»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! | Rakudo Star Released!
Set by diakopter on 6 September 2010.
Tene masak: I work at imvu. We're hiring. :) 00:02
masak :D 00:03
diakopter masak: I think that was a year ago or more 00:05
before sara ford & sam ranji left 00:06
ramji
00:07 timbunce left
masak diakopter: I'm not familiar with those two names. 00:08
diakopter codeplex folks
open-source advocates
masak ok.
still feel I'm missing some context. 00:09
diakopter (I'm guessing they were his contacts)
00:09 kulp left
Tene We very much need another sysadmin on the ops team here, so if anyone's looking for work in sfbay, lemme know. 00:09
diakopter Tene: hey um u not fair; we need like 4 00:10
(halfway between sf & san jose)
Tene diakopter: we have 4 total right now.
diakopter $1x0k/year & great bonuses/benefits to the best-qualified 00:11
00:11 timbunce joined 00:12 donaldh left
TimToady there is a long history in English of referring to a set of requirements as either 'the specification' or 'the specifications'; requiring one or the other is mere shibbolethery, in my opinion 00:20
see for instance definition 2a of education.yahoo.com/reference/dicti...cification 00:21
or google for "the specs" and ignore the rock band :)
masak I see. then I'll stop kvetching about it.
diakopter keeps trying not to read shibbolethery as shibbolechery, but fails 00:22
TimToady a spec is a fractal group of self-similar specs, it the semantic problem...
*is
at least we can have a spec if we want, unlike a "semantic" :)
masak I got the feeling, up until I saw the repository name, that the community tradition was to refer to "the synopses" (plural) but "the spec" (singular). 00:23
00:23 mkramer joined, risou_ joined
sjohnson TimToady: maybe p6 should use a "plan" instead 00:25
00:26 XaRDaX left, risou left 00:28 bluescreen joined, bluescreen is now known as Guest3023, wamba left 00:29 mkramer left 00:34 timbunce left
colomon masak: have you tested rakudo with the latest parrot? 00:34
masak colomon: I believe so. will perl Configure.pl --gen-parrot test with the latest parrot? 00:35
because I just ran the spectests that way.
colomon huh. 00:36
no, --gen-parrot is getting parrot 2.9.1
we need parrot 2.10.1, right?
TimToady did they really just make a new release for "perl Configure.pl" noise? Somehow this seems...sorry...laughable, given all the rest of the spewage 00:39
00:39 raw_ joined 00:40 raw_ left
TimToady wonders if 1 => 2 => 3 -> ... should be taught to iterate like a lisp programmer would expet 00:46
*ct
seems like it would be pretty trivial
sjohnson just another perl 6 hacker 00:49
just another lisp hacker actually 00:52
01:09 redicaps joined 01:11 justatheory left
masak 'night, #perl6. 01:23
01:24 jaldhar joined, masak left 01:25 risou joined 01:27 risou_ left 01:30 ggoebel joined 01:33 envi left 01:39 envi joined 01:52 justatheory joined
dalek kudo: 32c08e3 | colomon++ | src/Perl6/Actions.pm:
Allow operators like infix:<..> to autocurry when passed a WhateverCode argument.
 02:03
kudo: 019c864 | colomon++ | src/Perl6/Actions.pm:
Rewrite right-hand WhateverCode currying to work around the bizarre $counter + $right.arity error.
lichtkind rakudo: @a = 1..9; @b = @a[3&5&8];@b 02:10
p6eval rakudo 015d77: OUTPUT«===SORRY!===␤Symbol '@a' not predeclared in <anonymous> (/tmp/altLX4rbrg:22)␤»
lichtkind rakudo: my @a = 1..9; @b = @a[3&5&8];@b
p6eval rakudo 015d77: OUTPUT«===SORRY!===␤Symbol '@b' not predeclared in <anonymous> (/tmp/Rn2OV6eJXF:22)␤»
lichtkind rakudo: my @a = 1..9; my @b = @a[3&5&8];@b
p6eval rakudo 015d77: ( no output )
lichtkind rakudo: my @a = 1..9; my @b = @a[3|5|8];@b
p6eval rakudo 015d77: ( no output )
lichtkind rakudo: my @a = 1..9; my @b = @a[3;5;8];@b 02:11
p6eval rakudo 015d77: ( no output )
lichtkind rakudo: my @a = 1..9; my @b = @a[3,5,8];@b
p6eval rakudo 015d77: ( no output )
lichtkind rakudo: my @a = 1..9; scalar @a
p6eval rakudo 015d77: OUTPUT«Could not find sub &scalar␤ in main program body at line 22:/tmp/_tMbFRG1xM␤»
02:11 risou left
lichtkind rakudo: my @a = 1..9; + @a 02:11
p6eval rakudo 015d77: ( no output )
02:17 whiteknight left
colomon rakudo: my @a = 1..9; my @b = @a[3,5,8]; say @b 02:25
p6eval rakudo 015d77: OUTPUT«469␤»
02:26 drbean left
colomon anyone out there know how the build/PARROT_REVISION file works now? 02:27
RELEASE_2_9_1-261-g495feaf is what it is currently set at 02:28
the RELEASE_2_9_1 part seems clear enough, and I assume g495feaf is the beginning of the git hash for the revision 02:29
but what's the 261?
oooo, wait, what's config/auto/git_describe.pm? That sounds promising.... 02:30
02:31 molaf_ joined 02:32 drbean joined 02:34 jaldhar left, ch3ck3r joined, molaf left 02:39 ch3ck left 02:44 risou joined 02:48 satyavvd joined
colomon wishes he understood the new build/PARROT_VERSION version number system. What a stupid obstacle to testing the latest parrot with rakudo.... 02:52
02:57 satyavvd_ joined, satyavvd left, satyavvd_ is now known as satyavvd 02:59 agentzh joined
colomon rakudo: sub a(&b) { say &b.arity }; a(*+1) 03:06
p6eval rakudo 015d77: OUTPUT«1␤»
03:08 risou left
dukeleto colomon: what is your issue? 03:16
colomon I need to update build/PARROT_REVISION to the latest Parrot release. But after the git changes, I've got no idea how to do that. 03:17
dukeleto colomon: the 261 is how many commits past the tag it is, and 495feaf is the beginning of the sha1, g is for 'git'
colomon: moritz_ wrote the code, he probably has it already
colomon alas, I suspect moritz_ is asleep
dukeleto colomon: he wrote the rakudo side of things
colomon: are you release manager?
03:17 Guest3023 left
colomon nope, that's masak, also asleep 03:17
but getting rakudo up on the Parrot release is an important pre-release activity. 03:18
dukeleto colomon: so what are you trying to do with PARROT_REVISION ?
colomon figure out how to bring it to 2.10.1
is there a way to automatically generate the version string? 03:19
.... 03:20
or should it be RELEASE_2_10_1-0-hash?
that seems easy enough
so maybe the question becomes, what's the hash of the release?
RELEASE_2_10_0-0-c4477dd2 does not work. :\ 03:25
03:27 envi left 03:29 meppl left 03:30 envi joined 03:35 bluescreen joined, satyavvd_ joined, bluescreen is now known as Guest31423, satyavvd left, satyavvd__ joined 03:39 zibri joined 03:40 satyavvd_ left 03:41 jaldhar joined
dukeleto colomon: i don't understand what you are asking 03:41
colomon: you want the sha1 of our latest release? 2.10.1 ?
colomon that would be great 03:42
03:42 lichtkind left
dukeleto $ git rev-parse RELEASE_2_10_0 03:42
80d156ef8b867e29750211ec97d675ee1529d40b
someone forgot to push a tag for 2.10.1, it seems. I will look into that. 03:43
colomon: git rev-parse is your friend
03:43 rgrau left
colomon dukeleto++ 03:44
dukeleto colomon: you can also just do "git rev-parse HEAD" to get the sha1 of whatever master points to in the parrot repo 03:46
colomon: you can give rev-parse any branch name, tag or "commitish" 03:48
04:00 plobsing_ left
colomon rakudo: sub a(&b) { say &b.count }; a(*+1) 04:02
p6eval rakudo 015d77: OUTPUT«1␤»
colomon rakudo: sub a(&b) { say &b.count }; a(*+*) 04:03
p6eval rakudo 015d77: OUTPUT«2␤»
colomon rakudo: say 4 xx 5 04:07
p6eval rakudo 015d77: OUTPUT«44444␤»
04:10 gavv\w left 04:11 gavv\w joined 04:23 plobsing joined
colomon phenny: tell masak I wasn't able to figure out how to set Rakudo to use the latest Parrot. On the plus side, I have a patch to make @a[*-3..*-1] work spectesting now. Please don't release until I can check it in. :) 04:23
phenny colomon: I'll pass that on when masak is around.
