6.2.7 released! | pugscode.org/ <Overview Journal Logs> | pugs.kwiki.org | paste: sial.org/pbot/perl6
Set by autrijus on 13 June 2005.
Guest53185 a 01:18
nothingmuch a 01:26
01:27 Guest53185 is now known as Weems
nothingmuch seen Alias 01:40
jabbot nothingmuch: I havn't seen Alias , nothingmuch
nothingmuch uh, he spells his name with an underscore, right? 01:41
Juerd seen Alias 01:47
jabbot Juerd: I havn't seen Alias , Juerd
Juerd jabbot: STUPID BOT, REMOVE THAT SPACE BEFORE THE COMMA. IT IS MISLEADING.
jabbot Juerd: 所以？
Juerd and s/havn/haven/
gugod alas 01:55
my bad
theorbtwo wonders what he's missed. 03:08
Khisanth nothing! 03:10
theorbtwo Is the SVN still down? 03:14
hcchien I guess so 03:17
nothingmuch hackathon:
"monkey butts evaluate to doodoo"
or "monkey buts evaluate to do do" 03:18
what fun
theorbtwo Nothimgmuch, you remind me of me, but smarter and more immature. 03:22
At the moment, however, I haven't a clue what you are talking about.
(God, bandwidth is nice.) 03:23
nothingmuch Obj.new but { $obj.method }; 03:29
my $obj = Obj.new;
$obj.method;
$obj
these are the same things
do given Obj.new { .method; $_ }
at some point someone mentioned 'do d'
'do do'
and 'but' on a closure is known as monkey but
see also monkey_but.t when it gets checked in
(that is to say, when openfoundry is back up) 03:30
theorbtwo Looks like a reasonable equilency, but I don't see where do do comes into it. 03:31
Whatever. 03:32
Hackathon is fun, but I can't get anything done there.\
Eeep. I can't remember my perlmonks password. Either of them. 03:36
...and it's my freaking site. 03:37
theorbtwo sshes in and asks. 03:38
Oh, there we go, found the godly one. 03:42
Brain so molten.
brentdax Is Perl 5 interop known to [not]? work on AMD64? 03:58
mugwump blast, still no SVN huh? 04:44
all those in favour of a temporary public SVN server for the meantime say aye 04:47
svnbot6 r4946, mugwump++ | r15293@speights: samv | 2005-06-25 20:59:17 +1200 06:46
r4946, mugwump++ | Disable Date::Gregorian, even minimal class fails to load (?)
r4947, mugwump++ | r15294@speights: samv | 2005-06-25 21:08:17 +1200
r4947, mugwump++ | t/00load.t passes
r4948, mugwump++ | r15295@speights: samv | 2005-06-25 21:47:10 +1200
r4948, mugwump++ | Make t/01sanity.t pass
r4949, mugwump++ | r15296@speights: samv | 2005-06-25 22:44:38 +1200
r4949, mugwump++ | some tests for attribute mutators
r4950, mugwump++ | r15297@speights: samv | 2005-06-25 23:27:21 +1200
r4950, mugwump++ | Make a stub test to test Test script output
r4951, mugwump++ | r15298@speights: samv | 2005-06-25 23:42:52 +1200
r4951, mugwump++ | Make t/02last_day.t fail gracefully
r4952, mugwump++ | r15299@speights: samv | 2005-06-25 23:59:35 +1200
r4952, mugwump++ | make a few tests pass and comment out the rest
r4953, mugwump++ | r15300@speights: samv | 2005-06-26 00:02:24 +1200
r4953, mugwump++ | TODO a broken test, check in changes to make partial 03components pass
r4954, mugwump++ | r15301@speights: samv | 2005-06-26 00:03:12 +1200
r4954, mugwump++ | Add :depends flag to Test ok(), etc
arcady hey hey, openfoundry is back 06:54
time to rebuild pugs and see if it still works!
