»ö« | perl6-projects.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot: perl6: say 3;' | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by moritz_ on 27 June 2009.
00:00 Chillance left 00:01 yahooooo left 00:03 ssm joined 00:10 skids left, Hale-Bopp left, gfldex left, simcop2387 left, PZt left, asciiville left, cognominal_ left, FurnaceBoy left, Ehtyar left, pugs_svn left, unitxt left, frew left, elmex left, jauaor left, dLux left, synth left, drbean left, ibrown left, PacoLinux left, eiro left, meppl left, bigpresh_ left 00:11 PerlJam joined, elmex joined, PZt joined, jauaor joined, asciiville joined, eiro joined, Hale-Bopp joined, gfldex joined, simcop2387 joined, skids joined, cognominal_ joined, dLux joined, FurnaceBoy joined, meppl joined, unitxt joined, synth joined, Ehtyar joined, pugs_svn joined, frew joined, bigpresh_ joined, drbean joined, ibrown joined, PacoLinux joined, jiing joined, patmat joined, tarbo2 joined, buu joined, IRSeekBot joined, jjore joined, RonOreck joined, Lorn joined, mj41 joined, StephenPollei joined, he_ joined, Gothmog_ joined, edenc joined, shachaf joined, ascent_ joined, jiing_ joined, yves joined, hexmode joined, scrottie joined, ashizawa joined, c9s joined, cosimo joined, kcwu joined, spinclad joined, rhr joined, krunen joined, BinGOs joined, felipe joined, kolibrie joined, irc.freenode.net sets mode: +o PerlJam, nnunley joined, Aisling joined, stepnem left 00:12 bigpresh_ left, bigpresh_ joined, yahooooo joined, bigpresh_ is now known as Guest24823 00:18 kborer left 00:19 stepnem joined 00:24 ihrd joined, ihrd left 00:29 meppl left 00:33 frew_ joined 00:35 cspencer left 00:39 dukeleto joined 00:43 cspencer joined 00:44 dukeleto_ joined, cspencer left 00:59 FurnaceBoy left 01:02 meppl joined, meppl left 01:08 dukeleto left 01:28 Whiteknight left 01:47 justatheory left 01:49 justatheory joined 01:52 dukeleto joined, dukeleto left, dukeleto_ left 02:03 kate21de left 02:14 justatheory left 02:27 eternaleye left, kane___ joined 02:32 simcop2387 left, colomon joined, kane___ left 02:43 ihrd joined, simcop2387 joined, kane_ left
colomon I love Perl6! 02:43
sjohnson me too 02:50
TimToady it's okay in spots 02:52
colomon Did what feels like my first real programming in it today -- modifying a Perl 5 program to be Perl 6ish.
It's beautiful.
TimToady I'll settle for cute and adorable...
colomon I'm not sure cute really fits, actually. 02:53
It's all elegant and such.
02:54 azawawi joined
azawawi hi 02:54
sjohnson some cute things about Perl 02:56
colomon Time for bed. Just had to get that out of my system. Thanks, all.
02:56 colomon left
sjohnson $_ is cute 02:57
02:57 justatheory joined 02:59 azawawi left 03:01 kane_ joined
pugs_svn r27588 | lwall++ | [camelia] put back TM 03:09
sjohnson （　｀ー´） 03:14
03:17 Limbic_Region left 03:20 donaldh left, donaldh joined 03:23 ihrd left 03:37 frew_ left 03:39 skids left 03:40 justatheory left 03:52 Khisanth left, Khisanth joined 03:53 justatheory_ joined 03:57 szabgab joined 04:15 synth left 04:17 synth joined 04:24 FCO joined 04:29 araujo left 04:30 SmokeMachine left, jauaor left 04:41 eternaleye joined 04:42 frew|work joined 04:45 alester joined 04:47 frew_ joined 04:49 szabgab left 04:50 szabgab joined 05:01 szabgab left, szabgab joined 05:05 justatheory_ is now known as justatheory 05:14 ironcamel1 joined
szabgab ironcamel1, why 1 ? 05:15
05:15 DemoFreak joined
szabgab can't you use the same nick on two channels with the freenode web interface ? 05:16
ironcamel1 this account is view the web client... my ironcamel account i use a client that i wrote
as a plugin for vim
szabgab oh I see
ironcamel1 i keep 2 clients running for error checking/debugging etc
szabgab I have enough errors in one client :-) 05:17
ironcamel1 szabgab, one suggestion for your website is don't make users have to register to leave a comment. just my opinion. 05:18
szabgab anyway we still have to see if the Perl 6 regexes are going to become ugly because of the extensive escaping or not
