»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.
00:08 ajr_ left 00:21 FROGGS left 00:28 rurban left 00:32 jnap left 00:34 rurban joined 00:36 rurban1 joined 00:37 fridim_ joined 00:39 rurban left 00:44 vael joined 01:06 rurban1 left 01:10 SmokeMachine joined 01:15 risou_awy is now known as risou, pecastro left 01:17 pecastro joined, SmokeMachine left 01:18 SmokeMachine joined 01:23 SmokeMachine left 01:33 risou is now known as risou_awy
xenoterracide so is the jvm version able to use libraries from java? or is that pending the nativecode for java 01:33
(unsure whether nativecode jvm means C* code or jvm bytecode) 01:34
01:37 rurban joined
xenoterracide also, this is probably in a FAQ somewhere, but why is there a separate repo for star? isn't star just fundamentally tagged points in time and pulls in additional useful libraries? 01:39
01:39 rurban1 joined 01:42 rurban left 01:44 rurban1 left
colomon I believe nativecode jvm means C code 01:52
but there already is some other java interop stuff
(for rakudo jvm)
dunno the exact details, I've been bit out of the loop 01:53
benabik Nativecall on JVM calls native (C ABI) code. JVM interop is built-in to Rakudo/JVM. Don't know the magic incantations for it. The threading code is actually pretty readable and uses the interop, so you could look there. 02:03
02:09 grondilu joined
BenGoldberg Why doesn't this work right? 02:15
r: say "abcdefgh" x .5
:)
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«␤»
BenGoldberg is in a silly mood today 02:16
benabik BenGoldberg: It does.
r: say Int(.5)
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«0␤»
BenGoldberg I was reading perl.plover.com/IAQ/IAQlist.html, and found it under the question "How can I get just the first half of a long string?" 02:18
benabik Erm. 02:19
lue r: say "abcd".comb[0..*/2].join 02:20
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«abc␤»
benabik r: say "abcd".comb[0..^*/2].join
lue r: say "abcd".comb[0..*/2 - 1].join
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«ab␤»
benabik r: say "abcd"[^*/2]
camelia rakudo-parrot 995207: OUTPUT«Index out of range. Is: 0.5, should be in 0..0␤ in method gist at gen/parrot/CORE.setting:11965␤ in method gist at gen/parrot/CORE.setting:1023␤ in sub say at gen/parrot/CORE.setting:12865␤ in block at /tmp/Z37u6sRza6:1␤ in any at /tmp/Z…»
..rakudo-jvm 882e33: OUTPUT«Index out of range. Is: 0.5, should be in 0..0␤ in block at /tmp/P6KQVWXF8m:1␤ in any eval at gen/jvm/stage2/NQPHLL.nqp:1084␤ in any evalfiles at gen/jvm/stage2/NQPHLL.nqp:1290␤ in any command_eval at gen/jvm/stage2/NQPHLL.nqp:1194␤ in any c…»
benabik r: say "abcd".comb[^*/2].join 02:21
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«c␤»
lue
.oO(Str.bifurcate)
benabik r: say "abcd".comb[^(*/2)].join 02:23
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«ab␤»
benabik Oy. The ^ binds tighter than /, so it's doing (0..^4)/2, which is 2. 02:24
Took me a moment to figure that out.
BenGoldberg It's good to see real answers... but if you take a look at the web page I mentioned... well, $str.bifurcate would probably be the "right" perl6 answer for that question on that type of list. 02:26
diakopter p5eval: print $\ x 1_000_000 02:32
p5eval diakopter: 1
diakopter rnp: print $\ x 1_000_000
rp: print $\ x 1_000_000
camelia rakudo-parrot 995207: OUTPUT«===SORRY!=== Error while compiling /tmp/Nc2lgUDHLa␤Unsupported use of $\ variable; in Perl 6 please use the filehandle's :ors attribute␤at /tmp/Nc2lgUDHLa:1␤------> print $\⏏ x 1_000_000␤»
benabik r: say ("abcd" ~~ /(.*)(.*) <?{$0.chars == $1.chars}>/)[0] 02:33
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«｢ab｣␤␤»
benabik Splits only exactly in half. :-D
02:35 rurban joined
benabik r: say ("abcde" ~~ /(.*).?(.*) <?{$0.chars == $1.chars}>/)[0] # Or, rounding down 02:36
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«｢ab｣␤␤»
lue BenGoldberg: too bad .bifurcate doesn't exist :)
r: say "abcde".comb[0..^*/2].join 02:37
camelia rakudo-parrot 995207, rakudo-jvm 882e33: OUTPUT«abc␤»
lue ^ more correct, IMO. round(2.5) = 3 02:38
xenoterracide does the async stuff work on parrot as well or jvm only 03:13
benabik JVM only.
Nobody's figured out how to map Parrot's parallelism to what Rakudo wants.
03:16 risou_awy is now known as risou 03:17 woosley left
lue I thought it was that we haven't figured out Parrot's parallelism yet :P 03:18
benabik Parrot has the ability to launch parallel tasks. It does not use the standard shared memory model of threading though 03:19
03:19 wbill left 03:21 odoacre joined
benabik niner.name/Hybrid_Threads_for_the_Parrot_VM.pdf 03:23
03:25 woosley joined 03:26 xinming_ joined 03:27 xinming left 03:32 risou is now known as risou_awy 03:42 telex left 03:44 telex joined
xinming_ df 03:45
03:48 johnny5_ joined 03:58 Timbus joined 04:22 rurban left 04:44 rurban joined 04:55 preflex left 04:56 preflex joined, ChanServ sets mode: +v preflex 04:57 fridim_ left 04:58 grondilu left 05:17 risou_awy is now known as risou 05:19 BenGoldberg left 05:30 SamuraiJack_ joined 05:33 risou is now known as risou_awy 05:55 rurban left 06:02 wbill joined 06:04 xenoterracide left 06:13 xenoterracide joined 06:19 dmol joined 06:26 stevan__ joined, rurban joined 06:28 xinming_ left, vael left, stevan_ left, timotimo joined 06:30 perigrin joined, xinming joined 06:31 vael joined, rurban left, revdiablo joined, japhb__ joined 06:32 ribasushi joined, yeltzooo joined 06:34 [Sno] left 06:36 [Sno] joined
moritz rakudo.org/downloads/star/rakudo-st....10.tar.gz uploaded, lue++ 06:56
06:56 rindolf joined
moritz I'll try to get around to publishing the announcement on rakudo.org later today 06:57
07:02 rindolf left 07:03 rindolf joined 07:05 kaleem joined 07:06 FROGGS joined 07:18 risou_awy is now known as risou 07:27 rurban joined 07:29 darutoko joined 07:31 rurban left 07:36 risou is now known as risou_awy 07:38 cognominal__ left, denis_boyun joined 07:44 rindolf left 07:53 iSlug joined 08:10 odoacre_ joined 08:13 odoacre left 08:16 odoacre_ left 08:19 leo2007 joined 08:21 leo2007 left
FROGGS o/ 08:24
08:25 odoacre joined
FROGGS timotimo: I am planning to fix run() today btw 08:27
timotimo great! :)
08:27 fhelmberger joined 08:28 rurban joined
FROGGS that is, fixing it on linux@parrot 08:29
(windows@parrot works)
tadzik so many platform combinations 08:30
FROGGS yeah :/
timotimo got woken up by construction work apparently 08:31
FROGGS timotimo: at what time? 08:32
timotimo began half an hour ago
FROGGS not too bad
timotimo so ... i did get an acceptable amount of rest, but it still gave me a headache :|
FROGGS (I have kids, you know)
08:33 rurban left
FROGGS and these little things tend to wake up at 6 (daylight savings time) 08:33
so, I AM AWAKE FOR 4.5 HOURS ALREADY! 08:35
>.<
nwc10 and you have made how many commits? :-/ [yes, real life doesn't work like that] 08:36
FROGGS nwc10: none, kids wanted to play, eat, get to the kindergarten...
moritz so, I tried my hands at implementing "no strict", gist.github.com/moritz/7210995 08:38
you can see that even though I wrote $decl.decl('contvar'); the AST dump says :decl(static)
any idea why?
tadzik timotimo: at least you're not woken up by windows vibrating because of the bass of the asshat below you playing his music on 7:30 08:39
tadzik grumbles
timotimo o_O 08:40
tadzik setting a new wifi network named "what a &$!%*&$^%!* plays hiphop at 7" doesn't help
FROGGS moritz: my v5 has some sort of 'no strict' 08:41
moritz: it just autodeclares vars instead of complaining that they are not declared
moritz FROGGS: that's what I'm trying to do, yes 08:43
but somehow they end up as :decl<static>, and that makes it kinda hard to assign to them
08:44 bloonix joined
moritz oh, because $*W.install_lexical_symbol sets .decl to 'static' 08:44
seems there is some fun with static lexpads and actual values etc. going on that I don't understand 08:45
or maybe install_lexical_symbol is simply the wrong method to call 08:46
masak moarning, #perl6
FROGGS morning masak
hoelzro morning #perl6 08:47
bloonix morning
masak tadzik: passive-aggressive neighbourly communication through wifi network names. ingenious.