Darren_Duncan in case it wasn't noticed, I'm happy to report that openfoundry.org seems to be up now 07:02
arcady it's just been noticed in fact 07:03
arcady rejoices
Darren_Duncan I also have a commit to make, preferably before the release, but I won't be able to do it until about 15 hours from now
bed time and all
also, given what svn up brought, its obvious that the Toronto hackathon's stuff hasn't been put on openfoundry yet
the stuff from the last 2 days that is 07:04
now, looking at the IRC log ... noticed something journal worthy 07:07
Juerd's yelling at a bot, colabti.de/irclogger/irclogger_log/...2005-06-26 near the top 07:08
good quote for Autrijus' journal, to join his other quote collection 07:09
anyway, back in 14-15 hours, good night
svnbot6 r4955, autrijus++ | * minor source cleanups. 07:13
autrijus greetings. 07:14
luqui !!
autrijus luqui: a minimal example?
or commit and tell me what to run? :) 07:15
luqui class Foo { method bar() { 50 } } sub baz($x) { say $x.bar } baz(Foo.new)
autrijus 'k
luqui (untested, so it might work ;-) 07:16
autrijus that prints 50
luqui indeed
hmm
so it's deeper
luqui is reducing to a real minimal example
it's a two-level thing 07:20
svnbot6 r4956, autrijus++ | test was full of crap about precedence
r4957, autrijus++ | clearer failure boundries for hash subscript reduction
r4958, autrijus++ | &as tests can work in an eval
r4959, autrijus++ | * mark two tests as todo.
luqui class Foo { method foo() { say "HI" } } -> $x { -> $y { $y.foo }.($x) }.(Foo.new)
hi kakos 07:21
autrijus luqui: you know... it's exactly the same bug. 07:22
just put `is rw` in either $x or $y. 07:23
zero deref etc.
luqui yep
i figured as much
autrijus so the same fix for transparent reference will fix this.
why did we convince larry to adopt zeroderef again? ;)
(because it's saner and easier to reason about, yes, but it really uncovers all the design glitches that was hiding underneath) 07:24
luqui hehe
maybe that's why too 07:25
or do you mean language design glitches
svnbot6 r4960, autrijus++ | * initial cut at P5AST frontend for pugs!
r4961, autrijus++ | * pairs are always evaluated in lvalue context for binding name purposes.
autrijus language design of course :) 07:26
we're now regressing like mad because of this misassumption :)
luqui don't run the test suite. it's just depressing, and it will get fixed up once nm and stevan are back on the job 07:28
:-)
autrijus right. :p
mugwump so, in Date::Gregorian, I'd like to be able to present read/write accessors for properties which are actually derived... 07:40
svnbot6 r4962, autrijus++ | * first check in of Perl-Compiler, a PIL representation in Perl 6.
r4963, andras++ | Introducing WTemplate, the widget based template engine for Perl 6
r4964, autrijus++ | * remove empty files
r4965, autrijus++ | Re-added Perl-Compiler
r4966, autrijus++ | Delete stupidity.
r4967, autrijus++ | Adapt to lvalue pair fix.
r4968, autrijus++ | Added PIL tests.
r4969, autrijus++ | Added Load test.
r4970, autrijus++ | * `for Int { ... }`, `when Int { ... }` etc are now always
r4970, autrijus++ | parsed correctly, instead of grabbing the trailing block as
r4970, autrijus++ | the sole argument for the list operator.
mugwump eg, $date.month will return the month where 1 = January
$date.month_0 will return the month where 0 = January 07:41
internally only $.month exists
but I want to be able to go, eg, $date.month_0 = 6 to set the month to July
luqui and you can assign to both of these?
mugwump that's what I'd like 07:42
luqui okay... I'm not sure if pugs does that
luqui is looking up what it *would* look like
mugwump I had a look in the synopses .. the only mention of r/w accessors mentions making a sub return an lvalue
luqui it's in A06 07:43
pasteling "luqui" at 206.47.164.228 pasted "Proxy lvalue" (10 lines, 330B) at sial.org/pbot/11352 07:44
mugwump thanks, I'll fix t/oo/attributes/mutators.t 07:46
svnbot6 r4971, autrijus++ | * parse the type literals and variables for the condition
r4971, autrijus++ | parts more correctly.
r4972, autrijus++ | * rename Trans to CodeGen (inspired by luqui)
r4973, autrijus++ | * finish the renaming
r4974, autrijus++ | * finish s/trans/codegen/
luqui when you set attributes in a class, Pugs currently copies, not binds, right? 07:48
(which is the correct behavior)
mugwump sure, if you used = 07:49
luqui er, when you pass the attributes in as named params to new
mugwump well, it definitely doesn't use the lvalue accessor :) 08:01
svnbot6 r4975, autrijus++ | Eine Kleine Documentation.