ironcamel1 well, of course they will be
they will be full of backslashes everywhere 05:19
szabgab regarding the comment system, if that can be done with the disqus thhing then I can try to do that
ironcamel1 unless you can do something clever with the literal q{} strings in the regexes
05:21 ironcamel joined 05:28 kate21de joined 05:32 szabgab left 05:47 sparc joined, riffraff joined 05:48 frew|work left 05:49 frew_ left
Khisanth I guess parsing html will look ugly ... 05:50
ironcamel1 yes indeed 05:57
06:02 frew_ joined 06:04 finanalyst joined 06:07 frew|work joined 06:11 rfordinal3643 left, rfordinal joined
Khisanth well at least it matches 06:15
06:19 tann joined 06:30 Su-Shee joined
Su-Shee good morning. 06:30
tann 'moring, pretty pirl 06:31
06:31 rfordinal left 06:32 justatheory left 06:35 barney joined 06:39 kane_ left, japhb joined 06:44 szabgab joined 06:45 rgrau joined 06:49 ihrd joined, ihrd left
moritz_ std: sub a { }; say a ~~ 4 06:53
p6eval std 27588: OUTPUT«##### PARSE FAILED #####␤Preceding operator expects term, but found infix ~~ instead at /tmp/CpslxfakGq line 1:␤------> sub a { }; say a ~~ 4␤FAILED 00:02 36m␤»
07:00 rfordinal joined 07:03 rfordinal left
Matt-W Morning 07:07
sjohnson hello
moritz_ oh hai 07:08
sjohnson does anyone here hate programming in Javascript, PHP, CSS, and HTML all in one? 07:10
Su-Shee sjohnson: I don't besides PHP.
moritz_ sjohnson: yes ;-)
Su-Shee actually, I really like the combi js, css and html/xul :) 07:11
07:11 synthEEEE joined
sjohnson moritz_: me too 07:11
i think i'm gonna have to quit my job
this is getting far too boring and painstaking
Su-Shee sjohnson: you forgot the additional window with the database/sql stuff. ;) 07:12
sjohnson yes, i did forget that :) 07:13
i become very happy programming Perl scripts at work to do routine things
i almost jump at the opportunity 07:14
contrasted with the unbelievable pain of coding in PHP
... PHP and friends, i mean
Su-Shee sjohnson: until today I managed to avoid to get any php job but the one I've tried once and hated.
sjohnson Su-Shee: this may seem like a no brainer, but you like Perl, right? 07:15
Su-Shee sjohnson: totally.
sjohnson me too 07:16
Su-Shee sjohnson: I had a perl crisis a few years ago and let myself be seduced by python and a little ruby, but went back. 07:17
sjohnson please tell me how that went
as i also had a crisis
but came back once i found that Perl 6 does all that Perl 5 didn't do
jeekobu I'm in the middle of such a crisis I guess 07:18
For my research scripts anyway. Python has easy memoization and built-in big int support, P5 doesn't
Su-Shee well, from the outside it looked like perl wouldn't make it into more contemporary programming, so I checked other things out. then came a) perl6 and b) cool new shiny things in perl5 and I jumped back immediately.
sjohnson jeekobu: do you know if Perl 6 can handle it? 07:19
Su-Shee jeekobu: check higher order perl, there's plenty about memoization.
sjohnson Su-Shee: i checked out ruby too
i think David Flanagan's book is confusing as hell
obviously will get 5 stars on amazon, but an example of the worst kind of technical manual i have ever read
the Camel book is so much nicer
and a hell of a lot funnier to read
Su-Shee sjohnson: it's mostly that I read many amazing details about perl 5 (like HOP) which convinced me even more. I got in touch with perl's sexiness again.
07:20 donaldh left
Su-Shee sjohnson: and if you check out projects like moose for example - it's a-ma-zing how flexible perl is. 07:20
jeekobu sjohnson: You mean the current implementations? I don't know
07:20 donaldh joined
jeekobu sjohnson: Sort of why I drifted back to checking out P6; it's been a few years 07:20
Su-Shee sjohnson: and there's nothing like CPAN.
jeekobu sjohnson: OTOH I'm vary wary of jumping on a still-rapidly-evolving platform. That's been a huge part of Python's problem. (And various other software platforms.) 07:21
sjohnson righto 07:22
my biggest gripe in Perl 5 is no trim function
Su-Shee trim?