04:40 ggoebel left 05:10 ggoebel joined 05:16 ggoebel left
dalek d: bb56d4e | TimToady++ | CORE.setting:
[CORE] add roll, set, and bag
 05:27
05:36 kjeldahl joined 05:37 Guest31423 left 05:40 ch3ck3r left
colomon phenny: tell masak The @a[*-3..*-1] patch is pushed. 05:46
phenny colomon: I'll pass that on when masak is around.
dalek kudo: f8f556e | colomon++ | src/core/Any-list.pm:
Support code blocks with arity > 1 in Any.postcircumfix:<[ ]>.
 05:48
06:06 molaf_ left 06:44 _xiaomo_ joined 06:47 _kaare joined 07:01 wtw joined
sorear good * #perl6 07:18
07:23 QinGW joined
dalek ast: 12982c8 | TimToady++ | integration/weird-errors.t:
time() will eventually fail at compile time

missing subs need to be in eval_dies_okay		 07:24
07:40 kjeldahl left 07:44 dual left
sorear Ick, eval 07:45
moritz_ eval is a very useful tool. 08:01
sorear is thinking that the way to a faster tryfile start is to slim down CORE.syml 08:09
08:13 kjeldahl joined
sorear TimToady: What was the connection between $¢.<foo>:delete and multidimensional array subscripts? 08:22
08:22 cjk101010 joined 08:27 aesop joined 08:33 _xiaomo_ left 08:36 justatheory left 09:00 s1n joined
sorear STD keeps breaking for no reason when I try to remove stuff from $ALL :/ 09:15
dalek ecza: 048dc16 | sorear++ | v6/tryfile:
[v6] Prune sub pads from CORE.syml for a much faster start
 09:25
09:25 david____ joined
david____ i know i am not the first that ask this question 09:26
but when will be the official release of perl 6
which year ?
the first production release
moritz_ when it's done
david____ :-)
ok
moritz_ I'm serious. There's no good way to plan volunteer's work in an unexplored territory 09:27
david____ yes
you are correct
moritz_ that said I find rakudo usable for some tasks already
and I fully expect it to become more and more usable each month and year
and usable for a broader audience too 09:28
09:28 redicaps left
david____ and are there any plans to merge more perl 6 features into perl 5 09:28
like switch or state
which i use in every day programming
?
moritz_ given/when, which is the perl equivalent of switch, is already backported to perl 5.10 09:29
which was released in 2007
likewise say(), smartmatching, // (defined-or) and a few other things 09:30
09:30 redicaps1 joined
moritz_ and state variables too 09:30
the upcoming 5.14 release will have another feature backported, but with different syntax 09:31
$obj.?method vs. $obj&&->method
09:34 tzhs joined
sorear there will not be an offical release of Perl 6, since there is no official implementation 09:36
moritz_ there can be still official Perl 6 releases 09:37
every implementation that passes the test suite my call itself an official Perl 6 release, just not the official Perl 6 release
sorear looks like most of my time is being spent in the first line of ws, and in canonicalize_name 09:49
09:49 whiteknight joined
moritz_ ws is called often. 09:50
ENOphenny :( 09:53
david____ thanks a lot for your info 09:56
09:56 david____ left 09:57 wamba joined
sorear right 09:59
so
next order of business I guess I'll be speeding up autovivification somehow
what. 10:10
std: $_foo
p6eval std a194beb: OUTPUT«===SORRY!===␤Variable $_foo is not predeclared at /tmp/tHVrzjjqyF line 1:␤------> $_foo⏏<EOL>␤Check failed␤FAILED 00:01 118m␤»
sorear perl6: say '_' ~~ /<alpha>/ 10:11
p6eval rakudo 015d77: OUTPUT«_␤»
..pugs: OUTPUT«Error eval perl5: "if (!$INC{'Pugs/Runtime/Match/HsBridge.pm'}) {␤ unshift @INC, '/home/p6eval/.cabal/share/Pugs-6.2.13.16/blib6/pugs/perl5/lib';␤ eval q[require 'Pugs/Runtime/Match/HsBridge.pm'] or die $@;␤}␤'Pugs::Runtime::Match::HsBridge'␤"␤*** '<HANDLE>' trapped by operat…
10:12 QinGW left
dalek ecza: fe78c53 | sorear++ | / (2 files):
[v6] fix 1, and $_
 10:17
10:18 scottp joined, scottp left
sorear starts another roast run, which should be about 30 minutes faster and might even pass more tests 10:19
moritz_ how long was the previous run? 10:20
sorear I don't know
the time command failed. horribly.
sorear tries it differently
sorear -> sleep 10:21
10:26 nadim joined 10:27 pmurias joined 10:38 snarkyb00jum joined 10:41 daxim joined, dakkar joined 10:42 snarkyb00jum left
dalek kudo: db14a08 | moritz++ | build/PARROT_REVISION:
[release] bump parrot revision to 2.10.0 release

Note that parrot 2.10.0 has a known bug that prevents it from being installed if git is not installed. That's unfortunate, and a 2.10.1 release exist that fixes it, but it's nowhere to be found in the git repo.
gerd++ promised to put the stuff into the parrot repo soon, when that's done PARROT_REVISION should be bumped to the new tag		 10:45
kudo: 16b7466 | moritz++ | build/lib/Rakudo/CompareRevisions.pm:
[configure] at a release tag, "git describe" returns just the tag. Fix a regex accordingly
kudo: 227feae | moritz++ | build/PARROT_REVISION:
bump PARROT_REVISION to 2.10.1
10:49 [particle] left 10:50 ggoebel joined, _xiaomo_ joined 10:53 muixirt joined, [particle] joined 10:57 bbkr joined 11:00 pmurias left 11:07 knewt2 left 11:08 knewt2 joined 11:23 XaRDaX joined 11:45 _xiaomo_ left 11:47 satyavvd__ left 11:51 masonkramer left
tadzik ~~ 11:54
12:05 knewt2 left 12:12 knewt2 joined 12:14 sftp joined 12:19 whiteknight left 12:22 meppl joined
colomon moritz_++ 12:23
12:27 drbean left 12:29 orafu left, orafu joined 12:30 masak joined
masak oh hai, #perl6! 12:30
Rakudo release today :)
moritz_ oh hai masak 12:31
12:32 drbean joined
colomon masak: I've got @a[*-3..*-1] working for you now. phenny was supposed to tell you... 12:32
masak I see, I'm backlogging now...
colomon++
moritz_ rakudo: say (1 ... 10)[*-3..*-1]
p6eval rakudo 015d77: OUTPUT«Method 'succ' not found for invocant of class 'WhateverCode'␤ in 'RangeIter::reify' at line 4771:CORE.setting␤ in 'Any::postcircumfix:<[ ]>' at line 1␤ in main program body at line 1␤»
moritz_ that revision is quite old. 12:33
ah, I didn't remove the parrot svn checkout 12:35
12:39 phenny joined
masak moritz_++ # PARROT_REVISION work 12:48
phenny masak: 04:23Z <colomon> tell masak I wasn't able to figure out how to set Rakudo to use the latest Parrot. On the plus side, I have a patch to make @a[*-3..*-1] work spectesting now. Please don't release until I can check it in. :)
masak: 05:46Z <colomon> tell masak The @a[*-3..*-1] patch is pushed.
masak phenny: oh, *now* you spring to life? :) 12:49
ah, she wasn't here before.
moritz_ rakudo: say (1 ... 10)[*-3..*-1] 12:53
p6eval rakudo : OUTPUT«8910␤»
moritz_ colomon++
masak \o/ 12:54
awesome.
moritz_ rakudo: say (1 ... 10)[2..*-2] 12:55
p6eval rakudo : OUTPUT«3456789␤»
12:56 bluescreen joined
moritz_ acually I had considered making a challenge out of that feature 12:56
12:57 bluescreen is now known as Guest70621
moritz_ because I didn't want to tackle it myself :-) 12:57
in the end I didn't do it, because I wasn't quite sure how much effort it would be
takadonet morning all 12:58
MindosCheng Morning 12:59
masak takadonet: morning.
