»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend!
Set by sorear on 4 February 2011.
00:00 risou_awy is now known as risou 00:02 Chillance left
diakopter colomon: teehee sexpexted 00:03
00:07 risou is now known as risou_awy 00:09 ispeak joined 00:11 replore joined
Radvendii nom: my @a=1,2,3; for @a {$_++} 00:21
p6eval nom 1eadbb: ( no output )
Radvendii b: my @a=1,2,3; for @a {$_++}
p6eval b 1b7dd1: ( no output ) 00:22
araujo wonders what it is the status of the current available perl6 implementations
Radvendii b: my @a=1,2,3; for @a {say ++$_}
p6eval b 1b7dd1: OUTPUT«2␤3␤4␤»
Radvendii nom: my @a=1,2,3; for @a {say ++$_}
p6eval nom 1eadbb: OUTPUT«2␤3␤4␤»
Radvendii nom: my @a=rand(),rand(),rand(); for @a {say ++$_} 00:23
p6eval nom 1eadbb: OUTPUT«1.26594015590763␤1.88294592625195␤1.3494678012013␤»
diakopter nom: my @a=1,2,3; for @a {++$_}; say @a;
p6eval nom 1eadbb: OUTPUT«2 3 4␤»
diakopter wheh 00:24
I must have a profound misassumption about how for @a works
$_ points to the slot?! 00:25
Radvendii $_ points to the element of the array
what is the speed of perl compared to C?
does mutable arrays slow it down significantly? 00:26
diakopter $_ pointing to the slot is ... very strange. 00:27
how have I missed this up until now
Radvendii why?
diakopter nom: my @a=1,2,3; for @a {$_ = "a"}; say @a; 00:29
p6eval nom 1eadbb: OUTPUT«a a a␤»
diakopter that is very unexpected to me. I'm sorry.
geekosaur thought that was expected
Radvendii uh... yeah
what did you expect? 00:30
geekosaur (and it is certainly very Perl)
diakopter I thought $_ pointed to the object, the array element, not the container that holds the object.
Radvendii what?
diakopter what what?
like I said, I'm sorry for expecting the wrong thing. 00:31
Radvendii no no, it's fine, i just dont understand what you *were* expecting 00:32
diakopter I was expecting it to behave like every other iterator in every other language I've used. 00:33
geekosaur this is longstanding perl behavior, back to at least sometime in the 3.x series; iterators actually allow you to mutate the thing iterated over 00:34
00:36 ispeak left
diakopter somehow I have missed this. 00:37
00:40 dorlamm joined 00:50 snearch left
Juerd What's mono doing on feather, besides causing high loads? 00:53
colomon running niecza?
00:53 xaka joined
Radvendii is perl6 slower than perl5? 00:54
Juerd [Coke]: Could you nice(1) the mono processes, please, and perhaps figure out something to moderate their memory usage?
Radvendii in general?
Juerd Radvendii: Perl 6 is a language, not a program, so that question can't really be answered.
diakopter yes, in general the implementations are much slower. 00:55
Juerd Currently, implementations of Perl 6 are typically slower than Perl 5, but this is expected to change.
Radvendii oh i see, it's a matter of the implementations, rather than the specifications
that makes sense
diakopter who expects the implementations to improve dramatically? I don't know of anyone.
Juerd diakopter: You don't know me? :( 00:56
Radvendii: Of course, if you meant to refer to the languages, then I'd say Perl 6 is much faster because it has so many useful features built-in that Perl 5 doesn't, many of which will save the programmer a lot of time.
Radvendii i'm not talking about programmer time, i meant compile/run time 00:57
diakopter obviously :)
Juerd diakopter: I seriously expect nothing less than dramatic performance improvements from several implementations.
Perhaps not this year, but in 2012 and subsequent years, I'm convinced it'll be sped up by staggering ratios. 00:58
diakopter afaik, sorear doesn't expect dramatic improvements from niecza unless it's ported/rewritten in another environment, and same for jnthn/pmichaud.
Juerd It'll happen, one way or another, and not just because computers themselves are getting faster :) 01:00
01:00 slavik2 left
diakopter unless I was working on it myself, I wouldn't have such an expectation. 01:01
unless of course someone actually said it.
18:32 < Radvendii> no no, it's fine, i just dont understand what you *were* expecting
ack mispaste
Juerd I don't know exactly how and when, but I know why: it's necessary to have something fast, or a lot of work will have been a huge waste of energy...
diakopter of course that's the desire, but there is a vast dearth of resources available to work on it. 01:02
and apparently always will be 01:03
Juerd Just means that the work will either take longer, or die off prematurely. 01:04
Radvendii is larry involved in the coding much? 01:05
diakopter intentionally not, except for the standard grammar
Juerd I'm off to bed. Good night :) 01:06
01:06 raiph joined
Radvendii good night 01:06
raiph \o radvendii
Radvendii \o?
diakopter raiph is waving goodbye to you
not understanding that you were talking to Juerd 01:07
raiph diakopter: i believe jnthn expects rakudo to get way faster
Radvendii oh
raiph oops, goodnight juerd, hello radvendii
Radvendii haha 01:08
01:08 donri left
Radvendii what was C written in? 01:08
raiph c
diakopter <eye roll> 01:09
01:09 franek joined
Radvendii seriously though... 01:09
flussence B
en.wikipedia.org/wiki/BCPL 01:10
Radvendii oh... thats right! didn't B never actually come out because C came out first?
01:13 lestrrat left, dorlamm left
raiph diakopter: sorry. wikipedia says "assembly language". 01:14
diakopter cm.bell-labs.com/who/dmr/chist.html
01:14 lestrrat joined, Radvendii is now known as Radvendii_BRB
raiph wikipedia claims b did come out. i know bcpl did, because i spent 3 years writing in it. 01:15
01:15 packetknife joined
raiph 1985-87. no types, and only assembly level debugger. shudder. 01:16
01:16 colomon left
geekosaur ...and now you also know the early history of C. and why in K&R C types were largely optional (defaulting to int) 01:17
01:18 fieldsa joined
fieldsa I seek run() shell cmd in bg w/o wait, is a way from qx// directly? I attempt run("$cmd &"), but it waits on child. 01:22
I do not need to capture or send input via socketpair. I also notice there is no fork? 01:23
01:24 MindosCheng left, Radvendii_BRB left
sorear Juerd: was niecza causing a problem? 01:26
01:26 colomon joined
diakopter Juerd went to bed 01:26
01:27 xjiujiu joined
diakopter colomon: I was giggling at your sexpexted 01:27
sorear o/ colomon
colomon o/
sorear diakopter: that was mine
colomon diakopter: it was sorear's actually
diakopter what's a sexpex 01:28
colomon I was giggling at it too, actually. :)
sorear stringified_expected
diakopter oh, you typoed x for c then :P
sorear really is() should be using eqv
colomon that typo might have been me
sorear: I take it you're working on getting things like \c[ARABIC NUMBER SIGN] working? 01:30
sorear colomon: yes
colomon is drooling at the prospect of getting all those S05-mass tests running 01:31
sorear fetches a towel
01:32 xjiujiu left 01:33 xjiujiu joined
fieldsa pugs apparently has &fork(), but rakudo doesn't? 01:35
I don't want to rush them along, but I would consider fork() fairly good feature to have, perhaps it is in the pipeline, so to speak? 01:36
sorear eh. rakudo is rather chained by portability
also, we don't say "them" here
01:37 shinobicl_ joined
sorear I'm suprised run("$cmd &") doesn't work for you. that's what I would have tried 01:37
01:39 envi_ joined, franek left
dalek ast: 9fa9559 | (Solomon Foster)++ | S02-names-vars/names.t:
Fudge for niecza.
 01:41
sorear rakudo: say "\c[NO SUCH CHARACTER]" 01:42
p6eval rakudo 1eadbb: OUTPUT«===SORRY!===␤Unrecognized character name NO SUCH CHARACTER at line 1, near "]\""␤»
01:44 xjiujiu left
fieldsa sorear: running strace, seems there is a wait4() - perhaps has to do w/ my config 01:45
dalek ecza: d9f1c7b | sorear++ | lib/ (2 files):
Add name->character inverse mapping support
01:53 alvis left
dalek ecza: 8c76823 | (Solomon Foster)++ | t/spectest.data:
Turn on S02-names-vars/names.t.