tadzik: I bet a really nice short story could be written on that topic. :) 08:48
FROGGS o/ hoelzro, bloonix 08:49
tadzik masak: active-positive doesn't work either :( I had "Bring cookies to 256" for a few months, nothing happened
I'm afraid that no one reads those :( 08:50
FROGGS ó.ò 08:51
tadzik but I'm amused at neighbours setting names like "Mommoth-Dadmoth"
(or Mamut-Tatut in Polish)
FROGGS hehe 08:52
ahh
tadzik ah, it's mammoth in english. Oh well
masak: I'm wondering if I could sketch some interesting plot and write it during vacation 08:53
like an Orwell-ish society, never able to leave their homes, only able to communicate by setting essids ;) 08:54
08:55 tipdbmp left
FROGGS tadzik: or you hack your router firmware so it broadcasts your facebook status via SSID 08:56
tadzik hahaha
ooh
or something that instead of regular wifi traffic just streams music 08:57
now that's beyond my wifi-fu
moritz tadzik.wifi.fu.upgrade 08:58
tadzik I wonder if someone will pick up my YAPC talk idea, I saw that it got published in survey results. "How wifi works and why we can never get it right" 08:59
08:59 tipdbmp joined
tadzik (on the conferences, that is) 08:59
hoelzro hahaha
09:01 zakharyas joined 09:02 xinming left, pernatiy joined
moritz well, the second part (why we can never get it right) is quite easy to answer 09:03
consumer hardware and uplink doesn't like more than a hand full of users
and enterprise stuff is way too expensive
timotimo ah, yes, i think "install lexical symbol" is exactly that: symbols. like ... i guess classes for example? 09:04
09:04 dansamo joined
moritz timotimo: is there a helper method for installing lexical variables? 09:04
tadzik timotimo: classnames, I think
timotimo hm, something the compiler does when it sees vars with the right delc? 09:05
decl
but you're already before the compielr
moritz well, decl isn't the only thing
that doesn't install it into the the block's symbol table, for example 09:06
or does it?
timotimo it could be you have to do it like line 5593 of Actions.nqp 09:07
er, not as in create a container descriptor, just the push part with the cur_lex_block? 09:08
cur_lexpad, even
moritz is $*UNIT a QAST::Block? 09:10
timotimo hm, i seem to remember something about that 09:11
in the way the optimizer finds the unit
yes, unit is a block 09:12
FROGGS moritz: $block.symbold(:name('$foo'), :decl('var')) does it
09:12 darutoko left
FROGGS without the :decl it is just a lookup 09:12
symbol*
09:13 fhelmberger_ joined
FROGGS I think you'll see everything you need in variable_declarator 09:13
09:14 cognominal joined 09:16 fhelmberger left
moritz FROGGS: so $block.symbol does the pushing to $block[0] for me? 09:17
09:19 donaldh left, risou_awy is now known as risou
moritz now I got it to actually generate a :decl(contvar) 09:22
FROGGS moritz: no, you would have to push something like QAST::Var.new( :name('$_'), :scope('lexical'), :decl('var') ) to $block[0]
moritz and it still dies with "Cannot modify an immutable value"
FROGGS because block[0] is a statementlist of variable declarations, every block has two statementlists... 09:23
09:23 donaldh joined
moritz yes, I know 09:24
FROGGS k
maybe you an steal parts from github.com/rakudo-p5/v5/blob/maste...#L329-L360 09:25
09:28 rurban joined
FROGGS moritz: method create_container_descriptor($of, $rw, $name, $default = $of, $dynamic = nqp::chars($name) > 2 && nqp::substr($name, 1, 1) eq "*") { 09:33
09:33 rurban left
FROGGS you passed something true-ish as second arg? 09:33
moritz I didn't even create a container :( 09:34
09:34 daxim joined 09:35 risou is now known as risou_awy
moritz will try later, but have to $work first 09:42
09:47 dakkar joined 09:48 yannickberthoud joined, iSlug left
dalek Heuristic branch merge: pushed 357 commits to nqp/nqp_spawn by FROGGS 09:59
Heuristic branch merge: pushed 200 commits to rakudo/nqp_spawn by FROGGS 10:03
10:04 cognominal left, cognominal joined 10:05 spider-mario joined
dalek p/nqp_spawn: aca76d0 | (Tobias Leich)++ | src/vm/parrot/ops/nqp.ops:
silence type cast warnings
 10:09
arnsholt Spawn?
FROGGS used underneath run() 10:10
arnsholt Ah, right
10:13 rindolf joined 10:18 denis_boyun left
dalek p/nqp_spawn: 6ac28c2 | (Tobias Leich)++ | src/vm/parrot/ops/nqp.ops:
strip accidentally added duplicate nqp_spawn op
 10:19
10:24 denis_boyun joined 10:27 xenoterracide left 10:29 Vendethiel joined, rurban joined
FROGGS huh, it compiles O.o 10:33
10:34 rurban left 10:38 xenoterracide joined
FROGGS hmmm, it even seems to work 10:38
timotimo: is there a simple one-liner that shows how run() is broken? 10:41
except handling %*ENV I mean
timotimo except handling %*ENV? but that's the only way i know it's broken? 10:42
FROGGS yeah, you are right 10:43
timotimo i guess you could ./perl6 -e '%*ENV<GIT_DIR> = "nqp/.git"; run <git show a764846>'
FROGGS backlogged right now
timotimo and that should show nqp's 2013.08 tag
if, indeed, your nqp checkout is in nqp/
FROGGS nqp::spawn(["echo", "\$PWD"], '/tmp', {}) worked already 10:45
have to recompile though
10:51 denis_boyun left
FROGGS hmmm, somehow I does not take the 2013.10 tag into account in my nqp checkout 10:56
I see the tag, but git describe tells me something about 2013.09-390... 10:57
11:00 sqirrel joined
timotimo that may be okay 11:04
the only interesting thing is that the environment variable properly gets set 11:05
FROGGS dev/rakudo$ perl6 -e '%*ENV<GIT_DIR> = "../nqp/.git"; run <git show a764846>' 11:06
commit a76484663e96b374e1425509d9cc3283be4cdd14
Author: Moritz Lenz [email@hidden.address]
perl6 -e '%*ENV<HURZ> = "42"; run <echo $HURZ>' 11:07
42
moritz FROGGS: try git describe --all
FROGGS git describe --all 11:08
heads/nqp_spawn
moritz that could happen if the tag wasn't annotated
FROGGS k, I'll hope that the tag appears then when I merge my branch into master 11:10
11:11 denis_boyun joined
FROGGS hmm, or maybe I just copy the source files, and create new clean commits 11:13
11:19 spider-mario left
jnthn moritz: (no strict) you may be able to just pass the right kinds of things to sub declare_variable and have it work out, but the install_ method on Perl6::World you want is install_lexical_container or so 11:19
11:20 risou_awy is now known as risou
moritz jnthn: ok 11:20
xenoterracide can I put a class or role inside of a package or a module and then strip off the outer namespace on importa somehow? so package Foo { class Bar {} } in Foo/Bar.pm could be used like use Foo::Bar; Bar.new (or some such) 11:27
11:27 pernatiy left
moritz xenoterracide: on the user side, you can alias with 'constant' 11:27
constant Bar = Foo::Bar;
jnthn class Bar is export { } # or so 11:28
moritz though if you want to do it always, I don't see the point in nesting the classes at all
11:30 rurban joined
xenoterracide hmm 11:33
k
FROGGS hmmm, my run() implementation seems to trigger the following regression: 11:34
t/spec/S06-other/main-usage.rakudo.parrot (Wstat: 0 Tests: 22 Failed: 14)
Failed tests: 1-6, 8-12, 18-20
11:34 rurban left
FROGGS ahh, maybe not 11:35
I think it just tries to execute ./perl6 when it is called perl6-p
11:36 risou is now known as risou_awy
FROGGS yeah, creating ./perl6 fixes it! 11:37
11:37 denis_boyun left
moritz p: say $*EXECUTABLE_NAME 11:38
camelia rakudo-parrot 995207: OUTPUT«./rakudo-inst/bin/perl6-p␤»
moritz j: say $*EXECUTABLE_NAME
camelia rakudo-jvm 882e33: OUTPUT«perl6␤»
moritz shouldn't it use that one?
FROGGS perl6-p -e 'say $*EXECUTABLE_NAME' 11:39
perl6-p
but:
not ok 1 - Parse error contains line number
# got err: "sh: 1: ./perl6: not found\n"
11:39 colomon left
xenoterracide is there something that does a fantastic job of explaining why module or package 11:40
11:40 colomon joined
moritz $cmd = $perl6 ~~ m:i/java/ ?? "./perl6 " !! "$perl6 "; 11:40
from t/spec/packages/Test/Util.pm 11:41
dalek ast: 7f925f9 | moritz++ | packages/Test/Util.pm:
try to fix invocation of rakudo-jvm
 11:42
moritz FROGGS: try with that patch agian
*again
11:43 denis_boyun joined
FROGGS moritz: line 97 would strip the -p, no? 11:44
xenoterracide Method 'username' must be implemented by User because it is required by a role # error message sucks a little? which role
jnthn flight &
FROGGS jnthn: I wish you a good one :o)
moritz FROGGS: aye
FROGGS timotimo: /o\ I still used parrot's spawn in my test 11:45
moritz FROGGS: so better to write it as $perl6 ~~ s{^perl6}{./perl6};
11:45 yannickberthoud left
FROGGS yeah 11:46
dalek ast: 61a7f63 | moritz++ | packages/Test/Util.pm:
another attempt to fix perl6 invocation in tests
moritz Unsupported use of brackets around replacement; in Perl 6 please use assignment syntax 11:50
at least the error message is awesome
./perl6-j -e 'say $*EXECUTABLE_NAME' 11:52
perl6
that totally sucks.