r4976, autrijus++ | Delete Perl-Compiler in the root. How'd that get there?
r4977, autrijus++ | * in-class `does` and `is` now works.
r4978, autrijus++ | Moved Perl5.pm to Perl5_str.pm
r4979, autrijus++ | monkey butts are fun to play with
r4980, autrijus++ | * the `do STMT` form.
r4981, autrijus++ | * ...implementing the `do STMT` form.
r4982, autrijus++ | thunky-do form implemented in Perl::Compiler and made stuff better.
r4983, autrijus++ | NameGenList -- abandoned skeleton
r4984, autrijus++ | It parses again, baaaaby.
r4985, autrijus++ | Autrijus's journal transparent reference no verb.
coral svn is up 08:07
svnbot6 r4986, autrijus++ | * Repair smart match on types and `does` calls from the zero
r4986, autrijus++ | dereference regime by explicitly fetching their metaobjects
r4986, autrijus++ | and walking the inheritance/mixin tree ourselves.
r4987, autrijus++ | * parse for `use perl5:DBI`.
r4987, autrijus++ | * `is 1,2,3` is now always parsed as function call; only the
r4987, autrijus++ | `is ClassName;` form is parsed as declaration.
r4988, autrijus++ | * remove the no longer used author part.
r4989, autrijus++ | * oops, fix `use`.
luqui you mean openfoundry? 08:11
yeah, I think that's pretty clear
here come all our commits for the weekend 08:12
(or rather, there go)
coral reads
autrijus journal up
svnbot6 r4990, autrijus++ | * `use perl5:DBI` form now works -- the old `use DBI--perl5` form
r4990, autrijus++ | is no longer supported.
r4991, autrijus++ | Decloudify my type thinking.
r4992, autrijus++ | * &foo in module Foo now looks up &Foo::foo as well.
autrijus I'll sleep soonish :)
enjoy
luqui hey autrijus 08:13
where is Object::new defined? 08:14
autrijus like everything else, in src/Pugs/Prim.hs
only it's called Any::new
because the relationship was not clarified
luqui no wonder I couldn't find it
I'm going to (try to) have it carp when it recieves any positionals 08:15
autrijus arg, I should not have pushed --verbatim
now it's all my fault.
luqui heh
well it is
we're just your minions
autrijus luqui: oh, trivial. line 966 has a _ 08:16
replace it with a "positional" variable
check that is always VList [] (or VUndef? not trace it a bit)
s/not//
you can trace anything with
trace (show foo) $ do 08:17
by its one line in the monad block
if you want to fail in there, just
fail "hey you passed in positional"
luqui okay
autrijus so it's like
positionals <- fromVal p 08:18
svnbot6 r4994, autrijus++ | * also look at the compile time package.
r4995, autrijus++ | Seemingly fixed some stuff.
autrijus unless (null positionals) $ fail "blah"
with p being the bound positional slurpy array 08:19
should work, but untested :)
luqui I'll try
svnbot6 r4996, autrijus++ | Fixed bugs in test program. 08:24
r4997, autrijus++ | * remove temp file
r4998, autrijus++ | * switch to the `use perl5:Digest::MD5` form.
mugwump almost r5000! 08:36
whee! r5000! Do I win a prize? :D 09:09
svnbot6 r4999, iblech++ | Usual svn properties added to 30+ new files :)
r5000, mugwump++ | Add documentation for :depends
r5000, mugwump++ | make Test.pm report unexpected successes
r5000, mugwump++ | support multiple test runs in the same process
r5001, iblech++ | Fixed chomp($str) to not edit $str inplace. 09:18
r5001, iblech++ | * Pugs.Prim -- Fixed the implementation of &chomp.
r5001, iblech++ | * examples/, ext/, t/, src/perl6/Prelude/PIR.pm, Makefile.PL -- Test fixes.
r5001, iblech++ | * Added note about chomp's new behaviour to docs/quickref/data.
r5001, iblech++ | Renamed ext/WTemplate/readme to ext/WTemplate/README (consitency).