 01:54
01:55 alvis joined
sorear I guess Phase 2 can be started now. :> 01:57
colomon btw, I find it interesting how many times niecza and rakudo are both fudging the same tests
01:57 alc joined
fieldsa run("$cmd &") gets stopped on waitpid(-1,) even though the child is already fork(), certainly I'll investigate further 01:58
01:58 raiph left
colomon sorear: "Action method charname not yet implemented"? 01:58
sorear colomon: that's part of phase 2 01:59
colomon ah
sorear so far I've been working at the Niecza::UCD level
now I need to 1. add unidata to the repository/build system 2. link stuff into the compiler 02:00
sorear wonders how best to note the origins of the 400kb binary file ey is about to check in 02:02
02:07 s1n joined, xjiujiu joined 02:09 xjiujiu left
dalek ecza: e0d294d | sorear++ | / (3 files):
Embed unidata into Kernel.dll, check in a working quasi-finalized version
 02:17
sorear now, it will be possible to "use Niecza::UCD" out of the box
rakudo: say "SS" ~~ /:i ß/ 02:18
p6eval rakudo 1eadbb: OUTPUT«#<failed match>␤»
sorear rakudo: say "ß" ~~ /:i SS/ 02:19
p6eval rakudo 1eadbb: OUTPUT«#<failed match>␤»
sorear rakudo: say "SSß" ~~ /:i ßSS/
p6eval rakudo 1eadbb: OUTPUT«#<failed match>␤»
sorear b: say "SSß" ~~ /:i ßSS/
p6eval b 1b7dd1: OUTPUT«␤»
sorear b: say $( "SSß" ~~ /:i ßSS/ )
p6eval b 1b7dd1: OUTPUT«␤»
sorear b: say ?( "SSß" ~~ /:i ßSS/ )
p6eval b 1b7dd1: OUTPUT«Bool::False␤»
02:19 xaka left 02:23 shinobicl_ left 02:44 oZ] joined, fieldsa left, oZ] left 02:46 Radvendii joined, xaka joined
sorear o/ Radvendii 02:52
Radvendii hey 02:56
whats up?
02:57 Radvendii left, Radvendi1 joined 02:58 Radvendi1 left, Radvendii joined 02:59 xjiujiu joined, Radvendii left, Radvendii joined
sorear Radvendii: was trying to greet you, although you seem to be having connection troubles 02:59
Radvendii no... i'm trying to figure out how to remove automatic connections from irssi 03:00
:P
hey
sorear I usually do that by /quitting and deleting stuff from .irssi/config
there's probably a more elegant way :D
Radvendii yeah, you're not supposed to mess with that file 03:01
i think i will anyways though...
yay tabbed terminals!
(i dont even have to /quit
anwyays, how is life going for you?
brb 03:03
03:03 Radvendii left, Radvendi1 joined
Radvendi1 it works :D 03:03
does panda work with niecza? 03:05
sorear very doubtful
(my life) classes are over for the year; I'm madly hacking in an attempt to drown out the existential angst 03:06
Radvendi1 sweet 03:07
:P
can i help?
sorear with what?
Radvendi1 "madly hacking"
namely, niecza?
sorear yes, niecza.
Radvendi1 (does it work on mac, by the way)? 03:08
sorear niecza: use Niecza::UCD
p6eval niecza v12-82-g8c76823: OUTPUT«Rebuild in progress␤»
sorear Radvendi1: yes it does
Radvendi1 sweet
sorear I'm sure you can help 03:11
Radvendi1 damnit. irssi got my username wrong again! 03:13
what is *wrong* with it 03:15
sorear at the instant you logged in, Radvendii was taken 03:16
try: /nick Radvendii
Radvendi1 it says "Nick Radvendii already in use"
im going to restart irssi...
brb
03:16 Radvendi1 left 03:17 Radvendi1 joined
Radvendi1 there we go 03:17
sorear Radvendi1: it's not irssi's fault that it can't use Radvendii
Radvendi1 no
wtf??
sorear Radvendi1: STOP
Radvendi1: you have two irc clients open
Radvendi1 i do?
sorear yes
/whois Radvendi1
it's a current connection from your computer.
/whois Radvendii 03:18
and Radvendii even responds to pings, so it's not a ghost conn
03:19 Radvendii joined, Radvendi1 left
Radvendii yay 03:19
sorear Radvendii: could you nopaste your .irssi/config?
Radvendii nopaste? 03:20
sorear we need a better verb
put it on a pastebin somewhere so I can point out the mistake
Radvendii gah! i used to be able to do it in github. now that i have an account, i cant figure out how! 03:22
i think i may have solved it... 03:23
let me try again
03:23 Radvendii left, Radvendi1 joined
colomon 'night, all. 03:23
Radvendi1 nope
good night 03:24
but now it's happening to all of my channels...
sorear Radvendi1: install App::Nopaste and run nopaste -x ~/.irssi/config
cpanm -Si App::Nopaste
hey, Radvendi1, how good are you with written Hebrew? 03:26
Radvendi1 uh...
i'm better reading than writing
i'm better understanding spoken though... 03:27
why?
sorear perl 6 i18n support; I want to know who I can direct questions to
Radvendi1 i dont see how that has to do with hebrew... 03:28
sorear it's a language with a peculiar orthography
like if I have questions about ß, I go to moritz 03:29
Radvendi1 oh
i could probably help on that...
i could ask other members of my family if i didnt know, too 03:30
is nopaste a bash command or run from perl?
brb 03:32
03:32 Radvendi1 left, Radvendii joined
Radvendii win :D 03:33
it worked
sorear \o/
nopaste is a shell command, yes
Radvendii then how do i install it with cpan? doesn't that do perl modules?
sorear it's a Perl module that comes with a shell command
Radvendii GAH! So many dependencies!
sorear 1. use cpanm 03:34
2. you need those depenedencies anyway
Radvendii and now it's telling me i dont have YAML installed! i specifically installed it to get rid of those errors
or warnings, as they may be
is there no ./config(ure)(.PL) for niecza? 03:37
sorear correct 03:39
Radvendii make is giving me an error
03:39 Patterner left
sorear what kind of error 03:39
sorear guesses: mono: command not found 03:40
Radvendii wget --no-check-certificate -Oboot/niecza.zip $(cat FETCH_URL)
wget: unrecognized option '--no-check-certificate'
Usage: wget [OPTION]... [URL]...
Try `wget --help' for more options.
gnumake: *** [.fetch-stamp] Error 2
nope
sorear wow
what version of wget do you have?
try removing the offending option
Radvendii 1.13
sorear wget -V | grep https ? 03:41
Radvendii why is vim making a ding noise every time i quit out!! grr....
03:41 Psyche^ joined, Psyche^ is now known as Patterner
Radvendii why the capital v? just to be not standard? 03:42
+digest -https +ipv6 -iri +large-file -nls -ntlm +opie -ssl
do i have to recompile wget?!
sorear nah, that *is* standard
Radvendii i've always seen lowercase
sorear -V is short for --version 03:43
Radvendii i know
sorear -v is short for --verbose
Radvendii oh
right
i did know that
back to the point, i dont have https compiled in, do i have to recompile it?
sorear no, you should be able to work around it somehow. 03:45
some ideas:
Radvendii is it better(easier) to just recompile it? :P
sorear - removing the s from FETCH_URL might work
I've never tried to recompile wget on Mac OS X
- or download manually and then fake the Makefile directions 03:46
Radvendii i think i had to do it a while ago...
why dont they compile with these things by default?
ugh
sorear You're the first person I've heard to have this problem.
Radvendii great. 03:47
sorear You're not the first OSX user I've had.
