eternaleye left 18:44 lizmat joined 18:51 eternaleye joined 18:52 pecastro joined
lizmat good * from Cologne 18:54
at the NR.pm meeting, the question came if Perl 6 has something like the "around" functionality in Moose 18:56
I'm aware of Callable.wrap
but somehow I miss how you would do this with roles in Perl 6 18:57
PerlJam I don't think you would do it with roles. 19:01
lizmat that's my thought also, but folks here seem wildly confused about this 19:02
PerlJam you might ask stevan or someone more in-the-know, but I think that Moose roles + around is more of a misfeature because of the ordering issues that can arise during composition. 19:03
19:04 eternaleye left
stevan_ yes, yes it is 19:04
lizmat hehe, the ordering issues was exactly what people wanted me to tell here: how does Perl 6 solve the ordering issues 19:05
jnthn I already did in one of my talks a demo of how to implement around style stuff in a module. 19:06
That's where I think it belongs.
lizmat jnthn: link ?
ribasushi stevan_: what about "ze mopz" ?
stevan_: does it have it / does it plan to have it
jnthn The wrap primitive and all the meta-programming support are in place to do it in pure Perl 6 :)
ribasushi or will Moose implment it outside of the core mop >? 19:07
19:07 eternaleye joined
jnthn lizmat: In jnthn.net/papers/2012-gpw-meta-programming.pdf I think 19:07
19:08 denis_boyun left
stevan_ ribasushi: there is no method modifiers in the mop 19:08
lizmat thank you all for instant clarity about this 19:09
:-)
#perl6 is the best!
jnthn Note that *any* method in Perl 6 can be an "around"; it just calls callsame or callwith... :) 19:10
ribasushi jnthn: so... this raises the very same question then 19:12
jnthn: basically how does "diamond" role composition works in this case?
PerlJam ribasushi: are yo asking who wins? 19:13
ribasushi I mean obviously it is "whatever the author of the module did"
it's more about "what is the right thing to do"
academically
DFS, C3-like, customizeable, exception...? 19:14
jnthn Ambiguity = exception. 19:15
PerlJam jnthn++ very succinct.
jnthn The *entire point* of roles is to not do MRO-based resolution.
moritz we even have a passing test for that :-)
jnthn Well, an entire point... :)
moritz t/spec/S12-class/mro.t 19:16
19:20 FROGGS joined 19:21 skids joined
dalek ast: 0fed404 | moritz++ | S12-class/mro.t:
simplify MRO test a bit
 19:22
skids r: multi sub a(:$i where { $_ > 0 } = 1) { $i.say }; a(:i(2)); a();
camelia rakudo-jvm 413650: OUTPUT«Cannot modify an immutable value␤␤»
..rakudo-parrot 413650: OUTPUT«Cannot modify an immutable value␤ in sub a at /tmp/RQAmDGX7yW:1␤ in any at gen/parrot/BOOTSTRAP.nqp:1075␤ in any at gen/parrot/BOOTSTRAP.nqp:1066␤ in any at gen/parrot/BOOTSTRAP.nqp:1052␤ in sub a at /tmp/RQAmDGX7yW:1␤ in block at /…»
19:23 jeffreykegler_ joined, dwarring joined
skids r: role A [ :$i where { $_ > 0 } = 1 ] { $i.say }; class B does A[ :i(2) ] { }; B.new(); 19:23
19:23 jeffreykegler left
camelia rakudo-parrot 413650, rakudo-jvm 413650: OUTPUT«===SORRY!===␤None of the parametric role variants for 'A' matched the arguments supplied.␤Cannot modify an immutable value␤» 19:23
moritz seems that named and constraints don't mix in rakudo 19:24
n: multi sub a(:$i where { $_ > 0 } = 1) { $i.say }; a(:i(2));
camelia niecza v24-108-g17d73e4: OUTPUT«(timeout)[auto-compiling setting]␤»
skids Was working before. This I found making sure Sum module worked with new star. 19:25
(which it doesn't, sigh)
jnthn r: multi sub a(:$i = 1 where { $_ > 0 }) { $i.say }; a(:i(2));
camelia rakudo-jvm 413650: OUTPUT«===SORRY!=== Error while compiling /tmp/Yxk8W13kpx␤Missing block␤at /tmp/Yxk8W13kpx:1␤------> multi sub a(:$i = 1 ⏏where { $_ > 0 }) { $i.say }; a(:i(2));␤ expecting any of:␤ postfix␤ infix…»
..rakudo-parrot 413650: OUTPUT«===SORRY!=== Error while compiling /tmp/_J7zTqrteg␤Missing block␤at /tmp/_J7zTqrteg:1␤------> multi sub a(:$i = 1 ⏏where { $_ > 0 }) { $i.say }; a(:i(2));␤ expecting any of:␤ postfix␤ in…»
jnthn r: multi sub a(:$i where { $_ > 0 }) { $i.say }; a(:i(2)); 19:26
camelia rakudo-parrot 413650, rakudo-jvm 413650: OUTPUT«2␤»
jnthn It's something to do with the default
skids yep.