jnthn o/ 13:00
13:01 snearch joined
moritz_ \o 13:02
btw I did a small test release locally, and it seemed to obtain the right parrot version
it just produces a slightly weird version output:
$ ./perl6 --version 13:03
This is Rakudo Perl 6, version 2010.10 built on parrot 2.10.0 RELEASE_2_10_1
Copyright 2008-2010, The Perl Foundation
it reports the wrong version, but the right tag
I guess I can live with that
not rakudo's fault though - it's what parrot_config VERSION returns 13:04
bbkr rakudo: say +q; say q.Num # is this a bug that forcing numeric context on q works only with prefix:<+> ? 13:11
p6eval rakudo : OUTPUT«===SORRY!===␤Confused at line 22, near "say q.Num "␤»
bbkr rakudo: say +q;
p6eval rakudo : OUTPUT«Could not find sub &q␤ in main program body at line 22:/tmp/vpXZn42hT_␤»
13:11 wamba left
bbkr hmm, it parses on * .10 release... weird... 13:12
moritz_ std: q.Num
p6eval std a194beb: OUTPUT«===SORRY!===␤Couldn't find terminator . at /tmp/cUlc9O4qt9 line 1 (EOF):␤------> q.Num⏏<EOL>␤Parse failed␤FAILED 00:01 117m␤»
bbkr std: +q
p6eval std a194beb: OUTPUT«Invalid class name (STD::stop::) at /usr/local/lib/perl/5.10.1/Class/MOP.pm line 88␤ Class::MOP::load_first_existing_class('STD::stop::') called at /usr/local/lib/perl/5.10.1/Class/MOP.pm line 136␤ Class::MOP::load_class('STD::stop::', undef) called at
../usr/local/lib/perl/5.10.1/Mo…
moritz_ bbkr: it parses it as q/.../ quote, except with dots as delimiters
MindosCheng rakudo looks cute when confused.
moritz_ bbkr: which is correct 13:13
bbkr rakudo: q.Num. 13:14
p6eval rakudo : ( no output )
bbkr indeed :)
moritz_ rakudo: say q.Num.
p6eval rakudo : OUTPUT«Num␤»
moritz_ you can't have everything AND a pony
jnthn Ponies are outside of the set of everything? :) 13:15
MindosCheng lol
Pony is everything. 13:16
moritz_ jnthn: the definition of "everything" is a bit tricky. If "everything" encludes the set of sets that don't contain themselves, it's not decidable
s/encludes/includes/
and ponies are a bit tricky too :-) 13:17
jnthn Hmm, did we just reduce everything to the halting problem? :)
Oh wait, we can't decide...
:)
masak oh, we're on the topic of naive set theory now? 13:18
this is why Zermelo and Frankel decided to go have a cup of coffee one day :)
moritz_ masak: actually we just finished before you joined the discussion :-) 13:19
masak s/Frankel/Fraenkel/
moritz_: dang! :)
but you only pointed at the *problems*, whereas the solutions are what's interesting...
you'll probably know whether you want a pony if you adopt the Axiom of Choice... :P 13:20
moritz_ and if I can't decide? :-)
13:20 muixirt left
masak that's just life. :) 13:20
13:22 fhelmberger joined 13:23 Patterner left 13:26 PerlPilot joined 13:28 MayDaniel joined 13:29 PerlJam left 13:34 Psyche^ joined, Psyche^ is now known as Patterner 13:39 nymacro joined 13:40 masonkramer joined 13:41 MayDaniel left
masak moritz_: after having read through the list of uses of 'local' in perl.plover.com/local.html -- I'm starting to think you're right about 'temp' not having many remaining uses in Perl 6. :) 13:42
dalek ast: e4acb32 | (Solomon Foster)++ | S02-builtin_data_types/whatever.t:
Remove fudging on tests that now work.
masak I basically saw two uses in that article: glob filehandles and dynamic scoping. Perl 6 has scalar filehandles and star-twigil variables, respectively. 13:43
moritz_ even perl 5 has scalar filehandles :-)
13:46 redicaps1 left
jnthn Maybe it's just a temp feature then. ;) 13:46
masak moritz_: I know. maybe it didn't at the time of that article's writing, though. 13:47
in some ways, I have much more hope in 'let' becoming a cool/useful feature. it abstracts away some backtracking boilerplate. 13:48
moritz_ fwiw lexical file handles are a rather old feature, but in ancient version of perl 5 they didn't autovivify, so <c> open my $var, ...</c> wouldn't work 13:49
masak I don't understand that use of the word 'autovivify'. 13:50
moritz_ my $var; creates a lexical varaible in perl 5. it contains undef. 13:51
open FOO, ...; otoh references a typeglob
with a slot for file handles
masak right.
moritz_ so in "modern" perl 5, 'open my $var, ...' automatically puts all the stuff into $var that's required to store a file handle 13:52
that's what "autovivification" means
masak ah. yes. 13:53
I didn't think of file handles as requiring any special machinery.
they don't in Perl 6, far as I'm aware.
moritz_ (p5 guts hacker will probably cringe in horror at that explanation :-)
masak: afaict open() is much more magic in p5 too 13:54
in many, many ways
for example
open Foo;
is the same as open Foo, $Foo; (assuming that $Foo is a package var, not a lexical) 13:55
masak ick. 13:57
jnthn wow
That's...sort of cute.
moritz_ $ perldoc -f open|wc -l
383
jnthn :)
Juerd If it did open Foo, 'Foo' it'd be kinda useful for many cute oneliners :) 14:05
jnthn slurp('filename') and lines('filename') tend to serve that role well in Perl 6 though :) 14:06
Especially as with the latter you can set up a pipeline.
moritz_ isn't it lines('filename'.IO) these days? 14:07
jnthn please no
but maybe.
moritz_ because words() acts on the string, and doesn't interpret it as file name
jnthn ah, h
mmmmmm 14:08
masak I remember some spec changes in that area. 14:09
without remembering their exact details, I'd hope that &words and &lines have multis with the same signatures. 14:10
(and that, apart from the specific thing they do differently, they have the same semantics) 14:11
14:11 nymacro left
moritz_ that was the point of the spec change 14:11
masak \o/
MindosCheng wanna know cute one-liners. 14:12
masak MindosCheng: you should watch jnthn's talk about small and large Perl 6.
MindosCheng: www.nuug.no/aktiviteter/20100914-li...arge-apps/ 14:13
MindosCheng masak++ 14:15
tadzik o/ 14:17
takadonet tadzik: o/ 14:18
14:19 felipe joined 14:23 Lorn left, Lorn joined
jnthn masak: Looking at that video though, you may note that it makes my examples less cute. ;) 14:27
masak oh?
takadonet jnthn++ 14:28
jnthn I'm not sure what can really be done about it though. I agree having lines and words have different siggies is wrong.
masak: well, lines('events') ==> grep /beer/ ==> say 14:29
Is cuter than
lines('events'.IO) ==> grep /beer/ ==> say
That .IO feels kinda noisy.
masak well... yes and no.
the .IO feels noisy in $file.IO ~~ :e as well
especially since we started out at :e $file
er. -e $file 14:30
jnthn aye
I understand why, just not sure it ain't a false minima. 14:31
masak come to think of it, &slurp is arguably also in that group with &lines and &words. 14:33
jnthn $string.slurp is identity though :)
How...useless. :)
masak er, except that slurping a Str would then be a... right.
so never mind.
jnthn :) 14:34
phew
14:36 pmurias joined
colomon jnthn: can you please review my patches last night, if you haven't already? thanks. 14:47
jnthn colomon: Didn't get chance yet. Unfortunately, I'm still trying to fight off some illness, which is leaving me mostly wiped out on an evening after $dayjob. So I'm not getting to much at all. I'll try and remember to look at them soon, but hassle me again at the weekend if I forget. :) 14:49
14:49 wtw left
MindosCheng pat jnthn 14:50
14:50 nymacro joined 14:51 PerlPilot left 14:52 redicaps joined, PerlJam joined
jnthn MindosCheng: :) 14:53
14:54 blongden joined
PerlJam So, I guess the relationship between parrot and rakudo is still somewhat broken. 14:56
masak PerlJam: in what sense? 14:57
this sounds important, since I'm to make a release today.