FROGGS *nod* 11:53
lizmat agree
moritz patches
11:53 GlitchMr joined
FROGGS timotimo: so I guess parrot's spawn was only broken on windows? (wrt env) 11:56
I know that the autoquoting was totally off 11:57
dalek kudo/nom: 55ff4fc | moritz++ | src/core/terms.pm:
on the JVM, we are perl6-j
moritz t/spec patch forthcoming too
but first I want to test in on parrot also 11:58
12:00 GlitchMr left 12:02 GlitchMr joined 12:03 xenoterracide left
donaldh thinks that on JVM, executable name should be a system property that is configured by the shell script wrapper. 12:05
12:05 benabik left
dalek ast: 4a3de07 | moritz++ | packages/Test/Util.pm:
actually fix perl6 invocation

  ... on my machine, at least. Tested with perl6-p and perl6-j		 12:05
moritz donaldh: sounds plausible. I'll accpet such patches :-) 12:06
donaldh moritz: ack 12:08
donaldh really must get around to posting a signed CLA.
arnsholt has probably submitted a CLA 12:09
I have a Rakudo commit bit, but can't actually remember doing the CLA thing =)
donaldh does anyone know a FAX number for TPF ?
moritz arnsholt: maybe you did a release once, and snuck in like that
arnsholt Possibly 12:10
[Coke] diakopter: `locale -a` says otherwise. 12:11
12:11 odoacre_ joined
[Coke] diakopter: oops. sorry, "i tried setting that and perl5 doesn't think so." 12:11
timotimo computer says no. 12:15
[Coke] suspects he should start opening rakudo-jvm tickets for things that are failing the spectests so we don't get complacent. 12:17
moritz [Coke]: +1
FROGGS Method 'compose' not found for invocant of class 'Attribute 12:20
:o(
moritz FROGGS: what are you doing?
FROGGS adding class Proc::Status to src/core/control.pm 12:21
colomon [Coke]: +1 12:22
12:22 sqirrel left
dalek rl6-roast-data: 6eed77e | coke++ | / (5 files):
today (automated commit)
 12:22
12:31 rurban joined 12:35 rurban left
dalek ast: 5bb2991 | coke++ | S02-types/baghash.t:
Opened a ticket for this rakudo-jvm failure
 12:36
[Coke] ok, there's one. :P
I wonder if rakudo-p is implicitly relying on some parrot Hash behavior. 12:37
-> $dayjob
mail
ww !
12:40 sqirrel joined 12:45 pernatiy joined
FROGGS this is what I am doing about run() if someone is interested: gist.github.com/FROGGS/7350a48de9d920072d90 12:56
testing on windows now
12:58 Vendethiel left 13:00 ajr joined, ajr is now known as Guest59326 13:02 Guest59326 is now known as ajr_ 13:03 sqirrel left 13:06 wbill left, kaleem left 13:08 jnap joined 13:12 fhelmberger joined 13:14 fhelmberger_ left 13:17 Vendethiel joined 13:20 bluescreen10 joined, risou_awy is now known as risou, brrt joined 13:21 bluescreen100 joined, logie joined 13:22 bluescreen__ joined, bluescreen10 left, BitPuffin joined 13:25 bluescreen100 left 13:26 bluescreen__ left 13:31 rurban joined 13:33 benabik joined 13:34 bluescreen10 joined 13:36 rurban left, risou is now known as risou_awy 13:40 araujo joined 13:45 raiph joined
dalek p/run: 54f6198 | (Tobias Leich)++ | src/vm/parrot/ (2 files):
implemented our own nqp::spawn()
 13:48
kudo/run: ea5c22f | (Tobias Leich)++ | src/core/ (2 files):
use nqp::spawn()
colomon FROGGS: so that just works on Parrot so far? 13:54
FROGGS colomon: with micro-tweaks of the spectests, yes 13:55
13:55 logie left
colomon FROGGS++ 13:55
FROGGS testing on windows now again though
but last time it worked
hmmm, I could even make it work so that no spectest-changes are needed 13:56
but this would be against spec 13:57
lizmat commute to AmsterdamX meeting& 14:03
14:03 lizmat left
masak .oO( Amsterdam.pm and AmsterdamX.pm -- now with twice as much Amsterdam! ) 14:11
14:15 colomon left
moritz and next year: generaten AmsterdamY 14:17
*generation
typo'd jokes are only hal as funny
FROGGS C:\rakudo>perl6-p -e "%*ENV<HURZ> = "42"; say run(<echo %HURZ%>) == 0"
42
True
moritz \o/ 14:18
FROGGS++
timotimo sounds great!
FROGGS C:\rakudo>perl6-p -e "%*ENV<HURZ> = "42"; say ?run(<echo %HURZ%>)"
42
True
seems to work...
I am just thinking....
timotimo well, thank you for doing this :) 14:19
FROGGS currently shell() returned posix style value, whereas it would return the exit value only atm
:o)
timotimo the posix-style value was the one you had to +> 8 first? 14:20
FROGGS to get the exit values, yes 14:21
and +& 0xFF to get the signal 14:22
which always zero on windows
that is the spectest diff btw: gist.github.com/FROGGS/7350a48de9d...ctest-diff 14:24
moritz / timotimo / all: would be nice to get a +1 / -1 14:25
moritz FROGGS: if the spec says so, +1 14:26
masak what moritz said :) 14:27
14:27 colomon joined 14:30 colomon left
timotimo yes, spectest-diff seems sane (if the spec says so, and it would make sense that it does) 14:32
14:32 kaare_ joined 14:34 telex left 14:35 colomon joined 14:37 FROGGS[mobile] joined, colomon left 14:38 telex joined 14:39 sqirrel joined 14:47 Vendethiel left 14:59 btyler joined 15:02 colomon joined 15:04 wbill joined
timotimo tadzik: hellsongs is coming to my town to play a concert. do you think i should go? 15:05
15:05 telex left
tadzik timotimo: oh, I know I would :) 15:05
jnthn survived his flight :) 15:06
15:06 telex joined
tadzik bah, I don't remember what was the last concert I've been on 15:06
Arkona, most likly 15:07
colomon 's last concert was Brian Miller and Randy Gosa just 10 days ago, and he's listening to their album right now. ;)
timotimo i've only been on one concert so far, that was immogen heap. i think i should probably go
21 euros is probably a regular price for a concert like that?
tadzik I usually pay about that much for concerts 15:08
timotimo now all i need is to find some people to accompany mi 15:09
jnthn tadzik: Arkona as in the Russian pagan metal band?
tadzik aside from the daily ones in Pirates' Nest
jnthn: yep
jnthn oh, cool. I'd not mind seeing them :)
tadzik it's absolutely amazing on concerts
jnthn can imagine
colomon believes most music is better live
tadzik although I think you have to put "folk" somewhere in the genre name to give them justice :) 15:10
jnthn Well, dream theater will be playing in Copenhagen sometime early next year, so I'm hoping to catch that :)
tadzik www.youtube.com/watch?v=Tt6_65LDZ60 for a glimpse
15:11 sqirrel left
TimToady I don't think shell/run should return a number with reversed boolean sense; I think it should return an object that is natively true/false but with methods to retrieve things like status if you want them 15:13
moritz +1
TimToady "natively" rather
primarily, I guess 15:14
jnthn yes, "natively" is a bit confusing there ;)
TimToady the current dualvarish approach gives me the willies
moritz "innatively" 15:15
timotimo great, because it seems like that's exactly what FROGGS just implemented :)
TimToady is also wondering about 'no strict' semantics; autodeclaring is not the same semantics as Perl 5, and we should have a discussion of whether 'no strict' really makes package vars PLUS allows lookup in the current package without OUR:: 15:16
that's what a P5 programmer would expect
15:17 odoacre_ left, odoacre left
diakopter TimToady: autodeclaring declares it in the first scope in which it appears or the outermost in that compilation unit, or the global a la JS 15:17
TimToady oh, btw, I don't think ::('$!x') should necessarily work, but maybe HAS::('$!x') would
15:17 odoacre joined
TimToady diakopter: yes, there are various autodecl schemes in various languages, but none of them map directly to what P5 does 15:18
15:19 rurban joined
diakopter TimToady: I'm asking, what do p5/p6 do 15:19
TimToady r: { our $foo; $foo = 42 }; say $foo; 15:20
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/0c2Prsw5fA␤Variable '$foo' is not declared␤at /tmp/0c2Prsw5fA:1␤------> { our $foo; $foo = 42 }; say $foo⏏;␤ expecting any of:␤ postfix␤»
..rakudo-parrot 55ff4f: OUTPUT«===SORRY!=== Error while compiling /tmp/zMilVJsNvU␤Variable '$foo' is not declared␤at /tmp/zMilVJsNvU:1␤------> { our $foo; $foo = 42 }; say $foo⏏;␤ expecting any of:␤ postfix␤»
TimToady this works in p5
p5eval: { our $foo; $foo = 42 }; say $foo;
p5eval TimToady: 421
moritz that's kinda scary
diakopter that's autodeclare? that's our...