FROGGS star: multi sub a(:$i where { $_ > 0 } = 1) { $i.say }; a(:i(2)); a();
camelia star 2013-09: OUTPUT«Cannot modify an immutable value␤ in sub a at /tmp/jw8VCu6Vcl:1␤ in sub a at /tmp/jw8VCu6Vcl:1␤ in block at /tmp/jw8VCu6Vcl:1␤␤»
jnthn r: multi sub a(:$i where $i = 1) { $i.say }; a(:i(2));
camelia rakudo-jvm 413650: OUTPUT«Cannot assign to a readonly variable or a value␤␤»
..rakudo-parrot 413650: OUTPUT«Cannot assign to a readonly variable or a value␤ in sub a at /tmp/7ySvzdPvu5:1␤ in any at gen/parrot/BOOTSTRAP.nqp:1075␤ in any at gen/parrot/BOOTSTRAP.nqp:1066␤ in any at gen/parrot/BOOTSTRAP.nqp:1052␤ in sub a at /tmp/7ySvzdPvu5:1␤ …»
jnthn I *think* it's actually parsing the = as an assignment operator as part of the where clause... 19:27
moritz n: multi sub a(:$i where { $_ > 0 } = 1) { $i.say }; a(:i(2));
camelia niecza v24-108-g17d73e4: OUTPUT«Unhandled exception: Writing to readonly scalar␤ at /tmp/Il2QCmmfiU line 0 (ANON @ 1) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at /tmp/Il2QCmmfiU line 0 (a @ 1) ␤ at <unknown> line 0 (ExitRunloop @ 0) ␤ at <unknown> line 0 (&a @ 0) ␤…» 19:28
jnthn Looks liek it's not the only one :)
Which may well mean that STD is doing it too...
FROGGS yeah, that kinda proves that the parser is faulty
jnthn Well, or at least surprising :)
But yeah, we may want to tweak the preclimit there. 19:29
skids is there an "is" for initializers?
19:29 Alina-malina left
skids (as a workaround) 19:29
jnthn No; best workaround is probably to declare a named subset type. 19:30
PerlJam I thought we had an "is default(...)" or something.