PerlJam If I build parrot from the tarball, there is no git_describe key
(and thus I can't build rakudo from the tarball)
Everything works fine as long as you have a repo around
masak oh, but that doesn't affect the --gen-parrot instll path, right?
masak tries that path from scratch 14:58
moritz_ PerlJam: uhm, it should fall back to comparing the versions if no git_describe output is available 15:00
oh.
and the installed parrot identifies itself as 2.10.0
not 2.10.1
so
two solutions
masak :(
moritz_ 1) a fixed parrot release
or
2) require version 2.10.0 15:01
2) isn't as bad as it sounds, because every time you manage to build a parrot 2.10.0 you get the same functionality as with 2.10.1
masak I'd like to once again request a suitable .pm group for this month's release name.
moritz_ just that 2.10.1 has a higher chance to build
so build/PARROT_REVISION could like like this: 15:02
PerlJam Here's what happened when I tried to build rakudo using a freshly compiled parrot from the parrot-2.10.1 tarball: gist.github.com/705084
moritz_ RELEASE_2_10_1 2.10.0
jnthn masak: I should found Lund.pm or something ;) 15:03
masak jnthn: :)
jnthn But ENOTAPPROPIRATE due to lack of current existence :)
moritz_ we could just invent .pm names
and see if somebody notices it :-) 15:04
PerlJam Hogwarts.pm
;)
masak Silesia.pm
GothamCity.pm 15:05
Metropolis.pm
15:05 _kaare left
PerlJam Gondor.pm 15:06
Moria.pm
Rivendell.pm
masak Ankh::Morpork.pm 15:07
moritz_ Ankh/Morpork.pm please :-)
masak right.
NewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewYork.pm 15:08
PerlJam Which pm group sponsored the NPW? (or was there a pm involved?) 15:09
huf that city'd wrap around the earth a few times
each new adds about an atlantic ocean's worth of offset west 15:10
masak huf: if it were on the Earth, yes.
mathw but in Futurama, New New York is built upon the ruins of (old) New York
masak TwinPeaks.pm
mathw so wouldn't NewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewNewYork be somewhere in the upper atmosphere?
reached by a long, spiral ramp
masak mathw: the New**19 York is a Dr Who reference. 15:11
mathw and inhabited by bird people
oh that one
I didn't like that episode
It was from the 'oh no it's still awful' era
masak :)
15:15 cjk101010 left
moritz_ masak, PerlJam: I've started the parrotgit2 branch 15:15
I've tried to fix Configure.pl when dealing with a parrot-from-tarball
PerlJam: it would be very nice if you could test that
masak moritz_++
PerlJam will do 15:16
moritz_ tests --gen-parrot
I didn't make --gen-parrot smart
colomon rakudo: say "alive" 15:17
p6eval rakudo : OUTPUT«alive␤»
PerlJam I think --gen-parrot doesn't need to be smart since it grabs the parrot repo
colomon ummmm.... what happened to the version hash in the p6eval output?
moritz_ so if you have a parrot-from-tarball installed, and use --gen-parrot, it'll build a redundant parrot
colomon: I wiped the old rakudo builds, and with it the hash file 15:18
colomon: a cron job should eventually trigger a rebuild that creates it again
colomon rakduo: my @a = 1..10; say @a[*-4..*-2]
moritz_ if it does not by tomorrow, I'll investigate
colomon rakudo: my @a = 1..10; say @a[*-4..*-2]
p6eval rakudo : OUTPUT«789␤»
PerlJam patiently waits for his other tests to finish 15:19
15:25 MayDaniel joined 15:28 snearch left
jnthn afk 15:28
15:31 mkramer joined
dalek kudo: e7a20ed | moritz++ | build/PARROT_REVISION:
change PARROT_REVISION again

We require only parrot 2.10.0, but use 2.10.1 for --gen-parrot. The background is that 2.10.1 identifies itself as 2.10.0. OTOH the only difference is that 2.10.1 is more likely to build, so if somebody presents us an installed 2.10.0, there's no reason to reject that.		 15:33
15:33 mkramer1 joined 15:35 mkramer left
PerlJam moritz_: parrotgit2 seems to work fine with my parrot-built-from-tarball. (at least it compiles, I didn't run through all of the tests) 15:36
moritz_ PerlJam: thanks, that's all I'm interested in
PerlJam: feel free to merge :-)
15:37 agentzh left, justatheory joined
dalek kudo: b469741 | moritz++ | Configure.pl:
an attempt to fix Configure.pl with a parrot-built-from-tarball
 15:38
15:42 alester joined 15:43 redicaps left 15:48 kjeldahl left 15:51 risou joined 15:57 tzhs left 16:04 icwiener joined 16:08 hercynium joined 16:11 pmurias left
masak moritz_: sorry, I haven't been following along too closely. would I be able to go into release mode with the current master? 16:14
moritz_ masak: yes 16:15
masak: assuming you find a release name :-)
masak considers going with Ankh/Morpork.pm and see how far he gets with it 16:16
I mean, even though nobody here *went* to the YAPC::Discworld this year, their Perl 6 track looked really nice. 16:19
moritz_ indeed :-) 16:20
masak compared to BristolBath (and as far as I know), they've done a lot.
16:21 kjeldahl joined
masak just learned that each Discworld half-year has 13 months 16:23
Offle, February, March, April, May, June, Grune, August, Spune, Sektober, Ember, December, and Ick.
colomon Ick and Offle for winter, eh?
masak dunno. got it from here: wiki.lspace.org/wiki/Discworld_calendar 16:24
MindosCheng saw GURPS Discworld once...
16:26 risou_ joined 16:28 risou left 16:33 notjack joined
notjack is there an official perl6 news site or page? the ones I googled seem out of date 16:34
masak perl6.org 16:35
moritz_ notjack: perl6.org is our website. planetsix.perl.org has nearly all interesting Perl 6 blogs
masak as for news, see rakudo.org for latest Rakudo releases.
Teratogen will perl 6 eventually be written in perl 6? 16:36
moritz_ Teratogen: it already is, in big parts
Teratogen is that the ultimate goal?
also, early on, I heard some crazy story that the perl 6 compiler was going to be one huge long regular expression! 16:37
notjack thanks
moritz_ most Rakudo code I look at is written in nqp-rx, which is a subset of Perl 6. Or in rakudo's Perl 6 dialect.
Teratogen: we parse Perl 6 with Perl 6 grammars. Which are collections of regexes and methods.
Teratogen: github.com/rakudo/rakudo/blob/mast...Grammar.pm 16:38
16:39 _kaare joined
notjack I know someone wrote an APL parser in Perl6 (or some old flavor of Perl6) - is that person here, and can s/he comment on how long it took? 16:39
I'm thinking of writing a J parser, but I'm not much of a parser guy - I just want an easy way to make an open-source J interpreter (the current one isn't open source and so I'm a bit hamstrung) 16:40
(J is very like APL, for those who don't know)
moritz_ notjack: I don't know how long it takes, but once you get the hang on it, it's no rocket science 16:41
perlgeek.de/en/article/debug-a-perl-6-grammar might be useful at some point :-)
notjack I've heard that once you get the hang of it, rocket science isn't exactly rocket science either - but I get your thrust :)
moritz_ notjack: a professor once said "that's no rocket science... actually it's *MUCH HARDER* than rocket science" :-) 16:42
it was about density functional theory in solid state physics 16:43
masak Teratogen: what moritz_ said. the "one huge long regular expression" is in some sense true, because grammars are that. but it's just as strange to describe it as that, as it would be to describe a class hierarchy as "one huge long jumble of variables and statements". 16:44
moritz_ masak: I'll be offline for about 2.5H... if you need any fixes for the configure system, contact your local p5 dealer, or wait for my return :-) 16:45
arnsholt I'd disagree with grammars being just a long regex, but only a bit =)
masak moritz_: I won't start relenging until in about 2.5 h anyway.
arnsholt: well, it's not a perfect fit. 16:46
arnsholt: just as there's a bit of runtime polymorphism in methods, there is in grammars as well.
