Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2005-05-31

Perl 6 | Reference Documentation | Rakudo

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

All times shown according to UTC.

Time Nick Message
00:00 luqui Darren_Duncan, you can do that in Perl 5.  And yes, you can do that in perl 6.
00:01 Darren_Duncan I guess I never really tried to do that in either language
00:01 Darren_Duncan it smacks of using something before its declared
00:02 Darren_Duncan which seems to be what 'use strict' should prevent but doesn't
00:02 Jonathan_ autrijus: Any chance you can make Pugs look for Parrot in the current working directory, not just in the path? If you can, then I should be able to get this Pugs binary distribution thing working...
00:03 Darren_Duncan doesn't the path include the cwd?
00:03 jf3 has quit IRC ("Leaving")
00:03 luqui I use autovivi all the time
00:03 Jonathan_ Darren_Duncan: It appears not.
00:03 obra 84
00:04 Darren_Duncan odd, I saw it there before
00:04 obra mischan
00:04 Jonathan_ Darren_Duncan: Have a parrot.exe anywhere in the path, it works out fine. Drop one in the CWD instead, I get errors.
00:04 autrijus obra: yes, I can test on whiteg's laptop
00:04 Darren_Duncan maybe there are multiple paths
00:05 autrijus Jonathan_: problem is that the "current directory" is ~/src/pge/
00:05 Darren_Duncan when I 'use' a module that doesn't exist, the error message includes the path, which ends with '.'
00:05 autrijus (For PGE)
00:05 clkao autrijus: fix av return before release! :)
00:05 autrijus Jonathan_: and it has to be that way
00:05 obra autrijus: cool
00:05 autrijus clkao: p5_apply _always_ return a av?
00:05 autrijus clkao: even when GIMME_SCALAR?
00:06 clkao autrijus: ya. when multiplevalues are returned.
00:06 clkao should i make it scalar directly?
00:06 autrijus clkao: nah... that is fine
00:06 autrijus Jonathan_: so I don't know... place parrot.exe inside ~/src/pge
00:06 autrijus should fix it
00:06 clkao right, i thought it'd be handled up there. since it's already generated it doesn't matter
00:06 Jonathan_ autrijus: I did, and
00:06 Darren_Duncan as an example, *** Can't locate Fafdsfd.pm in @INC (@INC contains: /System/Library/Perl6/darwin-thread-multi-2level /System/Library/Perl6 /Library/Perl6/darwin-thread-multi-2level /Library/Perl6 .).
00:06 Jonathan_ pugs> /\w+/;
00:06 Jonathan_ *** Cannot parse PGE: \w+
00:06 Jonathan_ *** Error: Cannot find the parrot executable in PATH
00:06 Jonathan_ undef
00:06 Jonathan_ pugs> system('src\pge\parrot.exe');
00:06 Jonathan_ parrot -[abcCEfgGhjprStvVwy.] [-d [FLAGS]] [-O [level]] [-o FILE] <file>
00:06 Jonathan_ undef
00:07 Darren_Duncan the CWD is at the end of the list
00:07 autrijus oh well...
00:07 clkao autrijus: btw i'm diving into monad this week i think
00:07 autrijus clkao: cool. I think you'll like what you'll find.
00:07 clkao is there a taichi glyph ?
00:07 clkao can we overload that when defining monad?
00:07 Darren_Duncan back in a bit
00:08 autrijus ☯
00:09 clkao orz
00:09 autrijus but monad is more like ☮
00:10 autrijus the identity supported by an endofunctor and natural transofmration
00:10 autrijus to form a triad
00:10 autrijus Jonathan_: hm.
00:11 clkao not dual before that?
00:15 Jonathan_ autrijus: Gotta run, I'll look at it again soon...
00:15 Jonathan_ has quit IRC ("Leaving")
00:17 autrijus jonathan: I attempted a fix. let me know if it works when you backlogeth
00:19 justatheory has quit IRC ()
00:19 IshaCocoa_ has joined #perl6
00:19 clkao autrijus: what happens if p5 land is returning a raw hash not hashref?
00:20 autrijus clkao: how do you do that.
00:20 autrijus (without a lvalue sub)
00:20 clkao ? just return %foo
00:20 clkao as multiple value
00:20 autrijus I thought that explodes %foo?
00:20 autrijus (into k,v,k,v)
00:20 clkao yes. so can this be somehow easily reconstructed to p6 hash?
00:21 clkao so you can just add such wrapper around those sick function
00:21 autrijus it's not what we want.
00:21 autrijus in p6 when you say return %foo
00:21 autrijus er
00:21 autrijus in p5 when you say return %foo
00:21 jql no. it explodes into :k<v>, :k<v>
00:21 autrijus it means return *%foo in p6
00:21 jql I thought we went over this last night. :)
00:21 clkao ok
00:21 autrijus jql: we're talking about p5 man.
00:21 luqui depends on context in p6
00:21 jql ahh. p5. true
00:21 justatheory has joined #perl6
00:22 clkao alright. it's fine then.
00:22 luqui return %hash when want(Scalar);  # returns \%hash
00:22 clkao next. getopt.
00:22 autrijus luqui: p5!
00:22 luqui okay
00:22 * jql sees the confusion
00:22 clkao sorry for brining in perl6 here. but it's a necesary evil
00:22 clkao autrijus: \$foo in perl6.
00:22 clkao should became a tied evil in p5
00:22 luqui I just assumed you were discussing perl 5 for some analogy to perl 6
00:22 luqui but I guess you guys are doing the p5 bindings for haskell?
00:23 autrijus luqui: ah. no. clkao is writing SVK in perl6
00:23 autrijus reusing existing cpan modules
00:23 clkao luqui: we are trying to make pugs looks lovely.
00:23 clkao perl5 isn't lovely without pcan
00:23 clkao cpan
00:23 autrijus CPAN is the language, Perl5 just the syntax
00:23 autrijus (samtregar++)
00:23 clkao bwaha
00:23 luqui CPAN is the vocabulary
00:23 clkao autrijus: so, scalarref?
00:24 autrijus clkao: a Ref in p6 when passed in p5 land is just another wrapped thing
00:24 clkao yes. but it needs to be flag
00:24 clkao flagged
00:24 autrijus clkao: but we need to defined ${}
00:24 clkao so i can tie it
00:24 autrijus for it
00:24 luqui btw, what does svk buy me if I'm always online
00:24 autrijus via overload
00:24 clkao ${} ?
00:24 autrijus or some such.
00:24 clkao luqui: faster
00:24 autrijus luqui: no .svn/
00:24 autrijus luqui: incremental merges.
00:25 * luqui doesn't understand that last one
00:25 autrijus luqui: say you have a branch and trunk
00:25 clkao you don't have to, justjump on the wagon
00:25 autrijus you can say "merge new changes from trunk to branch"
00:25 autrijus and then some 10 days later
00:25 luqui clkao, that's what I did :-)
00:25 autrijus "merge new changes from trunk to branch"
00:25 clkao autrijus: what overload? that should be just tied no?
00:25 luqui ooh
00:25 autrijus and svk will just pick up the changes in the last 10 days
00:25 autrijus can't do that with svn.
00:25 luqui you can, but you have to be explicit
00:26 autrijus that is the same to say you can't.
00:26 luqui heh ok
00:26 autrijus clkao: you can approach it with tie or with ov
00:26 clkao ya being explicit = too much invasion from the tool to real work!
00:26 clkao i would avoid overload
00:26 clkao it's bringing lots of problem
00:27 luqui autrijus, where should I look to implement BUILD ($.foo) ?
00:27 clkao plus if someday people are crazy enough to support io::all...
00:27 luqui we're doing that in p6...
00:27 autrijus luqui: what does that do again?
00:27 luqui it automatically initializes $.foo
00:27 autrijus clkao: alright.. so if I pass in a [] from p6 you'll TIEARRAY
00:27 autrijus and a \$ you'll TIESCALAR
00:27 autrijus etc
00:28 clkao *nod*
00:28 autrijus clkao: that's not for this release ;)
00:28 clkao wait
00:28 autrijus I'll look at handling av returns
00:28 clkao sure av.
00:28 clkao but
00:28 clkao pugs_MkValRef
00:29 clkao do the same isa thing and bless them into pugs:scalar and pugs:array
00:29 vcv-- has joined #perl6
00:29 clkao i'll do the rest
00:29 clkao maybe this can get finished tomorrow.
00:29 autrijus luqui: it automatically initialized in what way?
00:29 clkao when's the release?
00:29 vcv- has joined #perl6
00:29 autrijus clkao: in 12 hours or so
00:29 luqui new Bar(42);  # $.foo becomes 42
00:29 luqui I mean Bar.new(42), of course
00:29 autrijus luqui: oh. wow.
00:30 autrijus sick.
00:30 clkao sucks. i am busy in the morning i think. but it should be trivial right?
00:30 clkao autrijus: can i use ref() rather than doing multipl &isa call to pugs?
00:30 autrijus clkao: sure, you can just call &ref
00:30 autrijus and stringify the rv
00:30 clkao good.
00:30 misc has quit IRC (Read error: 60 (Operation timed out))
00:30 clkao want to make a stub and i continue when i wake up?
00:31 clkao if getopt actually works i can do real first step svk in p6  :))
00:31 larsen has quit IRC ("later")
00:31 autrijus so in pugs_MkValRef
00:31 autrijus you have
00:31 autrijus    sv_setref_pv(sv, "pugs", val);
00:31 autrijus and you don't want that
00:31 autrijus instead you want to call &ref
00:31 autrijus first
00:31 autrijus and analyse the rv
00:31 autrijus and bless accordingly
00:31 ImustDIE has quit IRC (Connection timed out)
00:31 autrijus I don't see why it needs anything at hs side
00:31 autrijus it's all C
00:32 clkao it still has to be blessed as the entrace back to haskell
00:32 clkao ya, it's a call. how to stringy return value from &ref
00:32 autrijus "bless accordingly"
00:32 clkao no, blesseed and then wrapped in closure.
00:32 autrijus eh.
00:32 autrijus you just clal pugsapply
00:32 autrijus then SvPV_nolen
00:32 autrijus piece of cake
00:32 clkao otherwise you can't get back
00:32 clkao ok. good.
00:32 autrijus yeah yeah
00:33 clkao i'll do it tomorrow.
00:33 autrijus clkao++
00:33 clkao \
00:33 clkao now zz &
00:33 clkao make av return work!
00:33 awwaiid has joined #perl6
00:33 autrijus (sigh) sure
00:33 clkao and maybe write some tests for ref accessing
00:34 clkao (or find people to write tests :)
00:34 autrijus luqui: I'm not sure how to attack this.
00:34 autrijus luqui: I guess you can hack Bind.hs
00:35 luqui shouldn't it go right where the parameters are bound from the signature
00:35 autrijus but is it just BUILD?
00:35 autrijus method blah ($.y) {}
00:35 luqui I think it's everywhere
00:35 autrijus it will automagically become set-accessor to $.y?
00:35 autrijus instead of just lexically set $.y ?
00:35 autrijus and how does it play with the "param is ro by default"?
00:35 luqui right, because $.y isn't a valid parameter name anyway
00:35 autrijus surely you can always write to $.y?
00:36 luqui yeah, that has to be an exception
00:36 autrijus ($:y) too?
00:36 autrijus yay exceptions
00:36 luqui I think ($.y) is just short for ($dot_y) { $.y = $dot_y }
00:36 luqui maybe with a : before the =, not sure.
00:36 autrijus except it happens at binding time
00:36 luqui oh, right
00:36 autrijus method foo ($.y, ?$z=$.y) {}
00:37 autrijus having it to be := will be terribly surprisin.
00:37 luqui I expect $:y works too
00:37 autrijus because you end up with random aliasing.
00:37 autrijus foo(@x[1])
00:37 autrijus $obj.foo(@x[1])
00:37 svnbot6 r4181, autrijus++ | * VType to p5 land becomes a String, so clkao can typecase from C.
00:37 svnbot6 r4182, autrijus++ | * try finding parrot in cwd, for Jonathan.
00:37 autrijus $obj.y = 4; # touches @x[1]
00:37 autrijus oops
00:38 luqui good point
00:38 autrijus so line 1167 Eval
00:38 autrijus is where you want to attack
00:38 autrijus analyse "name" for $. and $: forms
00:38 autrijus and turn it into assignment somehow.
00:38 autrijus I need to run -- $work and $food
00:38 autrijus good luck!
00:39 luqui okay thanks
00:43 meppl gute nacht
00:43 svnbot6 r4183, autrijus++ | * delay the ifdef call in Parrot.hsc
00:47 justatheory has quit IRC ()
00:47 Darren_Duncan autrijus, I want to revisit that test suite thing where uppercase file names have special meanings ...
00:47 Darren_Duncan if it wasn't already done, that hack should be removed where files with uppercase chars are not executed
00:48 Darren_Duncan I want to name my tests using uc chars and have them execute
00:48 Darren_Duncan I think a better way to say 'don't run this' in the short term is to put the non-running files in a different dir name, such as 'todo_t' rather than 't'
00:48 Darren_Duncan its much more self-documenting too
00:49 luqui why do you want uppercase chars?
00:49 Darren_Duncan it reads better
00:49 luqui than lowercase with underscores?
00:49 Darren_Duncan for example, one of my desired test file names is 'LKT_00_Compile.t'
00:50 luqui oh I get it
00:50 Darren_Duncan the hack requires it to be 'lkt_00_compile.t' which is harder to read
00:50 luqui yeah I agree with you
00:50 Darren_Duncan besides, module names are mixed case
00:50 luqui maybe non-running tests should just be named .T
00:50 Darren_Duncan that's fine too
00:51 jql works like \s and \S in rules
00:51 jql I like
00:51 Darren_Duncan I just want the hack gone that prevents me from using uc chars
00:52 Darren_Duncan since the next release is 12 hours away, is it reasonable to get this change in for that?  I hope so
00:53 luqui the next release is 12 hours away?
00:54 Darren_Duncan autrijus: clkao: in 12 hours or so
00:54 Darren_Duncan twas said about 30 mins ago
00:56 IshaCocoa_ Anyone here built Parrot on win32? I'm getting a "Expanded command line '...' too long error whenit tries to "lib -out:blib/lib/libparrot_s.lib". Workarounds?
00:56 Darren_Duncan no win here
01:03 Darren_Duncan I believe the problem is this portion of the Makefiles: @ARGV = grep !/[A-Z]/, map glob, @ARGV;
01:04 Darren_Duncan right now I'm trying to figure out where that is added to them
01:06 Darren_Duncan actually, we can't use T for the alternative, unless the whole test suite is turned off
01:06 Darren_Duncan because some platforms that Pugs has to run on don't have case insensitive file systems
01:06 Darren_Duncan so we can't have both a 't' and a 'T' in parallel on them
01:08 luqui right
01:08 luqui .tx?
01:08 Darren_Duncan sure
01:08 luqui nobody likes textans
01:08 luqui texans
01:09 Darren_Duncan it just has to be different characters
01:11 theorbtw1 has joined #perl6
01:16 castaway_ has joined #perl6
01:27 theorbtwo has quit IRC (Read error: 110 (Connection timed out))
01:28 castaway has quit IRC (Read error: 110 (Connection timed out))
01:36 Darren_Duncan okay, I never found it
01:37 Darren_Duncan where is this being put into the Makefiles, so we know what to edit so this doesn't happen: @ARGV = grep !/[A-Z]/, map glob, @ARGV;
01:37 Darren_Duncan ?