FROGGS PerlJam: not there
is default(42)r: multi sub a(:$i is default(42) where $i) { $i.say }; a(:i(2));
r: multi sub a(:$i is default(42) where $i) { $i.say }; a(:i(2));
camelia rakudo-jvm 413650: OUTPUT«===SORRY!=== Error while compiling /tmp/Iw9ITmF6kg␤Can't use unknown trait 'is default' in a parameter declaration.␤at /tmp/Iw9ITmF6kg:1␤------> ␤ expecting any of:␤ rw␤ readonly␤ copy␤ requir…»
..rakudo-parrot 413650: OUTPUT«===SORRY!=== Error while compiling /tmp/anIa4hcY5z␤Can't use unknown trait 'is default' in a parameter declaration.␤at /tmp/anIa4hcY5z:1␤------> ␤ expecting any of:␤ rw␤ readonly␤ copy␤ req…»
skids thx. 19:31
19:32 takadonet left 19:37 jeffreykegler_ left
skids Heh. Works with a newline after the where. 19:38
19:38 estrabd joined
lizmat hmmm... do we have any tests for MAIN ? 19:39
doesn't seem like it 19:40
19:40 jeffreykegler_ joined
lizmat duh, user error :-( 19:40
PerlJam lizmat: We don't have any tests for those ;) 19:41
jnthn shoves some tests between keyboard and chair 19:43
19:43 Rotwang joined 19:45 raiph left 19:47 jeffreykegler_ left, jeffreykegler_ joined, kivutar joined
lue hello world o/ 19:49
19:50 Alina-malina joined 19:56 tgt left, tgt joined
masak lue! \o/ 19:57
19:57 denis_boyun joined 19:59 fhelmberger joined, rurban1 left 20:00 tgt left 20:02 woolfy joined 20:03 zakharyas joined 20:04 jeffreykegler__ joined, jnap1 joined 20:05 jeffreykegler_ left, jnap left
lue tadzik: so you know, I'm currently in the process of testing fork vs. vfork . (Most of the testing is going to be recompiling everything :/) 20:06
20:06 rurban1 joined 20:07 fhelmberger left 20:09 xenoterracide left, xenoterracide joined
timotimo weird. ohloh found my contributions to ufo and druid and rakudo star and moarvm, but not to rakudo and nqp 20:09
jdv79 planeteria works now
timotimo cool, thanks
20:12 jeffreykegler__ left 20:17 jeffreykegler joined
woolfy lizmat and ribasushi and Sno are totally absorbing the Niederrhein Perl Mongers meeting in Cologne with a discussion about parameter passing in Perl 6. 20:20
dalek kudo/moar-support: aa9592d | (Tobias Leich)++ | src/core/terms.pm:
provide S*OS and $*OSVER
FROGGS Btw, one must reconfigure more for this ---^ 20:21
woolfy Lots of Perl 6 code flashes over the beamer towards the big screen. Fiery words.
Way over my head.
timotimo heh
20:22 jeffreykegler left
woolfy Somebody mentions "poost". Parameter passing in Perl 6 should be more like it happened in poost. I am very confused about the whole discussion. 20:22
lizmat Boost
PerlJam woolfy: are you sure they aren't just making stuff up? ;) 20:23
20:23 rurban1 left
timotimo hahaha :) 20:23
FROGGS maybe someone just said *prost* :o)
woolfy Eh no, I never make things up
lizmat there appears to be a C++ library called Boost
and someone here suggested porting it to Perl 6 calling it Poost 20:24
timotimo yeah, it's a big collection of libraries actually
woolfy Oh we are doing proost as well. I brought a bottle of wine from the city of Perl, a bottle of Belgian beer (Leffe Royal), liquorice liquor (Salmiakki), and some more stuff. We did proost.
lizmat Poo++ ?
woolfy Dump your Poo in the potty. Potty++
ribasushi and if it smells right - Poo# 20:25
FROGGS www.boost.org/doc/libs/1_37_0/libs/...parameters
lue
.oO(It should really be called Boost6, and hope the original Boost doesn't get to v6)
FROGGS that does look very ugly
ribasushi lue: that's ok they can always go to Boost7
PerlJam lue: boosix? bix?
ribasushi b00bsix 20:26
FROGGS hehe
ribas00shi++
lue maybe we could call it not-so-much-of-a-boost-round-these-parts :)
woolfy Oh oh, it is getting poornographic here
skids eww that is ugly. I bet the underscore is mandatory too.
[Sno] www.boost.org/doc/libs/1_55_0/doc/h...tions.html
FROGGS skids: yes, I fear so 20:27
like foo_ is a named
>.<
good that we have Larry++
lue The Boost6 parameter library: the functionality already exists in Perl 6. ← /me hopes that's most of Boost's libraries when ported to Perl 6 :D 20:28
skids would be funny if it ended up being like a 10 line slang to implement.
woolfy Comment from another attendee: this looks like Perl-code from 15 years ago. 20:29
[Sno] where does that stuff is in perl6?
lue assumes "this" equals "Boost code" :)
[Sno] MAIN() isn't
woolfy Ah, we are just teasing you, you big baby!
FROGGS woolfy: that sounds about right
[Sno]: what is wrong with MAIN() ? 20:31
woolfy OK, back to Perl 6, y'all! Nothing to see here, move on, move on...