notjack let me ask for a concrete example then
masak but some people, diakopter for example, have been experimenting with compiling grammars down to one huge regex. I think. 16:47
arnsholt notjack: But as people have pointed out, writing grammars is much like writing regexes. Only quite a bit easier in a few key aspects I'd argue
notjack let's say I have a language where foo bar and baz are functions of ambivalent arity: they can each be called as either foo arg1 or arg0 foo arg1
masak notjack: the best single advice I can give you on writing grammars (besides reading moritz_++' article) is "use TDD". 16:48
notjack and now I want to add a grammar rule that says "(foo bar baz) arg1" <-> "(foo arg1) bar (baz arg1)" and similarly "arg0 (foo bar baz) arg1" <-> "(arg0 foo arg1) bar (arg0 baz arg1)"
does that seem like it would be hard to do?
masak no. 16:49
notjack masak: oh yes, TDD definitely.
masak though you seem to be suggesting internal re-writing in some prior phase.
notjack masak: can you should me the rule in broad strokes?
masak: sorry, I missed that: what prior phase?
masak notjack: well, I'd rather you went and experimented for a while and came back with an even more concrete question :) 16:50
notjack: I'm not sure I have the energy to do design right now.
notjack that's not concrete enough? (that's an actual - and useful - grammar rule in J, btw)
masak: understood
masak it's not syntax I'm familiar with.
notjack maybe I'll cop the APL example and start from there 16:51
masak that is, it looks neither like (E)BNF nor like Perl 6.
notjack <-> means "equivalent to"
masak start simple. don't you dare write code before you've written a test :)
yes, I got that part.
16:51 mila_ joined
notjack the rest is actual syntax 16:51
masak ...except that arg0 and arg1 are probably metasyntactic? 16:52
notjack no, they're arguments
data
masak ok.
notjack let's say foo=- and bar=* and baz=- (same as foo)
then "(- * -) 5" <-> "(- 5)*(-5)"
masak I wouldn't do a rewriting phase, I'd probably parse into an AST, and then do manipulation on the AST.
notjack similarly "4 (- * -) 5" <-> "(4-5)*(4-5)" 16:53
masak notjack: right, ok.
notjack: bit too many details to take in, I'm afraid. if it were my problem, I might make more of an effort to dissect it. as it's yours, I encourage you to do it. 16:54
notjack: have you skimmed S05?
notjack ah, the much feared AST. Fine, fine ... I'll go learn what parsing actually means.
yeah, I've skimmed most of the Ss and As
masak wow, nice.
notjack I just haven't typed a line of code :) 16:55
arnsholt ASTs aren't too hard either. It's just a fancy word for data-structure-that-represents-my-program
16:55 mila__ left
masak exactly. 16:56
and the reason I'd recommend manipulating those rather than strings, is that the AST already embodies the structure of the program.
strings are flat, and therefore it's much messier to handle them.
arnsholt Exactly. Objects (which is what you'll probably use for the AST) have structure, which is so much nicer =) 16:57
TimToady though I suspect notjack++ is thinking of strings of tokens, not chars
notjack correct 16:58
I mean, I could actually write a regex to do what I want -- but that doesn't feel official :)
oh yeah - and lexing. well, that's not too hard. I have the FSM already. 16:59
masak strings of token is also fine, and perhaps better for this purpose. it also has more structure than just strings of chars. 17:00
notjack strings of chars add no value
masak no, it's numbers that add value :P 17:04
notjack 4+"5" ; # :)
masak I'm going to go make food now.
notjack later - thanks
masak if anyone has a better name than Ankh::Morpork for the upcoming release, please let me know. 17:05
would be nice to stick with real pm groups for Rakudo, and maybe save the fictional ones for Yapsi...
masak bows out
17:05 masak left
dukeleto how is the Rakudo release going? 17:16
which parrot sha1 is it targetting?
17:18 LionMade0fLions is now known as MilLionsOfLions 17:22 Axius joined 17:24 Axius left, Bzek_ joined, Axius joined 17:26 risou joined 17:27 Bzek left 17:28 MayDaniel left, risou_ left 17:30 Axius_ joined 17:31 [particle] left 17:32 Axius left 17:36 nymacro left 17:42 wamba joined 17:54 Chillance joined 18:00 cdarroch joined, cdarroch left, cdarroch joined 18:01 Axius_ left 18:03 impious joined 18:04 tylercurtis joined 18:06 dakkar left
tylercurtis Should $*EXECUTABLE_NAME and $*PROGRAM_NAME be failures in modules that are being loaded by "use"? 18:06
18:12 MayDaniel joined 18:13 MayDaniel left
TimToady I don't think we can separate benign uses of those variables from malign; for instance, they could be useful for error reporting. 18:13
18:13 kjeldahl left 18:14 notjack left 18:15 daxim left, impious left
TimToady and we tend to go with innocent until proven guilty here, most of the time *cough* Durations *cough* 18:16
18:18 [particle] joined 18:26 ch3ck joined, risou_ joined 18:27 dual joined 18:28 risou left
mkramer1 ahh, priority one troubleshoots are out of my box, now to begin my day's work.... 18:48
whoops
wrong channel again
(if only I were working in perl6, I wouldn't have these problems!) 18:49
18:57 risou joined 19:01 risou_ left
moritz_ back 19:06
diakopter wb 19:07
Teratogen will most perl 5 code work in perl 6? 19:10
diakopter depends... most of ... code what size? 19:11
:)
(affects how much you'll have to port) ;)
s/ll/d/
flussence someone, please put a "FAQ" link on the front of perl6.org 19:12
diakopter flussence: you can :) 19:13
would you like to be that someone? 19:14
colomon dukeleto: the good news on the Rakudo/Parrot interface is that moritz_++ stepped in to make it work. :)
19:17 MayDaniel joined
flussence diakopter: I dunno, maybe later if nobody else volunteers though. Do we even have a FAQ to link to? 19:18
moritz_ notreally 19:19
but you can start /faq/ page if you want :-)
flussence there's a page for one on the wiki, it's a bit sparse right now though...
tylercurtis TimToady: is that a no to $*EXECUTABLE_NAME and $*PROGRAM_NAME failing in loaded modules? That appears to be Rakudo's current behavior. 19:30
On a related note, is there any dynamic variable I can use to know whether a module is being executed directly or just being imported?
flussence you could look at the code that does MAIN() for a hint, that needs to know the same thing. 19:33
19:33 MayDaniel left
dukeleto colomon: good to hear :) 19:39
sjohnson Teratogen: it will not unless you give it a directive to at the start, like use v5; or something 19:59
colomon which doesn't work in Rakudo.
Teratogen ok 20:00
sjohnson that is to be done in the future i believe.. so you might have to hold your horses 20:01
tylercurtis On the other hand, Perl 5 code still works in Perl 5. And if you want, you could even use Blizkost to access that Perl 5 code from Perl 6. 20:03
20:10 wamba left 20:13 WebDragon joined
WebDragon tips proverbial hat 20:13
recently noticed that Rakudo Star is part of Fedora 14's release 20:14
first off, _nice_ 20:15
congratulations are, methinks, in order 20:16
colomon thank you
WebDragon has anyone here been working with that particular release (F14) ?
I have yet to upgrade from F12 but have it planned for the next weekend or two, due to the EOL coming up in december for F12 20:17
more or less wanted to know what, if any, interesting things I should be looking for in order to start playing with it (admittedly for the first time)
I'd noticed somewhat ago that I'd, as a necessary part of my web workload at $WORK, been spending way too much time in php, and my perl skills had atrophied considerably from where I'd managed to get them 20:19
colomon The only Fedora-like thing I have around here is an old copy of CentOS, but hopefully someone else is around who is using it.
WebDragon been working towards remedying that fact, when I discovered this little gem
colomon What do you know about Perl 6?
WebDragon only what little I'd played with as part of the various porting of bits back into perl5 and the various exegesis and apocalypse postings over the past few years 20:20
20:20 rindolf joined
WebDragon nothing's *really* stuck in my brain yet, but that just gives me more incentive to play 20:20
:D
rindolf Hi WebDragon
WebDragon heya rindolf :)
TimToady rosettacode.org/wiki/Category:Perl_6 has 240 examples of working Perl6 code
WebDragon oh, _lovely_
colomon woah, you guys are up to 240 examples now?! 20:21
rindolf BTW, does Parrot already have bignum types?
WebDragon TimToady: do they share corresponding examples of what it would have looked like were it written using perl5 ?