TimToady you wanted "no strict" :)
diakopter p5eval: { $foo = 42 }; say $foo; 15:21
p5eval diakopter: 421
diakopter that's like JS
masak p5eval: { $foo = 42 }; $foo
p5eval masak: 42
TimToady r: { our $foo; $foo = 42 }; say $OUR::foo;
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«42␤»
moritz fwiw I'd vote for making p5's 'no strict;' autodecl lexical
TimToady P6 requires the OUR:: but P5 doesn't
15:21 risou_awy is now known as risou
moritz and putting the declaration in the outer-most scope of the current compilation unit 15:21
diakopter TimToady: I thought you were talking about a-declaration 15:22
moritz that's mostly analog to how settings work
TimToady moritz: I think that's silly; if you want autodeclare, that's something else, but we shouldn't pretend it's anything like P5
moritz p5eval: $x = 42; { use strict; $x }
p5eval moritz: ERROR: Global symbol "$x" requires explicit package name at (eval 7) line 1.
diakopter heh.
TimToady r: { our $foo; $foo = 42 }; say $MAIN::foo;
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«(Any)␤»
TimToady r: { our $foo; $foo = 42 }; say $Main::foo; 15:23
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«(Any)␤»
moritz I don't think there's a default package/module specced
TimToady er
moritz which is why Main:: doens't give you anything
TimToady r: { our $foo; $foo = 42 }; say $GLOBAL::foo;
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«42␤»
TimToady there we go
moritz ok
15:23 rurban left
diakopter r: package Boo { our $foo; $foo = 42 }; say $Boo::foo; 15:24
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«42␤»
TimToady is very opinionated this morning :)
diakopter r: package Boo { our $foo; $foo = 42 }; say $GLOBAL::foo;
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«(Any)␤»
timotimo FROGGS: is your code in a branch so i can try it out? or do i have to get the diff from the gist?
diakopter
.oO( .. and this is the opinionated chapter... )
 15:25
TimToady r: package Boo { our $foo; $foo = 42 }; package Boo { say $foo }; # would work under P5 semantics, which assume "OUR::", unlike P6
camelia rakudo-parrot 55ff4f: OUTPUT«===SORRY!=== Error while compiling /tmp/rL2SjnGivQ␤Variable '$foo' is not declared␤at /tmp/rL2SjnGivQ:1␤------> foo; $foo = 42 }; package Boo { say $foo⏏ }; # would work under P5 semantics, wh␤ expecting…» 15:26
..rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/DnbxLpjBKM␤Variable '$foo' is not declared␤at /tmp/DnbxLpjBKM:1␤------> foo; $foo = 42 }; package Boo { say $foo⏏ }; # would work under P5 semantics, wh␤ expecting an…»
TimToady r: package Boo { our $foo; $foo = 42 }; package Boo { say $OUR::foo }; # would work under P5 semantics, which assume "OUR::", unlike P6
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«42␤»
TimToady P5ers will carp unless we also assume OUR::
r: package Boo { $OUR::foo = 42 }; package Boo { say $OUR::foo }; # really more like this 15:27
camelia rakudo-parrot 55ff4f, rakudo-jvm 882e33: OUTPUT«42␤»
15:27 raiph left
moritz TimToady: so, no lexical alias, straigt lookup in the symbol table? 15:27
TimToady assumes OUR:: on both ends
unless already declared lexical, eys 15:28
*yes
moritz ss/symbol table/HOW/
TimToady WHY
WHO, WHATEVER
colomon WHEREFORE
moritz erm, WHO
yes
timotimo ah, there's a "run" branch 15:29
TimToady this interpretation of non-lexical variables has the, er, advantage of being largely orthogonal to lexical scoping, unlike the various autodeclare schemes, which interact in various weird ways 15:30
timotimo hm, i can't seem to --gen-nqp=run, i wonder why 15:31
TimToady thinks package globals are much more straightforward than audodecls
15:31 Vendethiel joined 15:32 raiph joined
moritz timotimo: if your last copy of moarvm is too old, you need a 'make realclean' in MoarVM first 15:33
15:33 Vendethiel left 15:34 Vendethiel joined
TimToady diakopter: sorry I wasn't very clear before my first cup of coffee--did you figure out what I was trying to say? :) 15:35
diakopter hm, I didn't perceive that you were answering my question, so no 15:36
15:37 risou is now known as risou_awy
diakopter TimToady: if today is an opinionated one, maybe it's a good day to add a billion github issues to perl6/spec 15:37
moritz s/eval/EVAL/ ! 15:38
diakopter or does opinionated not necessarily correlate with indefatigable?
timotimo moritz: sorry. that was in rakudo, not moarvm
moritz timotimo: oh, I somehow read --gen-moar 15:39
15:39 Vendethiel left
timotimo :) 15:39
moritz timotimo: can you please nopaste command + output?
TimToady Well, it's true that I'm not very opinionated when I'm asleep.
moritz is usually very opinionated which side of the bed to sleep on
timotimo well, i run perl Configure.pl --backends=parrot,jvm --gen-nqp=run and it checks out a different commit 15:40
diakopter moritz: I like to sleep on the top side of the bed
timotimo parrot,jvm rather
hm. now it works
moritz timotimo: does it check out the one from tools/build/NQP_REVISION?
timotimo perhaps because my nqp-p is now of the right revision
yes, it does
masak I don't think the specialness of &eval rises to the level of, say, BEGIN or MONKEY_PATCHING. if I have a vote, I vote for keeping it as lower-case. 15:42
moritz masak: then you don't have a vote :-)
masak heh.
15:42 telex left
TimToady BEGIN is uppercase because it's a phaser, and runs at a weird time 15:42
15:42 telex joined
TimToady eval doesn't really run at a weird time, only what's inside it :) 15:43
flussence
.oO( maybe it should be small-capitals )
dalek kudo/nom: c71db36 | moritz++ | Configure.pl:
[Configure.pl] try to fix --gen-nqp=branch
moritz timotimo: please pull and try again 15:44
uhm, not sure if that patch as a good one 15:45
let me think a bit more about that
diakopter stops the presses
timotimo i'm not sure how to reproduce the error case, tbh 15:48
it was one that also caused "wrong or missing dependency of ...QRegex.nqp"
TimToady well, even the insides of eval run at the normal time, they're just compiled at a weird time :) 15:51
dalek kudo/nom: 6f5ceb8 | moritz++ | / (3 files):
[Configure.pl] fix previus "fix"

do not discard the desired minimum version so early
moritz TimToady: so, everything inside the eval needs to be upper-cased? :-)
TimToady no, just the call to .PARSE :P 15:52
timotimo huh, shouldn't this give me some kind of output? 15:53
Compiling lib/File/Find.pm to pir
build stage failed for File::Find: Failed building lib/File/Find.pm
maybe panda just can't handle the Proc::Status objects properly? 15:54
tadzik possibly 15:55
I should do something about panda error reporting
it sometimes tries to make them too user-friendly and then they don't tell anything
timotimo yes, indeed
it has shell "foo bar" and fail "failed building" 15:56
tadzik I think I'd rather have it the other way around
yeah, that's a bit bad
timotimo :)
tadzik I'll have tuits soon, I'll give it some love
timotimo it used to work, so how do i fix this properly?