FROGGS :o)
lizmat given this MAIN: multi MAIN ( *@p, *%n ) { say "positional = @p, named = {%n.perl}" }
FROGGS I'm spec testing, so I have a few minutes to talk :o)
timotimo lizmat: @p[] 20:32
FROGGS true
or {@p} to be consistent
lizmat given this MAIN: multi MAIN ( *@p, *%n ) { say "positional = {@p}, named = {%n.perl}" }
should there be a difference between "--x foo" and "foo --x" ? 20:33
[Sno] FROGGS: why it's meant being wong when multi MAIN() isn't equivalent to po::parse_options()
lizmat foo --x 20:34
positional = foo --x, named = ().hash
--x foo
positional = foo, named = ("x" => Bool::True).hash
this feels wrong to me 20:35
FROGGS [Sno]: I didn't say that, I just said that boost is ugly... might be fine enough for C though
timotimo yeah, the analysis code can't tell if x is supposed to be a Str or Bool flag
lue lizmat: depends on the application (`eix` is one such "order matters" program), though I personally don't think order should matter by default.
[Sno] FROGGS: the code meant not be beautify in your eyes - that's never been the question
the question was whether there're comparable features in p6 20:36
FROGGS lizmat: yes, feels wrong-ish
[Sno]: I can't tell that quickly 20:37
[Sno] FROGGS: I discussed that with lizmat here at nr.pm
I think Liz can meanwhile tell :) 20:38
lizmat from what I know of Boost (now) and MAIN, there is no match
FROGGS [Sno]: but I know how powerful P6's signatures are, so I doubt these miss anything
lizmat it's all possible, to make, I would think
[Sno] FROGGS: that's not the question, either ;)
lue from quickly glancing at the Boost link provided, I'd say they're comparable. Doesn't mean the comparison shows mostly-similar designs :)
lizmat burt it doesn't match MAIN's specced or implemented capabilities
*but
lue lizmat: Perl 6's main requiring named parameters first feels like an unnecessary restriction in syntax. There're many times where putting the option after the positionals makes frequently editing that option easier. 20:40
lizmat indeed my point, I don't think there should be difference 20:42
skids There are a good many CLI that use --flags as adverbs on a subsequent ordered argument then use them again on the next. e.g. imagemagick. Not that that's good design. 20:44
lue skids: yeah, like the `eix` I mentioned. But IMO that tends to be something you know you're doing, not something done by default. 20:45
(also, Perl 6 doesn't follow the "positioning of nameds matter" rule; it just ignores the misplaced ones.) 20:46
20:47 eternaleye left
skids being informed he should hit the road before more snow falls. 20:48
segomos_ appropriate name is appropriate
20:52 zakharyas left 20:53 skids left
jnthn segomos_++ # best thing I read today 20:53
20:54 eternaleye joined
masak :) 20:58
21:04 denis_boyun__ joined 21:05 denis_boyun left, dmol left 21:06 dmol joined
lue tadzik: I... I don't know. Somehow last night I got it working again, because using parrot/nqp/rakudo with fork() works now. This issue is a lot more evil than a simple switch between fork and vfork. 21:07
timotimo wow, we can fork parrot?
and it doesn't explode horribly?
fork is scary to me. 21:08
lue timotimo: a fork happens in parrot code every time you qx// . A fork in NQP every time you shell() . 21:09
timotimo ah, of course
21:09 raiph joined
lue switching them to vfork last night seemed to fix the problem, but apparently it actually didn't. So I have panda now (yay! \o/) but no idea what causes those random build failures. 21:11
FROGGS lue: see? that is _exactly_ what happened to me 21:12
I guess some force wiped our memory after we found out that *they* are to blame... 21:13
FROGGS .oO( &$/§=%)&% )
hi, what's up? 21:14
lue
.oO(curse you, non-deterministic failures!)
 21:15
21:18 dmol1 joined 21:20 dmol left
lue FROGGS: it's *so* weird, that's about all i know at this point :) . 21:21
FROGGS true
tadzik lue: yeah, that's what I've found :/ it's a heisenbug 21:22
dalek kudo/moar-support: 48a2a1e | jnthn++ | tools/build/Makefile-Moar.in:
Don't run t/00-parrot on Rakudo on Moar.