01:38 DNA` has joined #perl6
01:39 sekimura has joined #perl6
01:42 meppl has quit IRC (Remote closed the connection)
01:45 DNA` has left
01:50 mugwump Darren_Duncan: didn't you say you were going to rewrite ExtUtils::MakeMaker?  :)
01:50 svnbot6 r4184, autrijus++ | * handle multiple return values.  It segfaults still --
01:50 svnbot6 r4184, autrijus++ |   clkao, please take a look.
01:51 Darren_Duncan I thought about it and then changed my mind
01:51 Darren_Duncan also it sounds like someone else was going to do it
01:52 mugwump that statement has conflicting tenses
01:53 Darren_Duncan wasn't there an announcement recently where someone else said they would do it?
01:53 mugwump anyway, makefiles are generated there.  I think ingy played with them last, but first he was possessed by a demon.  We later revived him and now he's OK
01:54 putter has quit IRC (Read error: 60 (Operation timed out))
01:54 Darren_Duncan I tried doing a number of string searches and couldn't find what I wanted in either the Pugs or Perl distro
01:56 Darren_Duncan I figured the ignore-uc thing was added explicitly for Pugs as a temp measure
01:56 Darren_Duncan normal CPAN makefiles certainly don't act that way
01:56 Darren_Duncan I just did a search for "!/[A-Z]/" and it only appears in the Makefiles themselves, not the makefile generators
01:57 Darren_Duncan still, I'm hoping that someone who wrote that functionality can speak up
01:57 Darren_Duncan seen ingy
01:57 jabbot Darren_Duncan: ingy was seen 1 days 9 hours 5 minutes 1 seconds ago
02:00 Darren_Duncan mugwump, re makemaker rewrite, more to the point it seemed like since I last discussed it someone already had did what needed doing, a month or 2 ago
02:01 Darren_Duncan ingy probably
02:01 Darren_Duncan so I thought it was done
02:07 eric256 has joined #perl6
02:29 vcv-- has quit IRC (Read error: 104 (Connection reset by peer))
02:37 vcv- has quit IRC (Read error: 60 (Operation timed out))
02:46 IshaCocoa has joined #perl6
02:50 eric256 quite in here
02:50 eric256 err quiet
02:56 Darren_Duncan tis
02:57 IshaCocoa_ has quit IRC (Read error: 110 (Connection timed out))
03:01 luqui AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA​AAAAAAAAAAAAAAAAAAAAAAAAAAAAAA'
03:01 mugwump what's all this racket about!?
03:02 luqui contrast?
03:02 Darren_Duncan the A key is stuck down
03:02 luqui no, I was screaming
03:03 Darren_Duncan I'm writing a quick note to p6c re the test file name issues so it gets more attention
03:03 luqui That or you just have to be here when somebody who has any idea what's going on is here
03:04 Darren_Duncan yes
03:04 Darren_Duncan I'll try both
03:10 mugwump Darren_Duncan: just hack a s/// into lib/Perl6/MakeMaker.pm, I think
03:11 Darren_Duncan could do, but isn't there a more elegant solution?
03:11 mugwump NO
03:11 mugwump not unless you want to rewrite ExtUtils::MakeMaker ;)
03:15 Darren_Duncan I get the impression that Perl6::MakeMaker isn't being used anymore, but rather inc::Module::Install is, so the latter should be hacked, correct?
03:19 mugwump Somewhere like inc/Module/Install/Makefile.pm (sub fix_up_makefile) perhaps
03:19 Darren_Duncan that's what I thought
03:20 hlen_ has quit IRC (Read error: 110 (Connection timed out))
03:20 hlen__ has joined #perl6
03:20 mugwump That might be copied from the Module::Install distribution, but in either case a commit there will communicate more clearly than a p6c post methinks :)
03:21 Darren_Duncan I'm deferring the p6c post
03:21 Darren_Duncan it may never go up
03:21 Darren_Duncan commit instead
03:25 rafl has joined #perl6
03:27 Darren_Duncan how does this line look to you ...
03:27 Darren_Duncan $makefile =~ s|\@ARGV = grep !/\[A-Z\]/, map glob, \@ARGV;|\@ARGV = map glob, \@ARGV;|;
03:27 IshaCocoa Does (GNU) make support inline files?
03:27 Darren_Duncan that's in fix_up_makefile() after all the other substitutions
03:27 mugwump use \Q ... \E please :)
03:27 mugwump oh, wait, no that wouldn't work either would it
03:27 Darren_Duncan what does that mean?
03:27 mugwump would interpolate
03:28 Darren_Duncan I thought my backslashing would prevent interpolation
03:28 Darren_Duncan all characters in regexp should be treated literally
03:28 mugwump \Q disables pattern metacharacters, but not interpolation
03:29 mugwump extremely handy with untrusted search strings
03:29 mugwump or when you've got a lot of brackets etc in your pattern
03:30 mugwump that change looks sufficiently harmless to me, commit and I'll smoke it for you
03:30 Darren_Duncan okay
03:31 mugwump jabbot: seen svnbot6 ?
03:31 jabbot mugwump: svnbot6 was seen 1 hours 40 minutes 25 seconds ago
03:32 * mugwump smokes r4184 as a control
03:32 jql hmm
03:34 Darren_Duncan committed 4185
03:35 jql I think literal syntax is <'@ARGV = grep !/[A-Z]/ ...'>
03:35 mugwump that's perl6 rules, though, innit?  :)
03:35 jql yes
03:35 mugwump This is a Perl5 module
03:35 jql I'm trying to remember p6
03:35 mugwump good
03:35 Darren_Duncan the literal text is: @ARGV = grep !/[A-Z]/, map glob, @ARGV;
03:36 Darren_Duncan I removed the grep
03:36 mugwump you could use \Q${\( '@ARGV = grep !/[A-Z]/, map glob, @ARGV;' )}\E :-)
03:36 mugwump but I think at that point it starts to defeat the purpose ;)
03:37 jql heh
03:37 jql @ doesn't interpolate, at least
03:37 jql you just need to escape [
03:37 mugwump it doesn't?
03:37 * mugwump tries
03:37 luqui what would it interpolate as?
03:37 Darren_Duncan while I could find nothing else in Pugs containing just '!/[A-Z]/', I thought a s/// with the longer string was more future proofing
03:37 svnbot6 r4185, Darren_Duncan++ | updated inc/Module/Install/Makefile.pm so that generated Makefiles do not ignore t/*.t scripts having uppercased chars in file names -- if you want your scripts to be ignored during 'make test', place them in a different dir than t/, such as tX/
03:37 jql well, that was a guess
03:38 * jql didn't try it
03:38 jql ok, you can laugh at me
03:39 luqui hahahahaha
03:39 mugwump perl -le 'my @a = qw(1 2 3); my $qr = qr/@a/; print $qr'
03:39 mugwump (?-xism:1 2 3)
03:39 luqui really
03:39 luqui that's unexpected in an expected sort of way
03:39 jql s/@// doesn't interpolate @, s/@@// doesn't interpolate. s/@foo// interpolated
03:39 jql damn perl
03:40 * mugwump notes that the perl5 embedding is causing a compile failure on 4184, and switches it off
03:40 jql even though @@ is a perfectly legal variable
03:40 Darren_Duncan is it now?
03:40 jql I know and love @@
03:40 jql perl -le 'push @@, 1; print @@'
03:41 jql %% is also good
03:41 jql I love me some @@
03:42 luqui it doesn't mean anything, right?
03:42 luqui '
03:42 jql define 'mean'?
03:42 luqui it's not magical
03:42 jql no, not magical
03:42 IshaCocoa has quit IRC ()
03:42 jql it's legal because $@ is legal
03:42 luqui right
03:43 mugwump but what about %%?
03:43 luqui that's legal too, I believe
03:43 jql %% is legal because $% is legal
03:43 Darren_Duncan how about @$ or %$
03:43 jql && is probably not legal, for obvious reasons
03:43 jql those are all fine
03:43 luqui Darren_Duncan, those are dereferencers
03:43 mugwump &{'&'} :)
03:43 jql well, of course. :)
03:43 Darren_Duncan by themselves?
03:44 luqui oh my god, it looks like @$ is fine by itself
03:44 jql sure it's fine
03:44 luqui just don't put \w after it
03:44 Darren_Duncan I always use braces for derefs for readability
03:44 Darren_Duncan eg, @{$foo}
03:44 mugwump yuck :)
03:44 luqui yeah, I always avoid braces when I can
03:45 luqui but, tmtowtdi, teho, etc.
03:45 mugwump I use them when ambiguous
03:45 Darren_Duncan of course, I'm often doing this: @{$foo->{'bar'}}
03:45 jql perl -le 'print %++%%
03:45 * jql hides
03:45 luqui in which case you have no other choice
03:45 Darren_Duncan you can't take the braces out then
03:45 luqui well, yeah you do I guess
03:45 luqui @{$$foo{bar}}
03:45 luqui but, uh, no
03:46 mugwump I use $foo->bar and let the accessor and context return a list :)
03:46 luqui and people say perl is linenoise
03:47 Darren_Duncan the result of $foo->bar is an array ref, which is a scalar
03:47 jql @{ my $bar = $foo->{bar}; my $baz = $bar->{baz}; $baz->{blurfl} }
03:47 Darren_Duncan I'm talking p5 here
03:47 * jql takes indirection to the Nth
03:47 luqui jql, I'm not even sure that's legal
03:47 jql it's legal
03:47 mugwump I like my array accessors to return lists in list context.
03:47 luqui no, it is
03:48 jql the {} is a do{} block which has its return-value array-derefed
03:48 jql except for a bareword, which is special-cased
03:48 jql ala. hash
03:49 Darren_Duncan question ...
03:49 Darren_Duncan do you know what the insanta/outsanta files are for?
03:50 mugwump elves?
03:50 jql sleds?
03:50 Darren_Duncan they seem related to Pugs' compilatoin
03:50 Darren_Duncan eg, insanta.988.59
03:50 Darren_Duncan but I don't know why they stick around
03:51 * jql knows not the ways of GHC
03:53 Darren_Duncan it seems my file change didn't work
03:53 Darren_Duncan trying again
03:54 Darren_Duncan mugwump, hold off the smoke for the moment
03:54 * mugwump puts down the crackpipe
03:54 Darren_Duncan I mean my version; continue with the control one
03:55 mugwump yup
04:01 Khisanth has quit IRC (Remote closed the connection)
04:05 Khisanth has joined #perl6
04:07 vcv-- has joined #perl6
04:07 vcv- has joined #perl6
04:14 mugwump Darren_Duncan: there seem to be lots of test scripts that make foo.$$.NNNN temps..
04:14 svnbot6 r4186, mugwump++ | force_todo / un-TODO various tests for release - round 1
04:15 Darren_Duncan I'm continuing to isolate the problem
04:15 Darren_Duncan I think the Makefile is post processed after Build is done with it, and that's where the A-Z etc is added, hence my change had no effect
04:17 mugwump ok, well why don't you commit something that works for you first ;)(
04:18 Darren_Duncan working on it
04:20 Darren_Duncan meanwhile, my change didn't break anything
04:20 Darren_Duncan twas a no-op
04:21 Shachaf has quit IRC ("[<>]")
04:23 Southen[a] has joined #perl6
04:32 Darren_Duncan I have confirmed a change following when fix_up_makefile() is run
04:33 luqui mugwump, by the way, I warnocked the message that you mentioned got warnocked because I didn't understand it
04:33 luqui There aren't too many people on p6l who grok Haskell, and I couldn't understand the message by ignoring the haskell
04:33 mugwump fair enough.  I barely understood it myself ;)
04:35 Darren_Duncan found the problem
04:36 Darren_Duncan inc/Module/Install/Pugs.pm is what's adding the offending line in the first place
04:36 Darren_Duncan I failed to find it before because my grepping didn't account for the target having the slashes escaped
04:36 Darren_Duncan will commit momentarily
04:40 Southen has quit IRC (Connection timed out)
04:40 svnbot6 r4187, mugwump++ | :todo / fix various tests for release - round 2
04:44 Darren_Duncan committed
04:46 Darren_Duncan mugwump, smoke away
04:47 svnbot6 r4188, Darren_Duncan++ | undid r4185's change to Makefile.pm and redid the fix properly, by altering inc/Module/Install/Pugs.pm to not add the offending section in the first place
04:47 mugwump I'm still smoking 4184 :)
04:47 Darren_Duncan no prob
04:47 mugwump or, rather, I'm marking the failures as known
04:47 Darren_Duncan meanwhile, I'm doing a clean smoke on my machine
04:47 Aankhen`` has joined #perl6
04:50 Darren_Duncan quickie ...
04:50 Darren_Duncan is there a svn command to rename a file, or do I have to delete + add?
04:50 Darren_Duncan nevermind
04:50 Darren_Duncan found it
04:52 nothingmuch good morning
04:52 Darren_Duncan time to rise and shine
04:57 * luqui feels like watching a movie
04:57 luqui any suggestions?
04:57 mugwump The Triplets of Belleville
04:57 luqui yeah, I loved that one
04:57 mugwump Shaolin Soccer
04:58 luqui oh right
04:58 luqui I saw kung fu hussle
04:58 luqui okay, thanks
04:58 mugwump if you've seen that, ... ah, beat me to it :)
04:59 mugwump what about Snatch, or Two Hands
05:02 nothingmuch Battleship Potemkin
05:07 Darren_Duncan so now the LKT test file names are back to what they should be
05:07 svnbot6 r4189, Darren_Duncan++ | LKT test file rename stage 1 of 2
05:07 svnbot6 r4190, Darren_Duncan++ | LKT test file rename stage 2 of 2
05:17 mugwump Darren_Duncan: sadly I've run out of time to smoke the new version - those tests took 47m to run on my system!  Must be due to the NFS directory :)
05:17 svnbot6 r4191, mugwump++ | :todo / fix various tests for release - round 3
05:18 Darren_Duncan mugwump
05:19 Darren_Duncan that's fine, I'm doing a full run on my system
05:19 Darren_Duncan realclean, make, test
05:20 mugwump not only that, but the smoking I did didn't even cover testing the modules ... but hopefully it will free up autrijus' time a bit :)
05:20 Darren_Duncan yep
05:21 Darren_Duncan now, you only updated the test files, right, no haskell files?
05:22 mugwump that's right
05:27 Darren_Duncan okay, I'm on 4191 and beginning make test
05:28 misc has joined #perl6
05:31 nothingmuch has quit IRC ()
05:36 simcop2387 has quit IRC ("Client exiting")
06:00 joepurl is now known as joepurl28730292
06:00 vcv-- has quit IRC (Read error: 54 (Connection reset by peer))
06:00 svnbot6 r4192, mugwump++ | Document Set.pm unicode operators as requested on p6l
06:01 eric256 has left
06:08 joepurl28730292 is now known as joepurl
06:09 vcv- has quit IRC (Read error: 60 (Operation timed out))
06:17 spinclad has quit IRC (Remote closed the connection)
06:17 svnbot6 r4193, autrijus++ | * allow calling Perl 5 methods with named arguments
06:17 svnbot6 r4193, autrijus++ |   (which are flattened into two positional arguments each).
06:18 xern has quit IRC ("Lost terminal")
06:18 larsen has joined #perl6
06:24 saorge_ has quit IRC ("Leaving")
06:24 * mugwump sidles off home
06:24 svnbot6 r4194, mugwump++ | Add some more unicode operators for the hell of it ;)
06:30 jql good one
06:30 nothingmuch has joined #perl6
06:32 Darren_Duncan finished 'make test' for 4190 ...
06:32 Darren_Duncan Failed 1/319 test scripts, 99.69% okay. 498/6489 subtests failed, 92.33% okay.