TimToady Util++ has been going to town on it lately
colomon rindolf: I believe it has some sort of depreciated bignum type from the early days
TimToady WebDragon: generally, just page up from each Perl 6 example 20:22
though a few of them have no corresponding Perl 5
rindolf colomon: ah.
WebDragon sweet. that'll definitely help with the mental transitions
colomon Perl 5 version may not be the same approach, right?
rindolf colomon: maybe I'll work on binding GNU MP
colomon rindolf: if it were easy, I'd have done it already myself. :)
TimToady colomon: sometimes it is, sometimes it isn't
some of the P6 examples give something like the P5 approach, and then an alternative FP-oriented approach 20:23
WebDragon colomon: just use monofilament. (i.e. binary rope)
:D
WebDragon hides under his rock
colomon rindolf: not that you should let it stop you! I'd certainly love to have bigints working in Rakudo. 20:24
TimToady funny you should mention mono...
WebDragon I know comparatively little regarding the various bits (such as Parrot) 20:25
TimToady niecza: say "Am I running under mono?"
p6eval niecza fe78c53: OUTPUT«Am I running under mono?␤»
WebDragon I've read bits over the years but have never gone on to try installing any of it due mostly to various time constraints involving large chunks of RealLife™ which keep invading my space
sjohnson hi
colomon WebDragon: you can try simple things right here, you know. 20:26
rindolf WebDragon: what's up?
WebDragon things have been settling down for a while now to where I've actually started work on rearranging my apartment once again, seeing as things have approached stability
colomon rakudo: my @fib := 1, 1, * + * ... *; say @fib[^40].perl
p6eval rakudo : OUTPUT«(1, 1, 2, 3, 5, 8, 13, 21, 34, 55, 89, 144, 233, 377, 610, 987, 1597, 2584, 4181, 6765, 10946, 17711, 28657, 46368, 75025, 121393, 196418, 317811, 514229, 832040, 1346269, 2178309, 3524578, 5702887, 9227465, 14930352, 24157817, 39088169, 63245986, 102334155)␤»
TimToady perl6: say 1..3 X~ <a b c>
p6eval pugs: OUTPUT«123␤»
..rakudo : OUTPUT«1a1b1c2a2b2c3a3b3c␤»
WebDragon once that's done I'll have a more comfortable working environ once again, at home as well as at work 20:27
oh, that's just nifty
how did the Rakudo Star bit come about? 20:29
flussence rakudo: say *.WHAT 20:30
p6eval rakudo : OUTPUT«Whatever()␤»
WebDragon I recall hearing about Rakudo but don't recall exactly when the "Star" business happened, as I was out of touch with Perl for way too long
TimToady twit.tv/floss140 if you don't mind a 1.5 hour interview of Patrick Michaud by Randal Schwartz 20:32
20:33 Trashlord left
TimToady well, not 1.5, it's 1:21:40 20:33
20:36 risou left
WebDragon nice, links in with google reader 20:38
20:45 masak joined
masak hey there, #perl6 people. 20:45
tylercurtis Hi, masak. 20:47
20:47 Guest70621 left 20:49 AABMEJ7 joined
rindolf Hi masak 20:52
masak I'll just backlog a little, and then we'll do a release, OK? 20:53
WebDragon: hi! welcome! 20:59
WebDragon heyaz :)
WebDragon passes around the carafe of DragonCoffee™ 21:00
masak WebDragon: besides the scripts at Rosettacode, modules.perl6.org/ contains over 80 modules of Perl 6 code to look at. 21:01
shortcircuit TimToady, WebDragon: Give me a minute or five, and I should be able to create pages that will quickly list the RC tasks which have P5, but no P6, and P6, but no P5.
masak WebDragon: also, Rakudo itself is chock full of good Perl 6 code.
WebDragon it's still quite new to me. I haven't even got F14 installed yet (but it will be soon) 21:02
I don't even yet know how to poke at it to make it sit up and hello, world me
masak rakudo: say "Hello WebDragon!" 21:03
p6eval rakudo : OUTPUT«Hello WebDragon!␤»
WebDragon well I can see how to do it in the channel
masak WebDragon: if you have installed it on F14, my guess is that you have a 'perl6' binary somewhere in your $PATH 21:04
21:04 Guest70621 joined
masak pokes Guest70621 with a feather 21:04
WebDragon but I'm reasonably sure you don't want me writing and testing my programs in the channel, one line of code at a time considering that even this as an interactive shell is somewhat lacking in that individual line invocations aren't going to recall variables from one line to the next
use.perl.org/~masak/journal/38279 <-- ala
masak WebDragon: it does nowadays :) 21:05
...I think
WebDragon chuckles
masak yes it does!
> my $a = 52
52
> say $a
52
pmichaud_++ pmichaud_++ pmichaud_++
WebDragon doesn't want to make a nuisance of himself doing interactive shell invocations here in #perl6
shortcircuit rosettacode.org/wiki/User:Short_Cir.../SomePerls 21:06
masak WebDragon: by the way, I've moved blogs: strangelyconsistent.org/blog/preten...rl-virtues
colomon what, no mandelbrot in p6 there yet?!?
masak I should put up a sign at the old place saying I've done so. 21:07
colomon: go for it!
oh right, I *can't* post on use.perl. 21:08
now I remember why I left.
masak goes into releng mode
frettled :)
shortcircuit OIC, Slashcode. 21:09
WebDragon indefinite hiatus
WebDragon wonders what pudge is doing these days
21:09 kjeldahl joined
WebDragon dates back to the days when Matthias Neeracher had put together MacPerl for pre-OSX macintoshes 21:10
it ultimately led me on to RHL and thence to Fedora to have a native environment to play with, and THEN they come out with OS X once I was no longer in a position to afford a new Mac, and by then I liked linux enough to stick with it 21:11
or rather, by the time i had the cash to spare to buy a new Mac with, I liked linux enough to stick with it
masak you know what? I'm going to go with "Melbourne" for the release name. 21:12
WebDragon read: "it had improved enough by then that I could live with it" :D
frettled Y2K: Windows for work, Linux for play. Y2K.01: Linux for work, Windows for play.
masak and if anyone asks why, I'm going to say that it's because I know a guy there who writes great Exegeses.
frettled masak: good enuff
WebDragon heheh
masak I think so.
maybe it'll make Damian finally go public with his Perl 6 code, too :P 21:13
WebDragon frettled: indeed, I dual-boot my laptop and the only reason I use the windows boot is to A> test work in MSIE8, B> use my MagicJack which doesn't yet have a working linux version, and C> play games
21:13 molaf joined
WebDragon everything else happens on the linux side 21:13
incluing a few games that work properly under wine ever since 1.0 release of it
frettled masak: TheDamian has Perl 6 code waiting in the wings? *drool* 21:14
WebDragon shouldn't the bottom of rosettacode.org/wiki/Category:Perl_6 have some link to the SomePerls page ?
masak frettled: I've seen one of the modules. it was nice.
WebDragon one thing that's always frustrated me about wiki's is the nonstandard methods of navigating the information found there 21:15
masak Hm, ChangeLog needs updating. 21:16
masak does that
WebDragon I really wish I had the tuits to graft a front-end on top of a wiki that required various crosslinking relationships between the data at the top level before drilling downwards
masak WebDragon: write up a clear description of your vision somewhere, and someone might jump on it. 21:17
frettled WebDragon: I'm not sure you _want_ to create a hierarchy of it.
WebDragon i.e. you can't publish newly written articles til you've identified the relationships to the existing information and identified how you get there from elsewhere
frettled What if there is no relationship to the existing information, except something that is extremely contrived? 21:18
flussence put it in an "Other" category and let someone else fix it :)
frettled Let's say that you have a wiki containing information about animals. Then you want to write an article about a programming language. Please explain how you want those to be linked.
masak frettled: Python! :P 21:19
frettled masak: that's a road that's nearly as humpy as a camel's back!
masak but you can't deny it's a connection... :)
frettled It's in the «contrived connections» category. 21:20
21:20 AndreasX joined
frettled (actually, that's the term flussence probably was looking for, rather than «Other» ;)) 21:20
masak why does 'git status' sometimes say "Your branch is ahead of 'origin/master' by 36 commits." even when I've just pulled and it fast-forwarded? 21:21
flussence you haven't pushed
frettled yup
masak huh.
moritz_ 36 local commits?
that's... a lot.