 21:23
lue tadzik: I still suspect it's a system issue, considering I was getting ENOMEMs out of it. Which tells me that sometimes your computer just doesn't feel like being helpful :/ 21:24
21:24 kaare_ left
tadzik seems that it happens for everybody's computers 21:24
21:25 ajr joined 21:26 ajr is now known as Guest39393, Guest39393 is now known as ajr_ 21:28 xenoterracide left
lue tadzik: yeah, and unless there's a C-level way to do a better fork, I don't see a solution :/ (I see only a sort of "try again if it failed, up to X times" deal atm) 21:30
21:30 kolmurm joined
lue finally has Pod::To::HTML, after all this time :) 21:32
21:34 xenoterracide joined 21:36 tgt joined 21:38 kolmurm left
lue tadzik: would it be on the Pod parser or the Pod::To::HTML module to *not* cut off the leading spaces on implicit code blocks? 21:38
21:41 tgt left
tadzik I think it's specced to do that 21:43
so, the former
21:44 dmol1 left
lue tadzik: the POD parser preserves that space, which I think looks better. S26 demonstrates the difference between explicit code blocks and implicit ones. (The implicit ones are the code blocks that are horribly aligned with normal text ☺) 21:45
21:45 dmol joined
tadzik lue: by POD you mean Perl 5 POD, right? 21:45
21:45 denis_boyun joined
lue Yes. P6's is Pod or Pod6 in my spelling :) 21:45
tadzik right 21:46
well, in that case implicit blocks are parsed differently than explicit ones, I'm not sure I like that
21:46 denis_boyun__ left
lue (That space removal thing also broke the WHY/WHEREFORE diagram in S26, though that feels more "classical bug" than "design issue") 21:46
raiph perl6advent.wordpress.com/2013/12/1...mment-3478 # commenter asks what does "winner" mean? does "await" mean the opposite of its meaning in C#? 21:47
21:47 dmol left 21:48 thou left 21:49 dmol joined
masak 'night, #perl6 21:49
tadzik good knight masak
lue tadzik: I would go for "use CSS to create that pleasing left-side space", but the HTML output makes no distinction between explicit and implicit code blocks, so you'd end up adding too much LHS space to the explicit blocks.
♞ masak o/
tadzik lue: yup 21:50
21:50 tgt joined
tadzik that's another reason why I don't really like it :) 21:50
it'll result in renderers trying to guess what the longest leading space is, bleh
21:50 colomon left 21:51 colomon joined 21:52 FROGGS left, denis_boyun__ joined
lue I personally think "indented code blocks" is a fair assumption to make on how most people want blocks of code to be output. (i.e. preserve those extra spaces from the normal text's left margin rightward) 21:53
21:53 denis_boyun left, Khisanth left
tadzik but I think they'll also assume the explicit code blocks to be indented in the same way 21:54
lue tadzik: all they have to do is not indent the =begin code line :) 21:55
21:55 eternaleye left, dmol left 21:56 dmol joined
lue I see this more as a special-handling of implicit code blocks issue, than a "make both kinds of code blocks consistent" issue. 21:56
The other solution is for renderers to make distinctions between the kinds of blocks for formatting purposes, e.g. <pre style="implicit"> and <pre style="explicit"> 21:57
tadzik special handling is what I'd rather avoid
this assumes that the resulting AST remembers what kind of blocks they were 21:58
lue Some would say the parsing text differently based on how much space is in front is already special handling :)
21:59 eternaleye joined 22:00 denis_boyun joined, nnunley left
lue I'd say that implicit v. explicit is special enough to warrant some extra info in the AST. So long as the grammar parses their contents differently, at least. 22:00
22:00 denis_boyun__ left, woolfy left
lizmat decommute& 22:01
22:02 bibifuc left, lizmat left, BenGoldberg joined
lue I should file a rakudobug on that WHY/WHEREFORE diagram though. From the HTML source it seems that it just interpreted the diagram as a series of consecutive implicit code blocks :) . (Seeing a new code block whenever the next line's indentation was less than the codeblock's set indentation) 22:05
22:07 spider-mario left 22:08 Khisanth joined
dwarring here's my day 18 advent draft - perl6advent.wordpress.com/?p=2129&...eview=true 22:09
all being good, I'll publish in a couple of hours
22:11 [Sno] left 22:16 rurban1 joined 22:20 eternaleye left, FROGGS joined
lue dwarring: some general points: the whole post is formatted as a blockquote (if you go into source mode there'll likely be a <quote> or something at the top), and double-check your code to make sure all the <rule> stuff isn't missing. 22:21
(those angle brackets tend to get eaten by wordpress.)