06:32 autrijus lkt?
06:32 Darren_Duncan failed test: t/rules/Disabled/rules.t              683  498  72.91%
06:32 autrijus oh, all tests
06:32 autrijus the "Disabled" test is not supposed to run anyway
06:32 Darren_Duncan I didn't do anything with LKT since yesteday except rename the test scripts
06:33 Darren_Duncan yes
06:33 Darren_Duncan so Disabled should be moved out of t/
06:33 autrijus I thought our harness already ignored it. hm
06:33 Darren_Duncan it did, but I fixed that
06:33 autrijus heh, "fixed".
06:34 Darren_Duncan now the harness will run test files no matter what their name is
06:34 luqui \did someone try to implement that "describe" program that was mentioned on the list?
06:34 autrijus Darren_Duncan: please svn mv the disabled tests somewhere else then.
06:34 autrijus t_disabled
06:34 autrijus or something. improvise.
06:34 Darren_Duncan will do
06:34 autrijus darren++
06:35 autrijus <- in $work... doing a Class::DBI based perl6 script
06:35 Darren_Duncan and thats a much more elegant solution than the old hack
06:35 autrijus Darren_Duncan: sure, I agree
06:35 Darren_Duncan funny thing is, we discussed this before, but such a remedy didn't come up
06:36 luqui autrijus, oh, I just saw the six in @autrijus[-2]
06:36 luqui lucky you, wish I could use p6 in my job
06:36 Darren_Duncan autrijus, r4185,4188 contained my fix ... it even says in the checkin comment that a good alternate way to disable is to move out of t/
06:36 Darren_Duncan so your suggestion is in line with that
06:36 autrijus luqui: the six?
06:36 autrijus Darren_Duncan: cool, please proceed then
06:37 luqui Class::DBI based perl >>6<< script
06:37 autrijus ahh.
06:37 luqui hmm.. we don't have a hyper 6 operator yet...
06:37 luqui so you don't know if anybody tried to write "describe" ?
06:38 autrijus no.
06:38 luqui okay.  time to re-invent if someone did, then :-)
06:39 Darren_Duncan autrijus, there are a lot of other files in Disabled, and they did not fail
06:40 autrijus Darren_Duncan: do you have parrot in path?
06:40 Darren_Duncan no
06:40 autrijus then those tests didn't run.
06:40 Darren_Duncan probably the reason
06:41 Darren_Duncan will move entire folder
06:41 autrijus nod.
06:45 castaway_ is now known as castaway
06:56 Darren_Duncan autrijus, is it fine to 'svn mv' a whole folder, or should I do each file inside separately?
06:56 autrijus Darren_Duncan: the whole folder.
06:56 autrijus (that will preserve history)
06:56 Darren_Duncan ok
06:57 Orkun_TR has quit IRC (Read error: 110 (Connection timed out))
06:58 Darren_Duncan I wasn't sure, cause when I did a trial run the info spat out just said "A + folder" but didn't mention each file, whereas it did mention each file for the "D" part
06:59 Darren_Duncan actually it just said "A   folder"
07:00 autrijus the "+" is right.
07:00 autrijus ah.
07:00 autrijus don't worry. it's fine
07:03 spinclad has joined #perl6
07:04 scook0 has joined #perl6
07:06 Darren_Duncan committed
07:07 autrijus cool. bbl &
07:15 Aankh|Clone has joined #perl6
07:16 svnbot6 r4195, Darren_Duncan++ | moved tests from t/rules/Disabled/* to t_disabled/rules/*; removed old Disabled
07:16 Darren_Duncan fyi, assuming that 6.2.6 will be going up in a few hours, I'll defer any further LKT work until afterwards ... its midnight here
07:16 iblechbot has joined #perl6
07:18 Darren_Duncan fyi, all of my commits today are related -- they are all specifically about letting test files have uc chars in their names, allowing those to run, and making sure files that should be disabled stay that way
07:19 Darren_Duncan I didn't do anything else
07:19 Darren_Duncan oh yeah, and actually renaming some tests to uppercase
07:24 Aankhen`` has quit IRC (Read error: 145 (Connection timed out))
07:24 Aankhen`` has joined #perl6
07:30 SM_ax has joined #perl6
07:35 dada has joined #perl6
07:36 vcv-- has joined #perl6
07:36 vcv- has joined #perl6
07:36 Darren_Duncan is there a dedicated ChangeLog updater, or does anyone do that?
07:37 castaway you can volunteer :)
07:38 castaway (generally someone does, or gets volunteered by autrijus)
07:42 Aankh|Clone has quit IRC (Read error: 110 (Connection timed out))
07:44 Darren_Duncan in any event its, obviously, out of date right now
07:47 castaway quite likely
07:47 castaway someone usually compiles it from the journal entries iirc
07:54 Darren_Duncan good night
07:54 svnbot6 r4196, Darren_Duncan++ | some MANIFEST.SKIP updates
07:54 Darren_Duncan has left
08:09 osfameron has joined #perl6
08:25 masak castaway: I think "gets volunteered" is a quite blatant contradiction in terms
08:27 vcv- has quit IRC (Read error: 110 (Connection timed out))
08:27 arcady say, does perl6 allow underscores in numeric literals?
08:28 arcady because apocalypse 2 says it does, and pugs doesn't allow it
08:31 scook0 arcady: probably just unimplemented--care to write some tests?
08:34 arcady I kind of don't have any free time at the moment
08:34 * castaway grins
08:34 castaway I think its correct tho masak :)
08:34 arcady eval("42_000") should be 42000
08:35 castaway ?eval 42_000
08:35 evalbot6 Error:  unexpected "_" expecting digit, fraction, exponent, term postfix, operator, postfix conditional, postfix loop, postfix iteration, ";" or end of input
08:35 * castaway pats evalbot on the head
08:36 scook0 I've just written up that one, any other useful test cases?
08:36 arcady with operators
08:36 castaway umm.. 42_000_000 and 42_000.31 ?
08:36 castaway right
08:37 arcady and 42.0_32
08:37 scook0 castaway: good thinking
08:37 autark-jp and 4_2000
08:37 arcady and possibly 42__0 not being valid
08:38 scook0 arcady: 42__0 is valid P5
08:39 arcady oh, is it?
08:39 arcady didn't know
08:39 castaway umm, whats 42.0_32 ?
08:39 arcady underscores in a number are ignored
08:40 arcady btw, this goes for all kind of numbers
08:40 arcady float as well as int
08:40 arcady and octal and hex too
08:40 castaway oh, sounds like an easy implementation then
08:40 autark-jp throw away all '_' except at the beginning or something maybe.
08:41 arcady yes
08:41 castaway right
08:41 svnbot6 r4197, bsmith++ | Added an export list.
08:41 svnbot6 r4198, bsmith++ | Moved some parser types to Pugs.Parser.Types.
08:42 autark-jp anyone on good terms with the haskell type system btw?
08:42 scook0 autark-jp: depends, what's your issue?
08:43 autark-jp scook0: in Pugs.Lexer.hs, interpolatingStringLiteral
08:43 autark-jp scook0: it takes in something called 'RuleParser x'
08:44 scook0 autark-jp: aye, do you understand what that is supposed to mean?
08:44 integral autark-jp: A (RuleParser x) is a just a parser than returns some unspecified type
08:45 Jouke has joined #perl6
08:45 autark-jp scook0: it says (in comments) that it is a delimiter (for strings)
08:45 integral autark-jp: notice that where endrule is used lower down we don't use it's value
08:45 Jouke has left
08:45 autark-jp integral: ok, but I want to use it's value ;)
08:45 autark-jp integral: and then the type system starts yelling at me :/
08:45 integral hmm, you'll probably need to change it to RuleParser Exp then
08:46 integral what type of value do you expect it to have?  an expression?
08:46 scook0 the type system should be telling you the type it's expecting
08:46 autark-jp scook0: it says something about rigid variable x can't be match to 'char' or 'string' (depeneding on whether delim is a single char or mulit-char)
08:47 autark-jp ie. <<foo>> vs. <foo>
08:47 scook0 hmm
08:47 * autark-jp wants q(foo(bar)baz) to be legal.
08:47 scook0 so, people are already giving it both chars and strings?
08:48 scook0 or is that what you're trying to do?
08:48 integral this is harder.   You'll have to change where it's called so that they always use strings
08:48 castaway hmm, that got discussed a while back , but I cant remember what the outcome was
08:48 autark-jp it is being given both both string (symbol?) and char it seems.
08:48 autark-jp integral: hm, I should maybe try that.
08:50 clkao bad autrijus can't write c
08:51 castaway who wants to write C :)
08:57 luqui autark-jp, I think we've said that the delimiters nest unless they're backwhacked
08:57 svnbot6 r4199, clkao++ | Fix perl5_apply returning multiple values and the use in mkval generating closure.
08:58 autark-jp luqui: sounds reasonable.
08:58 theorbtw1 is now known as theorbtwo
09:08 fireartist has joined #perl6
09:08 svnbot6 r4200, scook0++ | Tests for numeric literals with underscores
09:10 fireartist warnocked :) Could someone take a look and set me straight?...
09:10 fireartist http://www.mail-archive.com/perl6​-language@perl.org/msg20241.html
09:10 fireartist (question about class construction)
09:13 luqui lookin
09:13 luqui g
09:14 luqui Oh, I read that
09:14 luqui I'm not really sure what you were asking
09:14 luqui But in any case, if you write your own new, you have to call BUILD yourself
09:14 fireartist oh, ok, I'll try again now,
09:14 fireartist hmm, didn't realise I'd have to call build myself
09:15 luqui the recursive BUILD call is something that is done by the default new
09:15 luqui we generally don't write our own news (no pun intended) in perl 6
09:15 fireartist is build not called by BUILDALL? Do I have to call that too?
09:16 luqui maybe that's the one you call
09:16 fireartist the reason I defined my own 'new', is the object's usually constructed with only one arg
09:16 fireartist I wanted a shortcut, rather than having to pass it my name everytime
09:17 luqui oops
09:17 luqui I'm wrong
09:17 fireartist (by name everytime)
09:17 luqui apparently bless calls BUILDALL for you
09:18 fireartist the email was trying to clarify if that's what is supposed to happen, because I was seeing some strange behavour with pugs...
09:19 luqui what were you seeing
09:20 fireartist just a minute, I'm opening my editor to make sure I get this right
09:20 luqui Couldn't you just write:
09:21 luqui submethod BUILD (Str $.date, +$.time, +$.offset)
09:21 fireartist my explicit new method was being run, when I was calling 'new' with a named pair of args
09:21 luqui oh, you want your multi to "append" behavior to the default new?
09:22 luqui hmm
09:22 fireartist and in my build, the arg was undef
09:22 * luqui writes up a little test case to see what you're talking about
09:23 fireartist I wanted the option of either a single arg, or all named args
09:23 fireartist http://www.fireartist.com/foo.pl shows the behaviour
09:24 IshaCocoa has joined #perl6
09:25 fireartist it prints 'ok', showing my new is called, rather than the new inherited from Class
09:25 fireartist it also prints date: ''
09:25 fireartist meaning the arg in build is undef
09:25 luqui hmm, I think it should call your new.
09:26 luqui but the fact that date is undef looks wrong
09:26 fireartist I thought Foo.new(date => 'blah') wouldn't match my 'new' signature
09:27 luqui that depends on whether the default new is a multi
09:27 luqui well, and whether that makes any sense with our multi system
09:27 luqui a couple of us (including me) think that the multi system needs a rework
09:27 fireartist I'm not sure whether the design I'm going for is perl6'ish
09:28 luqui what I find perplexing is the fact that your "return $class.new(...)" doesn't fall into infinite recursion
09:29 fireartist if I change it to Foo.new, it does!
09:29 evalbot6 has quit IRC (Remote closed the connection)
09:29 Juerd For some reason, pugs locks itself
09:29 luqui But $class should be the Foo object
09:29 Juerd So I had to kill evalbot6 in order to be able to upgrade pugs
09:29 luqui I mean Foo class object
09:29 fireartist If I add 'say $class.perl' in 'new', it reads \{obj:Class}
09:29 Juerd autrijus: Any idea how to avoid this?
09:29 luqui right, not very informative
09:30 Juerd 1;0 juerd@feather:~/auto/pugs$ cp pugs ../bin
09:30 Juerd cp: cannot create regular file `../bin/pugs': Text file busy
09:30 fireartist Should it not be \{obj:Foo} ?
09:30 luqui something like that
09:30 clkao how's mod_pugs?
09:30 luqui If you just "say Foo.perl"
09:30 luqui you get the same thing
09:31 luqui okay, this is definitely not what is supposed to be happening
09:31 luqui I don't grok the object internals though
09:31 fireartist outside of the class, after I've constructed $foo, is reads \{obj:Foo}
09:31 luqui maybe you should post the example and behavior to p6c
09:32 luqui Not for me
09:32 fireartist I'll do that thanks, I'll try to be more clear what I'm asking though!
09:32 luqui $foo.perl does, but not Foo.perl.
09:32 vcv has joined #perl6
09:32 luqui as far as how you would do it in Perl 6, though
09:32 luqui I think you're best off putting all the logic into BUILD
09:32 luqui and avoid writing your own new altogether
09:32 vcv how do i create a list of 26 items (the letters of the alphabet)? i feel stupid, but my brain isnt working
09:33 luqui 'a'..'z' ?
09:33 vcv doesnt seem to be working.
09:33 luqui works for me
09:33 luqui what codE?
09:33 fireartist thanks for your help luqui, I'll start on that email now.
09:33 luqui thanks for the email ;-)
09:34 vcv pugs> @chars »~« '+'
09:34 vcv 'a b c d e f g h i j k l m n o p q r s t u v w x y z+'
09:34 vcv that cant be right
09:34 vcv @chars >>~<< '+' works
09:34 luqui huh?
09:35 luqui it looks like its working to me
09:35 luqui arrays join on spaces when interpolated (just like in p5)
09:35 scook0 vcv: maybe your pugs is hosed
09:36 luqui the example that he just showed looks like it's working fine
09:36 vcv ok. why would @chars »~« '+' and @chars >>~<< '+' return different results
09:36 luqui because one is stringifying @chars and then appending '+'
09:36 luqui the other is appending '+' to each element of @chars, and returning that list
09:36 luqui the same reason @chars + 1 and @chars >>+<< 1 are different
09:36 vcv i thought >><< was ascii version of »«
09:37 scook0 vcv: »~« worksforme
09:37 vcv i just cant understand why »~« and >>~<< are working differently
09:38 luqui so, "@chars >>+<< 1" means "map { $_ + 1 } @chars"
09:38 scook0 vcv: I'm not seeing any difference when I try it
09:38 luqui whereas "@chars + 1" means [email@hidden.address] + 1"
09:38 luqui vcv, are you a perl 5 programmer?
09:38 vcv i understand what it does luqui
09:38 vcv im tellin you »~« and >>~<< are doing different things
09:38 scook0 vcv: maybe you need to update & rebuild
09:38 luqui than what they're supposed to, or than each other
09:38 vcv just did
09:38 vcv than each other
09:38 luqui they're supposed to
09:39 luqui oh!
09:39 luqui I must be unicode-stupid
09:39 luqui You're probably typing the unicode version of >>~<< and I'm just seeing ~
09:39 luqui okay
09:39 vcv oh! haha
09:39 vcv i bet i looked like a real idiot ;)
09:39 luqui but now it is I hahaha!
09:39 luqui :-)
09:40 luqui are you doing this on the pugs command line?
09:40 vcv yes
09:40 luqui have you tried writing it in a script?