21:21 WebDragon|laptop joined
masak I'm not aware of anything I've neglected to push... 21:21
frettled moritz_: Hey, it's masak.
moritz_ git diff origin/master
WebDragon|laptop kicks his frakking router and asks where he left off 21:22
21:22 WebDragon left, WebDragon|laptop is now known as WebDragon
flussence WebDragon: irclog.perlgeek.de/perl6/today#footer 21:22
frettled Arghle, bedtime. No soup for me. 21:23
21:23 AndreasX left
WebDragon <frettled> What if there is no relationship to the existing information, except something that is extremely contrived? 21:23
because the only way I seem to navigate a wiki is by already knowing what I'm searching for and somehow finding it via google
then you identify the contrived pathway, and the various people editing the wiki come up with additional paths (or correct your assumptions :D )
masak moritz_: thanks. there are indeed lots of differences, none mine.
masak pulls --force
WebDragon so the 'wikiness' is preserved, but also the mind-mapping, so you can always go back and forth along whichever pathways lead there. 21:24
frettled WebDragon: That depends on the wiki. But you're not answering the question, you're just raising a different one.
WebDragon using multiple breadcrumb-trails as tags, something like that. admittedly I've only given this a modicum of thought, but the _relationships_ between the information seem to me to be as important as the information itself
frettled WebDragon: You don't «identify the contrived pathway», that's a contradiction in terms.
WebDragon frettled: indeed you have nailed it in a nutshell
frettled WebDragon: You _contrive_ a pathway where there is none.
And that's very problematic in terms of organization of content. 21:25
WebDragon yes but you still have to identify the pathway you've contrived, TO the wiki so that others can see it
frettled No.
I think you misunderstand my point entirely.
WebDragon are you still thinking in terms of how existing wikis work?
frettled Which means I'm not communicating clearly.
WebDragon either or both of us.
frettled WebDragon: I'm thinking in terms of how to organize information.
WebDragon I'm off in wonderland thinking about an idea that's not fully crystallized yet, mind you :)
frettled WebDragon: Have a look at my example again. Why would you contrive a connection between programming and animals? 21:26
WebDragon chuckles and hides a grin in his coffee
21:26 rindolf left
WebDragon 'contrive' is perhaps the wrong word to use to describe the various pathways one would 'mental-leap' from to get from point a to point b in the process of information discovery 21:27
although it's close
frettled Aha, I see where you're missing the point now.
Or where I'm miscommunicating. 21:28
WebDragon straighten me, I'm ready. (as Lord Buckley would say)
frettled Let me rephrase: why should you enforce a connection between animals and programming?
diakopter contrive has a negative connotation; it can be used pejoratively to describe it as artificially (or mis-)constructed
frettled Why does there need to be?
diakopter: as intended.
WebDragon frettled: if there's a connection, one makes it
from whatever source derived 21:29
diakopter brain->link
WebDragon precisely, diakopter
frettled WebDragon: yes, but you quite clearly stated that you wanted everything to be clearly connected, to _require_ that new entries have a connection to an existing one.
WebDragon frettled: exactly. answering the question "how did I get here?" as well as "where do I go from here?"
frettled WebDragon: which, in my not so humble opinion, is ill-advised both in terms of information theory, and in managing people
What's the point? 21:30
WebDragon wiki's inevitably leave me wondering where to go next
or how to backtrack
so I just click around
frettled WebDragon: the back button in your browser, as $deity intended?
WebDragon which usually leads me to google, not more of the wiki's information
:P
which was my point 21:31
21:31 MayDaniel joined
WebDragon invariably I'm led to the wiki by deep-links from somewhere OTHER than the wiki 21:31
frettled You're not going to solve the problem by enforcing links where there is no "natural" connection between the pieces of information
WebDragon which is now missing my own point
frettled Sure, it's relevant that "antimatter" is linked to "matter", and vice versa, and probably to a bunch of other articles regarding physics etc.
WebDragon what about the _natural_ connections?
frettled Sorry, AFK, alarm at work. 21:32
WebDragon I'm not saying it's required to enforce arbitrary and unnatural connections between information, but surely if you're creating information there would be _natural_ links to follow that map how you get there from somewhere else, and how you'd continue onward
masak smoke reports looking good: smolder.parrot.org/app/projects/smoke_reports/5 21:33
I believe even those two failing tests on darwin might have gone away since yesterday.
maybe.
21:35 y3llow_ joined, y3llow left
WebDragon your typical website is going to be reaosonably linear within the plateau of information presented therein, usually as trees or tags 21:35
21:35 pothos_ joined 21:36 y3llow_ is now known as y3llow 21:37 pothos left 21:38 pothos_ left, pothos joined
WebDragon but the progression and iteration and mental-mapping of the information therein is, while inclusive, also much more clearly defined. very rarely will you have pages existing in isolation 21:40
on the other hand NONE of this is really relevant to perl6 21:41
:)
which is what I REALLY came here about
WebDragon follows some more of the links and keeps on ready
s/ready/reading/ .. wow dunno how I made that typo
21:41 MayDaniel left
WebDragon summons up visions of his old Model III Trs-80 and the ubiquitous READY> prompt 21:42
21:47 fhelmberger left
masak ChangeLog commit coming up. 21:47
let me know if I missed any major change since last release. or even any minor change that still might be worth mentioning. 21:48
21:49 donaldh joined
masak dalek? 21:49
diakopter takes up to a minute I think 21:50
dalek kudo: 7b94fc9 | masak++ | docs/ChangeLog:
[docs/ChangeLog] updated

Condensed the git log down to these things.
masak there we go.
diakopter 5% done porting PAST::Regex from .pir to nqp 21:51
er
porting nqp-rx itself
anyways
good thing pmichaud did all the hard (and easy) work already 21:52
WebDragon is very much looking forward to having F14 installed and being able to really play with this stuff
masak grr! t/spec/S32-io/IO-Socket-INET.rakudo fails one subtest here!
masak wants to remove that file from spectest.data 21:53
moritz_ you can comment it out for the release.
masak good idea.
moritz_ or you can fudge the test
masak I already fudged one test in that file.
seems it's very stochastic. 21:54
I'll comment out the test file, and then I might "forget" to uncomment it again.
21:55 donaldh left
WebDragon makes a bunch more bookmarks, tips his hat, passes around more DragonCoffee™ for all and sundry, and congratulates all on the most excellent milestone 21:58
but alas I must bid you adieu; time to bike home before the rain starts xD 21:59
21:59 WebDragon left, Guest70621 left
dalek kudo: 78b3fcf | masak++ | t/spectest.data:
[t/spectest.data] t/spec/S32-io/IO-Socket-INET out

This file is notoriously misbehaving (at least on certain platforms). Commenting it out for the release.		 22:00
22:02 risou joined
masak wow, cpanm is rocking my world. 22:04
miyagawa++
ok, announcement coming up. 22:06
now, it's *important* that there are not any unnecessary typos in there, so if you want to help by doing a simple task, now's yer chance :) 22:07
any and all proofing is appreciated.
also suggestions for specific things that might be worth mentioning. 22:08
Tene offers editing services.
frettled is back.
frettled offers editing services as well.
masak the review link should be here any moment now. 22:09
or you can just go to github directly and find it there.
I already pushed it.
Tene Ah, that announcement.