tadzik change it to 'or'? :D
now, that'd be shotgun coding
timotimo haha
yes, i've done that. i mean in the broader sense
i guess i'll wait for run to be merged into both masters and then push the "or" change upstream 15:57
i think i have a panda commit bit and i think i could fix all the usages of shell and run to accept the new Proc::Status objects
tadzik if you do, please tag the previous panda revision as 2010.11
erm, 2010.10
moritz we're not that far in the future yet :-)
tadzik gah! 2013.10 15:58
you know what I mean :P
timotimo :) 15:59
All tests successful. test stage failed for File::Find: Tests failed
:P
tadzik wat :D 16:00
then maybe it's not "or"
maybe File::Find is indeed broken
timotimo no, or is correct 16:01
it was "and" again, this time in the Tester instead of the Builder
tadzik ah
timotimo curiously the fetcher still has that and and didn't seem to b0rk 16:02
probably because it didn't have to git clone
and with the run changes, perl6-bench runs again :3 16:05
i'm happy
FROGGS++
jnthn FROGGS++
masak FROGGS++
timotimo now i'll verify that my changed components.json are correct to build all the thing 16:07
all the things
and then i have to fix another file to change how the things are run and it should be good to push :)
16:08 denis_boyun left 16:15 brrt left 16:18 ajr_ left 16:20 rurban joined 16:24 rurban left 16:31 darutoko joined 16:36 zakharyas left 16:41 spider-mario joined 16:42 jnap left 16:46 colomon left 16:52 jnap joined 16:55 xinming joined 16:58 raiph left 17:01 sqirrel joined, raiph joined 17:02 sqirrel left, sqirrel joined 17:13 ssutch left 17:18 [Sno] left 17:20 rurban joined 17:21 [Sno] joined 17:22 risou_awy is now known as risou 17:25 rurban left 17:27 ajr joined, ajr is now known as Guest29274 17:28 Guest29274 is now known as ajr_
masak a really nice blog post about the type model of JavaScripdt: vijayan.ca/blog/2012/02/21/javascri...ype-model/ 17:29
17:31 raiph left
masak ...and this is a kind of follow-on, in literal-programming code: github.com/stevekwan/experiments/b...ction.html 17:35
17:39 risou is now known as risou_awy
moritz I like the circular relation between Object and Function 17:40
masak yes, very MOP-y. 17:41
17:47 fhelmberger left 17:52 dansamo left
FROGGS timotimo: awesome! that is the best test I could imagine! 17:58
18:01 denis_boyun joined
dalek p: 54f6198 | (Tobias Leich)++ | src/vm/parrot/ (2 files):
implemented our own nqp::spawn()
 18:05
18:05 dakkar left 18:06 lizmat joined
lizmat at the AmsterdamX meeting, seeing Flavio Glock giving a presenation about Perlito 18:06
dalek kudo/nom: ea5c22f | (Tobias Leich)++ | src/core/ (2 files):
use nqp::spawn()
kudo/nom: 0eda3d3 | (Tobias Leich)++ | src/core/ (2 files):
Merge branch 'run' of github.com:rakudo/rakudo into nom
kudo/nom: c27c7db | (Tobias Leich)++ | tools/build/NQP_REVISION:
bump nqp revision (nqp::spawn)
18:07 xinming left
dalek ast: 3c43ace | (Tobias Leich)++ | packages/Test/Util.pm:
bring in line with rakudo's run()/shell() changes

This should still work on niecza and pugs.		 18:08
18:09 colomon joined, xinming joined 18:18 stevan__ left 18:21 rurban joined 18:23 pernatiy left 18:25 rurban left
TimToady masak: but it looks like he never wrote the third article... 18:27
"Function is the only given metatype in the system, but you can’t really do much with it. It sits there elegantly, being elegant." 18:28
FROGGS they should have called it Kitty instead 18:30
18:30 daxim left
FROGGS (or Minka) 18:31
18:35 stevan_ joined 18:36 stevan__ joined 18:40 stevan_ left 18:41 sqirrel left 18:49 lizmat left 18:57 SamuraiJack_ left, lizmat joined
colomon nothing quite as disturbing as discovering you had a big piece of the logic in your code EXACTLY BACKWARDS and the vast majority of the tests still passed… 18:59
timotimo colomon: like "shell foo bar and fail "failed the command!" 19:00
colomon timotimo: in this case I said "if there is no filename, write the file out with that name" 19:01
timotimo %) 19:02
colomon (with the implied "if there is a filename, don't write out the file") 19:03
19:03 darutoko left
timotimo oooooh, of course 19:03
FROGGS well, that saves disk space
timotimo yeah, i tend to not want to test things that touch the filesystem :|
FROGGS: files with no filename are super space-saving, eh? :) 19:04
colomon hard to delete if they actually get written. ;)
the sad thing is, I'm not very careful about cleaning up files generated by tests, so part of what was going on is the tests were trying to write a file and failing, and then reading a file of the same name written by something else. 19:05
there are probably all sorts of lessons to be learned from this.... 19:07
lizmat check the return value of say / print ? :-) 19:08
19:08 ssutch joined
colomon lizmat: well, not as such. It was my own file export routine that wasn't actually writing a file, and it returned SUCCESS even when it didn't. 19:10
19:11 nfg joined
lizmat but inside that export routine, was there a say/print that was executed and not checked ? 19:12
colomon nope 19:13
the code to actually write the file to the disk was in an if-then with the logic of the if reversed. So there never was an error of any sort.
as far as the code is concerned, I mean. obviously there was a huge error on my part. 19:14
Util #ps in 15m 19:15
19:15 nfg left 19:18 nofks joined 19:19 fhelmberger joined 19:22 rurban joined 19:23 risou_awy is now known as risou 19:26 fhelmberger left, rurban left
lue hell world o/ 19:27
*hello
FROGGS hi lue :o) 19:28
lizmat lue /o
lue ✔ ⑫ upload tarball to rakudo.org (thanks to moritz++) 19:29
colomon lue++
lizmat lue++ 19:31
moritz oh right, I was supposed to publlsh the annoucement on rakudo.org
lizmat AmsterdamX shutting down, socializing and decommiute&
19:31 lizmat left
lue aw shucks, you guys :) 19:35
moritz rakudo.org/2013/10/29/rakudo-star-2...-released/ 19:37
lue++
19:39 rindolf left 19:41 risou is now known as risou_awy
Util lue++ 19:43
jnthn lue++
lue (mails sent; just Wikipedia left (manual moderation of the parrot-lists announcement may be required))
19:48 nofks left
lue moritz: should it be my name or yours on the rakudo.org citation? 19:50
moritz lue: it should, but I have no idea how to make that happen with wordpress :( 19:51
lue moritz: I meant the wiki citation on the Rakudo page. I'll put my name there then :) 19:52
moritz lue: yes, do that
lue ✔ ⑬ Publicize release 19:53
dalek kudo/moar-support: a1b98b0 | jonathan++ | src/vm/moar/ (2 files):
Stub p6store.
 19:55
jnthn Hm. Seems we now successfully make it into the first BEGIN block of CORE.setting, and then attempt the trait_mod:<is> call that it makes, and fail inside of there :) 19:56
20:00 xenoterracide joined
dalek ar: bcd881c | lue++ | tools/star/release-guide.pod:
Add name to release guide [RG 14], some fixes.

Aside from my name, I decided to PODify the sublist in 13.		 20:00
lue ✔ ⑭ Update release guide with name.
✔ ⑮ Celebrate the end! \o/ (in progress; considered done) 20:01
<== STAR RELEASE MODE DE-ACTIVATED 20:02
moritz lue.eor
TimToady lue++ 20:04
(for finishing more things than I've finished lately :) 20:05
lue :) 20:06
moritz also, tools/build/create-moar-runner.pl doesn't exist
20:08 uvtc joined
uvtc Hi #perl6. Anyone have a rough idea on when Rakudo will be running on Moar? 20:09
diakopter uvtc: eh, before the end of the year I'd bet 20:10
moritz this year, mostl likely
uvtc Oooh. That's soon. Thanks.
Will it be included with an R* release?
diakopter uvtc: well, the real question is what in particular you're expecting/hoping from that rakudo... any particular features you're expecting? 20:11
moritz uvtc: that depends on a bit more than just rakudo, like having a native call interface
uvtc Just the basic language features, to try Perl 6 out again.
diakopter uvtc: it disappointed in the past? (if so, how in particular?) :) 20:12
uvtc I pop in every few months or so and see if any tutorials are around, and if I can figure out the difference between lists and parcels. :)
diakopter uvtc: I gave up on that long ago 20:13
uvtc Heh. Maybe early on I accidentally ascribed more importance to it than I should have. :)
diakopter I'm like, so what's the parcel sigil
jnthn argh, why does "make spectest" with Rakudo JVM re-build stuff? :( 20:14
TimToady parcels are just lazy tuples :)
lue might do the release thing again, assuming it doesn't take days next time :) 20:15
TimToady we just won't change anything between now and then, so it'll be easy 20:16
jnthn we...we won't? :P
dalek p: 292dfa6 | moritz++ | tools/build/gen-moar-runner.pl:
beautify tools/build/gen-moar-runner.pl

I do not like overdose of backslashes		 20:17
20:17 pernatiy joined
uvtc diakopter, I'm currently using Clojure for hobby projects, but am not a fan of (A) slow start up time, and (B) asking friends to install stuff from Oracle to run my programs. So, I'm not too terribly interested in Rakudo on JVM. So, waiting for Rakudo+Moar. (Though I recall Rakudo + Parrot having very fast startup time.) 20:18
TimToady, are Parcels not-yet-realized Lists? 20:19
20:22 rurban joined
moritz doc.perl6.org/type/Parcel tries to explain what a Parcel is 20:24
uvtc lue, I like your release-stage-numbering. ① , ②, ③, ... (makes me want to hum "final countdown"). :)
lue :) 20:25
masak uvtc: except this was a count-up... 20:27
20:27 rurban left
masak lue++ 20:28
uvtc masak, Yeah. Actually, seeing them reversed for the Christmas release would be like watching the New Years Eve ball drop. :) 20:32
moritz, thanks for the link. 20:33
lue r: my $a = "①"; for ^25 { say $a; $a.=succ } 20:34
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«①␤②␤③␤④␤⑤␤⑥␤⑦␤⑧␤⑨␤⑩␤⑪␤⑫␤⑬␤⑭␤⑮␤⑯␤⑰␤⑱␤⑲␤⑳␤①①␤①②␤①③␤①④␤①⑤␤»
lue r: my $a = "①"; for ^25 { say $a; $a=$a.ord.succ.chr } 20:35
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«①␤②␤③␤④␤⑤␤⑥␤⑦␤⑧␤⑨␤⑩␤⑪␤⑫␤⑬␤⑭␤⑮␤⑯␤⑰␤⑱␤⑲␤⑳␤⑴␤⑵␤⑶␤⑷␤⑸␤»
uvtc lue, How would you count down instead of up?