07:22 synth left
bionoid sjohnson: for real? it's like a 0,2 liner 07:22
Su-Shee *hihi* 0.2 liner :)) 07:23
07:23 dakkar joined
sjohnson yeah, having to reinvent the wheel for the 50 perl scripts i have that use it 07:23
gets annoying fast
Matt-W yeah that bugs me too
It's so common it just feels like it should be a builtin
sjohnson Matt-W, i couldn't agree more 07:24
having to refine my own trim function every single time i write a script that uses it, or install String::Strip with CPAN...
on every system I migrate to.. 07:25
moritz_ maybe you could sneak it into Scalar::Util, which is a core module?
sjohnson i would like to, but how sneaky would i have to be?
i want to do whatever it takes to get it backported into Perl 5
i could live with a core module 07:26
moritz_ no idea really
sjohnson until Perl 6 comes out
Su-Shee (oooh, please let me get the job where the ceo says "oh, yes, let's have an eye on perl 6, we're commited to use perl the next decade..") 07:27
sjohnson moritz_: maybe i could get in touch with the Core Module author?
i convinced the ANSIColor author to inclide a stripColors function 07:28
so maybe it might be all that is needed
include*
moritz_ Su-Shee: when will that be decided?
sjohnson: it's dual lived, so you'll find the author by a quick CPAN search
Su-Shee moritz_: they said, they'll contact me "end of the week - monday the latest.."
moritz_ Su-Shee: then I wish you the best of luck 07:29
sjohnson moritz_: i am wondering if you think it might work, not so much who the author is
if you have faith, then i will give it my best
moritz_ sjohnson: it's worth a try.
I've never tried to get something into a core module (except documentation), so I really can't tell
sjohnson perhaps if i make a good case for it
about how frustarting it is getting an @INC error 07:30
on migrating systems
when emergencies happen
... based on a true story!
07:30 alester left
sjohnson perhaps matt-w might be kind enough to let me use his quote 07:31
:3
07:33 asciiville left
sjohnson *puppy dog eyes* 07:33
Su-Shee sjohnson: just find 2 good arguments why it's missing and offer the already written code. 07:34
07:37 payload joined
Matt-W My quote? 07:43
Su-Shee It's so common it just feels like it should be a builtin 07:44
that one.
Matt-W Oh 07:47
yes
sure
07:52 beggars joined 07:56 pmurias joined 08:02 sri_kraih joined 08:05 payload left 08:11 tann left 08:16 pmurias_ joined 08:18 unitxt left 08:32 agentzh1 left 08:33 agentzh joined 08:37 payload joined 08:46 payload left 08:55 sri_kraih left 08:56 Ehtyar left, sri_kraih joined 08:59 kate21de1 joined 09:07 zamolxes joined, synthEEEE left 09:11 athomason left 09:12 payload joined 09:13 payload left, payload joined, DemoFreak left 09:14 athomason joined
szabgab calling die() now gives me a Null PMC access in get_pmc_keyed() 09:14
moritz_ rakudo: die('a')
p6eval rakudo 9a7a1d: OUTPUT«a␤Null PMC access in get_pmc_keyed()␤current instr.: 'perl6;Perl6;Compiler;eval' pc 289208 (src/gen_actions.pir:24161) (src/gen_setting.pm:3340)␤»
moritz_ bah. 09:15
rakudo: 1/0
p6eval rakudo 9a7a1d: OUTPUT«Divide by zero␤»
moritz_ bah, no line number
09:16 kate21de left
Matt-W who broked it? 09:23
moritz_ probably the parrot folks
Matt-W entirely possible 09:25
I do look forward to the day when it's a bit more of a stable footing
Although I'm in no rush, because Parrot can become much more awesome before then
09:26 szabgab left, PZt left
dalek kudo: 3d5f23c | moritz++ | docs/architecture.svg:
[docs] shiny chart of Rakudo's architecture (svg)
 09:29
Matt-W moritz_: I liked that architecture chart 09:31
09:32 szabgab joined 09:33 explorer joined
spinclad std: 2++4 09:38
p6eval std 27588: OUTPUT«##### PARSE FAILED #####␤Syntax error (two terms in a row?) at /tmp/TYzGLCkRqI line 1:␤------> 2++4␤ expecting any of:␤ POST␤ infix or meta-infix␤ infix stopper␤ postfix␤ postfix_prefix_meta_operator␤ standard stopper␤ statement
..modifier loop␤ terminator␤ whitespac…
spinclad (goes on backlogging) 09:39
09:41 payload left, payload joined 09:52 szabgab left
pugs_svn r27589 | hinrik++ | Move perlintro.pod6 to perlintro.pod 09:55
09:55 PZt joined
pugs_svn r27590 | hinrik++ | [docs] Fix some formattting in perlintro.pod, replace Pugs with Rakudo 10:06
r27591 | hinrik++ | [docs] perlintro.pod: Use semantic Pod blocks, move Author section to the end, more reformatting 10:11
10:17 frew_ left, frew|work left 10:19 pmurias_ left
pmurias moritz_: re inclusion into Scalar::Util doesn't it contain magical functions? 10:21
moritz_ pmurias: dunno, the name doesn't imply that
pmurias i just read through the function list and most of the functions seemed deeply tied into the perl5 language/implementation 10:23
moritz_ C<Scalar::Util> contains a selection of subroutines that people have 10:24
expressed would be nice to have in the perl core, but the usage would
not really be high enough to warrant the use of a keyword, and the size
so small such that being individual extensions would be wasteful.