You have an exceptionally old/new/weird OSX, I'm afraid
Radvendii ugh. 03:48
03:48 alc left
Radvendii can i wget wget? 03:48
:P
yeah, i definetily had to install wget 03:49
my mac didnt come with it
benabik OS X comes with curl instead of wget
Radvendii ah
sorear Niecza OSX folks! What did y'all do to make the wget call work? 03:50
Radvendii you know... inevitably this install of nopaste is going to end in some minor error
benabik I recommend homebrew
Radvendii and then i'm going to cry
homebrew?
sorear Radvendii: I supposed your Perl install is so virgin it doesn't even have Moose yet?
benabik mxcl.github.com/homebrew/ 03:51
Radvendii sorear: i have no idea. i just install things on a need basis. but *so* often it just doesnt work. like with SDL.
i dont remember if i got SDL working in the end
03:51 xjiujiu left
benabik It's a way to install utilities. 03:51
Radvendii benabik: are you kidding me? another macports? 03:52
benabik Radvendii: I've been through macports and fink, and homebrew has been the least painful option.
Radvendii does it let you add options to the install?
benabik It does allow formula to define options. 03:54
Radvendii :O nopaste... it just... worked...
im blown away
now to see if i can actually use it... 03:55
what did you say to use? cpanm?
dalek ecza: 8533895 | sorear++ | / (4 files):
Add <:Foo>, <:gc(/L/)> syntax.
sorear I wanted you to use cpanm to *install* nopaste
it's a smaller faster and less breakable version of the 'cpan' instlaler 03:56
now: nopaste ~/.irssi/config
Radvendii what?
so i shouldnt be using cpan period?
03:57 s1n left
sorear I don't think you should. 03:57
Radvendii sweet, so nopaste worked (i already solved the problem though
i dont have cpanm...
should i cpan cpanm? 03:58
or maybe macports?
or homebrew?
or wget
gah!!!
:P
benabik cpanm is a CPAN module, I think. :-D
sorear cpan -i App::cpanminus or cpanmin.us/
Radvendii speaking of which, homebrew is not installing
03:59 envi_ left, xjiujiu joined
Radvendii share/man/: Cannot extract through symlink share 03:59
share/man/man1/: Cannot extract through symlink share
share/man/man1/brew.1: Cannot extract through symlink share
tar: Error exit delayed from previous errors.
Failed during: /bin/bash -o pipefail -c '/usr/bin/curl -skSfL github.com/mxcl/homebrew/tarball/master | /usr/bin/tar xz -m --strip 1'
benabik Radvendii: If you already have macports, I probably wouldn't bother with homebrew unless you want to switch.
04:00 s1n joined, s1n left, xjiujiu left
Radvendii well, it's always another thing to try if there isn't a macport available 04:00
...
04:02 xjiujiu joined
Radvendii except that it's not working... 04:03
GAH 04:04
now port selfupdate isn't working!!
no, i think i got it by adding sudo :P
good old sudo :-D
wow. i just freed about 48 gb of my hard drive space from emptying my trash 04:11
sorear perl6: say "foo2bar" ~~ /<:gc(/N/)>/ 04:14
p6eval rakudo 1eadbb: OUTPUT«===SORRY!===␤regex assertion not terminated by angle bracket at line 1, near "(/N/)>/"␤»
..pugs b927740: OUTPUT«*** ␤ Unexpected "N"␤ expecting term postfix, operator, ":" or ","␤ at /tmp/i3SpiaXKvp line 1, column 25␤»
..niecza v12-84-g8533895: OUTPUT«#<match from(3) to(4) text(2) pos([].list) named({}.hash)>␤»
sorear er? I thought rakudo added that already
04:17 jaldhar joined
sorear niecza: say (2**256)- + ~(2**256) 04:31
p6eval niecza v12-84-g8533895: OUTPUT«0␤»
sorear niecza: say (3**256)- + ~(3**256)
p6eval niecza v12-84-g8533895: OUTPUT«0␤»
sorear niecza: say ~(3**64) 04:32
p6eval niecza v12-84-g8533895: OUTPUT«3433683820292512484657849089281␤»
sorear niecza: say +~(3**64)
p6eval niecza v12-84-g8533895: OUTPUT«3.43368382029251E+30␤»
04:35 sivoais left
sorear niecza: say ~(3**64+1) 04:35
p6eval niecza v12-84-g8533895: OUTPUT«3433683820292512484657849089282␤»
sorear at least one direction is lossless 04:36
04:39 replore left, replore joined 04:41 sivoais joined 04:43 araujo left 04:54 jaldhar left 04:58 molaf joined 05:01 jaldhar joined 05:07 alvis left 05:09 alvis joined 05:12 xjiujiu left, xjiujiu joined 05:17 sftp left 05:21 orafu left 05:27 molaf left 05:42 kaleem joined 05:44 baest joined 05:47 baest left
dalek ecza: 0030424 | sorear++ | test.pl:
Fix some tests that didn't like the is change
 06:00
ecza: 0a6396e | sorear++ | / (3 files):
Refactor number handling in the parser and add \c
sorear perl6: say "foo2bar" ~~ /<:gc(/N/)>/
p6eval rakudo 1eadbb: OUTPUT«===SORRY!===␤regex assertion not terminated by angle bracket at line 1, near "(/N/)>/"␤»
..niecza v12-84-g8533895: OUTPUT«#<match from(3) to(4) text(2) pos([].list) named({}.hash)>␤»
..pugs b927740: OUTPUT«*** ␤ Unexpected "N"␤ expecting term postfix, operator, ":" or ","␤ at /tmp/QaEgCdW6zi line 1, column 25␤»
sorear niecza: use Unicode::UCD; say value("na", "\x3BB") 06:01
p6eval niecza v12-84-g8533895: OUTPUT«Unhandled exception: Unable to locate module Unicode::UCD in @path␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 880 (die @ 2) ␤ at /home/p6eval/niecza/src/NieczaPathSearch.pm6 line 23 (NieczaPathSearch.load_module @ 11) ␤ at /home/p6eval/niecza/sr…
sorear niecza: use Niecza::UCD; say value("na", "\x3BB")
p6eval niecza v12-84-g8533895: OUTPUT«===SORRY!===␤␤Q:CgOp not allowed in safe mode at /home/p6eval/niecza/lib/Niecza/UCD.pm6 line 4:␤------> { (ucd_get_ranges {$table} {$matcher}) ⏏}␤␤Q:CgOp not allowed in safe mode at /home/p6eval/niecza/lib/Niecza/UCD.p…
sorear moritz: Helmut W. returned my mail 06:05
06:16 overrosy_ left
moritz sorear: "returned" as in "delivery failed"? 06:17
phenny moritz: 19 Dec 21:48Z <diakopter> tell moritz typo in your last spec commit
moritz tries to sleep again
sorear moritz: no, a human replied. 06:18
06:18 overrosy joined
sorear will probably have to memorize what the W. stands for eventually 06:18
dalek ecza: 78dbb2f | sorear++ | src/ (2 files):
Remove unneeded [0]s
 06:41
06:50 Trashlord left 06:51 Radvendii left 07:00 koban joined
dalek ecza: 746a6bc | sorear++ | src/ (3 files):
Mergeback
 07:10
07:14 wtw joined 07:23 fridim_ joined
sorear hmm, niecza's parser has gotten a lot slower and it's not clear why 07:31
just took 1:50 to compile STD.pm6 instead of the usual 0:50
I need to investigate later
07:32 jrockway left 07:33 araujo joined
sorear -> sleep 07:37
07:38 koban left 07:40 zby_home_ joined 07:43 fridim_ left 07:51 xaka left, am0c left 07:56 krunen joined 08:00 baest joined 08:03 am0c joined 08:04 krunen_ joined 08:05 mkramer joined 08:06 zby_home_ left 08:07 mkramer left 08:27 agentzh joined, am0c left 08:31 jlaire left 08:33 jlaire joined 08:44 mj41 joined 08:58 krunen_ left 08:59 am0c^ joined 09:00 mj41 left 09:01 am0c^ left 09:06 mkramer joined 09:14 mj41 joined 09:38 dakkar joined 09:41 cooper left 09:43 mkramer left 09:46 replore left 09:48 mj41 left 09:54 jerome_ left 10:01 jrockway joined 10:08 jerome_ joined 10:14 domidumont left 10:18 mj41 joined 10:20 domidumont joined 10:25 xjiujiu left 10:31 X99 joined 10:32 daxim joined 10:37 snearch joined 10:46 underline_ joined 10:47 snearch left 10:48 underline_ left 10:56 envi_ joined 11:09 mkramer1 joined, mkramer1 left 11:19 lichtkind_ joined 11:20 lichtkind left, lichtkind_ is now known as lichtkind 11:28 jerome_ left 11:38 jerome_ joined 11:46 cedric joined
cedric Hello #perl6 11:47
Is there someone that could help me on regex modifiers? 11:51
phenny, tell moritz S05:623 explains how to apply the :pos modifier to m//-like calls, how can one do the same with a call to a subrule within a grammar? 12:04
phenny cedric: I'll pass that on when moritz is around.
cedric std: rx:pos(3) /hello/ 12:05
p6eval std dc62e1d: OUTPUT«ok 00:01 121m␤»
cedric phenny, tell moritz viv and std don't complain when a "call-modifier" (like :pos) is applied to a definition. This is unexpected according to S05:657. 12:06
phenny cedric: I'll pass that on when moritz is around.
lumi___ phenny: tell sorear I'm a native Hebrew speaker and reader, ask me anything 12:09
phenny lumi___: I'll pass that on when sorear is around.