lue r: say "⑳".ord; say "㉑".ord
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«9331␤12881␤»
lue no wonder 21..25 are in a different font to me :P
r: my $countdown = "㉕"; for ^25 { say $countdown; $countdown=$countdown.ord.pred.chr; $countdown = chr(9331) if $countdown eq chr(12880);} 20:37
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«㉕␤㉔␤㉓␤㉒␤㉑␤⑳␤⑲␤⑱␤⑰␤⑯␤⑮␤⑭␤⑬␤⑫␤⑪␤⑩␤⑨␤⑧␤⑦␤⑥␤⑤␤④␤③␤②␤①␤»
uvtc Neat. :) `pred` == "predecessor"? (and `succ` == "successor"?) 20:38
lue yep. 20:39
uvtc Nice. If there's one thing Perl 6 wins hands-down at, IMO, it's naming. 20:40
lue there's even a ⓪!
r: say "⓪".ord 20:41
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«9450␤»
moritz say "⓪".ord.base(16)
p: say "⓪".ord.base(16)
camelia rakudo-parrot c27c7d: OUTPUT«24EA␤»
moritz p: say "⓪".ord.fmt: '%x'
camelia rakudo-parrot c27c7d: OUTPUT«24ea␤»
lue uvtc: also note that I had to do a bit of fiddling because Perl 6 knows to treat the "①" range as a vigesimal number system :)
moritz p: say "⓪".ord.fmt: '%X'
camelia rakudo-parrot c27c7d: OUTPUT«24EA␤»
20:42 stevan__ left
FROGGS p: say "⓪".ord.fmt: '%#X' 20:42
camelia rakudo-parrot c27c7d: OUTPUT«0X24EA␤»
lue uvtc: whoops, not vigesimal, but base 19 (⑳++ becomes ①①, not ①⓪) 20:43
uvtc Weird. Can't find that circled 0 anywhere... 20:44
flussence that's weird, ①..⑩ work for me, then 11..20 show as missing font boxes, then 21+ show as doublewidth chars...
lue flussence: look at this: 20:45
r: my $a = "①"; for ^25 { say $a; $a=$a.ord.succ.chr }
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«①␤②␤③␤④␤⑤␤⑥␤⑦␤⑧␤⑨␤⑩␤⑪␤⑫␤⑬␤⑭␤⑮␤⑯␤⑰␤⑱␤⑲␤⑳␤⑴␤⑵␤⑶␤⑷␤⑸␤»
20:45 kaare_ left
uvtc I see ① .. ⑳ as codepoints U+2460 .. U+2473. 20:45
Dunno where you're finding the others. :)
lue flussence: do you still see what you described?
flussence r: say ord '⑸'
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«9336␤»
geekosaur I get those and then I get (1) (2) ... except as chars
flussence it only shows up to 10 when you do that...
geekosaur (that is as single chars, not separate open-paren digit close-paren) 20:46
lue flussence: then your fonts aren't giving you 11..20 :(
flussence hm, I have fonts for the others but not in monospace :(
lue geekosaur: those last five at the end are the next codepoints in the sequence. ㉑..㊿ are in another part of Unicode all-together 20:47
r: my $a = "㉑"; for ^50 { say $a; $a=$a.ord.succ.chr } 20:48
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«㉑␤㉒␤㉓␤㉔␤㉕␤㉖␤㉗␤㉘␤㉙␤㉚␤㉛␤㉜␤㉝␤㉞␤㉟␤㉠␤㉡␤㉢␤㉣␤㉤␤㉥␤㉦␤㉧␤㉨␤㉩␤㉪␤㉫␤㉬␤㉭␤㉮␤㉯␤㉰␤㉱␤㉲␤㉳␤㉴␤㉵␤㉶␤㉷␤…»
flussence those all render fine using multiple fonts, which is even weirder :/
lue sorry, ㉑..㉟ are in another part. Where's 36? 20:49
masak ㉟ ought to be enough for anyone.
lue r: say "㉟".ord.fmt("%#X"); say "㊱".ord.fmt("%#X")
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«0X325F␤0X32B1␤»
lue r: my $a = "㊱"; for ^30 { say $a; $a=$a.ord.succ.chr } 20:50
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«㊱␤㊲␤㊳␤㊴␤㊵␤㊶␤㊷␤㊸␤㊹␤㊺␤㊻␤㊼␤㊽␤㊾␤㊿␤㋀␤㋁␤㋂␤㋃␤㋄␤㋅␤㋆␤㋇␤㋈␤㋉␤㋊␤㋋␤㋌␤㋍␤㋎␤»
lue (there probably is a (51), but (50) is where my XCompose draws the line)
geekosaur mrrr. that is not what my font shows... 20:51
flussence I don't see a 51 nearby but I wouldn't be surprised if it went up to 99 20:52
lue I'm guessing it'll go up to 99 in a future revision of Unicode :)
geekosaur: what's your font not showing?
geekosaur www.dropbox.com/s/79g42wdhx1mop1c/....52.31.png
lue r: say "\c[FACE SAVOURING DELICIOUS FOOD]" # :D 20:53
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/f7s5CsWKG4␤Unrecognized character name FACE SAVOURING DELICIOUS FOOD␤at /tmp/f7s5CsWKG4:1␤------> say "\c[FACE SAVOURING DELICIOUS FOOD⏏]" # :D␤»
..rakudo-parrot c27c7d: OUTPUT«😋␤»
hoelzro fglock++ # perlito talk at AmsterdamX
I was going over the notes on the MOP from the NQP internals talk; are the MOP objects meant to be exposed to the Perl 6 runtime so that users can create their own MOP objects that implement interesting behaviors? 20:54
lue geekosaur: that first series of boxes would be 36..50. The three at the end are scientific symbols of some sort (in order, "Hg", "erg", "eV")
hoelzro or is that something that will stay in NQP land?
moritz hoelzro: there is a rakudo MOP and an NQP MOP. The Rakudo MOP is certainly exposed to the user 20:55
hoelzro ok 20:56
I'm just trying to figure out how to create a custom metaobject that does something special for type checking
jnthn There are already some Perl 6 modules that do custom MOP stuff.
hoelzro adds TODO list item: "Read more Perl 6 code" 20:57
jnthn hoelzro: type_check and accepts_type
hoelzro right, I just don't know how to "register" the object with the runtime
20:57 stevan_ joined
hoelzro I suppose that it's probably simpler than that? 20:57
20:57 xinming left
masak today's autopun: twitter.com/quietfanatic/status/39...2227909632 20:58
hoelzro I should just create a regular class, and create types with it?
20:59 xinming joined 21:00 felipe left 21:01 mtj_- joined
lue masak: I still would really like it if autopun was defined as a "play on situations", just so the defintion of autopun was a pun on the definition of pun :P 21:02
.oO(And would that be an autopun?)
 21:03
21:03 [Sno]_ joined
moritz think it's an autounpun 21:03
uvtc What's an autopun? Closest I get is wikipedia and an autopen. :) 21:05
masak uvtc: masak.org/carl/gpw-2012-autopuns/talk.pdf
21:05 raiph joined
jnthn An autopen is what you write an autopun with. 21:05
masak :P 21:06
lue: "play on situations" isn't such a bad first approximation. but autopuns definitely have a language aspect.
21:06 bluescreen100 joined
moritz recalls that presentation with pleasure 21:06
masak recalls the plesentation with pressure 21:07
21:07 bluescreen__ joined
masak pleasantation*, d'oh 21:07
21:07 ssutch left, [Sno] left, bluescreen10 left, benabik left, [Sno]_ is now known as [Sno]
diakopter repleasantation 21:07
masak with my luck, my one lasting legacy will be autopuns. 21:09
jnthn :D 21:10
21:10 bluescreen100 left
uvtc masak: that was a great talk, where did my autopen go, refrigerator. 21:11
:)
moritz masak: wouldn't be a bad legacy 21:12
compared to, say, Hitler :-)
diakopter replies to masak's non-pun
with triple-absurdity
21:14 ssutch joined, DarthGandalf joined, robinsmidsrod joined, ggoebel joined, tadzik joined, BooK joined, cxreg joined, szbalint joined, ingy joined
lue I tried to describe that as an autopun, but I was wrong. 21:14
masak moritz: I would never allow my legacy to be Hitler.
21:15 jnap left
lue
.oO(Poor Hilter, he never could get elected to British government.)
 21:16
jnthn masak: I really can nazi that happening...