10:25 szabgab joined
moritz_ seems to fit, although no black magic is required for trim() 10:25
szabgab std: 1+1
p6eval std 27591: OUTPUT«ok 00:02 36m␤»
10:25 szabgab left
literal hm, perldoc.perl.org has a new design 10:26
moritz_ wohoo
literal very neat 10:27
moritz_ imho a bit dark, but very nice
it's a strange irony that the one perl site which a very decent layout already got a new one, while others could use it much more 10:31
literal heh, yeah 10:32
perlmonks and use.perl have awful designs 10:33
pugs_svn r27592 | hinrik++ | [docs] perlintro.pod: Use K<> and =input blocks instead of C<> and =code where appropriate 10:34
10:37 clintongormley joined 10:40 ihrd joined, szabgab joined 10:41 ihrd left
pugs_svn r27593 | hinrik++ | [docs] perlintro.pod: Reformat the email L<>inks 10:43
Matt-W oh my 10:44
that's weird
going to take a while to get used to the new perldo
although the search engine isn't finding me anything at all 10:48
weird
and some nasty scrolling redraw effects too. No, I think I preferred the old perldoc. 10:49
szabgab rakudo: 1+1 10:50
p6eval rakudo 3d5f23: ( no output )
szabgab rakudo: 'aab' ~~ m/ a | aa /; say $/ 10:51
p6eval rakudo 3d5f23: OUTPUT«a␤»
moritz_ szabgab: how stand-alone is the stand-alone Padre download? do you still need a strawberry perl installed?
szabgab shouldn't that be aa ?
moritz_ longest token matching is NYI
pmichaud is working on it this summer
10:51 DemoFreak joined
szabgab the stand-alone includes Strawberry 10:52
moritz_ ok, thanks
szabgab actually it currently installs in c:\strawberry
moritz_ my girlfriend wants to learn perl
literal how big is the download?
moritz_ and has no internet connection, and only windows
so I looked around what to download for her
szabgab so how are the test starting at line 55 in pugs/t/spec/S05-metasyntax/longest-alternative.t passing ? 10:53
moritz_ takes a look 10:56
szabgab it is still a bit far from being really beginner friendly - the built in help is not that good yet
it is about 14 mb
oh I guess they are skipped in that external test skipper
not by fudge
I keep forgetting about that
moritz_ only a selection of spectests is run, yes
szabgab rakudo: 'aab' ~~ m/ a || aa /; say $/
p6eval rakudo 3d5f23: OUTPUT«a␤» 10:57
szabgab evalbot does not have a good day it seems
moritz_ she just needs an editor with syntax hilighting and a button that runs the script
hm, why?