12:16 Alias joined
dalek : 1e13cf9 | util++ | misc/perl6advent-2011/schedule:
[advent] Claim 23rd.
 12:22
12:35 cedric left
colomon phenny: tell sorear My wget appears to come from macports 12:38
phenny colomon: I'll pass that on when sorear is around.
12:42 X99 left 12:45 ispeak joined 12:47 Trashlord joined 13:18 overrosy left 13:19 overrosy joined 13:29 kaleem left
[Coke] Juerd: I typically do run things under "nice". I'll try to be more careful in the future. (and I'll see about the memory limiting also.) 13:32
Juerd: sorry about that.
13:33 bluescreen10 left
Juerd [Coke]: Shit happens and is allowed to happen on a dev box :) 13:33
13:33 bluescreen10 joined
Juerd Just not repeatedly :P 13:33
13:36 X99 joined 13:40 wolfman2000 joined 13:41 mkramer joined 13:42 X99 left 13:43 wolfman__ joined, wolfman2_ left 13:46 drbean joined, wolfman2000 left
[Coke] yah, my normal "spectest run" is under nice -20 for rakudo & niecza. 13:49
(if i run "make spectest" by itself, that's not.)
I should just nice my entire shell, honestly. ;)
Juerd The memory usage is a somewhat greater issue 13:50
I took away a lot of RAM from feather1 because it wasn't needed anymore, and we needed some elsewhere
[Coke] ok. and that I don't normally do anything with. I can fix that. 13:51
Juerd It only has 1200 MB left
So if a process starts to use 1500 MB, it'll start swapping.
Well, way before that of course
[Coke] I personally don't mind if I'm swapping, but can see where it would impact others. ;) 13:52
13:54 kaleem joined 13:56 sftp joined 13:57 GlitchMr joined
[Coke] Juerd: ok, I made my default run of niecza be niced, time limited, and memory limited. 14:03
Juerd That's great; thank you 14:04
[Coke] (my ./perl6 alias). so everything I run there should be a little less greedy going forward.
ping me if you see a problem, though.
I will also try to stick with a smaller number of TEST_JOBS.
running a typical spec test run for niecza now. 14:07
(it's building first, test harness starts shortly.) 14:08
14:08 risou_awy is now known as risou
[Coke] Juerd: oh, and I know there was one case when I nearly blew feather away, because I mistyped a TEST_JOBS parameter as 100 (!), but that was a onetime thing. ;) 14:09
I would posit that if niecza has skipped a roast due to "nonsensical test" perhaps we should think about killing that test. ;) 14:10
14:13 REPLeffect left, REPLeffect joined
moritz it's just that people don't always agree abotu whether such a test is nonesensical 14:21
phenny moritz: 12:04Z <cedric> tell moritz S05:623 explains how to apply the :pos modifier to m//-like calls, how can one do the same with a call to a subrule within a grammar?
moritz: 12:06Z <cedric> tell moritz viv and std don't complain when a "call-modifier" (like :pos) is applied to a definition. This is unexpected according to S05:657.
14:23 drbean left
[Coke] moritz: mmm. eventually I will open a ticket on specs or roast for clarification. 14:24
yay, niecza got another 350 passing tests since yesterday. 14:26
moritz [Coke]: in general I'm happy to review dubious tests; but today i'm rather ill, so it's a bad time
colomon [Coke]: ;) 14:27
moritz lichtkind: any objections to publishing your advent post now?
lichtkind moritz: no :) 14:28
i was going to ask you too :)
i made some final touches in the night
to make it spund more coherent after the deletions
its a bit pitty that the last slamdunk doesnt work out but well
s/spund/sound/ 14:29
benabik moritz: Hope you feel better soon.
lichtkind is he ill?
14:29 drbean joined
jnthn lichtkind: Read like 10 lines ago ;) 14:29
moritz: Get well soon!
colomon # Failed test 14:31
# got: WTF
# expected: אאא
14:31 agentzh left
colomon sorear++ 14:31
lichtkind jnthn: i think i came in after it
moritz thanks for the good wishes 14:32
14:32 s1n joined
lichtkind yes get well 14:33
moritz tries 14:34
[Coke] moritz: get well soon.
moritz: NOW! ;)
lichtkind i ask god and arrange something :) 14:35
14:36 cedric joined
cedric Hello #perl6 14:40
colomon \o
cedric o/ colomon
[Coke] regarding irclog.perlgeek.de/perl6/2011-12-20#i_4856073, were you talking about the questions on "call-modifiers"?
14:40 Alias_ joined 14:41 [particle]1 joined
lichtkind there are broken html entities at the end 14:41
dalek ast: 06087cb | (Solomon Foster)++ | S03-operators/autoincrement-range.t:
Fudge for niecza.
lichtkind did you fixed that or should i?
moritz: ? 14:42
14:42 Timbus|Away joined, c1sung_ joined 14:43 lestaway joined, kcwu_ joined 14:44 hillu_ joined
moritz lichtkind: please do 14:45
14:45 zamolxes_ joined
lichtkind moritz: done 14:45
[Coke] cedric: I actually don't know what test it was. I just saw it in a list of skip reasons.
nom: say abs(2000).WHAT 14:48
p6eval nom 1eadbb: OUTPUT«Parameter '' requires an instance, but a type object was passed␤ in method Bridge at src/gen/CORE.setting:2441␤ in sub infix:<<> at src/gen/CORE.setting:2371␤ in sub prefix:<abs> at src/gen/CORE.setting:2387␤ in block <anon> at /tmp/rA1FRy_aQ6:1␤ in <anon> at /tm…
[Coke] nom: say (abs(2000)).WHAT
p6eval nom 1eadbb: OUTPUT«Int()␤»
14:49 simcop2387 left 14:50 Alias left, lestrrat left, PacoLinux left, kcwu left, [particle] left, c1sung left, zamolxes left, hillu left, kfo left, TiMBuS left, Patterner left, jnthn left, mikemol left, Helios left, kranius left, DarthGandalf left, chra left, Tene left, jasonmay left, lestaway is now known as lestrrat, Patterner joined, jnthn joined, mikemol joined, kranius joined, Helios joined, DarthGandalf joined, chra joined, Tene joined, jasonmay joined 14:51 Timbus|Away is now known as Timbus 14:52 simcop2387 joined
moritz oh, that's the weird "abs is a prefix, not a listop" 14:53
pugs: say abs(2000).WHAT
p6eval pugs b927740: OUTPUT«Int␤»
14:53 GlitchMr left
dalek ecza: 39ccdd6 | (Solomon Foster)++ | t/spectest.data:
Turn on S03-operators/autoincrement-range.t.
 14:54
[Coke] moritz: yah, just saw a spectest skip with that.
gist.github.com/1476841 - colomon, good job, we're almost at 58%. ;) 14:55
with moritz sick, I bet you can make up even more ground! hurry!