I thought this was your mysterious upcoming announcement. 22:10
masak oh, sorry to disappoint :)
no, still 22 days left until that one :)
2 days until next clue... 22:11
dalek kudo: 0bf28d5 | masak++ | docs/announce/2010.11:
[docs/announce/2010.11] added
Tene I didn't notice any errors.
masak thanks. 22:12
flussence looks good to me (though I have no idea what I did worthy of mention there...)
masak "implement Cool.samecase, .subst(:samespace) and ss/// Patch partly by flussence++" 22:13
flussence oh, that :) 22:14
jnthn masak: Looks fine to me.
frettled masak: «can now index with both» -> «indexing is now possible with both
jnthn masak: I think explicator is a word. :P
frettled apart from that, very fine indeed.
masak frettled: thanks -- will fix.
jnthn: thanks. 22:15
jnthn: could you check whether it is a word, please? you got me worried :)
dalek kudo: 03e6f5a | masak++ | docs/release_guide.pod:
[docs/release_guide.pod] slight tweaks

The biggest improvement was probably that, since Parrot is now git-based, PARROT_REVISION contains slightly diff'rent information.		 22:16
22:16 Guest70621 joined
jnthn masak: The verb "to explicate" most certainly exists. 22:18
Looks like it's from Latin, lit "to unfold" :)
Cute. :)
masak yup, the 'ply' root again :) 22:19
love that root.
jnthn masak: ah, found a dictionary that has "explicator" listed as a noun :)
masak: It felt obvious in meaning, just wasn't sure I'd seen it before. :)
1530s, from L. explicatus, pp. of explicare "unfold, unravel, explain" 22:20
tylercurtis rakudo: multi foo (Str $s) { nextsame; say 'test'; }; multi foo ($s) { say 'next'; }; foo 's'
masak perfect.
dukeleto yes, explicator is definitely a word. it is often used to describe people that solve problems
p6eval rakudo : OUTPUT«Null PMC access in clone()␤ in 'foo' at line 1␤ in main program body at line 22:/tmp/lLVxsbODlb␤»
masak tylercurtis: I think that one's known and reported.
phew! :)
jnthn dukeleto: Oh? I'd have had it as somebody who explains things, but the latin root seems to fit that nicely too :) 22:21
dalek kudo: 09489cf | masak++ | docs/announce/2010.11:
[docs/announce/2010.11] improvement in wording

Suggested by frettled++
kudo: cbea8a8 | masak++ | docs/announce/2010.11:
[docs/announce/2010.11] added contributor frettled++
kudo: 58d81aa | masak++ | docs/announce/2010.11:
[docs/announce/2010.11] rewording

For the greater preposition good!
jnthn
.oO( I know what I want on my business card now... :) )
masak "Turns beer into code" ? :P
tylercurtis masak: I didn't realize that bug existed. I was just trying to remember which of [call|next][same|with] is which. :)
jnthn masak: :P 22:23
masak tylercurtis: bad news is it's broken. good news is that jnthn's 6model will likely fix it all.
dukeleto jnthn: "explicate" is often used to say that one person got someone else out of a bad situation, at least from what I have heard
masak dukeleto: that would be the more literal interpretation, then.
"to bend [someone] out"
22:23 kjeldahl left
masak dukeleto: or are you thinking of "extricate"? 22:23
tylercurtis masak: redispatch in general is broken? Or just in some cases? 22:24
masak or "extradite"? :)
tylercurtis: don't remember. tias?
dukeleto masak: think i was mixing it with extricate 22:25
masak I suspected that.
jnthn tylercurtis: It's fine for methods.
tylercurtis rakudo: class A { method a { say 'A::a' } }; class B is A { method a { say 'B::a'; callsame; } }; B.a
p6eval rakudo : OUTPUT«B::a␤A::a␤»
jnthn tylercurtis: It's never been implemented for subs. 22:26
masak that's 'trahere', "to pull".
tylercurtis jnthn: conveniently, that's what I need it for.
jnthn :) 22:30
tylercurtis Is CANDO implemented? 22:31
masak running the spectests again. might as well November-blog while I wait.
22:32 benabik joined
dalek kudo: d39f4f6 | masak++ | VERSION:
[release] bump VERSION
 22:32
22:39 spq1 left, benabik left
moritz_ tylercurtis: no 22:40
tylercurtis moritz_: thanks. 22:42
Is there a nicer way (in Rakudo or in the spec) to add a method of variable name to a single object than adding the method to an anonymous role with .^add_method and mixing the role in? 22:48
masak you can do it all in one step :)
mkramer1 thinks that's a nice way
masak doesn't get much nicer than that. 22:49
rakudo: my $x = 42 but role { method foo { say "OH HAI" }; say $x; $x.foo
p6eval rakudo : OUTPUT«===SORRY!===␤Unable to parse blockoid, couldn't find final '}' at line 22␤»
masak rakudo: my $x = 42 but role { method foo { say "OH HAI" } }; say $x; $x.foo
p6eval rakudo : OUTPUT«42␤OH HAI␤»
masak \o/
tylercurtis A method of variable name.
masak I'll have to mentally add quotes to that for it to make sense. 22:50
or a dollar sign.
tylercurtis A method whose name varies.
masak oh wait. the name varies... right.
it's not very common of them to do that, you know? we're well into metaprogramming if they do. 22:51
jnthn What's wrong with .^add_method? 22:54
oh, to a single object...
In theory
$obj by role Foo[$name] { method ::($name) () { 42 } } 22:55
gah
but
Anyway, sleeps
TimToady well, you'd have to pass the name somehow to instantiate it 22:58
23:02 Guest70621 left
sorear good * #perl6 23:06
moritz_: last run = 88m 13s
masak last run of what? 23:07
picking up jogging? :)
sorear 60.78%, wow, almost doubled for two fixes
masak: roast 23:09
niecza-compiled tryfile.exe
masak wow, niecza is running roast?! 23:10
diakopter even without fudging?
sorear masak: no 23:11
23:11 _kaare left
sorear tryfile is part of the STD.pm6 suite 23:12
masak oh, right. tryfile.
still. nice.
sorear++
all spectests are green here. 45 minutes. 23:13
masak proceeds to next phase
23:14 tylercurtis left
masak blog post! strangelyconsistent.org/blog/novemb...-joke-here 23:15
perigrin in soviet russia joke inserts you! 23:23
diakopter giggles
23:32 Khisanth joined
masak question: does it matter whether the release tarball contains the tags for the release or not? 23:34
23:34 Limbic_Region joined
masak the release_guide seems to suggest one should only do 'make release' once, namely before tagging HEAD. 23:35
23:38 hercynium left
masak oh; it doesn't matter, because the release tarball isn't a git checkout. 23:39
masak uploads
23:39 redicaps joined
masak upload complete. 23:40
colomon \o/
masak if anyone would like to build and test... github.com/rakudo/rakudo/downloads
wow. I hadn't looked at the download stats for Rakudo monthly releases before. 23:41
sharp rise after switch to ng, sharp drop after Rakudo Star release.
between Feb and July, inclusive, it was >200 downloads per month.
23:42 whiteknight joined
colomon now everyone wants R* instead? 23:42
masak think so.
would be interesting to see those figures as well.
masak looks
whoa. 23:43
colomon I'm kind of sad you didn't go with A::M for the release. ;)
masak there's some Slashdot effect involved here, I'm sure.
3k releases for the first Star tarball.
2.6k for the msi file.
diakopter releases?
masak 800 for the next Star tarball, 550 for the next, 500 for last month's... 23:44
diakopter: I don't understand the question.
colomon you meant 3k downloads.
diakopter oh
colomon we didn't release R* 3000 times, right? 23:45
;)
masak oh, sorry.
yes, 3k downloads.
TheHarlot oh, hey--the other day somebody mentioned something about Model 6... or something. I do not have this client set up to log... looking into it as we speak. 23:47
I just... uh... do not remember what it is I am looking into.
>.>'
colomon we didn't really do much publicity for the releases after the first, did we?
masak p6c email's away. 23:48
Tene TheHarlot: 6model
mkramer1 masak, are you counting git pulls?
Tene TheHarlot: github.com/jnthn/6model/ 23:49
masak I take it from the release instructions that I don't need to blog the release announcement.
mkramer1 after the first r* I just started tracking rakudo head
masak mkramer1: no, those figures only take downloads into account.
mkramer1: I know of no way to count git clones or pulls on github.
colomon shoot, I probably did a at least 200 git pulls all by myself since R*. :) 23:50
masak turns off the releng mode
TheHarlot thank you Tene
masak you're now free to commit again, you wonderful people. ^^
23:51 masonkramer left, masonkramer joined
TheHarlot ah... hey... this will allow me to not reinvent the wheel... unless the particular wheel was wearing a tight dress and no practically meant yes with reinventing it. 23:53
diakopter TheHarlot: you'd want to sync with mberends so as not to duplicate/conflict work :) 23:54
the jre branch has lagged the clr branch for a few weeks
but it shouldn't be too difficult to catch up
23:56 impious joined
TheHarlot diakopter, well, I still have Java Kitty to work on... this mostly simplifies anything I need to do CJerl6-wise for Java Kitty. 23:56

Powered by Cro and the Raku® Programming Language.

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

Thank you!