10:57 M_o_C joined
szabgab oh and in the stand-alone padre there still is no icon to launch padre :-( 10:57
but I hope there will be one by the birthday 10:58
moritz_ but a simple link on the desktop will do for starting it, no?
szabgab moritz_, why what ?
moritz_ 12:57 < szabgab> evalbot does not have a good day it seems 10:59
szabgab yes it will do
earlier it seemed to be dead and now it took a long time - though eventually answered 11:01
11:05 colomon joined 11:20 donaldh left, donaldh joined 11:29 synth joined 11:36 meppl joined 11:41 ejs joined
pmurias mildew: say "alive!?" 11:42
p6eval mildew: OUTPUT«alive!?␤»
11:43 s1n joined
szabgab how do you explain $/ is that a magic variable that is both a scalar and an array ? 11:45
why is the array part not called @/ ?
moritz_ it's an object.
and @ really only implies the Positional role, % implies the Mapping role 11:46
and $/ does both
szabgab oh so there is an @/ ? 11:47
moritz_ no
@($/) or $/.list
lambdabot Unknown command, try @list
moritz_ lambdabot: shut up
szabgab I think these roles and stuff are still way over my head
pmurias szabgab: $/ is not really magical 11:50
moritz_ for a Perl 5 programmer it is ;-)
pmurias it just gets set after you do a match
szabgab: when you keep an array in $foo "$foo = [1,2,3]" is $foo both a scalar and an array? 11:52
szabgab I think I have to go back to square one and try to understand why do we need to have arrays that have a $ as a sigil
rakudo: my $x = (1, 2, 3); say $x.WHAT; say $x[0]; 11:53
p6eval rakudo 3d5f23: OUTPUT«Array()␤1␤»
moritz_ actually it should be a Capture
pmurias szabgab: they are array refs from perl5
szabgab rakudo: my $x = [1, 2, 3]; say $x.WHAT; say $x[0];
moritz_ but $x = [1, 2, 3]; would create an Array
p6eval rakudo 3d5f23: OUTPUT«Array()␤1␤»
szabgab pmurias, yes I think I understand that but why do we need them ? 11:54
pmurias we don't
szabgab I thought the whole point was automatic reference/dereference
and I was hoping to get a good explanation 11:55
or that they will be removed :-) 11:56
pmurias arrays with the @ sigil could be removed without much harm
moritz_
.oO( then we'd get PHP )
 11:57
szabgab I meant to remove arrays with $
moritz_ and give up polymorphism?
szabgab I don't understand how are they consistent with the design
moritz_ I mean you couldn't create a variable that either holds a hash or an array 11:58
pmurias szabgab: Arrays are objects, you can hold objects in scalars
the meaning of a '@' is this container does the Positional role
szabgab which means ? 11:59
moritz_ that you can call .[] on it
szabgab that you can say @x[3] ?
moritz_ right 12:00
%hash[3] will fail.
pmurias in $x = [1,2,3] $x is a Scalar which contains an Array, in @x = 1,2,3 @x is an Array
szabgab but you can do that on $x as well if it holds an array
12:00 icwiener joined
moritz_ right, but only then 12:00
think of @ as an abbreviation of a type constraint
pmurias szabgab: there is a difference between holding and containing 12:01
moritz_ (at lesat for now)
12:01 eiro left, eiro joined
szabgab so you mean @x can hold only arrays while $x can hold both scalars, arrays and hashes 12:01
pmurias you need to use VAR($x) to get at the $x variable
$x can hold anything
szabgab ok, what about $y ? 12:02
;-)
pmurias szabgab: do you see the difference between binding and assigment?
szabgab not really :-) 12:05
pmurias it's obvious once you try to implement binding & assignment ;) 12:06
moritz_ my ($x, $y); $x := $y; # now if you assign to $x, the value is also in $y, and vice versa 12:07
pmurias binding is like *foo = *bar; assignment is like $foo = $bar; 12:08
Matt-W reads backlog 12:10
Ah yes that confused me to start with
szabgab back later & 12:15
12:15 szabgab left, Jedai joined 12:24 ejs1 joined 12:35 ejs left 12:42 riffraff left
moritz_ @seen masak 12:43
lambdabot I saw masak leaving #perl6 1d 15h 36m 44s ago, and .
moritz_ would like to know how many people follow planetsix.perl.org 12:46
pmurias moritz_: re %hash[3] would fail, wouldn't %hash := \(...);%hash[3] work?
moritz_ pmurias: maybe 12:47
12:48 ejs1 left
moritz_ phenny: tell szabgab I notice that all your Perl 6 examples still use parens after 'if', ie 'if ($str ~~ m/\=/) {' - is there a good reason for that? or plain old habit? 12:48
phenny moritz_: I'll pass that on when szabgab is around.