14:55 PacoLinux joined
[Coke] ... I'm going to have to switch the order of those dates soon. 14:56
14:57 Trashlord left
colomon [Coke]: When sorear++ finishes the stuff he's been working on, we're probably going to pull in about 2000 passing tests overnight, I think. 14:58
15:01 kfo joined
[Coke] colomon: sweet. 15:03
daxim I found today's advent door, practical work on data structures, really useful. more of that stuff, please 15:13
lichtkind moritz: needed some reformat too, due different line breaks and more, but fixed all that immediately
15:15 renormalist joined
[Coke] daxim: good to hear. 15:15
15:20 raiph joined 15:21 PNR joined
daxim www.phoronix.com/scan.php?page=news...px=MTAzMDk for you compiler people 15:22
15:23 PNR left
raiph lichtkind: nice post. i like that P5 is merely ugly whereas P6 is "hideous"! 15:26
lichtkind raiph: thank you :)
sorear good * #perl6 15:30
phenny sorear: 12:09Z <lumi___> tell sorear I'm a native Hebrew speaker and reader, ask me anything
sorear: 12:38Z <colomon> tell sorear My wget appears to come from macports
flussence this is probably wildly inaccurate, but I've just got it up to date: github.com/flussence/specgraphs/ra.../impls.png
colomon o/, sorear 15:31
sorear: just tried to implement srand and earned myself a Invalid IL code in Run.CORE:C806srand error. :( 15:32
15:33 Trashlord joined, wtw left
sorear colomon: wow, nice 15:33
colomon: btw, I finished \c and <:foo> 15:34
dunno if that's what you meant
o/ cedric
cedric o/ sorear
colomon public static void srand_time() { lock (rng) { rng = new Random(); } } // correct or insane? 15:35
cedric sorear, I tried to had support for Cursor.at() to Niecza without success, it seems I don't speak NAM fluently :)
15:35 tokuhirom joined
sorear colomon: I vote for insane 15:36
moritz
.oO( insanely correct )
colomon sorear: can't change rng when you have a lock on it? 15:37
sorear colomon: one of the more insane things about the CLR is that every object contains a mutex 15:38
colomon: you're replacing the mutex at the same time that you change rng, which almost certainly will mess up locking
kthakore sorear: I got SDL working at some point
can update the repo again if you are still interested
github.com/PerlGameDev/SDL6 15:39
sorear kthakore: ENOCONTEXT
kthakore I believe
you in my last log
< Radvendii> sorear: i have no idea. i just install things on a need basis. but *so* often it just doesnt work. like with SDL.
03:51 < Radvendii> i dont remember if i got SDL working in the end
colomon sorear: does that mean we need another variable to have a lock on, to indicate when it is okay to change rng?
sorear colomon: yes 15:40
colomon: rather, the new variable has to be used everywhere lock (rng) currently is
colomon right, that's what I'm trying now
sorear cedric: ehh... Probably my fault. What did you try? 15:41
raiph flussence: surprises me that # of spectests (22k ish) is roughly unchanged for 6 months 15:42
colomon sorear: suggested type for the new lock?
sorear object
15:42 pkkm_ joined
kthakore ok sorry 15:43
kthakore goes back to lurking
colomon sorear: done, and I still get the error. and I just realized it's because I'm not waving the "no return value" dead chicken at my CgOp call 15:44
rnull?
cedric sorear, IIRC it was: "method at($pos) { Q:CgOp { (rxsetpos (cast int (obj_getnum {$pos}))) }}". Mono stopped due to an illegal IL opcode. 15:46
*illegal IL opcode
moritz nom: class A { method a() { } }; role B { method a() { ... } }; class C is A does B { };
p6eval nom 1eadbb: ( no output ) 15:47
dalek ast: 47189d0 | moritz++ | S14-roles/stubs.t:
test inheritance + stubbed method from a role
 15:49
15:50 tokuhirom left 15:52 MindosCheng joined 15:53 kaleem left
dalek ast: 74f5151 | (Solomon Foster)++ | S32-num/rand.t:
Fudge for niecza.
 15:57
sorear cedric: hmm... your immediate problem is the '"no return value" dead chicken', your less-immediate problem is that rxsetpos only works in regexes 15:58
cedric: also, S05:1724 suggests that Cursor.at should normally be a zero-width assertion, not a movement command 15:59
cedric sorear, my attempt to do some cargo-cult programming in Niecza is definitively a failure :) Also I tried to implement the <at()> form (S05:4420), not the <?at()> one. 16:03
dalek ecza: 4f23627 | (Solomon Foster)++ | lib/ (2 files):
Implement srand.
ecza: 64301c7 | (Solomon Foster)++ | t/spectest.data:
Turn on rand.t.
colomon perl6: say "abc\x[a]def" ~~ m/\c[LINE FEED (LF)]/ 16:06
p6eval niecza v12-89-g39ccdd6: OUTPUT«===SORRY!===␤␤Unrecognized character name LINE FEED (LF) at /tmp/sITQRGU4p_ line 1:␤------> say "abc\x[a]def" ~~ m/\c[LINE FEED (LF)⏏]/␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting …
..rakudo 1eadbb: OUTPUT«=> <␤>␤␤»
..pugs b927740: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20111008/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HAND…
colomon perl6: say "abc\x[a]def" ~~ m/\c[LINE FEED]/ 16:07
p6eval niecza v12-89-g39ccdd6: OUTPUT«===SORRY!===␤␤Unrecognized character name LINE FEED at /tmp/hFlxrComoU line 1:␤------> say "abc\x[a]def" ~~ m/\c[LINE FEED⏏]/␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 880 (…
..rakudo 1eadbb: OUTPUT«===SORRY!===␤Unrecognized character name LINE FEED at line 1, near "]/"␤»
..pugs b927740: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.20111008/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HAND…
16:07 mj41 left
colomon perl6: say "\c[LINE FEED]" 16:07
p6eval pugs b927740: OUTPUT«␤␤»
..rakudo 1eadbb: OUTPUT«===SORRY!===␤Unrecognized character name LINE FEED at line 1, near "]\""␤»
..niecza v12-89-g39ccdd6: OUTPUT«===SORRY!===␤␤Unrecognized character name LINE FEED at /tmp/ttOUpdUQvA line 1:␤------> say "\c[LINE FEED⏏]"␤␤Unhandled exception: Check failed␤␤ at /home/p6eval/niecza/boot/lib/CORE.setting line 880 (die @ 2) ␤ at /ho…
colomon perl6: say "\c[LINE FEED (LF)]" 16:08
p6eval niecza v12-89-g39ccdd6: OUTPUT«Rebuild in progress␤»
..pugs b927740, rakudo 1eadbb: OUTPUT«␤␤»
sorear ehehee... I was hoping nobody would notice that case for a while
colomon sorear: it pops right out in S05-mass/named-chars.t
16:12 envi_ left, envi_ joined 16:13 mj41 joined
flussence I'm writing perl5 code at $dayjob and really wish pack("n/($arbitrary_stuff)") worked right now... 16:14
I have to glue together a network thing which goes: [length of actual data, actual data with variable length last field, null padding]... I've ended up using three nested pack() calls :( 16:16
colomon perl6: say "\c[ARMENIAN SMALL LETTER ECH]"
p6eval pugs b927740: OUTPUT«Õ¥␤»
..rakudo 1eadbb, niecza v12-91-g64301c7: OUTPUT«ե␤»
colomon sorear: what are the boundaries of that case you were hoping to sneak by? 16:17
16:17 molaf joined, drbean left
sorear colomon: <[\x00..\x1F]> 16:17
the characters aren't officially named by Unicode 16:18
unrelated: I'm trying to get the bootstrap working again. "Compilation unit cannot be anonymous" at CORE.setting line... 2 :|
colomon sorear: gotcha 16:19
arnsholt On the bright side, you won't have to scroll far to find the spot where your compiler stopped O:)
dalek ecza: 6ff0fee | sorear++ | / (2 files):
Small fixes to restore self-compilation
sorear those were some little changes from last night 16:20
which incidentally fix \cJ and \cM
ahahaha 16:21
it's obvious what broke :> 16:22
16:23 drbean joined
colomon sorear: ack, named-chars.t starts off with like 100 tests of LF, CR, NEL, and FF. 16:31
dalek ecza: 282244a | sorear++ | src/STD.pm6:
Remove some stray [0] from STD
 16:33
sorear colomon: ack 16:34
colomon sorear: the fudging is epic and ugly
sorear don't bother just yet 16:35
colomon actually, I just fudged it enough that it runs all the way through the test 16:36
but it's got 168 failures
and 116 skipped tests
16:37 drbean left, awoodland joined 16:39 MayDaniel joined, am0c joined 16:40 slavik1 joined
colomon sorear: it looks like the failures are most negative character class tests 16:41
dalek ecza: 1b83260 | sorear++ | lib/UCD.cs:
Add name recognition via na1 for control chars
sorear stefan@stefans:~/rel/niecza$ mono-sgen run/Niecza.exe -e 'say ord "\c[LINE FEED
(LF)]"'
10
16:42 drbean joined
cedric Does someone know how to set the match position when calling a subrule in a grammar? I mean something similar to "m:pos(3)//" but applied to a subrule call. I guess it is "<at($pos)>" but I'm not sure. 16:44
colomon sorear: looks like \c[LF] still fails? 16:47
rakudo skips all those tests, for what it is worth... 16:48
okay, current stats: 43 skipped tests, 196 failed, 180 okay 16:50
failures look to all be negative character class tests
though not all negative tests fail 16:51
example:
ok(!( "\x[c]" ~~ m/^ \C[FORM FEED (FF)]/ )
16:52 baest left
[Coke] cedric: just so you know not everyone is ignoring you, no I have absolutely no idea how to do that, and am wondering why you wouldn't just adjust your grammar so you didn't have to manually muck with the position matched so far. 16:57
dalek ast: 957b9f9 | (Solomon Foster)++ | S05-mass/named-chars.t:
Fudge enough that niecza can run the file without a fatal error.