r5002, iblech++ | * Test.pm -- Unbroke &pass. 09:32
r5002, iblech++ | * util/svnlog2graph.pl -- Minor cosmetic fix.
mugwump hmm, should \( %hash<foo> ) auto-vivify %hash<foo> ? 10:41
svnbot6 r5003, iblech++ | unTODOs. 10:42
r5004, mugwump++ | Add documentation for :depends
r5004, mugwump++ | make Test.pm report unexpected successes
r5004, mugwump++ | support multiple test runs in the same process
r5005, mugwump++ | Unwanted hash auto-vivification seems to be related to Param.isLValue; adding `is copy' trait solves problem
coral . 10:53
mugwump , 10:58
svnbot6 r5006, iblech++ | * Added tests for the new relevations: 11:13
r5006, iblech++ | t/var/my.t, t/oo/syntax-but.t, t/oo/roles/instantiation.t, t/oo/precedence.t
r5006, iblech++ | * Prelude, Pugs.Prim -- Updated =$obj to mean $obj.shift instead of $obj.next.
r5006, iblech++ | * Added the usual shebang line to tests lacking it.
elmex is the ./method() call in the language already? 11:14
would be horrible
masak a grep for it in the examples/ folder yields only one false positive 11:19
examples/pipe_open.p6:my $pugs = (($*OS ~~ any<MSWin32 cygwin msys>) ?? "pugs.exe" :: "./pugs"); 11:20
but i seem to recall it was once there
svnbot6 r5007, mugwump++ | Taking a ref to a hash slot shouldn't auto-vivify - perhaps this needs Proxy object support 11:38
r5007, mugwump++ | Test A06-specified Proxy trait
r5007, mugwump++ | Fix broken test - probable precedence problem
r5007, mugwump++ | un-TODO fixed tests, TODO new broken tests, add test for given { when true { } }
r5007, mugwump++ | removed bad test for op0 'true'
Alias_ seen autrijus? 11:48
jabbot Alias_: autrijus was seen 3 hours 29 minutes 29 seconds ago
Alias_ autrijus: poke?
Boogie Alias_: I think it's very late / early in America now. 11:57
Alias_ sleep is for the genetically inferior
I expect more from him
:)
Boogie Yeah, as he is taking his journey, I think he sleeps more than before. :) 11:58
Odin-LAP Sleep is for the non-clinically-depressive. :p 12:00
nothingmuch Alias_: ping 14:00
Alias_ blinks 14:01
nothingmuch := autrijus?
oh wait...
ping, not pong
yes, right
nothingmuch: pong
nothingmuch we're trying to compile perl6 to ppi 14:02
no
nothingmuc.id ~~ Unique;
okay
i'd like to use quasiquoting for this stuff
Alias_ you are doing, heh, what? :)
You realise of course, that you are insane :) 14:03
heh, but continue
Alias_ chuckles
nothingmuch that is, construct a generic node that is somewhat equivelent to what we are generating
by parsing some perl 5 code
Alias_ yeah, I get it :)
nothingmuch okay
i read some of Tokenizer
and some node impls
and I have found no clean way to do this 14:04
Alias_ I'm sure you haven't :)
the problem is that PPI is based on _syntax_, not function
nothingmuch that's not what I mean
Alias_ oh, ok
continue
nothingmuch what we want 14:05
given $node {
when PIL::Number {
parse("123");
err, put that into $x;
$x.value($_.value); 14:06
$x;
}
....
ofcourse, this is a stupid example, and we have more complex ones
which are actually simple
Alias_ given? when? remember my Perl 6 is non-practical 14:07
nothingmuch this is my perl 6, using ppi--perl56
don't worry about it, all I want is to parse some perl statements (not full documents)
Alias_ Perl 5 statements? 14:08
I'm sorry, I'm not entirely sure of exactly where you are in the entire parsing scheme of things
nothingmuch yes
forget the entire scheme ;-) 14:09
Alias_ aha!
I get it
nothingmuch my $expr = parse_something_small("10");
$expr.value; # 10
$expr.ref; # PPI::Token::Number
err, not value, content
Alias_ PPI would still see that as Document(Statement(Token(10))) 14:10
nothingmuch $expr.{content} = $my_other_number;
right, i know
which is why i'd like to ask: is there a way to get tokenizing to work only on the statement level?