12:49 takadonet joined
takadonet morning all 12:49
Matt-W moritz_: I follow planetsix.perl.org 12:50
takadonet: hi 12:51
moritz_ Matt-W: I guess the number is in the order of 1k or so, an IRC poll won't bring a reasonable result for that :-) 12:54
Matt-W moritz_: no probably not 12:55
13:00 explorer left 13:01 rgrau left 13:08 drbean left 13:10 colomon left, rjh left 13:11 rjh joined 13:25 drbean joined 13:28 szabgab joined 13:40 skids joined 13:43 dukeleto joined 13:58 colomon joined 14:01 PerlPilot joined 14:06 jan_ left 14:07 rfordinal joined 14:08 unitxt joined 14:15 PerlPilot left 14:18 lichtkind joined 14:24 nihiliad joined 14:26 rfordinal left 14:28 icwiener left, jlaire joined 14:32 araujo joined 14:35 c1sung left, c1sung joined
araujo morning 14:35
lambdabot araujo: You have 1 new message. '/msg lambdabot @messages' to read it.
Matt-W afternoon 14:37
14:39 colomon left 15:02 beggars left
lisppaste3 Lorn pasted "recompiling re-smop" at paste.lisp.org/display/83747 15:03
Lorn pmurias: help ^^
15:03 payload left
Lorn pmurias: every 'svn up' i got a new error :P 15:03
15:05 kane_ joined, payload joined 15:11 Su-Shee left 15:20 donaldh left, donaldh joined
cj [Coke]: enough out of you! 15:22
15:23 hercynium joined 15:25 nihiliad left 15:26 payload left 15:31 jan_ joined
pugs_svn r27594 | pmurias++ | [re-smop] unbreak build 15:32
r27595 | pmurias++ | [re-smop] disable SMOP_PROFILE and SMOP_MOLD_DEBUG
r27596 | pmurias++ | [re-smop] remove debuging statements from mildew
r27597 | pmurias++ | [re-smop] make tools/ri support %yeast 15:33
pmurias Lorn: there's a bug at runtime which causes smop to loop infitly
[Coke] whistles innocently at cj. 15:34
pmurias cycling& 15:35
15:43 KyleHa joined
pmichaud Good morning, #perl6 15:43
15:46 KyleHa1 joined
Lorn pmurias: Good afternoon, from Brazil 15:46
KyleHa1 constant $foo = 22; lives_ok { $foo *= 1 }, 'identity'; # pmichaud, what do you think?
Lorn pmurias: works fine now, re-smop and re-mildew
15:48 mhsparks joined, mhsparks left
pmichaud KyleHa1: that looks odd to me 15:49
I'd say it's not a valid spectest
KyleHa1 Okeydoke.
15:49 dakkar left 15:52 sri_kraih left
pugs_svn r27598 | kyle++ | [t/spec] Test for RT #64522 15:52
r27599 | kyle++ | [t/spec] fix the plan I forgot
15:53 sri_kraih joined
pugs_svn r27600 | kyle++ | [t/spec] Can one "modify" a constant to its original value? I assume so here, but maybe s/lives_ok/dies_ok/g 15:53
r27601 | kyle++ | [t/spec] fudge constant.t for Rakudo
r27602 | kyle++ | [t/spec] Nope, can't assign to a constant, even if it doesn't change it.
KyleHa1 This concludes the fruits of my train ride. Time to work... 15:54
15:55 kst left 15:59 ispy_ joined, molaf joined 16:02 kst joined 16:03 iblechbot joined 16:04 payload joined 16:08 KyleHa left, zamolxes left 16:12 nihiliad joined 16:15 donaldh left 16:22 lichtkind_ joined 16:25 KyleHa1 left 16:31 FCO left 16:32 lichtkind left, FCO joined 16:35 agentzh left 16:37 lichtkind_ is now known as lichtkind 16:38 icwiener joined 16:39 szabgab left, tann joined 16:47 cdarroch joined 16:48 FurnaceBoy joined 16:53 KyleHa joined 16:56 DeathAnchor joined 16:57 Patterner left 16:58 tann left 17:02 Psyche^ joined, Psyche^ is now known as Patterner 17:06 barney left 17:09 DeathAnchor left, finanalyst left
pmurias Lorn: re works fine, lucky you re-mildew loops here 17:11
17:11 frew_ joined
Lorn pmurias: no, i did not get any re-mildew loops, all tests ran ok in re-mildew 17:17
17:22 kane_ left 17:27 orafu joined 17:46 pmurias left, dukeleto_ joined, pmurias joined 17:47 tann joined 17:51 kane_ joined 18:00 hercynium left 18:02 dukeleto left 18:14 szabgab joined 18:22 sparc left 18:25 ispy_1 joined 18:27 unitxt left
dalek kudo: faf91f9 | pmichaud++ | docs/spectest-progress.csv:
spectest-progress.csv update: 415 files, 11802 passing, 2 failing

   S12-attributes/instance.rakudo 121 - Can initialize one attribute based on another (1)
   S12-attributes/instance.rakudo 122 - Can initialize one attribute based on another (2)		 18:28
18:30 clintongormley left 18:33 pmurias left 18:35 ispy_ left 18:37 tann left 18:40 justatheory joined