moritz cedric: probably YourGrammar.parse('yourrule', :pos(12)) 16:58
cedric: note that in a grammar, <foo>-style subrule calls are *always* ancored to the current position, it doesn't make much sense to start off somewhere different inside a normal rule call 16:59
16:59 kaare_ joined
colomon sorear: "Letter does not exist as a UCD table"? 17:05
cedric [Coke], moritz: I'm writting a grammar to parse ELF files. I managed to get Niecza parsing the header correctly (32 or 64-bit) and now I want the parser to "jump" directly to the program/section header tables. The location for these tables are described by the header as an offset in the file, hence my will to continue the parsing somewhere else. 17:11
moritz cedric: then you should do that by some outside, controlling code 17:13
cedric: ie you have a routine that parses the header, extractions the location of the tables, and does another call to parse that with a different grammar 17:14
that'll be much easier than doing everything in a grammar
17:14 carlin joined 17:17 wolfman__ left
cedric moritz: much easier but less impressive :) I would like to demonstrate to my co-workmates how the Perl6 Grammar is powerful. Seriously, only one "controlling" statement was needed so far (to detect 32/64-bit class), everything else is declarative. I'm pretty sure it could be possible to get the whole ELF format described as Perl6 Rules (thanks to something like <at()> or :pos for subrule). 17:23
moritz in theory you can write .*?<at($location)> 17:24
not sure if that works in niecza 17:25
niecza: say 'abc' ~~/<at(1)>.*/
p6eval niecza v12-94-g1b83260: OUTPUT«Unhandled exception: Unable to resolve method at in class Cursor␤ at /tmp/x4WXX8J8dx line 1 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1688 (Regex.ACCEPTS @ 3) ␤ at /tmp/x4WXX8J8dx line 1 (mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.…
cedric It could work with the ELF format, but some formats require to jump backward (most .mov files for instance, the header is actually a footer ;) 17:27
17:33 daxim left
cedric Since the Unicode level "bytes" isn't supported yet, the binary file is first converted to ASCII with the command ``hexdump -v -e '/1 "%02x"' /path/to/binary/file``, then I use: "token byte { .. }" 17:34
17:39 MayDaniel left
cedric moritz: I guess I misread you, did you mean "<at()>" should allow to move anywhere? 17:41
17:42 mj41 left
moritz cedric: no, <at> doesn't move anywhere, it's just an assertion 17:42
17:48 Timbus left, Timbus joined
cedric moritz: thanks for the clarification. Then, what's the difference between <at()> and <?at()>? 17:50
moritz cedric: nothing, afaict
cedric: you can call every rule <foo> as <?foo> to make it zero-width, but <at> already is zero-width 17:51
cedric moritz: it makes sense to me now, thanks! 17:56
moritz note that <?foo> does not capture either 18:02
18:03 dakkar left 18:05 MayDaniel joined 18:07 rindolf joined
rindolf Hi all. 18:07
moritz hi rindolf 18:08
rindolf moritz: what's up? 18:09
moritz not much 18:10
moritz is sick, and not much use for hacking it seems
dalek p: bceccc5 | moritz++ | src/QRegex/Cursor.nqp:
implement <at($pos)>
 18:11
p: 3ae3f57 | moritz++ | t/qregex/rx_subrules:
regex tests for <at>
rindolf moritz: oh, I hope you get well soon. 18:13
moritz: yet you're still hacking.
moritz occasionally. When I can sit :-) 18:14
18:14 ispeak left
rindolf moritz: ah. 18:14
moritz: what's preventing you from sitting?
18:15 ispeak_ joined
rindolf BTW, I liked this year's Perl 6 Advent Calendar less than last year's. 18:15
moritz rindolf: why?
dalek kudo/nom: 106783c | moritz++ | t/spectest.data:
run roles + stubs test file
kudo/nom: dcfd439 | moritz++ | lib/Test.pm:
give eval_lives_ok some diagnostics in the case of test failure
kudo/nom: 89f54dc | moritz++ | tools/build/NQP_REVISION:
bump to a NQP_REVISION that supports <at> in regexes
rindolf moritz: don't know.
18:15 ispeak_ left
moritz choice of topics? linguistic? 18:15
rindolf moritz: I guess they were less eye-popping features covered.
benabik Last year's seemed more tutorial, this one seems more technical. 18:16
rindolf moritz: things that make me go "Wow! That's Cool"
sorear colomon: er... niecza must be treating :Letter as :Letter<Y> and looking for a boolean property of that name
hello rindolf 18:17
rindolf sorear: hi.
sorear: what's up?
sorear rindolf: well, all the "Wow! Cool!" topic were already covered last year
rindolf sorear: hmmm... 18:18
moritz last two years, actually 18:20
I guess next year will cover much more modules 18:21
18:23 drbean left
cedric How can I respectfully ask to language designers to think about a writable $/.pos? (since it could be useful to some parsers and also for the sake of consistency regarding the :pos modifier) 18:24
std: say 'abc' ~~ /{ $/.pos = 2 }./
p6eval std dc62e1d: OUTPUT«ok 00:01 124m␤»
cedric perl6: say 'abc' ~~ /{ $/.pos = 2 }./ 18:25
p6eval niecza v12-94-g1b83260: OUTPUT«Unhandled exception: assigning to readonly value␤ at /tmp/X5cKMxcIZw line 1 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1688 (Regex.ACCEPTS @ 3) ␤ at /tmp/X5cKMxcIZw line 1 (mainline @ 1) ␤ at /home/p6eval/niecza/lib/CORE.setting line 228…
..rakudo 1eadbb: OUTPUT«Method 'pos' not found for invocant of class 'Any'␤ in block <anon> at /tmp/y9h6VvgJtT:1␤ in regex <anon> at /tmp/y9h6VvgJtT:1␤ in method ACCEPTS at src/gen/CORE.setting:7174␤ in block <anon> at /tmp/y9h6VvgJtT:1␤ in <anon> at /tmp/y9h6VvgJtT:1␤»
..pugs b927740: OUTPUT«*** ␤ Unexpected "}./"␤ expecting operator, "," or ":"␤ at /tmp/Oygmil1_Pv line 1, column 28␤»
moritz perl6: say 'abc' ~~ /{ $/.pos == 2 }./ 18:26
p6eval rakudo 1eadbb: OUTPUT«Method 'pos' not found for invocant of class 'Any'␤ in block <anon> at /tmp/nrHCrA328e:1␤ in regex <anon> at /tmp/nrHCrA328e:1␤ in method ACCEPTS at src/gen/CORE.setting:7174␤ in block <anon> at /tmp/nrHCrA328e:1␤ in <anon> at /tmp/nrHCrA328e:1␤» 18:27
..niecza v12-94-g1b83260: OUTPUT«#<match from(0) to(1) text(a) pos([].list) named({}.hash)>␤»
..pugs b927740: OUTPUT«*** ␤ Unexpected "}./"␤ expecting operator, ":" or ","␤ at /tmp/kmZHyTscXK line 1, column 29␤»
moritz evalbot rebuild nom
p6eval NOT OK (maybe a rebuild is already in progress?)