Alias_ hrm...
not very easily
because statements still always happen in a context
nothingmuch assume void 14:11
assume scalar
assume I know it or will provide it
Alias_ I mean a syntactic parsing context
nothingmuch oh, you don't need any
this is independenant
perl -e '10'; # full program
Alias_ right, so Document(Statement(Token(10))) 14:12
:)
nothingmuch it's my job to make sure the snippets make sense
right
Alias_ It still exists syntactically
nothingmuch what about 'perl -e '5; 10; 20;'
nevermind, i'll deal, i'm not stupid
Alias_ Document(Statement(Token(5)Token(;))Token(' ')Statment, etc 14:13
Have you use PPI::Dumper?
That will dump out a syntax tree for an arbitrary chunk of source
Are you trying to assemble your own tree manually?
nothingmuch but poking in the source code I still can't figure out some stuff
interface wise
Alias_ from scratch?
nothingmuch $t.{content} = 10;
that's irrelevant, I know the structure
what I'm trying to avoid is this:
and then shove that into a statement
PPI::Token::Number.new;
Alias_ ok
nothingmuch yes 14:14
what I'd like to do is not construct a dom, which i can
that's tedious, and not very readable
but everybody knows the string representation of that ;-)
and reading the tokenizer code (not thoroughly), i don't really know what to do
Alias_ I don't understand "what I'd like to do is not construct a dom, which i can" 14:16
nothingmuch i don'
t want to say ".new; .{content} = " and then put nodes in nodes in nodes
the tokenizer and lexer can do that for me
that's what they're good for
Alias_ ok
nothingmuch i'm good for writing languages that work well for humans
like perl5
Alias_ So you can generate chunks of raw perl 5 source, from the Perl 6 AST? 14:17
nothingmuch i'd like to write perl5 snippets
Alias_ ah, gotcha
ok, and then generate what?
nothingmuch ah!, but i want to go further
i could just generate strings, and print that out
Alias_ ok
nothingmuch but I don't want to, because the PDOM is easier to transform
so I want to write snippets
to get a mini dom out of them
Alias_ and turn the snippits into..
single statements? 14:18
or fragments...
nothingmuch to transform the snippet DOMs
so they contian real data, and not just empty stuff
construct a full document
and then call .save on it
Alias_ The Tokenizer produces a stream of tokens
that's all
nothingmuch lexer 14:19
Alias_ The lexer assembles them into a tree
nothingmuch yeah, i realized i was calling them both the tokenizer at some point
Alias_ hrm
So you want to make template tree fragments, and than inject real values 14:20
nothingmuch yep
Alias_ then
ok
So PPI::Document->new( 'sample source' );
suck out the only statement in it
nothingmuch fair enough 14:21
Alias_ It would be very very difficult to actually wire yourself into the lexer or tokenizer 14:23
esp the tokenizer :)
as long as you consider each snippit to be a legal and discrete syntactic chunk, just making a document for each one is far and away superior 14:24
suck out the statements, and then weld them together with newline whitespace tokens
pdcawley Arternoon all. 14:37
broquaint Hey, pdcawley. Nice summary btw. 14:38
pdcawley Thanks.
It's always hard to do summaries during summer 'cos so much is happening at conferences. Especially with the hackathons. 14:39
It's like irc, but more so.
Juerd pdcawley: I'd think it'd be easier, with the hackathons summarized by autrijus 14:40
pdcawley Just reading Autrijus's summary, and it sounds like there's enough in place that I can start writing a debugger.
At the very least 'trace' should be trivial.
Juerd: There is that. 14:41
It's just I feel kind of 'out of the loop'. And I can't afford to just go to the conferences and take part in the hackathons. With the mailing list it's 'all' there.