pmichaud TimToady: STD.pm:4479 has :my $sigspace is context<rw> = $*sigspace // 0; 18:40
doesn't that need to be CALLER::<$*sigspace> or something like that? 18:41
18:42 explorer joined 18:43 ispy_1 left 18:49 szabgab left 18:53 lichtkind left 18:55 Guest98203 joined 19:01 Guest98203 left, olafk joined
TimToady those were probably all $+sigspace at the time, which was (at the time) defined to imply one level of caller 19:04
it currently works because of the stupid translation that gimme5 does, and the resulting support from p5 semantics
so yes, but when I fix that I'll also have to fix gimme5 :)
pmichaud in the final version, is it likely to become CALLER::<$*sigspace>, or will we have a notation that implies one level of caller?
okay. 19:05
I guess the notation that implies one level of caller is CALLER:: :-)
TimToady I think it's better if it starts looking in this space
pmichaud oh, I agree
I'm looking to start refactoring PGE to be more STD-like and ran into that. 19:06
TimToady looks like there are a number of places that are wrong that way
pmichaud yes
it's okay if they're wrong -- perhaps we just want to XXX them for now 19:07
I don't mind if STD.pm is wrong in places, as long as we know where they are :-)
migrating PGE to handle some of the new semantics is going to be interesting. I'm starting to think it's easier to start over than to migrate what I already have. 19:08
(because then I don't have to deal with Parrot deprecation issues)
I wonder how hard it would be to write the operator precedence parser in NQP 19:09
TimToady well, Cursor does most of the regex semantics, but not all
pmichaud right
TimToady doesn't try to do & or && yet
pmichaud that's one of the significant differences from what PGE does now
so starting over seems easier
TimToady and it only fakes LTM with a trie 19:10
pmichaud is the faking not likely to be good enough?
TimToady oh, it's good enough, unless you mean "fast enoguh"
*enough
pmichaud got it.
grammars are essentially subclasses of Cursor? 19:11
TimToady yes
pmichaud okay, good. 19:12
TimToady the current cursor encapsulates the current Match state as well as the current language 19:13
in theory you could freeze a cursor to have everything you need to provide the setting for an eval 19:14
pmichaud okay. I like this design.
TimToady but currently the p5 implementation keeps a few things in localized globals just to keep things from getting too slow
pmichaud it fills in a lot of the "holes" that I've been struggling with for years.
TimToady and in theory each cursor is immutalbe, and you just continually derive new ones 19:15
but for ratchet code you can actually reuse them mostly
pmichaud I hadn't caught that aspect of it.... that's helpful to know. 19:16
19:16 unitxt joined
TimToady but immutable cursors plus lazy lists is a brain-dead easy way to do backtracking 19:16
the interesting thing is to see the translation that gimme5 does to STD.pm5
where it needs to do lazymap vs where it can just get away with conditionals 19:17
pmichaud PGE's model had been to compile regexes into explicit subroutines to do the matching.... is that consistent with STD/gimme5's approach?
TimToady they're always methods in STD
pmichaud where the methods are a compiled form of the regex?
TimToady always methods on the current cursor
yes 19:18
pmichaud fantastic
TimToady again, you can see how they translate to p5 semantics in STD.pm5
pmichaud for the past week or so I've been having trouble getting motivated to write Rakudo code... perhaps that's an indication I need to be doing regex stuff
TimToady and note that it's hardly using p5 regexen at all
if viv were a bit further along, we could actually emit the rakudo code from it 19:19
pmichaud every time I sit down to make a change in Rakudo, I think "yeah, but I'm going to throw it out once we start doing std-like stuff"
TimToady but I haven't quite got it to the place where it can replace gimme5
pmichaud on another topic, while you're here
did you see the discussion yesterday regarding binding?