18:28 brno0 joined
sorear niecza: say 'abcdef' ~~ / <.cursor_at(4)> . / 18:28
p6eval niecza v12-94-g1b83260: OUTPUT«Unhandled exception: Unable to resolve method cursor_at in class Cursor␤ at /tmp/dAD7fIMwuo line 1 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1688 (Regex.ACCEPTS @ 3) ␤ at /tmp/dAD7fIMwuo line 1 (mainline @ 1) ␤ at /home/p6eval/niecza/li…
sorear niecza: say 'abcdef' ~~ / <.cursor(4)> . /
p6eval niecza v12-94-g1b83260: OUTPUT«#<match from(0) to(5) text(abcde) pos([].list) named({}.hash)>␤» 18:29
18:29 drbean joined
sorear cedric: undocumented feature 18:29
cedric: if you want an Official Way, I'm afraid the only way I've ever seen work is to stay on and pester TimToady over several weeks 18:30
18:30 mkramer left
cedric :D 18:30
18:30 mkramer joined
moritz what's wrong with <at> as the official way? 18:31
18:31 localhost left
moritz or do you mean the overlapping matches problem? 18:31
sorear moritz: <at> is zero-width, as you mentioned earlier
cedric moritz: I don't known how to use <at> to "jump" backward 18:32
18:32 localhost joined
moritz you might be able to use the . ** -$large_number .. $large_number 18:32
sorear 1. nobody has implemented that 18:33
2. I think that syntax is insane
moritz thinks that too
sorear 3. jumping around in the input should not be O(n)
cedric sorear: about 3. it isn't if a StrPos is used or if the input ~~ Buf 18:35
moritz niecza: say 'abcdefg' ~~ /. <.cussor(4)> ./
p6eval niecza v12-94-g1b83260: OUTPUT«Unhandled exception: Unable to resolve method cussor in class Cursor␤ at /tmp/5tfONBnUmh line 1 (ANON @ 3) ␤ at /home/p6eval/niecza/lib/CORE.setting line 1688 (Regex.ACCEPTS @ 3) ␤ at /tmp/5tfONBnUmh line 1 (mainline @ 1) ␤ at /home/p6eval/niecza/lib/C…
sorear cedric: .*? <at($n)> has to do an O(n) amount of backtracking
moritz niecza: say 'abcdefg' ~~ /. <.cursor(4)> ./
p6eval niecza v12-94-g1b83260: OUTPUT«#<match from(0) to(5) text(abcde) pos([].list) named({}.hash)>␤»
sorear no matter _how_ you specify $n
benabik Jumping around in input really isn't what regexen were designed for. 18:36
sorear true, but they do it admirably 18:37
cedric sorear: you're right, I was thinking about a <set_pos()>-like feature (note: I didn't understand how to use <cursor()> yet)
18:37 mkramer left
sorear niecza: say 'abcdef' ~~ / ... <.cursor(0)> / # I wonder what this will do 18:37
p6eval niecza v12-94-g1b83260: OUTPUT«#<match from(0) to(0) text() pos([].list) named({}.hash)>␤»
sorear niecza: say 'abcdef' ~~ / def <.cursor(0)> / # I wonder what this will do 18:38
p6eval niecza v12-94-g1b83260: OUTPUT«Unhandled exception: System.ArgumentOutOfRangeException: Cannot be negative.␤Parameter name: length␤ at System.String.Substring (Int32 startIndex, Int32 length) [0x00000] in <filename unknown>:0 ␤ at Niecza.CtxMatchStr.Get (Niecza.Variable obj) [0x00000]…
sorear niecza: say ('abcdef' ~~ / def <.cursor(0)> /).chars
p6eval niecza v12-94-g1b83260: OUTPUT«-3␤»
18:39 forestial joined
cedric benabik: :pos is already there for Match calls, my point is to provide the same feature to subrule calls 18:40
niecza: say 'abc' ~~ /<.cursor(2)>(.)<.cursor(0)>(.)/ 18:45
p6eval niecza v12-94-g1b83260: OUTPUT«#<match from(0) to(1) text(a) pos([#<match from(2) to(3) text(c) pos([].list) named({}.hash)>, #<match from(0) to(1) text(a) pos([].list) named({}.hash)>].list) named({}.hash)>␤»
18:51 MayDaniel left
cedric sorear++ # for <cursor()> 18:56
18:57 mkramer joined
masak <sorear> 2. I think that syntax is insane 18:58
oh good, so it's not just me.
sorear o/ masak
masak evening, #perl6
sorear masak: 10:33 * moritz thinks that too
masak oh good, so it's not just moritz and me ;)
18:58 molaf left
masak starts a massive backlogging venture 18:59
moritz masak and me often seem to think the same thing, to the point where it gets almost scary
18:59 raiph left
masak I pondered writing something similar :P 18:59
moritz :-)
sorear is now trying to figure out why the bootstrap doesn't like <+> 19:00
masak ugh, my jetbrain is all lagged. 19:03
I meant to say, my jet is all lagbrained. 19:05
19:05 REPLeffect left 19:15 fridim_ joined 19:17 daxim joined 19:18 REPLeffect joined
dalek ecza: 2c78cd7 | sorear++ | src/NieczaActions.pm6:
Fix compilation of "+" in bootstrap
 19:22
sorear now it gets to about line 26
19:24 fridim_ left 19:25 awoodland left 19:33 wk joined 19:39 fieldsa joined 19:41 cedric left 19:43 ispeak_ joined 19:46 awoodland joined, carlin left 19:53 mkramer left 19:54 am0c left
moritz why does bootsrapping compilation regress? 19:55
I mean, is it a much different code path from normal compilation?
masak what prompted your question? I'm not sure I understand it. 19:56
without knowing the context, it seems to me that bootstrapping compilation is extra sensitive simply because it feeds into itself, so errors tend to propagate. 19:57
moritz masak: around release time, sorear++ often mentions that bootsrapping niecza fails
sorear moritz: sometimes I accidentally write the compiler to rely on niecza bugs
moritz sorear: ok 19:58
sorear moritz: other times the compiler depends on stuff that is inadequately spectested
it happens a lot more often than I'd like :/ 20:01
masak sorear++ # exposing weak points in roast 20:06
moritz masak++ also learned the hard way that roast and applications often cover quit different areas of the language 20:07
masak .oO( cheese ) 20:08
sorear right now a lot of what I'm dealing with is niecza assuming that $/.ast means $/.ast // ~$/ 20:09
20:11 forestial left
masak it doesn't? 20:11
jnthn No. 20:12
If you don't make, you have no .ast 20:13
make $ast; # not war
20:14 drbean left
sorear o/ jnthn 20:18
jnthn hi sorear 20:19
20:19 Chillance joined, drbean joined
dalek ecza: 586fbaa | sorear++ | src/ (3 files):
More misc bootstrap fixes, mostly undef ast stuff
 20:20
sorear yay, it gets a whole minute into building the setting before failing this time
colomon \o/ 20:21
masak 'night, #perl6 20:22
sorear bye 20:23
hope to see you again
jnthn fwiw, my onlineness is gonna be very sketchy over the next week.
[Coke] jnthn: like, you'll be connecting from the wrong side of town? 20:24
jnthn [Coke]: Nah, going to be with family who have...well actually I've no idea what sort of net connectivity they have.