Hmm... thinking about that 'trace' thing again, not as trivial as I thought because of context propagation... 14:43
(At least if I want to catch the return values)
$_ = some_func() but context(&?CALLER_CONTINUATION) 14:45
14:51 Maddingue__ is now known as Maddingue
pdcawley_ Reminder needed. How do I get pugs to run with the parrot back end? 15:39
wilx PUGS_EMBED="parrot" iirc 15:59
Juerd pdcawley: pugs -CPIR, and it requires that it be compiled with parrot embedding, as wilx indicates 16:11
Limbic_Region am I the only Pugs is blowing up building for on Win32? 16:18
brentdax I suppose the fact that every single test complains about a "strange object" means that Pugs doesn't work on AMD64. 16:35
castaway_ it should if ghc does.. 16:41
16:42 castaway_ is now known as castaway
Limbic_Region anyone out there have current Pugs working on Win32? 16:43
brentdax I think AMD64 GHC is experimental. 17:12
svnbot6 r5008, iblech++ | Syntax but ($obj but {...}). 17:28
r5008, iblech++ | * t/oo/syntax-but.t -- Fixed test.
r5008, iblech++ | * Prelude -- Added Pugs::Internals helper sub.
r5008, iblech++ | * Pugs.Eval -- Implemented syntax but in the interpreter core...
r5008, iblech++ | * Pugs.Compile -- ...and in the compiler part of Pugs.
wilx but? 17:36
iblech wilx: my $obj = SomeClass.new but { .property1 = value1; .property2 = value2 }; say $obj.property1; # value1
wilx Huh, weird keyword : 17:37
:)
iblech I like it :) 17:38
Juerd Don't we have .new( property1 => value1, ... ) anymore? 17:45
iblech That works too, I think 17:46
TIMTOWTDI++ 17:47
svnbot6 r5009, iblech++ | * t/oo/class/basic.t -- Added test for inlined "is" declaration. 17:58
r5009, iblech++ | * Pugs.Parser -- Hack to make "is eval(...), ..." parse again.
r5010, iblech++ | Merged t/unspecced/monkey_but.t's comments into t/oo/syntax-but.t. 18:03
Alias_ seen chromatic? 18:04
jabbot Alias_: chromatic was seen 6 days 18 hours 20 minutes 28 seconds ago
Shillo Hello, all! 18:51
nothingmuch autrijus: if you guys are done with the cabal meeting, I need some PIL help 19:55
i think i've been improvising a bit too far
and I mostly would like luqui's opinion 19:56
Alias_ heh 19:57
you can never find a cabal when you need one 19:58
Nelasar51840G hello
Are you a Perl developer, Alias? 19:59
nothingmuch Alias_: i think the perl6 to perl5 code emmitter is working pretty well 20:00
Alias_ really? :)
nothingmuch yup
Alias_ Nelasar51840G, in what sense. I'm more of a CPAN developer ...
nothingmuch i don't know if it's working yet, but we'
ll see ;-)
Alias_ Nelasar51840G, I wrote these search.cpan.org/~adamk/ 20:01
Nelasar51840G I am a Perl 5 programmer, and a CPAN developer myself. I am interested in Perl 6 as well.
Alias_ But if you mean am I working on Perl 6, then no
Nelasar51840G Ah yes
In my opinion it would be best to write the Perl 6 compiler in a higher level language such as Haskell then Perl 6 itself, this would allow us Perl 6 developers to be able to improve it more easily 20:02
C is not an easy language for many to deal with 20:03
nothingmuch Nelasar51840G: what we've been doing in the hackathon might interest you ;-)
C is completely out of the question, as far as we're concerned
Allison Randal supports writing a perl 6 compiler in PIR
and the pugs oriented people are now working on an intermediate language called PIL 20:04
and writing a perl 6 to PIL compiler in perl 6
as well as some emitters
in a sense, the pugs parser thingy, and the compiler backends are interchangable across language boundries
and what I'm working on is the code emitter, that models PIL in perl 5 20:05
Nelasar51840G That sounds very interesting
nothingmuch it is =)
wolverian give me my neural cannula already, sheesh
Nelasar51840G I am glad to hear it
nothingmuch you're more than welcome to join in, btw
;-) 20:06
uh, where would you like that cannula, wolverian?
Nelasar51840G: have you hacked some pugs? 20:07
Nelasar51840G I would like to help in some way
wolverian nothingmuch: I don't really care, as long as it lets me think in perl!