09:40 vcv ill try that
09:40 luqui the pugs command line might be unicode stupid too
09:40 arcady it works fine for me though
09:41 luqui terminal settings?  all that unicode bullshit that we knew we'd run across once we accepted the latin-1 variants
09:41 scook0 hmm, it must be the command line
09:41 theorbtwo ...or your shell might be unicode-stupid.
09:41 theorbtwo Mine was, until I started tweaking things.
09:41 arcady ?eval ("a", "b", "c") »~« (1, 2, 3)
09:41 scook0 Juerd had to kill evalbot for a bit
09:42 arcady oh whoops
09:42 arcady it's kinda early
09:42 arcady didn't notice that
09:42 castaway hmm? it was there just now..
09:42 arcady yeah, I think it's the unicode issue
09:42 castaway oh, missed that, so he did
09:42 vcv >pugs test2.p6
09:42 vcv a b c d e f g h i j k l m n o p q r s t u v w x y z,
09:42 vcv a,b,c,d,e,f,g,h,i,j,k,l,m,​n,o,p,q,r,s,t,u,v,w,x,y,z,
09:43 vcv code is: my @chars = 'a'..'z';
09:43 vcv say @chars »~« ',';
09:43 vcv say @chars >>~<< ',';
09:43 scook0 vcv: your terminal seems to be throwing away «»
09:43 arcady yeah, that really looks like it
09:43 vcv but im running it from a file darn it! :/
09:43 arcady because you get @chars ~ ',' for the first one
09:44 luqui vcv, it's got to be unicode, because it works for me
09:44 scook0 vcv: yeah, the file seems to be working correctly
09:44 vcv no that cant be it
09:44 vcv watch..
09:44 scook0 sorry, no, that's wrong
09:44 lumi vcv: Are you sure you're doing it in unicode and not latin1?
09:44 vcv abcdefghijklmnopqrstuvwxyz
09:44 vcv a b c d e f g h i j k l m n o p q r s t u v w x y z,
09:44 vcv a,b,c,d,e,f,g,h,i,j,k,l,m,​n,o,p,q,r,s,t,u,v,w,x,y,z,
09:44 vcv say @chars;
09:44 vcv say @chars »~« ',';
09:44 vcv say @chars >>~<< ',';
09:44 luqui that makes sense too
09:44 vcv errrr
09:44 theorbtwo vcv: I think you are outputting latin1 into the channel instead of utf8.
09:45 vcv whoops made a mistake
09:45 theorbtwo Could you give us a euro sign, please?
09:45 vcv alright. thanks. stupid latin1
09:45 vcv hm.. yeah
09:45 theorbtwo (A good test because the euro sign is not in latin-1, since it didn't exist at the time.)
09:47 vcv im a unicode newb.
09:47 vcv well i use alt+0171 and alt+0187 to get the characters
09:47 vcv if that helps
09:47 luqui you're on windows
09:47 luqui ?
09:48 vcv yes
09:48 * luqui doesn't know anything about windows unicode
09:48 * theorbtwo knows very little about it.
09:48 theorbtwo What IRC client are you using?
09:48 theorbtwo Somewhere, there should be a setting for what charset it expects data from the server to be in.
09:49 vcv mirc because there is no good client for windows. yet. :)
09:49 luqui xchat on cygwin ;-)
09:49 vcv ha!
09:49 lumi irssi
09:49 vcv irssi is nice, but some reason i cant stand it when im on windows.
09:49 osfameron xchat on windows it pretty good
09:50 luqui oh, there's an xchat for windows
09:50 luqui well, screw cygwin then
09:50 vcv Yeah. used it and didnt like it
09:50 vcv i like to customize my stuff a lot and mircs scripting is easy, so
09:50 luqui afaik xchat is scriptable with perl
09:50 luqui never used it though
09:51 arcady is there no irc client written in perl6 yet?
09:51 luqui hahaha
09:51 luqui no, I don't think so
09:51 kungfuftr you can, Net::IRC runs on pugs
09:51 arcady yeah, someone better write one
09:52 luqui does curses or gtk or ptk or any of those?
09:52 vcv i use this when im bored: http://weakmind.org/zephyr/zephyr.png
09:53 theorbtwo Looks like mIRC has no utf8 support.
09:53 vcv figures
09:54 luqui has quit IRC (Remote closed the connection)
09:54 vcv though it doesnt look like its mirc, since the file does it too
09:55 luqui has joined #perl6
09:55 IshaCocoa Anyone know there way around the Parrot makefile on Windows?
09:55 vcv doubt there is
09:55 vcv whats wrong with the makefile?
09:55 luqui heh, I was trying to see if I had a combining key and accidentally pushed ctrl-alt-backspace :-p
09:55 theorbtwo Swift there, luqui.
09:55 scook0 luqui: fantastic
09:56 Aankhen`` theorbtwo >> That is correct.
09:56 Aankhen`` There are many petitions and requests for Unicode in mIRC.
09:56 Aankhen`` The author has said he will add it... eventually.
09:56 theorbtwo xmodmap -pk|grep -i multi will tell you if you have one.
09:56 vcv No he wont. he's too busy sailing on his yacht
09:56 * luqui is unskilled in xmodmapfu
09:56 Aankhen`` vcv >> I'm just relaying what he said. :-P
09:57 vcv Yeah. he's said a lot of things :p
09:57 luqui theorbtwo, doesn't seem I do
09:57 theorbtwo I use xmodmap -e 'keycode 94 = Multi_key', which makes the 102nd key on my european keyboard in en_US layout be the compose key.
09:57 luqui how do I give myself one
09:57 luqui ok
09:57 vcv then i realize mirc sucks and is going no where, so i starting writing my own client
09:57 luqui what's the keydump utility so I can find the scancode of a key
09:58 theorbtwo Do xev, and hit the key until you figure it out.
09:58 theorbtwo It's not exactly nice, but it works.
09:58 luqui yep
09:59 luqui ?woo? unicode
10:00 luqui and that goes into .xmodmaprc?
10:00 Jouke has joined #perl6
10:00 Jouke has left
10:00 luqui vim thinks so, so I'll trust it
10:01 Aankhen`` vcv >> Are you seriously writing a new IRC client?
10:02 luqui that may be a bit hubristic
10:02 luqui got to make sure you're lazy and impatient about it, otherwise it will go nowhere
10:02 luqui (I've found that the three virtues only work well when you have all three together in one room)
10:04 theorbtwo luqui: Haven't quite figured out where it goes, sorry.
10:05 theorbtwo luqui: Since you probably write better then me, put that info in the unicode minihowto thingy?
10:05 luqui where's that?
10:07 theorbtwo docs/quickref/unicode
10:07 clkao summon autrijus
10:07 clkao VType typ   -> vstrToSV (showType typ)
10:08 luqui oh, it seems that there's already a url there that does it
10:08 clkao this doesn't seem to catch what ref() returns
10:08 clkao so it's doing infinite recusion on mkValRef
10:08 luqui nope, it doesn't
10:08 luqui it's lower level than what we need
10:08 luqui adding it
10:09 kolibrie has joined #perl6
10:09 Jouke has joined #perl6
10:09 Jouke has left
10:11 nothingmuch yum yum dum dee dum
10:11 * theorbtwo wonders if nothingmuch is going to break our bones to make his bread.
10:12 * nothingmuch instead eats a jewish chicken
10:12 nothingmuch eek, what horrible english.
10:12 Aankhen`` Racist!
10:12 Aankhen`` What's wrong with us normal people, huh?
10:12 Aankhen`` We're not good enough for you?
10:12 * Aankhen`` pouts.
10:12 nothingmuch well, i was paraphrasing
10:12 theorbtwo s/jewish/kosher/
10:12 nothingmuch nay nay
10:13 Aankhen`` Hang on a sec while I look up 'kosher' in the dictionary.
10:13 nothingmuch jewish as in jewish mother (or father, in my dad's case)
10:13 Aankhen`` Hrm.
10:13 * Aankhen`` goes back to calling nothingmuch a racist.
10:13 castaway the chicken had jewish parents?
10:14 nothingmuch http://www.amazon.com/exec/obidos/tg/sim-explo​rer/explore-items/-/067084487X/0/101/1/none/pu​rchase/ref%3Dpd%5Fsxp%5Fr0/104-4000845-8113544
10:14 nothingmuch no, it was just behaving jewishly
10:14 * nothingmuch goes to help something compile
10:14 theorbtwo No, just a jewish father.  (Which is nice, and all, but neither neccessary nor sufficent, at least according to some people.)
10:15 theorbtwo Of course, I don't think chickens can have souls, much less Jewish ones, so it's really a pretty moot point.
10:16 * Aankhen`` koshers theorbtwo.
10:16 luqui oh, now you're being politically correct for sure theorbtwo
10:16 Aankhen`` Why can't chickens have souls?
10:16 * scook0 nominates this for "most OT discussion EVER" :-)
10:16 luqui this will lose
10:17 luqui The jewish faith and perl 6 have tons in common
10:17 * theorbtwo waits for luqui to give an example.
10:17 svnbot6 r4201, luqui++ | Add unicode tips.
10:17 luqui exactly what is left as an exercise for the reader
10:17 * Aankhen`` waits too.
10:17 Aankhen`` Cop out!
10:17 * Aankhen`` glares at luqui.
10:17 theorbtwo Total cop-out.
10:17 luqui I'm american
10:18 luqui roll a dice
10:18 Aankhen`` Ah, that explains it. ;-)
10:18 luqui if it doesn't come up six, then I was being sarcastic
10:18 theorbtwo "They are both not-so-secretly run by the Cabalists."
10:19 theorbtwo "They both require Unicode."
10:19 luqui huh?
10:19 luqui oh
10:19 luqui haha
10:19 luqui there we go
10:19 luqui and we were just talking about unicode
10:19 luqui see, not OT
10:20 pjcj Old Testament?
10:20 theorbtwo The word Cabal comes from Kabbalah, a Jewish tradition of hidden knowladge.
10:20 luqui Apocalypses
10:20 luqui or Perl 5
10:20 luqui can't decide
10:20 * luqui steps back to hide in his ignorance
10:20 luqui and watch a movie
10:20 * luqui &
10:21 * castaway declares lunch
10:21 theorbtwo Blast, scared away the Great Summorisor.
10:22 castaway ?
10:23 theorbtwo luqui is Luke Palmer, who... oh, nevermind.
10:23 * theorbtwo was thinking of somebody else.
10:23 * theorbtwo is very good at confusing people.
10:24 castaway as long as you dont confuse me with someone :)
10:24 theorbtwo You're Jill, right.
10:24 castaway close
10:24 theorbtwo Jane?
10:24 castaway nope
10:25 theorbtwo Ann?
10:25 * theorbtwo kisses his Jess.
10:25 castaway :)
10:25 * castaway munches a salad sandwich
10:25 castaway soup time!
10:27 * theorbtwo wonders what the best way of making a haskell-level Param into a perl6-accessable scalar is.
10:28 scook0 theorbtwo: trying to expose params as perl6 values?
10:28 castaway hmm, dont bits already do that.. like $?PUGS_VERSION ?
10:29 * nothingmuch wants tea
10:29 * castaway passes the thermos
10:29 fireartist has quit IRC ("Leaving")
10:31 nothingmuch so anyway, in the stinky cheese man there's a giant, which says that he'll make bread out of several characters
10:31 nothingmuch and there is also an annoying chicken
10:31 nothingmuch which says that no one helped her make the bread or something
10:32 nothingmuch and then  the giant eats the chicken with the chicken's bread, IIRC
10:36 * Aankhen`` pokes nothingmuch and runs around laughing maniacally.
10:36 nothingmuch tee hee
10:36 nothingmuch anyway, i meant to get me some tea
10:36 theorbtwo Oh.
10:37 theorbtwo I was thinking of Jack and the Beanstalk, and didn't get the Jewish Chicken reference at all.
10:39 meppl has joined #perl6
10:40 nothingmuch lemon tea with  honey
10:40 * nothingmuch feels sick
10:41 hlafarge has joined #perl6
10:42 nnunley has joined #perl6
10:42 clkao my $rw = 1; warn \$rw; why is it still 1 ?
10:43 nothingmuch clkao: i think autodereferencing goodness, but I'm not up to date on that stuff
10:44 theorbtwo s/goodness/badness/
10:45 nothingmuch i think autodereferencing could have it's goodness, if it were not confusingly special cased
10:45 nothingmuch the problem is that sane semantics for autodereferencing, are.... well... tough.
10:45 nothingmuch i don't like having to explicitly dereference with @{ } and so on in p5
10:45 nothingmuch but at least if I do it slowly and surely, balancing everything beforehand, things turn out OK
10:46 theorbtwo I don't like having to explicitly dereference all the time either.
10:47 theorbtwo But it means that I know what's going to happen in any given situation, and can treat things generically without too much worries.
10:47 nothingmuch exactly
10:47 theorbtwo I can have a tree class, into which you can put references or nonreferences, without having to do anything special.
10:47 theorbtwo I'd like a simple rule that says when something gets autodereferenced and when it does not.
10:48 theorbtwo ...and I'd like the answer to be "does not" most of the time.
10:49 nothingmuch i think this is like lazyness and autothreading
10:49 nothingmuch the only place where it should be controlled is the absolute edge
10:49 nothingmuch either at the lowest level, or the highest level
10:49 lathos has quit IRC ("leaving")
10:49 theorbtwo Oh!  Lathos is Simon Cozens!
10:50 theorbtwo Now I get his references to being in the Cabal.
10:50 nothingmuch Oh!
10:50 nothingmuch He Is Indeed!
10:50 wolverian has quit IRC (Read error: 60 (Operation timed out))
10:51 nothingmuch I on the other hand googled yesterday, and didn't have to realize it in a boom
10:51 nothingmuch BUAHAHAHA
10:51 theorbtwo Oh.
10:51 theorbtwo I looked in AUTHORS, but he wasn't there.
10:51 nothingmuch ah
10:51 theorbtwo I didn't google, after assuming that "lathos" didn't sound very googleable -- there were probably too many.
10:52 autark-jp gah! any way to speed up 'make test'? splice.t just took 1m15s :/
10:52 castaway get a faster machine
10:52 castaway ,)
10:52 autark-jp ;)
10:52 nothingmuch autark-jp: use feather
10:52 autark-jp feather?
10:52 nothingmuch Juerd's machine
10:55 autark-jp ah, and suddenly the topc makes sense :)
10:55 * autark-jp <- slow
10:58 clkao how do i call assignment with a sub? &infix:<=> ?
10:58 nnunley theorbtwo:  You probably got a few too many references to Chrono Trigger.
10:58 wolverian has joined #perl6
10:59 theorbtwo That should do it, clkao.
11:00 clkao it doesn't work. *** Undeclared variable: "&infix:="
11:00 theorbtwo The emphasis may be on /should/.
11:01 clkao ok. if i want to write a failing test where does it belong?
11:01 clkao to
11:03 scook0 currently '=' only exists as a Syn, not a regular function...
11:03 clkao eek. will that change?
11:03 scook0 clkao: I don't know what the specced behaviour is, that's just the current implementation
11:03 scook0 t/operators/assign.t ?
11:04 theorbtwo scook0: I don't know of anything specced on = that would make it different from any other op.
11:04 theorbtwo (And every op should be a function with a strange name.)
11:05 scook0 the only difficulty is evaluating the LHS in LHS context, but I suppose that should be feasible somehow
11:07 theorbtwo Hm, good point.
11:08 theorbtwo It might need to be a macro (and we don't have those yet... something which I should work more on fixing).