Either way I'll probably be distracted doing family stuff. :) 20:25
sorear 'course, this is Europe, "sketchy" connectivity means something compeletely different than it does to coke and I :D 20:28
benabik A year or two ago I bought one of my in-laws a wifi router just so I wouldn't have to suffer over the holidays. :-D 20:29
colomon benabik: I did the same at my parents' house years ago. 20:30
In November, I was at my wife's grandfather's house. No internet at all there. :( 20:31
benabik The horror!
jnthn sorear: I feel like I'm missing a USism ;) 20:32
sorear jnthn: I live in a moderately big city and 10Mbps is considered fast
also, I think we still have ~no non-tunnel IPv6 support 20:33
jnthn sorear: Whoa. When I was in Blava they had salesfolk wondering aroud the city trying to sell random folks 100Mbps fibre to home. :) 20:35
*wandering
*around
benabik :-(
jnthn Well, :-( for me too 'cus I moved away from Slovakia shortly after that.
I've actually no idea how fast my connection here is. Mostly, "fast enough that I never think about how fast it is" :) 20:36
sorear and by "moderately big city" I mean "the population of Estonia"
benabik Wow. I'm actually getting 13/1 at the moment. That'll change when other people are home though. 20:37
jnthn lives in a relatively major city in Sweden, which measn 80,000 people :P
*means
I'm 10 mins by train from the 3rd biggest city...which has a massive population of just under 300,000 :) 20:38
20:38 rindolf left
fieldsa jnthn: hello from Canada, lots of space here 20:53
sorear benabik: Rochester?
benabik sorear: yupyup
20:54 daxim left
sorear jnthn: how is life in sweden these days? the US news media wants me to think that you've experienced the Apocalypse due to the greek situation 20:55
jnthn sorear: Well, for one Sweden isn't actually in the Eurozone. EU, but not the single currency.
sorear: That said, the Eurozone is a major trading partner, so it is having an impact. 20:56
(e.g. interest rates cut today)
20:56 drbean left
jnthn Most unusual thing that's happened this month is that all of my trains have run on time, but I'm not sure that's a sign of the Appocalypse. ;) 20:57
20:57 brno0 left
jnthn fieldsa: Plenty here too - if you go to the north of Sweden. ;-) 20:57
fieldsa: I'd love to visit Canada though. Please host a YAPC::NA. ;-) 20:58
fieldsa jnthn: Ottawa.pm meets this evening, I can propose it to them. 20:59
jnthn fieldsa: Are there any Perl workshops in Canada? :) 21:00
colomon is now dreaming about having YAPC::NA in St. John's.... 21:01
sorear sweet, CORE built without _error_ 21:02
21:03 drbean joined 21:04 cooper joined
fieldsa jnthn: I would suggest they are, but I don't think they get publication, probably in Vancouver,BC. YAPC::NA was held in Ottawa 2004 and in Toronto 2005, where atrj presented on pugs 21:04
21:04 ksi joined
jnthn fieldsa: Yeah, I'm aware it's happened...it's before I was really involved with Perl things though. 21:05
fieldsa I personally think the work in rakudo is great, I want to see community involvement 21:06
colomon is very fond of Toronto too, mind you. 21:07
sorear needs to get out more :|
jnthn sorear: Careful, I went out with $dayjob last night and $friend tonight and got no hacking done at all ;) 21:08
jnthn realizes he really should get on with packing stuff, given he has a flight tomorrow morning.
sorear family back in the UK? 21:09
jnthn sorear: Yeah.
colomon sorear: Toronto's the closest awesome city to us, and so I've spent a good bit of time there over the years, starting with train trips when I was little.
mikemol colomon: Hm. Toronto is 'close' in some sense? Where are you? 21:14
Toronto is only twice as far from me as Detroit is. 21:15
colomon Michigan. 21:16
sorear closest awesome city to me is SF, which is still a 10 hour drive, and I'm only ever there to meet family :|
colomon Google says it's 5.5 hours to Toronto from here, and this is the furtherest I've ever lived from the big TO. 21:18
21:23 bluescreen10 left
dalek ecza: 9d8f021 | sorear++ | src/NieczaActions.pm6:
More random fixes, now can compile CORE without warning or error
 21:25
21:29 pkkm_ left
mikemol colomon: Seriously? I'm over in Grand Rapids. 21:33
colomon mikemol: Midland
mikemol Heh. Neat.
You should do a presentation in front of the Ann Arbor Perl Mongers group. :) 21:35
21:35 bluescreen10 joined
dalek ecza: b353fb6 | sorear++ | src/NieczaFrontendSTD.pm6:
One more small bootstrap fix
 21:43
21:43 mj41 joined
colomon mikemol: the only contact I ever had with them was going to one of Damian's presentations on Perl 6 back in 2004 or so. It was what got me interested in p6. 21:48
mikemol colomon: Just about the only contact I've had with them was when I presented on Rosetta code. LiquidWeb wanted to show off their new datacenter and that they were a Perl shop, and I heard the AAPM meeting in Lansing via the Grand Rapids Linux Users Group. 21:50
Though I'm still on their mailing list. They're wanting for topics for regular meetings. Same as anyone these days, I guess; GRLUG shifted to an informal structure a couple years ago, and it looks like AAPL is moving in the same direction. 21:51
21:54 kaare_ left
colomon I'm not really sure why I never got more involved with AA.PM, as I lived there from 2001-2010 and had been using Perl since well before that. 21:55
mikemol s/AAPL/AA.PM/ 21:56
I could have sworn I made that correction once already.
21:56 zby_home joined
benabik One of the last messages on the Rochester.PM list is "Perl 5.8.0 has been released!" 21:56
colomon mikemol: you were using them both, I just grabbed one. ;) 21:57
sorear was at SanDiego.pm on Thursday 22:01
:| current niecza can compile itself. the new version, however, is unusably slow 22:17
10 minutes to get a few % into NieczaActions.pm6
lue What is meant by "lexical imports" in topics-brainstorming for the 2011 advent calendar? 22:21
22:21 snearch joined 22:24 drbean left 22:31 drbean joined
sorear { use Module; exports available here; } but not here; 22:33
22:34 zby_home left
sorear hey, it seems they made the mono profiler suck less since the last time I tried it. 22:36
lue Ah. 22:37
22:41 dch joined 22:59 thou left 23:02 mj41 left 23:05 ksi left 23:07 snearch left
tadzik phenny: ask jnthn if I understand it correcly, at this moment two precompiled modules with different source files are _never_ binary compatibile? 23:13
phenny tadzik: I'll pass that on when jnthn is around.
23:18 awoodland left 23:24 packetkn_ joined 23:25 Chillance left, aindilis left 23:26 cotto joined, packetknife left, aindilis joined 23:32 awoodland joined
jnthn tadzik: What do you mean by "different source files"? 23:32
phenny jnthn: 23:13Z <tadzik> ask jnthn if I understand it correcly, at this moment two precompiled modules with different source files are _never_ binary compatibile?
jnthn tadzik: If you mean different content, then your statement is correct.
tadzik well, let's optimize JSON::Tiny a bit, w/o changing functionality 23:33
jnthn I've pretty much concluded Rakudo itself is going to have to maintain the precompilation cache.
tadzik *relief* :) 23:34
jnthn It's becoming less effort for me to implement that than it is to deal with people being constanly surprised by the thing working as I designed it to... :)
tadzik I figured that in the current state, Panda would have to be very careful not to break itself during updates
:D
I'm sorry for constantly nagging you about that :) 23:35
jnthn tadzik: Well, in theory the module loader is best placed to know the dependencies and do the Right Thing.
tadzik: And I woulda somewhat preferred pre-compilation "magic" not to be in the Rakudo core. 23:36
Either way, we'll get problems, people unhappy with the way its working, etc.
tadzik possibly
well, I'm probably the only unhappy user right now, so maybe nevermind :P 23:37
jnthn already busted one production system today because it had if-you're-lucky-enough-to-make-just-little-changes-binary-compat
23:38 awoodland left
jnthn (My change was just a little too un-little, it turned out :)) 23:38
tadzik :D
jnthn Anyway, I'll see what I can do.
sorear ow! LTM shot up from <10% of runtime to 75% 23:39
profiling++
jnthn eek! 23:40
23:46 Timbus left, kfo left 23:51 Timbus joined, kfo joined 23:52 awoodland joined 23:54 awoodland_ joined 23:57 awoodland left, awoodland_ left 23:59 risou is now known as risou_awy

Powered by Cro and the Raku® Programming Language.

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

Thank you!