Nelasar51840G I havent, I dont know Haskell, Ive taken a look at the code however
nothingmuch pulls drape around wolverian, and streightens his nurse hat
Nelasar51840G: you can write tests, and you can look in the ext/ directory of the pugs repo for perl 6 modules 20:08
you can port your CPAN modules to perl 6
you can try to learn haskell, we have a bunch of tutorial links in the READTHEM file
Nelasar51840G Porting some CPAN modules ive already done is something I will do
nothingmuch so what are you waiting for? i can give you commit access, and you can start right now =)_ 20:09
Khisanth hmm
nothingmuch: don't you need openfoundry to be up for that? 20:10
nothingmuch it's back up
Khisanth ah it's back up :)
nothingmuch Nelasar51840G: come on, you know you want to 20:11
everyone does it
well, the cool kids, that is
(none nerdy geeks around here ;-) 20:12
Nelasar51840G Yes, I can help certianly 20:13
nothingmuch okay, i'll need an email ;-) 20:14
Nelasar51840G [email@hidden.address] 20:15
nothingmuch oh wait, the last line didn't sound like a drug pusher line 20:16
anyway, sent
welcome aboard
a safe way to start writing perl 6 is to look over the tests 20:17
you can try to see if failing ones are failing because the tests are buggy
pugs.kwiki.org/?SmokedPugs
Nelasar51840G Thank you, I appreciate it 20:20
nothingmuch thank you too
anyway, i think i'm going to go back to the conf hotel 20:21
whiteg is the conf hotel with network? 20:22
nothingmuch i don't know
i hope so
whiteg hum..me too
thank you 20:23
nothingmuch see you there?
whiteg no, I am at taiwan 20:24
nothingmuch ah 20:26
everyone who is going to YAPC and wants keysigning, please mark yourself here: wiki.yapctoronto.org/index.cgi?PGPKeySigningBOF 20:32
geographical diversity is *VERY* good for key signings
ow! 20:52
mugwump ßËtå§Ç®îþt... now that's ℒℇ℮Ⓣ 20:57
nothingmuch very leet 21:00
port gaal's l33t to give extra points for unicode symbols? 21:01
lightstep was that latin1?
mugwump well, the signoff looks like all Latin-1 characters. I cut and paste mine from the Unicode sliderule, pages 21 and 24 which are full of letterlike symbols 21:03
sburke++ 21:04
stevan autrijus: ping 21:15
mugwump hijacks stevan's icmp session to say hi
stevan mugwump: hello
how is life down under
mugwump I see in autrijus' journal you've got a metamodel prototype ... is that on cpan? 21:16
stevan nope
I am actually refactoring a large chunk of it right now
it is not really CPAN worthy either
it is really pretty hackish
mostly meant to be a proof of concept
mugwump ok. well, it can always be used by the use Foo--perl5;^Wperl5:Foo; support 21:17
stevan yes, I think autrijus had some (crazy) ideas to do something to that effect 21:18
and also maybe to use it for the PIL -> Perl5 compiler luqui is working on
with nothingmuch
Alias_ via PPI (I say they are nuts) :) 21:19
stevan me too :)
Alias_: I think maybe the PPI idea has been abandoned
although hard to say, the hackathon has now dispersed 21:20
mugwump I've been porting parts of Perldoc to Perl 6, it's been quite good, really - I've been coming across "missing" language features/implementations all the time :)
integral PPI's lacking something to parse operator precedence
Alias_ I hope so... because going via a syntax parser is ... iffy
stevan mugwump: yes I aswa that
Alias_ integral: Yes, that was the entire point...
in order to do that, you also need to know which param goes with which sub
which is impossible
stevan personally I think nothingmuch's FORTH version will be the best approach :P 21:21
Alias_ shrugs
integral Alias_: unless you're a perl compiler/interpreter...
nothingmuch shutup
Alias_ I understand a fucking think you guys are doing. I don't even know the acronyms :)
nothingmuch i'll see you guys soon
stevan nothingmuch: where are you now? 21:22
nothingmuch i'm at my friends' house 21:24
i'll be leaving as soon as I have a sandwich
and see you at 89
can't talk much, the kids are playing with my computer
stevan ok 21:25
seen: autrijus
seen? autrijus
mugwump jabbot: seen autrijus ? 21:26
jabbot mugwump: autrijus was seen 13 hours 6 minutes 54 seconds ago
stevan mugwump: thanks :) lack of sleep is starting to catch up with me
mugwump so. I negotiated my employment contract with my new employer on Thursday. 21:44
I told him about what happened for YAPC::Taipei
I told him that I paid for the flights, took time off, and was happy when I called from Hualien to ask for another week off there and got it granted 21:45
But when I got back, I was told exactly how far into arrears my leave account balance was 21:46
...ie, that I'd not get any more leave until Christmas...