11:08 kungfuftr perl6 is moving way too quickly for me to keep track... =0(
11:11 castaway ditto, kungfuftr
11:11 theorbtwo macros have been semispeced for a while.
11:11 theorbtwo They just haven't been implemented.
11:13 nothingmuch hah! in our product we have a method named dearHierarchyGiveMePleaseHeightOfMember
11:14 * theorbtwo laughs aloud at nothingmuch's product.
11:15 castaway .)
11:15 scook0 reminds me of INTERCAL...
11:15 nothingmuch i'm 90% sure I Know who wrote it too =)
11:16 theorbtwo You don't have a svn blame (or equiv) to tell you?
11:19 IshaCocoa_ has joined #perl6
11:24 theorbtwo Mornin, IshaCocoa.
11:25 nothingmuch theorbtwo: p4 can do that in an acward way
11:25 nothingmuch ackward
11:25 nothingmuch i have to trace back through branches manually
11:25 theorbtwo Oh, perforce, not perl 4.
11:25 theorbtwo Ugh.
11:25 * castaway wondered there too
11:25 nothingmuch and I have to convert changelist numbers to authors
11:25 theorbtwo You could write a script to do it... but it probably wouldn't be worth the effort.
11:26 nothingmuch and just to find out who has a nice sense of humor, it's too much work =)
11:27 clkao the accessing perl6 data as reference requires the assigment function to work.
11:27 clkao failed test committted.
11:27 clkao s/the/to
11:27 autrijus greetings.
11:27 autrijus clkao: I want auto TIEHASH
11:28 clkao autrijus: i'm having it.
11:28 clkao autrijus: doesn't work because i can't assign
11:28 clkao i'll finish the tie stuff later
11:28 IshaCocoa_ Howdy theorbtwo.
11:28 autrijus clkao: what can't assign?
11:29 clkao if you can fix the assignment, and ($rw = 1; ref(\$rw) eq 'Scalar') i'll be hapy
11:29 clkao autrijus: see my last comit
11:29 autrijus ref(\$rw) is Int
11:29 autrijus always is
11:29 clkao it returns '1' :)
11:29 autrijus pugs> my $rw = 1; ref(\$rw)
11:29 autrijus ::Int
11:29 autrijus ?eval my $rw = 1; ref(\$rw)
11:30 clkao not right
11:30 castaway evalbot is gone walkies
11:30 svnbot6 r4202, clkao++ | Failed test for assignment operator.
11:30 castaway (Juerd said recompiling pugs confuses it or something)
11:30 clkao it's same as ref($rw) wtf?
11:30 castaway why not, if $rw is an int?
11:31 clkao so reference is the same?
11:31 clkao because we have "is rw" so we don't use reference to scalar anymore?
11:31 castaway $rw = 'foo'; ref($rw); -> still an int?
11:31 castaway oops ref(\$rw)
11:32 autrijus pugs> my $rw = 'foo'; ref(\$rw)
11:32 autrijus ::Str
11:32 clkao and try print ref(\$rw)
11:32 theorbtwo say ref 1; say ref \1; say ref \\1;
11:32 autrijus pugs> my @x; (ref(@x), ref(\@x))
11:32 autrijus (::Array, ::Array)
11:32 theorbtwo At least one of those is wrong.  I'd argue that two of them are.
11:32 autrijus pugs> ref \(\1)
11:32 autrijus ::Int
11:32 theorbtwo Autoderef is ucky.
11:32 clkao autrijus: the value of that is wrong
11:32 autrijus clkao: the value of what?
11:33 castaway ok, I'm confused too now
11:33 autrijus theorbtwo: I don't disagree.
11:33 clkao my $rw = 1
11:33 clkao print \$rw
11:33 clkao this is what got passed into p5
11:33 clkao and fscked
11:33 autrijus ok.
11:33 autrijus I agree it's a bug.
11:34 autrijus I further agree that infinite dereferece, although at one time mandated by larry, is now out of fashion
11:34 autrijus so pugs will fix
11:34 autrijus but probably not today
11:34 ChanServ sets mode: +o autrijus
11:34 theorbtwo On a completely different note...
11:34 clkao eek, fix that before i come back from lunch
11:34 autrijus clkao: not possible...
11:34 clkao otherwuise perl5 accessing p6 ref won't work
11:34 clkao it invovles p6l?
11:35 evalbot6 has joined #perl6
11:35 autrijus sets mode: -o autrijus
11:35 autrijus clkao: no, it involves rewriting 30% of AST.Internals.
11:35 theorbtwo I'd like to have programatic access to the body of expressions.  Is doing instance Value Exp sane?
11:35 autrijus theorbtwo: go ahead. see DirStream code
11:36 autrijus use opaque and fromObj
11:36 castaway ?eval $?PUGS_VERSION
11:36 evalbot6 \'Perl6 User\'s Golfing System, version 6.2.5, May 24, 2005 (r4201)'
11:36 theorbtwo Cool!
11:36 autrijus ok, dinner now. bbiab &
11:39 IshaCocoa has quit IRC (Read error: 110 (Connection timed out))
11:40 broquaint has joined #perl6
11:43 theorbtwo Allo, bro.
11:47 broquaint Hey, theorbtwo.
11:47 broquaint What's the latest?
11:48 theorbtwo I'm trying to make expressions introspectable.
11:49 broquaint Huzzah!
11:49 theorbtwo LISP with syntax, baby.
11:49 broquaint More introspection would be very handy at the moment.
11:49 IshaCocoa has joined #perl6
11:49 broquaint LISP has syntax? ;)
11:49 theorbtwo No -- perl6 does.
11:50 broquaint Hehe.
11:50 * theorbtwo wonders what the difference between vCast and doCast is.
11:50 Odin-LAP Which is rapidly becoming the single difference between the two...
11:50 theorbtwo (They have the same signature.)
11:50 broquaint vCast is a lazy cast?
11:51 theorbtwo Ah.
11:51 theorbtwo Wait, no.
11:51 broquaint That was a guess.
11:51 theorbtwo Those would have different signatures -- the nonlazy version would have a !.
11:52 nothingmuch blah!
11:52 nothingmuch Class::Events takes 20 seconds to parse on feather
11:52 nothingmuch and then generates an error
11:53 theorbtwo Oh.  Everything in value has default instances anyway.
11:53 theorbtwo So I only need to implement vCast and castV and everything else hopefully just works.
11:56 * broquaint tries to track down a recursive infinite loop in some mjd code ...
11:57 * broquaint 's brain teh splode
12:02 meppl has quit IRC (Remote closed the connection)
12:02 broquaint Awesome. Tron.
12:06 IshaCocoa_ has quit IRC (Read error: 110 (Connection timed out))
12:18 sekimura is now known as sekimuraway
12:19 Arathorn has joined #perl6
12:20 saorge has joined #perl6
12:23 * luqui just finished watching shaolin soccer
12:23 luqui very much fun :-)
12:24 luqui thanks mugwump
12:26 luqui has quit IRC ("Leaving")
12:27 vcv has quit IRC (Read error: 60 (Operation timed out))
12:35 broquaint ?eval my $s = -> { say @_ }; $s.(q[b0rk])
12:35 evalbot6 Error: Undeclared variable: "@_"
12:36 broquaint That's not right, is it?
12:37 broquaint Yeah, that used to work. Hrm ...
12:38 IshaCocoa has quit IRC (Read error: 110 (Connection timed out))
12:38 theorbtwo Looks wrong to me, but I'm not sure.
12:39 autark-jp isn't -> {} saying, I take no variables?
12:39 autark-jp but $_ get set.
12:40 broquaint ?eval sub f(Code $a) { $a } my $s = f { say @_ }; $s.(1)
12:40 evalbot6 Error: Undeclared variable: "@_"
12:40 broquaint Using sub {} works, but -> {} and a bare {} doesn't work. It seems odd. It also means a program that was working, does so no longer.
12:42 castaway oops
12:43 theorbtwo ?eval [(sub {}).arity, (-> {}).arity, ({}).arity]
12:43 evalbot6 [1, 0, 0]
12:43 theorbtwo Yup.
12:43 castaway whats that?
12:43 broquaint Shouldn't they all behave the same?
12:44 scook0 broquaint: they used to, but it was explicitly changed recently
12:44 theorbtwo I seem to recall Autrijus mentioning the other day that they were clarified not to.
12:44 autark-jp so if you define f { say @_ }, f(1,2) will work, but not f(1)?
12:44 broquaint Could someone point me to where this was clarified?
12:45 fireartist has joined #perl6
12:46 scook0 I don't think it's on p6l...checking the irc logs...
12:46 Juerd re evalbot6: it can run on feather if it doesn't use the global pugs
12:46 Juerd Because for some reason, the pugs executable gets hard-locked when in use
12:47 Juerd And then the automatic upgrades can no longer take place
12:47 Juerd I hope autrijus has a solution for this, and that the locking isn't necessary
12:47 clkao autrijus: aarrg. :/
12:48 scook0 broquaint: start looking from http://colabti.de/irclogger/irclogger_log/p​erl6?date=2005-05-30,Mon&amp;sel=891#l1346
12:48 broquaint Thanks, scook0 :)
12:49 theorbtwo r4160 | autrijus | 2005-05-30 11:04:07 +0200 (Mon, 30 May 2005) | 2 lines
12:49 theorbtwo * according to luqui, bare {} and pointy -> both have no
12:49 theorbtwo  default argument at all, unlikes sub/method/coro that has *@_.
12:50 theorbtwo Whoops, sorry.
12:50 theorbtwo Too slow.
12:50 scook0 just checked S06
12:51 broquaint Thanks, theorbtwo, that makes sense quicker than reviewing IRC logs ;)
12:51 scook0 it implies that only sig-less (p5-style) subs get @_, but it isn't 100% clear
12:51 castaway hmm
12:54 SM_ax has quit IRC (Read error: 104 (Connection reset by peer))
12:54 broquaint That should probably be noted in autrijus' next use.perl update.
12:55 * castaway wonders where that is declared..
12:55 castaway I cant find it anywhere in S06 or 04
12:56 scook0 castaway: the @_ thing?
12:56 broquaint It makes porting functional perl5 code a little verbose.
12:56 castaway yes
12:56 scook0 in S06 under 'perl5ish subroutine declarations'
12:57 scook0 but it doesn't explicitly say that @_ doesn't show up for  {} & ->{}
12:57 castaway well under the stuff about pointy subs it says
12:57 castaway The arrow operator -> is almost a synonym for the anonymous sub keyword, except that the parameter list of a pointy sub does not require parentheses, it does not require a preceding comma when included in a list, and a pointy sub may not be given traits.
12:57 broquaint Seems odd you have to declare the arguments for what is meant to be huffmanized :/
12:58 castaway (hmm, those are to anon subs not real subs.. )
12:58 scook0 I think the 'almost a synonym for anonymous sub' thing is misleading
12:58 castaway the perl5ish bit doesnt mention anon subs not having @_ at all
12:59 scook0 In my mind, ->{} is closer to {} than sub {}
13:00 castaway ah but what about Larrys? :)
13:00 scook0 castaway: perhaps we need a p6l ruling
13:01 broquaint @_ should always be the default arg stack. *Having* to declare it every time seems oddish.
13:01 broquaint Oh well, another quirk to learn, or unlearn, as it were.
13:02 castaway sounds crazy to me
13:02 scook0 perhaps we need @?ARGS or something
13:02 scook0 then @_ can default to an alias to that
13:02 castaway surely thats just a name for the same thing?
13:02 scook0 (I'm trying to crystalise my thoughts at the moment...)
13:03 scook0 anyway, I think there's a difference between @_ as implicit argument, and @_ as an extra array that gets all arguments
13:04 scook0 we also need to figure out whether ->{} and {} take 0 args or unlimited args
13:05 broquaint I don't see why @_/@?ARGS shouldn't be the default for all codesque things that don't have an signature, no signature shouldn't imply no args/$_.
13:05 broquaint Pointy and bare subs should have 'unlimited' args.
13:05 scook0 broquaint: that's probably best
13:05 broquaint It just means that if the args aren't used then the compiler can just optimize them away.
13:06 broquaint Or something.
13:06 scook0 I think @_ should be an alias to (or copy of) all the arguments, rather than being an argument itself
13:06 broquaint For sure.
13:07 castaway makes sense to me.. isnt it now?
13:07 scook0 currently, sub {} effectively expands to sub (*@_) {}
13:07 scook0 which works, but is less general
13:07 scook0 (IIRC)
13:07 broquaint Where as -> {} and {} now have an implicit arity of 0.
13:08 broquaint &code.arity = Inf ;)
13:11 Limbic_Region has joined #perl6
13:11 scook0 so sub {}, ->{} and {} should all get an implicit slurpy param that doesn't have a name
13:11 scook0 and you access the arguments through @_
13:11 scook0 (which isn't a param, but an array of all the arguments)
13:12 broquaint a la Perl5 :)
13:12 svnbot6 r4203, autrijus++ | * Kwid::Event::Parser unbroken and test fixed.
13:13 arcady whoa, the unary + makes the multiple dispatch confused
13:13 scook0 I must remember to take this proposal further...
13:13 arcady or am I just confusing myself
13:14 arcady ah, so I am
13:14 broquaint A perl6-lang post perhaps, scook0?
13:15 arcady but I'd think that +$a is more like an Int than an Str
13:15 arcady (though it's really a Num)
13:18 nothingmuch has quit IRC ()
13:19 arcady specifically, if I have a multi sub foo(Int) and a multi sub foo(Str)
13:19 arcady foo(2) call the Int one
13:19 arcady and foo(+2) calls the Str one
13:22 iblechbot has quit IRC (Read error: 104 (Connection reset by peer))
13:23 ramzi has quit IRC (Read error: 104 (Connection reset by peer))
13:23 svnbot6 r4204, autrijus++ | * small random typo fix to hw2005 talk
13:23 svnbot6 r4205, autrijus++ | * &infix<=> can now be used as the asignment operator.
13:23 Limbic_Region having read the recent commits - are we preparing for a release today?
13:28 castaway AFAIK
13:29 scook0 broquaint: I'm writing up a p6l post, but I'm not getting very far--I'm just about to go to bed
13:29 svnbot6 r4206, autrijus++ | * unTODO and remove unneccessary unshift @*INC
13:29 svnbot6 r4207, autrijus++ | * guarantee access to blib6/lib in `make smoke`.
13:29 svnbot6 r4208, autrijus++ | * unTODO and fix bad :todo syntax.
13:29 svnbot6 r4209, autrijus++ | * declaring `sub Qulified::name` inside a package no longer
13:29 svnbot6 r4209, autrijus++ |   prepends the package name to the sub name.
13:29 castaway karma autrijus
13:29 jabbot castaway: autrijus has karma of 511
13:30 castaway (hmm, that should read Inf :)
13:30 broquaint I could take a stab if you like, scook0.
13:30 IshaCocoa has joined #perl6
13:31 scook0 broquaint: feel free--I probably won't get another chance in the next 24h :(
13:31 broquaint Alrighty.
13:31 scook0 well, goodbye all
13:31 scook0 has quit IRC ("Leaving")
13:32 castaway when do you start the job, bro?
13:32 broquaint June 13th.
13:32 castaway nifty
13:33 * castaway hopes they allow PM/IRC access :)
13:33 broquaint So two weeks of intense relaxtion :)
13:33 castaway *g*
13:33 broquaint I should think so. They seem a like a very sensible outfit.
13:33 castaway yay
13:34 castaway hmm, their list claims to still be looking for perl devs
13:36 castaway mm, "other european languages" :)
13:38 broquaint Is Haskell a European language?
13:39 castaway Umm.. probably?