21:16 jnap joined
lue masak: Your legacy is pretty good at winning elections though :/ 21:18
masak lue: autopuns are winning elections?
lue: "Winston Churchill" is not an autopun :) 21:19
lue General Betray-us^WPetrayus is, though :)
diakopter no, that's just a pun 21:20
masak what diakopter said :)
lue so would President Robert Trustmi be an autopun during his congressional trial? 21:21
diakopter it has to be an utterance 21:22
almost certainly intended
because being able to laugh with the writer/speaker is pretty necessary among friends
unless you like ridicule
"HAHAHA YOU FAIL, Failure. way to autopun!" 21:23
21:23 rurban joined
diakopter
.oO( "..but it was an intentional Freudian slip, not an autopun.." )
 21:24
21:24 risou_awy is now known as risou, BenGoldberg joined
lue how 'bout automatic autopun, or autoautopun, or diautopun then? :) 21:24
diakopter lue: I'll send you a dollar if you can make one of those 21:25
masak my current definition of autopun is something like "bringing use and mention together deliberately for purposes of humor".
uvtc Off for dinner. Great to see all the Moar activity. Looking forward to trying out Ramoardo, Moarkudo, Rakumoar, ... whatever it's called. ;)
lue <lue> so would President Robert Trustmi be an autopun during his congressional trial? (Description of theroretical diautopun)
masak but note that the "bringing use and mention together" part is pretty vital.
lue masak: autoselfdemo then? 21:26
21:26 uvtc left
lue (but autopun is a cooler name) 21:26
masak yeah, "autoselfdemo" doesn't quite have that ring to it. 21:27
lue: <patient> Doctor, people have completely stopped paying attention to me... -- <doctor> Next!
diakopter anagrams for rakudo-moar:
21:27 rurban left
diakopter u karma odor 21:27
masak :) 21:28
lue wouldn't be surprised if masak means an alternate definition of "auto" and a archaic definition of "pun" for the resultant word
door u, karma!
diakopter Our Am A Dork 21:29
dalek kudo/moar-support: 95476c2 | moritz++ | tools/build/ (2 files):
import and adapot gen-moar-runner.pl from nqp

pretty pointless so far, because we do not get that far in the compilation.
lue masak: <patient> Doctor, people have completely stopped paying attention to me... -- <doctor> don't worry, we'll go somewhere nicer as soon as I stop the daleks. 21:30
diakopter boggles at what "adapot" could mean
oh, the preceding import was clouding my imagination
moritz can't write straight commit messages anymore? 21:31
time for sleep&
lue still wants new yoleaux to be named cyberman for consistency
♞ moritz o/
diakopter Ok Maraudor
masak 'night, moritz. dream of specially imported adapots. 21:32
diakopter Ur A Dark Moo
masak .oO( "addapot!" is what you affectionately say to your pot after it fetches that stick )
lue adapot is where you put your Ada code :P
masak well, remember that Ada was a gambler. 21:33
so, an Ada-pot would be...
21:33 colomon left
diakopter moving on to rakudo-on-moar 21:34
Our Random Oak
Am A Round Rook 21:35
lue
.oO("GNU Day Casino & Spa! Stunning views of the C++! Visit our museum of priceless Rubys and Perls! Compete in many of our Adapots!")
lue &
masak lue++
diakopter Rank Odor Moar 21:37
lue masak: what for? 21:38
diakopter jnthn: you'll like this one.. 21:39
Do Mu Or Anorak 21:40
21:40 risou is now known as risou_awy
jnthn diakopter: It stinks :P 21:40
21:40 ajr_ left
jnthn does the anorak 21:41
diakopter omfg
jnthn :P 21:42
diakopter Dr No Am Our Oak
jnthn It's useful for keeping the rain off :P
diakopter jnthn: oh; I thought there was slang for obsessed on esoterica 21:43
masak lue: that thought bubble :)
diakopter: "anorak" is slang for "obsessed on esoterica"? 21:44
diakopter this online dictionary might have been lying to me o_O
masak I'm not doubting you, I just wanted to be sure :)
segomos i'm doubtful 21:45
masak *Informal* a socially inept person with a hobby considered by most people to be boring
well, I'll be.
segomos re: amateur etymologist
diakopter :P
segomos :-) 21:46
masak touché.
masak goes to hide under his big water-proof coat
segomos i guess that isn't really etymology, though
diakopter hey, I'm not amateur, I just cargo cult phrases like cargo cult
segomos haha
diakopter they say cargo culting is the highest form of flattening 21:47
masak I heard it was the highest form of flappery. 21:48
diakopter well, frippery, at least 21:49
segomos cargo cult programming wikipedia page reads like 99% of webscale javascript programmers
masak .oO( cargo culting is the most sincere form of Richard P. Feynman )
timotimo oh hey 21:52
i totally forgot that i ran some benchmarks
so there you go: t.h8.lv/p6bench/all_backends.html
huh, seems like no parrot benchmarks were run, that's odd 21:53
masak 'night, #perl6
21:53 Mouq joined
diakopter what kind of crockery is this!? 21:53
FROGGS .oO( What kind of sauce is this? ) 21:54
diakopter saucier?
Mouq moritz: did you mean to commit your work with undeclared variables in 6f5ceb? »[Configure.pl] fix previus "fix" «
(rakudo)
21:56 ssutch left
FROGGS ohh, that does not seem right 21:57
21:59 felipe joined
Mouq r: 1 22:04
camelia ( no output )
Mouq r: $abcd = 1234
camelia rakudo-jvm 882e33: OUTPUT«Cannot modify an immutable value␤ in block at /tmp/zGytaL9fVL: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:1196␤ in any command_eval at…»
..rakudo-parrot c27c7d: OUTPUT«Cannot modify an immutable value␤ in block at /tmp/y6bnnfUhe1:1␤ in any at /tmp/y6bnnfUhe1:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/stage2/NQPHLL.nqp:1133␤ in any evalfiles at gen/parrot/stage2/NQPHLL.nq…»
Mouq It doesn't even look like we keep track of whether we're in a strict setting yet. Would that just be adding a $*NOSTRICT var or some such? 22:07
std: v6; $a = 1; say $a
camelia std 8adbc60: OUTPUT«===SORRY!===␤Variable $a is not predeclared at /tmp/biureBRib7 line 1:␤------> v6; ⏏$a = 1; say $a␤Variable $a is not predeclared at /tmp/biureBRib7 line 1:␤------> v6; $a = 1; say ⏏$a␤Check fa…»
22:07 donaldh left
FROGGS Mouq: there is no switch implemented 22:17
TimToady STD doesn't attempt 'no strict'
FROGGS (yet)
22:19 colomon joined 22:24 rurban joined
TimToady I'm thinking we should probably break the idea that ① .. ⑳ is a cycle, and recognize that some ranges are open-ended and Unicode might add more of them someday 22:24
especially since that's what happened in this case
timotimo t.h8.lv/p6bench/all_backends.html - now also has some parrot lines ... not sure about some of them ... something might have gone wrong
BenGoldberg r: say "adapot".comb.sort.join 22:25
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«aadopt␤»
timotimo at least it does hello and zero very, very fast
BenGoldberg r: my $a ($b) = 1; say $b 22:27
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«(Mu)␤»
BenGoldberg r: my $a ($b) = 1; say $b; say $c
camelia rakudo-jvm 882e33: OUTPUT«(Mu)␤No such method 'gist' for invocant of type 'NQPMu'␤ in block at /tmp/Xee5OuXavy: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:1196…»
..rakudo-parrot c27c7d: OUTPUT«(Mu)␤No such method 'gist' for invocant of type 'NQPMu'␤ in sub say at gen/parrot/CORE.setting:12856␤ in block at /tmp/olu_wHe17F:1␤ in any at /tmp/olu_wHe17F:1␤ in any at gen/parrot/stage2/NQPHLL.nqp:1146␤ in any eval at gen/parrot/s…»
BenGoldberg r: my $a ($b) = 1; say $b; ++$b 22:28
camelia rakudo-jvm 882e33: OUTPUT«(Mu)␤java.lang.NullPointerException␤ in block at /tmp/xLO1SZz_eJ: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:1196␤ in any command_ev…»
..rakudo-parrot c27c7d: OUTPUT«(Mu)␤Cannot modify an immutable value␤ in sub prefix:<++> at gen/parrot/CORE.setting:1733␤ in sub prefix:<++> at gen/parrot/CORE.setting:1731␤ in block at /tmp/HspbquR4Dn:1␤ in any at /tmp/HspbquR4Dn:1␤ in any at gen/parrot/stage2/NQP…»
Mouq r: my $r = rx/\w+(\d)\w*/; say "abcd89asdf" ~~ /<$r>/ 22:29
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«｢abcd89asdf｣␤␤»
Mouq n: my $r = rx/\w+(\d)\w*/; say "abcd89asdf" ~~ /<$r>/
camelia niecza v24-98-g473bd20: OUTPUT«｢abcd89asdf｣␤␤»
BenGoldberg Surely both rakudo should say "Cannot modify an immutale value" when I do that?
22:29 rurban left
BenGoldberg r: my $a ($b) = 1; say $b; $b = 1; 22:29
lue TimToady: I'm not sure ⑳ leads to ㉑... let me check.
camelia rakudo-jvm 882e33: OUTPUT«(Mu)␤java.lang.NullPointerException␤ in block at /tmp/5UW1qT75Sa: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:1196␤ in any command_ev…»
..rakudo-parrot c27c7d: OUTPUT«(signal SEGV)(Mu)␤»
lue ⑳: U+2473 CIRCLED NUMBER TWENTY 22:30
㉑: U+3251 CIRCLED NUMBER TWENTY ONE
㉟: U+325F CIRCLED NUMBER THIRTY FIVE 22:31
㊱: U+32B1 CIRCLED NUMBER THIRTY SIX (turns out the pockets of circled numbers do connect)
TimToady well, the question is whether .succ should do that for you
lue does unicode define .succ-like things? 22:32
TimToady doesn't recall seeing any such
but there's a lot of stuff there I've never gotten around to reading
22:33 denis_boyun left
TimToady obviously chaining 20 to 21 would be much more useful to the programmer, whether or not Unicode supports it directly 22:33
Mouq TimToady: $r = rx/…/; Should /<$r>/ retain grouping information? There's no way to access the '(\d+)'
BenGoldberg As far as I can recall, unicode defines groups of things
lue r: say "①".pred; # ideally should be ⓪
camelia rakudo-jvm 882e33: OUTPUT«Decrement out of range␤ in block at /tmp/B7aOymziA5:1␤ 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:1196␤ in any command_eval at src/Perl6…»
..rakudo-parrot c27c7d: OUTPUT«Decrement out of range␤ in method gist at gen/parrot/CORE.setting:11956␤ in method gist at gen/parrot/CORE.setting:1014␤ in sub say at gen/parrot/CORE.setting:12856␤ in block at /tmp/CqqERascT1:1␤ in any at /tmp/CqqERascT1:1␤ in any …»
Mouq Actually, why is that matching the whole thing? I only put (\d) 22:34
BenGoldberg According to wikipedia:
The standard consists of a set of code charts for visual reference, an encoding method and set of standard character encodings, a set of reference data computer files, and a number of related items, such as character properties, rules for normalization, decomposition, collation, rendering, and bidirectional display order
TimToady the subrule matches the whole thing
Mouq r: my $r = rx/<-[\d]>+(\d+)/; say "abcd89asdf" ~~ /<$r>/ 22:35
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«｢abcd89｣␤␤»
Mouq n: my $r = rx/<-[\d]>+(\d+)/; say "abcd89asdf" ~~ /<$r>/
camelia niecza v24-98-g473bd20: OUTPUT«｢abcd89｣␤␤»
TimToady r: my $r = rx/\w+<(\d)>\w*/; say "abcd89asdf" ~~ /<$r>/
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«｢abcd89asdf｣␤␤»
Mouq I didn't realize \w matched 0..9.