TimToady yes, not really sure what to think about it yet
pmichaud okay. 19:20
I'll just re-register the notion that := having semantics similar to signature bind feels more "correct" the more I think of it.
that's the way I'm going to head for now, at any rate :-)
in particular, it matches more closely with the idea that my @a := $v; requires that $v does positional 19:21
TimToady yes, proxies enforcing the additional constraints
pmichaud correct
TimToady and for unadorned @a you presumably don't need the proxy
er 19:22
unadorned $a
since @a is adorned :)
pmichaud at any rate, that looks dead-simple to implement in rakudo, and resolves several of the binding issues I had asked about at NPW
it also unifies our binding semantics some what, which I think will likely simplify a few other sections of code that are kinda weirdish.
TimToady so we can get pure ref aliasing still as long as you do it with 'my $a :=' 19:23
pmichaud that works for me.
anyway, we'll head that direction in Rakudo, and if things change later I don't think it'll be a killer for us.
TimToady still think we need a way of telling primitive binding to allow trailing stuff 19:24
so that for and map don't have to count args
pmichaud yes, we need that also somehow
TimToady maybe it's just an out parameter where it can put the remainder
basically an implied capture parameter at the end 19:25
(or parcel)
pmichaud okay, thanks for the quick tips -- they give me some solid directions to move in 19:27
19:36 zamolxes joined 19:40 nbrown left 19:41 takadonet left 19:43 justatheory left 19:47 KyleHa left 19:50 nbrown joined 19:52 olafk left 19:53 justatheory joined 19:56 szabgab joined 19:57 sri_kraih_ joined 19:59 alester joined 20:07 M_o_C left 20:13 sri_kraih left 20:14 baest joined 20:17 icwiener_ joined 20:21 dukeleto_ left 20:22 nwc10 joined 20:32 icwiener left 20:35 molaf left 20:49 rfordinal joined 20:55 kate21de1 left 20:58 iblechbot left 21:09 kate21de joined 21:12 kate21de left 21:14 frew_ left 21:16 frew left
payload rakudo: 1.0.atan2(1.0) 21:23
p6eval rakudo faf91f: ( no output )
payload rakudo: say atan2(1.0, 1.0)
p6eval rakudo faf91f: OUTPUT«Could not find non-existent sub atan2␤»
payload why is atan2 only a method? 21:24
rakudo: say 1.atan2(1.0) 21:25
p6eval rakudo faf91f: OUTPUT«Method 'atan2' not found for invocant of class 'Int'␤»
21:28 dukeleto joined, skids left, dukeleto left 21:29 dukeleto joined 21:40 icwiener_ left 21:44 frew joined 21:46 davef joined 21:48 davef left 21:57 tarbo2_ joined 22:00 dukeleto left 22:03 tarbo2 left 22:16 Limbic_Region joined 22:26 payload left 22:30 alester left 22:33 sri_kraih_ left 22:45 payload joined, M_o_C joined 22:49 nihiliad left 23:01 explorer left 23:06 frew left 23:11 explorer joined
TimToady n 23:13
nm
sjohnson toady! 23:14
TimToady you were supposed to say np
sjohnson np 23:15
pugs_svn r27603 | lwall++ | [S06-signature/optional.t] nullterm allowed only after final comma 23:26
r27603 | lwall++ | also, eval_dies_ok would always have died anyway due to missing {...}
23:27 nwc10 left 23:30 tann joined
tann rakudo: my $x = "\tx\t\n"; $x.trim; say "[$x]"; 23:45
p6eval rakudo faf91f: OUTPUT«[ x ␤]␤»
TimToady rakudo: my $x = "\tx\t\n"; $x.=trim; say "[$x]"; 23:46
p6eval rakudo faf91f: OUTPUT«[x]␤»
TimToady rakudo: my $x = "\tx\t\n".trim; say "[$x]";
tann TimToady: will there be a left_trim/right_trim? 23:47
p6eval rakudo faf91f: OUTPUT«[x]␤»
23:47 frew joined
TimToady there is none specced currently 23:48
and the situation doesn't arise very often
tann neither is trim specced :)
TimToady and when it does s/\s+$// works
is so
S32-setting-library/Str.pod
23:50 charsbar_ left, charsbar joined
tann s32 .t file shows trim_start and trim_end but i guess they won't make it then 23:51
methinks, trim_start, trim_end are bad names as people would think the trimming starts and expect the trimming to end :) 23:54
23:54 |Jedai| joined, payload left

Powered by Cro and the Raku® Programming Language.

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

Thank you!