13:39 * castaway was thinking of german :)
13:42 jhorwitz has joined #perl6
13:44 wilx has quit IRC ()
13:50 osfameron has quit IRC (Read error: 60 (Operation timed out))
14:03 IshaCocoa has quit IRC ()
14:05 * Limbic_Region isn't sure but he thinks efforts have been taken to prevent users from using anything other than IE
14:05 * Limbic_Region had to go through strange steps to get FF working this morning
14:05 Khisanth with pugs?
14:05 * Limbic_Region hopes it is all just a bad dream and tomorrow things will work fine
14:05 Limbic_Region huh?
14:06 vcv has joined #perl6
14:06 Limbic_Region I was referring to sensible outfits allowing access to certain locations
14:06 Limbic_Region and commenting on the fact that I am not sure I work for a sensible outfit
14:06 Limbic_Region one could argue that "The U.S. Government" by definition could never be considered a sensible outfit
14:06 Limbic_Region but hey - who am I to judge
14:07 stevan has joined #perl6
14:07 Khisanth hmm slightly different imo :)
14:08 Khisanth anyone who pays taxes should be able to access those!
14:08 Khisanth what the hell are they doing that would require IE specific features anyway?
14:08 theorbtwo Not quite sensible -- the hemline on the skirt is /all/ wrong.
14:08 Khisanth (otoh a FF specific site is equally bad)
14:08 Limbic_Region completely wrong perspective Khisanth
14:08 Limbic_Region it is about control
14:08 Limbic_Region not features
14:09 theorbtwo Khisanth: Not IE specific features -- only allowing the computers to run IE.
14:09 Limbic_Region they can lock you down to a specific app with a specific set of plugins with a specific proxy with a specific set of configurations
14:09 Khisanth ah see THAT could be useful
14:09 Khisanth for a politician
14:10 Khisanth a press diverting tool ;)
14:11 Limbic_Region given the number of people and sub organizations and geographic dispersity - it makes sense to have security controlled at a cetral location
14:11 Limbic_Region you can't trust that people will do the right thing (either intentionally or accidently)
14:12 Limbic_Region anyway - back to being on topic
14:12 Limbic_Region so there is a release scheduled sometime in the near future?
14:12 Limbic_Region 6.2.6?
14:13 theorbtwo I think Autrijus wanted to release it a bit ago, but that didn't happen somehow.
14:14 Limbic_Region I was just a bit confused - normally the topic changes
14:14 Limbic_Region when I looked at the commit history I got excited
14:15 castaway afaik, yes there is
14:15 castaway (didnt I answer that just now?)
14:15 Limbic_Region a while back
14:15 mauke has quit IRC (Remote closed the connection)
14:16 Limbic_Region I have ulterio (sp?) motives castaway
14:16 castaway oh dear :)
14:16 Limbic_Region I need to go fight with the INS in a couple hours but was thinking about pinging a few people who have said they are waiting for a stable release of Pugs that has features x, y, and/or z
14:17 osfameron has joined #perl6
14:17 Limbic_Region I think I convinced blokhead to join the madness
14:18 nothingmuch has joined #perl6
14:19 castaway we need more cooks?
14:19 svnbot6 r4210, autrijus++ | * MMD tiebreaking: remove sub type and globalness from the criteria
14:19 svnbot6 r4211, fireartist++ | * updated porting_howto with dotted-patenthesis requirement on method calls
14:21 mauke has joined #perl6
14:21 Khisanth .()?
14:21 fireartist <whitespace>.()
14:21 fireartist S12
14:21 Limbic_Region castaway - yes, I think so
14:24 Limbic_Region has quit IRC ("trying something")
14:24 svnbot6 r4212, duff++ | Added whitespace sensitivity of subs/arrays/hashes to the porting guide
14:25 Arathorn cooks are a state of flux, anyway, surely :)
14:30 Limbic_Region has joined #perl6
14:38 clkao my @a = (1,2);my $ar = \@a;warn $ar.elems;
14:38 clkao is this right?
14:38 svnbot6 r4213, duff++ | Slight reformat of docs/other/porting_howto
14:38 clkao we i need to (@$ar).elems?
14:38 clkao (it works in pugs, i am just unsure about p6l)
14:39 Limbic_Region why wouldn't it work?
14:39 Limbic_Region that isn't meant to be flippant btw
14:39 * Arathorn (very) vaguely remembers something about autodeferencing
14:39 Arathorn uh, autodereferencing even
14:39 Limbic_Region you would expect $ar[1] = 42; # to work right?
14:40 Limbic_Region the . is -> afterall
14:40 Limbic_Region and so $ar.elems is dereferning
14:40 * Limbic_Region can't speel
14:41 sahadev has joined #perl6
14:43 clkao ok, so @ar[1] and $ar[1] will both work. any of them is recommended form?
14:43 Limbic_Region hang on as that is a bit confusing
14:43 Limbic_Region at least to me
14:43 autrijus clkao: depends on what you expect it to behave in slurpy context.
14:43 autrijus say @ar; # each elements
14:43 theorbtwo clkao: They both do something, but they do different things.
14:43 autrijus say $ar; # a single ref
14:44 Limbic_Region thanks theorbtwo and autrijus
14:44 autrijus clkao: think of sigils as a context hinter.
14:44 * Limbic_Region was hung up on the "both work" part
14:44 autrijus they probably only have that much use now.
14:44 saorge has quit IRC ("Leaving")
14:45 theorbtwo Eh?  @a[1] works on the array in @a.  $a[1] works on the array reference in $a.
14:45 theorbtwo Different things.
14:45 clkao autrijus: ya, but the problem is single ref odesn't work right?
14:47 autrijus clkao: single ref doesn't work right how?
14:49 Shachaf has joined #perl6
14:49 clkao my @a = (1,2);my $ar = \@a;warn $ar;
14:49 clkao i thought $ar should just give me a reference so i can pass it as a single value to perl5
14:49 castaway Im guessing it does, but warn is dereffing it?
14:50 castaway warn $a.[0] also works there
14:50 clkao hmmm. ok.. i will try
14:50 autrijus clkao: $ar is a reference.
14:51 autrijus you don't need the \, even.
14:51 castaway (annoying tho, how do you now tell if you have a reference?)
14:52 CapoeiraB has joined #perl6
14:52 autrijus castaway: I think the idea is you don't :)
14:52 castaway Ugg
14:52 castaway why not? how does one debug there?
14:52 autrijus well you can do
14:52 autrijus ?eval (1,2).perl
14:52 evalbot6 '(1, 2)'
14:53 autrijus ?eval my $x = (1,2); $x.perl
14:53 evalbot6 '[1, 2]'
14:53 castaway having @a and \@a printed looking the same sounds like it will just lead to confusion
14:53 PerlJam Who was it that was talking about transparent vs. opaque references on p6l?
14:53 PerlJam This is exactly why.
14:53 autrijus luqui was it.
14:54 castaway well thats something at least
14:54 clkao oh ok.
14:54 castaway hmm, and is there a code-ish way of telling whether $ref is a hashref or an arrayref?
14:55 clkao ref($ref) ?
14:55 autrijus ?eval my $x=(1,2); $x.ref
14:55 evalbot6 ::Array
14:55 autrijus ?eval my $x={1=>2}; $x.ref
14:55 evalbot6 ::Hash
14:55 castaway did you just say that gave Str etc?
14:55 castaway Ok
14:55 osfameron_ has joined #perl6
14:55 clkao autrijus: so if i use pugs_apply for @a, and i want to apply &infix:<=>, @a[1], 123
14:56 clkao how do i do that?
14:56 autrijus clkao: simple
14:56 autrijus cv = pugs_eval("{ @^x[1] = $^y }")
14:57 Limbic_Region PerlJam - the opaque vs transparent references has implication on the number of levels dereferences (1 vs Inf)
14:57 autrijus pugs_apply(cv, @a, 123)
14:57 clkao what what what?
14:57 Limbic_Region basically, it sounds like Luke sees both sides of the argument and wants both
14:57 autrijus clkao: use the ☯!
14:57 Limbic_Region but isn't sure how the syntax should work
14:57 clkao i don't see pugsS_eval is exported?
14:58 autrijus clkao: it's exported.
14:58 clkao ok.
14:58 clkao i'll try tonight
14:58 autrijus extern Val *pugs_Eval ( char *code );
14:58 autrijus in pugsembed.h
14:58 autrijus so just use ☯ to implement them
14:58 IshaCocoa has joined #perl6
14:58 clkao ok ok
14:58 clkao but what was the \$ thing.
14:59 clkao it was passed in as the int value
14:59 kolibrie well, I posted my time formatting proposal to p6l -- hopefully no flames erupt
14:59 clkao my $rw = 1, p5_code(\$rw)
14:59 PerlJam Limbic_Region: it's too bad we can't say something like $ref.*{$foo}  to mean infinite deref, while $ref{$foo} would be only one
15:00 clkao what's in p5_code is 1, not a pugs blessed thing
15:00 autrijus clkao: you need to declare the prototype is "is rw" I think
15:00 Limbic_Region PerlJam - I don't fully grasp the implications
15:00 Limbic_Region If you feel you do and have a good idea for solving, please post to Luke's thread
15:00 autrijus clkao: I can do that for all methods
15:00 autrijus clkao: in the p5 side thatis
15:00 autrijus ?eval role Class::Events::Publisher {};
15:00 evalbot6 undef
15:00 Limbic_Region I am sure that there are at least 1 pending decision by @larry that is going to mean a significant overhaul of pugs code
15:00 PerlJam Limbic_Region: nah, I'm straddling the fence right now and don't see what I would consider a good solution
15:01 Limbic_Region as autrijus has said in the past - thanks for having lots of tests
15:01 clkao autrijus: ok. let me know when it's done
15:01 autrijus clkao: so just method invocation for now?
15:01 jhorwitz autrijus: fully qualified subs have broken mod_pugs -- so now i'd like to extract the package & subname using breakOnGlue in Compile/Parrot.hs, but of course i'm running into monad issues, which still perplex me...
15:01 clkao autrijus: wait, if declareed as rw it's still not a refernece that p5 sees? because i need to tie it
15:02 autrijus clkao: if declared as rw, ref won't deref.
15:02 autrijus clkao: which is what you want I persume
15:02 autrijus clkao: i.e. $x = 1; perl5_sub($x)
15:02 autrijus is _always_ going to pass $x as ref
15:02 autrijus only perl5_sub(1) gets 1
15:02 autrijus as int
15:02 clkao what? gets 1 then it's not ref?
15:03 autrijus perl5_sub(1) # gets 1
15:03 autrijus $x = 1; perl5_sub($x) # gets VRef
15:03 autrijus of course 1 is not a ref
15:03 clkao haha.
15:03 autrijus I hope, at least
15:03 clkao ahah, i misread.
15:04 clkao ok, but does mkvalref already seess that ref?
15:04 autrijus clkao: yeah
15:04 clkao good then.
15:04 autrijus jhorwitz: hrm, sorry. :-/ what monad issue?
15:04 Southen has joined #perl6
15:04 Juerd autrijus: Do you know why pugs locks itself?
15:04 jhorwitz autrijus: sorry, got distracted...
15:04 autrijus Juerd: hm?
15:04 autrijus let (Just (package, name')) = breakOnGlue "::" name
15:04 autrijus that should do
15:05 Juerd autrijus: The automated upgrades broke, because evalbot kept ~juerd/auto/bin/pugs locked
15:05 jhorwitz autrijus: i've tried that, except for the "let".  will try that now.
15:05 Juerd autrijus: And it seems every running pugs locks itself
15:05 autrijus "locked" ?
15:05 Juerd 11:36 < Juerd> cp: cannot create regular file `../bin/pugs': Text file busy
15:05 autrijus as in flocked?
15:05 Juerd no, a flock can be ignored...
15:05 autrijus try "cp -f"
15:05 Juerd I have no idea what kind of lock this was.
15:06 Juerd Never saw this error message before, and I wonder why pugs is a "text file"
15:06 Juerd I'll try
15:06 hlafarge_ has joined #perl6
15:06 autrijus clkao: meanwhile focus on hash and array first
15:06 autrijus clkao: I'll get the scalar ref thing done
15:07 Juerd autrijus: That work
15:07 Juerd autrijus: That works
15:07 jhorwitz autrijus: that worked.  i'm still not completely grokking this though -- why did the let work where "<-" didn't?
15:08 autrijus jhorwitz: because <- takes, as rhs, a compuatation
15:08 autrijus of type (Eval a)
15:08 autrijus and the lhs is of type (a)
15:08 autrijus but breakOnGlue has type
15:08 autrijus [a] -> [a] -> Maybe ([a], [a])
15:08 clkao autrijus: yes, i've had something. i ugess i can finish them tonight if the cv pugs_eval thing works
15:08 autrijus so no Eval is involved
15:08 autrijus clkao: I think it'd work
15:08 jhorwitz autrijus: getting it, thanks
15:08 autrijus :)
15:09 autrijus clkao: nice... I got SB and Class::DBI working with pugs today
15:09 autrijus if passing hashref can work I wouldn't need the helper modules anymore
15:09 hlafarge_ has quit IRC (Remote closed the connection)
15:10 clkao autrijus: oh ?? but export doesn't work yet.
15:10 clkao autrijus: see the readme items.
15:10 autrijus clkao: both modules are OO
15:10 autrijus clkao: so export is irrelevant
15:10 autrijus (but yeah)
15:10 clkao alright
15:10 clkao i've got plans for export.
15:10 clkao need some help from you
15:11 autrijus cool. I'd love to hear about them and what I can do
15:11 autrijus but I need to take a nap now
15:11 autrijus stayed awake for far too long again
15:11 autrijus bbl :) &
15:12 hlafarge has quit IRC ("KVIrc 3.2.0 'Realia'")
15:12 osfameron has quit IRC (No route to host)
15:12 harold has joined #perl6
15:18 fireartist has quit IRC ("Leaving")
15:20 clkao autrijus: hear about them = i wrote them in perl5/README
15:22 Southen[a] has quit IRC (Client Quit)
15:30 bd_ has quit IRC (Read error: 110 (Connection timed out))
15:47 mrborisguy has joined #perl6
15:52 harold has quit IRC (Client Quit)
15:54 zeromode has joined #perl6
15:55 Limbic_Region has quit IRC (""time to fight with the INS"")
15:55 vcv has quit IRC (Read error: 60 (Operation timed out))
16:01 zeromode has quit IRC ()
16:02 zeromode has joined #perl6
16:07 vitaminmoo has quit IRC (No route to host)
16:09 zeromode has left
16:09 zeromode has joined #perl6
16:10 mrborisguy has quit IRC ("leaving")
16:15 gaal how do i get runEvalIO env & opEval to not dump the AST?
16:17 gaal and does the eval name argument have magic significance anywhere?
16:31 elmex has joined #perl6
16:35 ninereasons has joined #perl6
16:36 dada has quit IRC ("Schneeschlag")
16:37 Southen[a] has joined #perl6
16:46 IshaCocoa_ has joined #perl6
16:47 nothingmuch has left
16:52 IshaCocoa has quit IRC (Read error: 110 (Connection timed out))
16:55 Southen has quit IRC (Connection timed out)
17:01 Obadiah has joined #perl6
17:02 osfameron_ has quit IRC ("Leaving")
17:04 broquaint has quit IRC ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
17:13 zeromode has quit IRC (Read error: 60 (Operation timed out))
17:18 IshaCocoa_ has quit IRC (Read error: 110 (Connection timed out))
17:56 ilogger starts logging #perl6 at Tue May 31 17:56:00 2005
17:56 ilogger has joined #perl6
17:56 cognominal has joined #perl6
17:56 sorje has joined #perl6
17:56 kungfuftr has joined #perl6
17:57 IshaCocoa_ has quit IRC (Read error: 110 (Connection timed out))
17:58 lumi has joined #perl6
18:05 ninereasons msg evalbot6 eval 1+1
18:05 ninereasons oops
18:05 Arathorn :o
18:11 gaal where's the place to put a util i use in building pugs, but is itself in haskell? till now things like this were put in util, but they were all perl5.