TimToady and _ 22:36
lue TimToady: supporting non-western-arabic numerals is weird. Perhaps we should refuse succ/pred on things that can't be literal digits in P6 :)
TimToady r: my $r = rx/\w+(\d)\w*/; say "abcd89asdf" ~~ /<sub=$r>/
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«｢abcd89asdf｣␤ sub => ｢abcd89asdf｣␤ 0 => ｢9｣␤␤»
Mouq :D 22:37
TimToady lue: perhaps we should allow them as literal digits :)
TimToady is always more interesting in telling people what they can do than what they can't
lue base-51 number system GO!
22:37 spider-mario left
Mouq Still, TimToady, is that correct behavior? 22:38
TimToady Unicode distinguishes numerals that can function in base-10 from numerals that can't
is what correct?
lue r: say $_ for ⓪..^①⓪ # equivalent to ⓪..㊿
camelia rakudo-jvm 882e33: OUTPUT«===SORRY!=== Error while compiling /tmp/m4SMH3SCEZ␤Confused␤at /tmp/m4SMH3SCEZ:1␤------> say $_ for ⏏⓪..^①⓪ # equivalent to ⓪..㊿␤ expecting any of:␤ postfix␤ statement end␤ …»
..rakudo-parrot c27c7d: OUTPUT«===SORRY!=== Error while compiling /tmp/SjUSDBa_wd␤Confused␤at /tmp/SjUSDBa_wd:1␤------> say $_ for ⏏⓪..^①⓪ # equivalent to ⓪..㊿␤ expecting any of:␤ postfix␤ statement end␤ …»
22:39 raiph left
Mouq Not capturing with just <$r> 22:40
TimToady yes, that's absolutely correct, relatively speaking
lue How's about "You can use any literal numeral system in Perl 6. The internal storage format (and default output format) is base-10 'arabic numerals'. Number ranges with no clear bound (e.g. the circled numbers ①, ②, etc.) are not recommended for general use, as Unicode may change the largest single-digit value at any time." for starters? 22:41
Mouq Huh. 'k
lue gears up the gist machine
timotimo is no longer sure why he even started up the benchmarks 22:42
but it seems to work all right
TimToady S05:1412 explains that leading alphabetic inside <> is what enables capturing; the <foo=...> is just a variant on that, but still starts alphabetic 22:45
synopsebot Link: perlcabal.org/syn/S05.html#line_1412
22:47 xenoterracide left
Mouq OK 22:47
22:47 BenGoldberg left
TimToady in fact, <foo> and <.foo> are really the same thing, since the first is also a method call; the only purpose of the '.' is to turn of capture explicitly 22:48
*turn off
22:49 BenGoldberg joined 22:52 FROGGS left
dalek rl6-bench: e53541a | (Timo Paulssen)++ | nqp/richards:
added an implementation of the richards benchmark.
 22:55
rl6-bench: 689120e | (Timo Paulssen)++ | compilers.pl:
--omg-optimize
rl6-bench: 59fc0f8 | (Timo Paulssen)++ | bench:
replace deprecated use of Hash.delete
rl6-bench: d94d6e8 | (Timo Paulssen)++ | comp (2 files):
chase recent nqp&rakudo build tool improvements
timotimo i has makes a push!
jnthn timotimo++ 22:57
timotimo oh, i can tag panda 2013.10 and then push some more changes :) 22:59
23:00 panchiniak joined
dalek nda: 9f2591b | (Timo Paulssen)++ | lib/Panda/ (3 files):
shell now returns a Proc::Status object that bools.
 23:00
23:05 FROGGS joined 23:16 raiph joined
dalek ecs: 521e009 | larry++ | S03-operators.pod:
no cycles on rangechars that Unicode might add to
 23:21
[Coke] lue++ #star 23:22
23:24 xenoterracide joined 23:25 risou_awy is now known as risou
timotimo lue++ 23:25
i hope we will find the easybto fix bugs before the next release rather than directly after it... 23:26
[Coke] rakudo.jvm daily test seems to be getting slower. 23:28
(it's 7:30, and it's only on S02)
23:29 colomon left
[Coke] ISTR it finishing by 530ish in the past. 23:30
23:30 btyler left
timotimo hmm anything pathological? 23:30
23:31 FROGGS left
[Coke] rant: having the entire classpath in the executable makes for tough htop using. 23:31
timotimo hehe 23:32
i can imagine
how can we do better?
[Coke] Yes, something pathological. Looks like 5 hours on one test file.
S02-lexical-conventions/comments.rakudo.jvm
timotimo is there a take classpath from file flag?
[Coke] 5.5h, even. 23:33
timotimo that is bad in deed
I'll be at my computer in a quarter hour
[Coke] timotimo: there's %CLASSPATH%/$CLASSPATH.
timotimo that sounds like an option 23:34
[Coke] kills it by hand.
timotimo you run only one test in parallel?
[Coke] yes. 23:35
I'm a guest on the box, and memory is always a concern when running/building java stuff for rakudo
timotimo that explains
[Coke] though diakopter will no doubt tell me to go ahead 23:36
diakopter go ahead
[Coke] (it also helps with day to day comparisions of what changed)
doesn't help that the eval server was crapping out so I'm running the test suite without it.
diakopter I ugpraded jdk since then I think
23:41 risou is now known as risou_awy 23:42 jnap1 joined, pdurbin joined
TimToady lue++ for breaking the cycle conjecture on rangechars :) 23:44
23:44 jnap left
lue just the no-upper-limit series. Indic numerals e.g. should work as expected. 23:44
23:45 bluescreen100 joined
lue r: say "Ⅻ".succ 23:45
camelia rakudo-parrot c27c7d, rakudo-jvm 882e33: OUTPUT«ⅠⅠ␤»
lue (Also undo it for non-positionals, like roman numerals, please.) 23:46
and finally, a :wholecharset adverb (name can change clearly) for .succ/.pred on Strs, so that way I can forcibly go to the next codepoint.
TimToady adverbs are sometimes a design smell 23:47
github seems...busy... 23:48
23:48 pochi joined 23:49 xinming left
timotimo it's complicated to fudgeandrun with a different perl6 than your first backend 23:49
23:49 bluescreen__ left
diakopter nqp: say(nqp::chr(nqp::ordfirst("Ⅻ")+1)) 23:49
camelia nqp-moarvm, nqp-jvm, nqp-parrot: OUTPUT«Ⅼ␤»
diakopter yay, I think
23:50 arnsholt_ joined, baest_ joined
TimToady yes, and that's ... semantics, but not .. semantics 23:50
.. translation to ... implies *.succ
timotimo pff. of course i can run that file that blew up for coke just fine locally
lue hey look: gist.github.com/lue/7224781
TimToady I've got a git push clarifying that, but it's hung up
lue (based on earlier talks of the ranges)
23:51 [Sno]_ joined, xinming joined
lue Sometimes you just want a .succ/.pred on a Str that doesn't char about being (graphically|semantically|linguistically) correct. 23:52
eh. *care 23:53
23:53 JimmyZ_ joined
lue accidental pun :) 23:53
23:54 amkrankr1leuen joined, Woodi_ joined
dalek ecs: 065a718 | larry++ | S03-operators.pod:
clarify that range->seq translation implies *.succ
 23:54
TimToady there we go
you'll note that roman numerals are already considered non-carrying 23:55
lue The best thing for context-sensitive .succ/.pred would be for "㊿".succ eq "①" or "⓪", IMO (just to avoid redefinitions of ①⓪/①① on Unicode's part.) 23:56
timotimo in the new makefile system, how do i get a spectest with the evalserver?
23:57 eternaleye_ joined
TimToady that's just ('⓪' .. *) xx * 23:57
23:58 raiph left, [Sno] left, pochi_ left, Vlavv` left, amkrankruleuen left, [Sno]_ is now known as [Sno], eternaleye_ is now known as eternaleye
TimToady no need to build cycles in 23:58
lue boggles at the idea of a NOT infinite-elem ..* list :)
23:59 dayangkun joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!