FROGGS &lt; is your friend 22:22
dwarring lue: been there with the angle brackets! 22:23
lue dwarring: "The deal contains of one" should be s/contains/consists/, and "please consider using on or more" on should be one :)
That's all I noticed (think there was one more nitpick but can't find it back). dwarring++ for the post :) 22:24
(I only noticed the missing angle brackets in the first code block, with the grammar.) 22:25
22:25 ssutch joined
dwarring aarrgh that would make for any interesting post 22:25
lue++ FROGGS++ 22:26
22:30 hummeleB1 left 22:33 cooper joined, cooper left, cooper joined 22:35 rurban1 left
FROGGS dwarring: how does it deal with duplicates in the first grammar? 22:36
dwarring The first grammar is'n t checking for duplicates 22:41
that doesn't start until actions or assetions are added 22:42
i'll see if I can make that clearer
FROGGS just this sentence is weird in this regard: As expected, the first cut of the grammar cheerily parses hands and deals with duplicate cards. 22:43
I'd just strip the " and deals with duplicate cards"
22:44 fridim_ joined
dwarring FROGGS: yeah that's wrong and confusing 22:46
FROGGS and the "s" from "Disallowing Duplicates" is not bold :P
timotimo now there's &lt; and &gt; inside the <pre> that show up as such 22:52
22:53 jdv79 left
lue Wp is kind of a moron when it comes to angle brackets :) . 22:53
22:53 bluescreen10 left 22:54 bbkr left 22:57 [Sno] joined 22:59 Rotwang left, [Sno] left 23:00 johnmilton left 23:01 FROGGS left, FROGGS joined, xinming_ joined 23:03 [Sno] joined
dwarring timotimo: whoops, fixed 23:03
23:04 xinming left 23:05 dmol left 23:06 dmol joined
timotimo i'd call them "callees" rather than "descendants" for the dynamic variable thing 23:06
and there's ’%#’ but then there's '%' and '*' right after that 23:07
probably "intelligent quotes" from the wysiwyg editor or something
'hey there's an extra ...' looks unpleasant, what with starting with ' and having a ' in between 23:08
23:08 rurban1 joined 23:09 dmol left
timotimo and i'd also briefly mention the :i and .lc, just in passing 23:09
23:10 dmol joined
timotimo you could mention what exactly causes trouble with your code assertion when backtracking and how you can deal with it 23:10
(normally, i'd say deal with it with a cascade of dynamic variables that copy their outer data, but that doesn't work in this case, because the data has to travel sideways, rather than downwards; you could, however, let the hand call down to hand again and thus have a stack to work with. doesn't seem very good for this exact case, though)
it's a good post, thanks dwarring :) 23:11
bedtime :) 23:13
dwarring timotimo++ thanks gnite
23:14 xinming_ left 23:15 FROGGS left, colomon left 23:16 xinming joined 23:21 rurban1 left 23:27 colomon joined
dalek ecs: 2b0ef1b | lue++ | S15-unicode.pod:
[S15] Clarify values the uniprops hash can contain

Based off the value types defined in UAX#44, §5.2, the mapping is now thus:
   B → Bool
   N → Rat
   C, E, S, M → Str
Though perhaps the 'E'numeration types could actually be Enums, if the benefits of codifying them as enumerations is worth it.		 23:31
ecs: 51a72a0 | lue++ | S15-unicode.pod:
[S15] Modify "Final Considerations" on mismatching in operators.

Essentially, NFC ~ NFD -> Uni, instead of the previous "causes an error" definition. Note that detailing various operators and functions still needs doing.
23:39 FROGGS joined 23:47 colomon left, woolfy joined 23:48 grep0r left 23:53 colomon joined 23:55 grep0r joined 23:58 denis_boyun left

Powered by Cro and the Raku® Programming Language.

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

Thank you!