18:13 gaal yay! namespaces were fixed! who did that?
18:13 nothingmuch has joined #perl6
18:15 gaal okay, i have a working prelude on my box. it's still a little rought though.
18:15 gaal seen autrijus
18:15 jabbot gaal: autrijus was seen 3 hours 3 minutes 57 seconds ago
18:15 zeromode has joined #perl6
18:21 integral is there a spec for pack in perl6 yet?
18:21 gaal i don't think so, integral.
18:21 integral pity :-)
18:21 gaal inline c code and use a struct? *duck*
18:22 integral inline perl5 ;-)
18:25 zo1 has joined #perl6
18:26 Southen has joined #perl6
18:27 Southen[a] has quit IRC (Client Quit)
18:28 cdpruden has joined #perl6
18:29 zo1 has quit IRC ("Download Gaim: http://gaim.sourceforge.net/")
18:31 Southen[a] has joined #perl6
18:33 simcop2387 has joined #perl6
18:35 justatheory has joined #perl6
18:36 zuhn has joined #perl6
18:37 justatheory has quit IRC (Client Quit)
18:40 nothingmuch fucking beurocrats! i have so far found 10 pages telling me when to replace my "Travel Document in lieu of National Passport" but I still don't have the slightest bit of a clue what the fuck it is!
18:41 nothingmuch GAH!
18:41 gaal what you need, my friend, is an agent.
18:41 nothingmuch i have an agent, i just forgot to call her
18:41 nothingmuch today, that is
18:41 nothingmuch so i said "oh, i'll google instead"
18:42 Chewie[] Do Parrot's string functions support multi-byte characters?
18:42 gaal call your agent. it's like calling your mother, only it's more likely to get you a trip abroad.
18:42 Aankhen`` has quit IRC (Read error: 54 (Connection reset by peer))
18:43 Chewie[] If so, is there a way to take bytes out of a string individually? I'm trying to implement base64.
18:43 nothingmuch Chewie[]: i think so, since parrot has unicode, but we don't know much about it as a rule of thumb
18:43 nothingmuch #parrot on irc.perl.org is a better place to ask
18:43 Chewie[] ah.
18:43 nothingmuch afaik leo posted some string changes about 2-3 months ago
18:44 nothingmuch regarding the way strings are stored inline, i think
18:44 nothingmuch but i'm not sure
18:44 nothingmuch string register OTOH have
18:44 nothingmuch are supposed to have full unicode support
18:45 Chewie[] Hmm.
18:45 iblechbot has joined #perl6
18:47 Southen has quit IRC (Connection timed out)
18:50 gaal anyone with Makefile-fu around? Any reason why Config.hs isn't unshifted to @srcfiles?
18:57 gaal how do i locally turn off debug prints, the ones trifggered by Pugs.Eval.evaluate ?
18:58 gaal .... or is that a leftover debug message someone forgot to remove?
18:59 eric256 what is with all the temp files created?  why arn't they cleaned up with some sort of batch rm temp* after testing is done?
19:00 gaal are you on windows?
19:00 eric256 yea
19:00 gaal there's an unhappy condition there. it goes like this:
19:00 gaal they are unlinked
19:00 gaal but handles aren't closed at the right time yet
19:00 gaal and on windows you can't delete an open file
19:01 gaal so... junk. :(
19:01 eric256 why don't the tests explicitly close the handles? or is that the borked part you mean?
19:01 gaal however! i'm just about to check in standard pipe open, so you're welcome to refactor some of the tests not to use temp files at all :-)
19:01 Aankhen`` has joined #perl6
19:02 Odin- has quit IRC (Read error: 104 (Connection reset by peer))
19:02 Odin- has joined #perl6
19:02 eric256 Pipe.open?  that was part of the namespace thing you where waiting for right?
19:02 Odin-LAP has quit IRC (Read error: 104 (Connection reset by peer))
19:02 gaal they do close on the p6 level, but apparently there's a problem deeper down. Ask Corion next time he's around.
19:02 gaal yes. for now you need to call it Pipe::pipeopen (note the ::)
19:03 eric256 Pipe::pipeopen seems a bit redundant. why not juse Pipe::open?
19:03 gaal because namespaces are still a little bit flaky :)
19:03 gaal it will be Pipe.open
19:08 Southen[a] has quit IRC (Client Quit)
19:08 clkao autrijus: bah, pugs_Eval isn't working properly. i think it evals twice or something weird. i'm going to do some svk refactoring tonight instead
19:09 gaal r4216. if the debug prints bother you, don't svn  up, or better yet, make a patch  :-)
19:11 gaal ./pugs -e 'File::openfile(<AUTHORS>).slurp.print'
19:16 gaal ./pugs -e 'Pipe::openpipe("pugs -v", :r).slurp.say'
19:16 gaal hmm, the :r is required, i think it's a bug.
19:17 gaal when you say sub foo (+$r is Bool), but call &foo w/o :r, $r
19:17 gaal ESCdd
19:18 gaal my bad, i should 'is copy'.
19:21 nothingmuch has quit IRC ()
19:22 rafl has quit IRC (Read error: 113 (No route to host))
19:23 Odin-LAP has joined #perl6
19:29 gaal (fixed as 4217)
19:37 nnunley has quit IRC ("This computer has gone to sleep")
19:48 SamB has quit IRC (Read error: 104 (Connection reset by peer))
19:57 vcv has quit IRC (Read error: 145 (Connection timed out))
19:59 gaal has quit IRC ("goodbye")
20:00 gaal has joined #perl6
20:01 khisanth_ has joined #perl6
20:01 Khisanth has quit IRC (Nick collision from services.)
20:02 khisanth_ is now known as Khisanth
20:06 stevan has quit IRC ("Leaving")
20:09 SamB has joined #perl6
20:11 stevan has joined #perl6
20:13 gaal is eval(Str $code, +$lang = "perl") the correct signature?
20:16 gaal and: if it is, will this now break? is(eval 'some_code', "expected_result", "look ma, no parens")    ?
20:29 _shachaf has joined #perl6
20:30 _shachaf has quit IRC (Read error: 104 (Connection reset by peer))
20:32 Aankhen`` G'night.
20:32 Aankhen`` has quit IRC ("To err is human, to forgive is against FidoNet policy [Time wasted online: 1hr 30mins 38secs]")
20:34 Corion has joined #perl6
20:35 Corion Shouldn't @foo[-1] work? It's a parsefail for me ...
20:35 gaal Hey Corion! :)
20:35 Corion Hi gaal
20:35 gaal I'd imagine it should.
20:35 Corion Then my error is elsewhere I guess :)
20:36 gaal say, does concurrent smoke work for you? i want it back :-)
20:36 Corion gaal: No, doesn't work for me, but I didn't try it
20:36 Corion It says "Concurrent smoking disabled" or something like that
20:37 gaal ah :) does homegrown perl on w32 have fork?
20:37 gaal it says that because i put in that message.
20:37 Corion gaal: It has fork emulation. Fork emulation only works in very specific cases.
20:38 * eric256 wonders if concurrent smoking is anything like chain smoking
20:38 gaal hmm. where's the specific specified? :)
20:38 Corion eric256: Faster.
20:38 Corion gaal: In the minds of the people programming the fork emulation.
20:38 Corion And incidentially, the fork emulation only does what these programmers wanted/needed
20:38 gaal that is unfortunate. well, i'll try it, i guess.
20:38 Corion (sockets and filehandles don't really survive fork)
20:39 jhorwitz has quit IRC ("Chatzilla 0.9.67 [Firefox 1.0.4/20050511]")
20:39 Corion (pipes are also a bad idea)
20:40 gaal pipes? really!? beh
20:41 Corion I think that's what made me decide that parallel smoking won't be. But you're free to try. I'd recommend a pipe-free approach though.
20:41 gaal hmm, this debug message is annoying. i think i'll remove it.
20:41 Corion I think that open(F, "|-") mostly works.
20:41 gaal no, it forks now.
20:42 gaal i'm sure open |- works, because Test::Harness depends on it on all platforms.
20:42 fireartist has joined #perl6
20:43 gaal btw, the channel is curiously quiet today.
20:43 iwt has joined #perl6
20:43 mj41 has joined #perl6
20:44 Corion Not my fault I hope
20:44 gaal well, you could try making more noise.
20:45 gaal but it's funny considering autrijus said he was planning a release this local_morning
20:47 eric256 why does smoke testing require so many modules?
20:47 eric256 what ever happened to the KISS theory? ;)
20:47 gaal eric256, 'make test' is the kissy alternative
20:48 mj41 hello,
20:48 mj41 my $fh = open 'mjserver-dyn.p6';  my $code = slurp $fh;  print $code;  
20:48 mj41 Has slurp problem with long  (>550 B) files?
20:48 Corion It shouldn't
20:48 eric256 yea. since i can't get the modules needed for smoke running i guess i'll settle for that
20:49 gaal seeing as "make smoke" outputs nice HTML, not having it use external modules would require inlining a templating tool, thus violating KISS badly
20:49 gaal what can't you get?
20:50 gaal Test::TAP::Model requires a new Test::More. it's s a known bug that it doesn't say so.
20:51 eric256 Test::TAP was the problem.
20:51 eric256 in my world KISS normaly refers to the end users experience not the programmers. ;)  in this case endusers just happen to be programmers.
20:51 zeromode has quit IRC (Read error: 110 (Connection timed out))
20:53 eric256 thanks that seems to have fixed the problem
20:53 mugwump morning
20:53 gaal hmm, nothingmuch about?
20:53 gaal we really should update the module's META.yml and fix that already.
20:54 eric256 couldn't Test::TAP::Model be updated to show that as a requirment? or is that what you meant gaal?
20:55 mj41 perlbot nopaste
20:55 perlbot Paste your code here and #<channel> will be able to view it: http://sial.org/pbot/<channel>
20:55 gaal yes, it is.
20:56 Corion Oy. Is the build broken? Or do I need a clean checkout ?
20:56 pasteling "mj41" at 147.229.221.107 pasted "Should anybody verify slurp bug (for file >550B) ?" (18 lines, 300B) at http://sial.org/pbot/10624
20:57 Corion I get lots of loading errors for (for example) ext\CGI\t\params.t
20:58 gaal paste, Corion? those coudl be just the debug messages i was looking to eliminate
20:58 pasteling "Corion" at 84.176.33.233 pasted "Debug message?" (4 lines, 96B) at http://sial.org/pbot/10625
20:58 Corion Yeah - it could be  a debug message
20:59 Corion Yeah - it's a debug message. That'll screw with some of my tests ;)
21:00 gaal yes, it's that. gimmie a minute, it's on its way out.
21:07 gaal do you grok ControlEnv?
21:08 Corion Who, me?
21:08 cdpruden has quit IRC ("Leaving")
21:08 Corion I don't grok much Haskell
21:09 gaal np, i'll figure it out.
21:10 Corion BTW, I imagine reading earlier that autrijus wanted glob() in Pugs ... That'll get ugly again, because Win32 usually has spaces in filenames/directories, and Unix folks never consider that.
21:10 gaal well, at least now you can implement it in perl.
21:10 Corion (other than that, globbing is a problem of the shell - it should come as a module instead of a builtin)
21:11 gaal (see Prelude.pm -- it's loaded automatically now.)
21:11 Corion Ah
21:11 gaal http://www.livejournal.com/users/gaal/164263.html
21:13 saorge has joined #perl6
21:13 Corion gaal: Heh - maybe I should have given you %ENV<PUGSOPT> for that. %ENV<PUGSOPT> = '-MPrelude' should do the same, but PUGSOPT doesn't exist yet :)
21:14 gaal i can't imagine why anyone wouldn't want a way to open files :)     (much of prim is going there too)
21:15 Corion Yeah - it's cool to have the things that are easy in Perl6 directly available as Perl6 code :)
21:16 G2 has quit IRC ("oops")
21:16 gaal r4219
21:17 gaal i hope i didn't turn on debugging globally, but i'm sure i'll hear about it if i did. :-p
21:17 wolverian has quit IRC (Read error: 110 (Connection timed out))
21:17 eric256 shouldn't that be class Pipe { sub openpipe } so that  stuff can happily go after it?
21:17 gaal yes, but i think that still doesn't work.
21:18 gaal see t/pugsbugs/namespace.t
21:18 eric256 stubs of what needs implemented would be good too... like an S29 outline in there....maybe S29 could be pod in that document. dunno
21:18 gaal (which need more tests!)
21:18 gaal indeed! go for it, eric256.
21:19 gaal also, there's a big job waiting to be done to p6ify eval_*
21:19 gaal it's big because of t/.
21:20 iwt has quit IRC (Client Quit)
21:21 gaal sysinternals.com++ # tools that make w32 nicer to use
21:22 kolibrie has quit IRC ("leaving")
21:29 cognominal has quit IRC (Client Quit)
21:31 cognominal has joined #perl6
21:32 fireartist has quit IRC ("Leaving")
21:37 vcv- has joined #perl6
21:39 geoff__ has joined #perl6
21:40 Corion Does ext\Perl-MetaModel\t\00-compile.t Just Run very long or does it hang?
21:40 Corion Pugs has already taken 6 minutes running it I believe ...
21:42 gaal i'd forgotten to start a smoke... checking.
21:43 gaal are you sure you're optimized?
21:43 * eric256 has been running a smoke for a while now...
21:43 eric256 almost an hour in fact...that is insane
21:44 Corion gaal: "nmake && perl run_smoke.pl"
21:44 gaal if that's from clean, it should be optimized.
21:44 Corion I guess there is an infinite loop, as pugs.exe has amassed 10 minutes of CPU now. Killing it.
21:44 gaal but you may have a leftover unopt ast?
21:44 Corion gaal: I do a "make clean" before that always.
21:45 Corion ext\Perl-MetaModel\t\10_Perl_Meta_MetaClass_test.t crashes after a few seconds. That's better than running forever :)
21:46 ramzi has joined #perl6
21:47 geoff__ Is there a doc or url that lists all of the packages a pugs user should have installed to enable everything that pugs can do?  I must be missing something obvious, because that info seems very distributed.
21:47 eric256 you shouldn't need any modules installed for pugs to enable all the core stuff.
21:48 gaal you need parrot for p6 rules...
21:48 gaal that's pretty core imo
21:48 Corion geoff__: Parrot and Perl5
21:48 geoff__ eric256, the kind of stuff gaal is mentioning.
21:48 gaal geoff, that's a good document to have. start one?
21:48 Corion The rest is included with Pugs from what I know. And GHC of course...
21:48 geoff__ gaal, I will as soon as I have pugs running the first time.  :-)
21:48 Corion (and Parrot isn't as crucial yet :) )
21:49 gaal well, i'd say test::tap::htmlmatrix and dependencies (many!) for smokes, but that isn't required for pugs itself.
21:49 justatheory has joined #perl6
21:50 geoff__ My eventual goal is to port some of my SDL / OpenGL stuff to PUGS, so any stuff I might need to do that linkup would be cool to know about as well.  Inline for P5, at least, yes?