And my new employer says, "Alright, how about this... 21:47
The days you're at the conference, you don't have to take those as leave.
The day's you're in transit, you don't have to count those either.
s/'// :-}
The days that you're working on hackathons, sure we'll pay you for those too. 21:48
Oh, and we'll pay for your flights, too."
pdcawley mugwump: Result! 21:49
mugwump fuckin' aye! 21:51
pdcawley Friend of mine back in fat times once almost got a (new) employer to pony up for a $1000 sushi bill at LISA... 22:25
It was only when he confessed that everyone had paid their share that they realised he was only after a cut of it...
mugwump those were the days alright. We're definitely back on the upturn, but that initial high may never come back 22:28
pdcawley God, but the fat times were stupid.
And we knew it. 22:29
mugwump I didn't. 22:31
This is why I went from earning £48/hr at JP Morgan to £5/hr under the table at "Tony's Hemp Corner" in King's Cross overnight
pdcawley Yeah. 22:32
Um... who are you under that pseudonym?
mugwump Sam I am 22:33
pdcawley Vilain? 22:34
mugwump That's me. 22:35
pdcawley Ah. Hi.
mugwump So what are you up to these days Piers? 22:36
pdcawley Unemployed in Gateshead. About to take some teaching assistant work (I hope).
mugwump Gateshead! Thass wha' me birth certificate sez, liike. 22:37
pdcawley Why aye!
Who you working for now? 22:38
mugwump data mining outfit ... www.bnz.marketview.co.nz
pdcawley in nz? 22:39
mugwump yep. I'm back in Wellington
pdcawley Coo. enjoying it?
mugwump Refer previous mention of "new employer" ;)
pdcawley NZ's one of those places that's on my list of 'places to go' and was there before I saw LOTR. Point.
Dad was over there earlier in the year with a bunch of vintage car nuts. 22:40
mugwump "vintage cars" here? I thought in NZ a vintage car meant a Kingswood or a mini 22:42
pdcawley The shipped their Frazer-Nashes over.
mugwump solves his lack of vintage car knowledge with google 22:44
pdcawley Heh.
Vintage car: pre 1929 sportscar. 22:45
mugwump actually I've been using Wikipedia for general searches like that increasingly recently. There's an entry for Frazer Nash
it's like a Hitch Hiker's Guide 22:46
pdcawley Hmm... I shall point dad at that. Being a trivia nut he can probably help correct it :)
mugwump yes! Wikipedia is Foundation! 22:47
mugwump notes that Secondary teachers and University Lecturers are listed on the "Long Term Skill Shortage List" at www.immigration.govt.nz/migrant/str...rtunities/ 22:52
pdcawley Yeah. But my wife's not really up for the move. 22:55
What with the grandson living 300 yards away.
And the being over 50. And the diabetes and asthma... 22:56
mugwump And I guess the pressure to make "The Great Escape" is much less outside of the Greater London fallout area ;) 23:00
pdcawley cept there's work for programmers in London. Not found any here. 23:08
mugwump still, teaching can't be all that bad. Low wages, but huge karma payout if you can really inspire people. 23:15
pdcawley Got to stick to it first. Hence the Teaching Assistant work -- I need more classroom time before I really start to teach. 23:17
cognominal what PDD/MAD_SKILLS refers too in autrijus journal? 23:47
Alias_ Layy Wall's conversion of the Perl 5 parser to make it generate Perl 6 code, with comments and such intact 23:48
Larry
cognominal The Perl5 project is turning is some sort of global dynamic Rosetta Stone 23:49
s/5/6/ 23:50
Alias_ That was kind of the point... 23:56
one VM to rule them all...
Perl/Python/PHP/Ruby all sharing each other's libraries, etc etc
Haskell, in light of their contributions, may play too :)

Powered by Cro and the Raku® Programming Language.

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

Thank you!