21:50 gaal Corion, I'm worried a bit about w32, because fancy ghc stuff doesn't work there. have you looked at native ghc there? or is that as crazy as it sounds to me? :)
21:50 Corion ext\Set\t\overload.t is taking 2 minutes already ... Weird ...
21:51 Corion gaal: I haven't compiled GHC from source there, but I'm considering trying to build GHC for MSVC7 there so that it uses cl.exe instead of gcc
21:51 Obadiah has quit IRC (SendQ exceeded)
21:51 Obadiah has joined #perl6
21:52 Obadiah is now known as mml
21:52 gaal can anyone with a fast machine back to r4215 and see if the problems appear there too? that's before the Prelude patch
21:53 Corion I'm going to bed now, sorry :)
21:53 gaal night, Corion!
21:54 gaal won't hurt if i did that too actually, $work tomorrow...
21:54 * eric256 needs a faster machine to run smoke tests like this on....or concurrent testing
21:54 Corion has quit IRC ("Chatzilla 0.9.68a [Firefox 1.0.4/20050511]")
21:55 gmccreight has joined #perl6
21:56 eric256 827 todo tests.. thats huge even out of 5763 tests.
21:56 gaal that's TDD for ya.
21:57 gaal better 827 todo tests than 827 missing tests.
22:00 geoff__ is now known as geoffb
22:02 mrborisguy has joined #perl6
22:03 gmccreight has quit IRC ("Leaving")
22:04 mrborisguy i've spent all day pouring over the subroutine portions of the apocolypse and synopsis, and i think this is right, but i'd love for somebody to comment on these:
22:04 mrborisguy multi sub split (  Str $delimiter ,  Str $input = $CALLER::_ ) returns List
22:04 mrborisguy multi sub split ( Rule $delimiter ,  Str $input = $CALLER::_ ) returns List
22:04 mrborisguy multi sub split (      Str $input :  Str $delimiter          ) returns List
22:04 mrborisguy multi sub split (      Str $input : Rule $delimiter          ) returns List
22:04 mrborisguy &split<> := &split<Str>.assuming:delimiter(' ');
22:05 mrborisguy (for S29)
22:05 gaal i think you want Str ?$input = $CALLER::_
22:05 gaal ("?" cause it's optional)
22:05 mrborisguy oh yeah... yep, you're right.
22:06 * gaal zzzZZzz &
22:08 mml has quit IRC ("Leaving")
22:08 clkao summon autrijus
22:10 iblechbot has quit IRC (Read error: 110 (Connection timed out))
22:11 eric256 does that actualy do anything? or do they setup there client to alert them when it sees that?
22:13 mrborisguy i was wondering the same thing
22:14 mrborisguy it reminds me of the movie "The Anchorman"... "News team, ASSEMBLE!!!!"
22:14 mrborisguy (and other vague references in my head)
22:14 eric256 summon eric256
22:14 eric256 yea i don't think it acutaly does anything
22:14 mrborisguy do anything?
22:15 eric256 that is a BAD movie
22:15 mrborisguy summon eric256
22:15 clkao
22:16 clkao perlbot: status?
22:16 clkao evalbot6: uc('SUMMON')
22:16 clkao evalbot6: ?uc('SUMMON')
22:17 clkao ?eval uc('summon')
22:17 evalbot6 'SUMMON'
22:18 clkao evalbot6: $summon = {((uc(shift)~' ') x 4) ~ 'COME TO ME'}.('autrijus')
22:18 clkao evalbot6: $summon = {((uc(shift)~' ') x 4) ~ 'COME TO ME'}; $summon.('autrijus')
22:18 clkao ?eval $summon = {((uc(shift)~' ') x 4) ~ 'COME TO ME'}; $summon.('autrijus')
22:18 evalbot6 Error:  unexpected ( expecting expression
22:20 mrborisguy can somebody give me commit rights to perl6 so i can upload those split prototypes?
22:22 ninereasons ?eval my &summon = {((uc($^name)~' ') x 4) ~ 'COME TO clkao'}; summon('autrijus')
22:22 evalbot6 Error:  unexpected ( expecting expression
22:22 ninereasons hm
22:22 clkao ?eval $x = { ( ( uc ( $_[0] ) ~ " ") x 4) ~ "COME TO ME"}; $x.('autrijus')
22:22 evalbot6 Error: Undeclared variable: "$x"
22:22 clkao ?eval my $x = { ( ( uc ( $_[0] ) ~ " ") x 4) ~ "COME TO ME"}; $x.('autrijus')
22:22 evalbot6 'AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS COME TO ME'
22:22 clkao ?eval my $x = { ( ( uc ( $_[0] ) ~ " ") x 10) ~ "COME TO ME"}; $x.('autrijus')
22:22 evalbot6 'AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS COME TO ME'
22:23 clkao ?eval my $summon = { ( ( uc ( $_[0] ) ~ " ") x 10) ~ "COME TO ME"};
22:23 evalbot6 \sub {...}
22:23 clkao ?eval $summon.('autrijus')
22:23 evalbot6 Error: Undeclared variable: "$summon"
22:23 clkao hate
22:24 clkao ?eval $::summon = { ( ( uc ( $_[0] ) ~ " ") x 10) ~ "COME TO ME"};
22:24 evalbot6 Error: Undeclared variable: "$summon"
22:24 ninereasons my $summon = {((uc($^name)~' ') x 4) ~ 'COME TO clkao'}; $summon.('autrijus')
22:24 ninereasons ?eval my $summon = {((uc($^name)~' ') x 4) ~ 'COME TO clkao'}; $summon.('autrijus')
22:24 evalbot6 Error:  unexpected ( expecting expression
22:24 ninereasons ??
22:25 wolverian has joined #perl6
22:26 mrborisguy does anybody here have such authority to allow me to commit?
22:27 clkao you need to get a hold of one of: autrijus, bestian, hcchien, iblech, jesse, nothingmuch, Stevan, theorbtwo
22:27 cdpruden has joined #perl6
22:28 mrborisguy clkao: thanks
22:29 clkao damnit, autrijus keeps commiting but not answering to summon
22:29 obra clkao: hi. I'm here
22:30 obra what's up?
22:30 clkao ?eval my $x = { ( ( uc ( $_[0] ) ~ " ") x (70/$_[0].chars)) ~ "COME TO ME"}; $x.('autrijus')
22:30 evalbot6 'AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS COME TO ME'
22:31 clkao ?eval my $x = { ( ( uc ( $_[0] ) ~ " ") x (70/$_[0].chars)) ~ "COME TO ME"}; $x.('perl6')
22:31 evalbot6 'PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 PERL6 COME TO ME'
22:32 b6s has quit IRC (Remote closed the connection)
22:33 mj41 has quit IRC (Read error: 131 (Connection reset by peer))
22:34 mrborisguy clkao: 70/($_[0].chars + 1) because you add the space?
22:34 clkao right
22:35 clkao ?eval my $x = { ( ( uc ( $_[0] ) ~ " ") x (70/(1+$_[0].chars))) ~ "COME TO ME"}; $x.('chicks')
22:35 evalbot6 'CHICKS CHICKS CHICKS CHICKS CHICKS CHICKS CHICKS CHICKS CHICKS CHICKS COME TO ME'
22:35 clkao too many nails
22:35 obra mrborisguy: hang on
22:35 obra mrborisguy: has someone reviewed your patches?
22:36 mrborisguy i haven't patched anything
22:36 theorbtwo Somebody was looking for me?
22:36 mrborisguy and i just posted my split prototypes here
22:36 theorbtwo Ah, I see.
22:36 mrborisguy (or, i mean, i haven't submitted any patches)
22:36 theorbtwo mrborisguy: Email address?
22:36 mrborisguy mrborisguy at gmail dot com
22:38 larsen has quit IRC ("later")
22:41 b6s has joined #perl6
22:41 theorbtwo mrborisguy: Invitation is on it's way.
22:42 mrborisguy theorbtwo: thanks
22:42 * eric256 waits for someone to mod evalbot so you can bind commands to perl snippets ;)
22:43 bd_ has joined #perl6
22:43 eric256 where is the svn bot?
22:46 ero has joined #perl6
22:47 mugwump it's a bit slow, eric256
22:47 mugwump jabbot: seen svnbot6?
22:47 jabbot mugwump: svnbot6 was seen 8 hours 8 minutes 54 seconds ago
22:47 ninereasons perlbot metacommit?
22:47 perlbot ?eval my @meta= <autrijus bestian hcchien iblech jesse nothingmuch Stevan theorbtwo>; sub commit { return 'summon'~ ~@_ }; commit @meta;
22:47 autrijus_ g'morning
22:47 autrijus_ is now known as autrijus
22:48 clkao autrijus!
22:48 obra HAH
22:48 ninereasons hmm. experiment didn't work.
22:48 autrijus gaal++ # Prelude
22:48 clkao ?eval my $x = { ( ( uc ( $_[0] ) ~ " ") x (70/(1+$_[0].chars))) ~ "COME TO ME"}; $x.('autrijus')
22:48 theorbtwo Allo, autrijus!
22:48 autrijus yo
22:48 * autrijus just woke up and relenging
22:48 clkao autrijus: quick, ifx pugs_Eval
22:49 mugwump <evalbot6> 'AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS COME TO ME'
22:49 autrijus clkao: how is it broken?
22:49 clkao autrijus: it's evaling twice
22:49 autrijus clkao: test case?
22:49 autrijus clkao: can you export it to p5 space?
22:49 autrijus like bind it to xsinit
22:49 autrijus as eg. pugs::_eval("...")
22:50 clkao hang on
22:50 clkao i did
22:50 * gaal wakes up, checks flist, finds this: http://code.google.com/summerofcode.html
22:50 evalbot6 'summonautrijus bestian hcchien iblech jesse nothingmuch Stevan theorbtwo'
22:50 evalbot6 'AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS AUTRIJUS COME TO ME'
22:51 clkao autrijus:
22:51 clkao eval ({ $^x[0] })
22:51 clkao VError "Insufficient number of required parameters: 0 actual, 1 expected" [(MkPos "t/unspecced/p5/method.t" 126
22:51 clkao 5 126 26)]
22:51 clkao that's what gets into pugs_Eval
22:51 clkao it works in raw pugs
22:51 gaal what does 'is primitive' signify?
22:52 autrijus gaal: it signifies that it should not be compiled into Parrot (yet) :)
22:52 autrijus other than that it means its subtype is SubPrim
22:52 autrijus so doesn't add to caller stack
22:52 autrijus and you can't use return() there
22:52 ninereasons perlbot metacommit?
22:52 perlbot ?eval my @meta= <autrijus bestian hcchien iblech jesse nothingmuch Stevan theorbtwo>; sub commit { return 'PERLBOT can speak to EVALBOT6: summon '~ ~@_ }; commit @meta;
22:52 evalbot6 'PERLBOT can speak to EVALBOT6: summon autrijus bestian hcchien iblech jesse nothingmuch Stevan theorbtwo'
22:53 theorbtwo ?eval '?eval "Hi!"'
22:53 evalbot6 '?eval "Hi!"'
22:53 gaal ah! great, i was wondering how to do stacktraces and things like that without messing up the stack.
22:53 autrijus :)
22:53 gaal i want to add java-like response to SIGQUIT by default
22:54 clkao autrijus: so, any idea? can you make sure the pugs_eval thing work in haskell land? cause it's not used elsewhere
22:54 Forth has quit IRC (Nick collision from services.)
22:54 gaal now we can add caller() that returns a real object
22:54 Forth_ has joined #perl6
22:56 mrborisguy has quit IRC ()
22:56 autrijus clkao: wait
22:57 autrijus clkao:
22:57 autrijus $ ./pugs -e "say eval_perl5('pugs::guts::eval(q[{$^x[0]}])')"
22:57 autrijus CODE(0x8577910)
22:57 autrijus worksforme.
22:58 autrijus please check r4227.
22:58 clkao orz
23:00 Limbic_Region has joined #perl6
23:01 autrijus $ ./pugs -e "say eval_perl5('pugs::guts::ev​al(q[{$^x[0]}])->([9])')"
23:01 autrijus 9
23:01 autrijus seems it's not my problem
23:02 autrijus gaal: er.
23:02 autrijus pugs::guts:: is a perl5 side thing
23:02 gaal oh. oops. :) merging back.
23:03 autrijus not sure if we want to be consistent with that... seems to be more confusing
23:03 clkao weird
23:03 autrijus please reverse :)
23:03 Limbic_Region bit OT, but if anyone is a student and wants to earn $4500 USD this summer - see http://code.google.com/summerofcode.html
23:03 clkao Could not find module `Pugs.Prelude':
23:03 gaal clkao, perl Makefile.PL and make again
23:05 gaal autrijus, i should do 'svn merge -r4226:4227', right? not 100% used to this yet
23:05 autrijus gaal: if your change was 4227, do -r4227:4226
23:05 gaal ah :)
23:05 autrijus the bigger number comes first
23:05 gaal thanks.
23:07 autrijus :)
23:07 autrijus breakfast, bbiab &
23:14 mrborisguy has joined #perl6
23:14 poletti has joined #perl6
23:17 clkao my bug. hate
23:17 clkao it's when applying that eval result
23:17 clkao now it worked! ship it!
23:22 clkao autrijus: make things rw!
23:22 Chewie[] has quit IRC (Read error: 54 (Connection reset by peer))
23:22 clkao i'm almost done
23:22 puetzk has quit IRC ("Client exiting")
23:24 clkao where's svnbot?
23:24 clkao gaal: what's these pugs:guts open stuff for?
23:27 clkao how to make subs with placeholder varaible rw ?
23:27 clkao can't find in s06
23:28 gaal clkao, my mistake, it isn't p5-related
23:28 clkao ok..
23:29 gaal it's the glue functions between haskell openFile and perl6 open
23:30 stevan seen nothingmuch
23:30 jabbot stevan: nothingmuch was seen 4 hours 45 minutes 50 seconds ago
23:30 stevan hmm
23:30 stevan hey all :)
23:30 eric256 hello
23:30 clkao win42
23:30 gaal hey stevan :)
23:30 stevan hey gaal, how goes it?
23:31 mugwump do we have units in Pugs yet?
23:31 gaal very well! see src/perl6/Prelude.pm
23:31 stevan units?
23:31 mugwump eg, 1`m
23:31 gaal of all the p6 features, /me is scared of that the most
23:31 eric256 what?
23:32 stevan gaal: this is very very cool
23:32 autrijus where is units specced?
23:32 stevan this gets loaded into Pugs?
23:32 gaal stevan: yes
23:33 gaal autrijus, i don't thiung they were finalized, but there was a debate on p6-l
23:33 autrijus stevan: got time to catch up changelog a bit?
23:33 stevan so we can start writing Perl 6 in Perl 6 now ?
23:33 clkao autrijus: rw func! arrayref access is complte, pending the rw function in assignment only
23:33 mugwump autrijus: remember that discussion about a superposition of all possible unit types for numbers?
23:33 autrijus clkao: k k. do hashref too
23:33 stevan autrijus: I only have about 20 minutes right now, but I can find some time later
23:33 gaal stevan: exactly :)
23:33 autrijus mugwump: yes?
23:33 autrijus stevan: thanks!
23:33 clkao autrijus: can't do hashref tonight.... drank champagne
23:33 gaal however, limitations apply.
23:33 stevan when is the release scheduled for?
23:33 autrijus heh
23:33 mugwump well, this is what I was referring to, so I thought you saw a post from one(@Larry) about it..
23:33 theorbtwo Units is unspecced, AFAIK.  I'd like to see the assumption that they are an external module.
23:34 gaal 1. can'y use "return()"
23:34 simcop2387 has quit IRC (Remote closed the connection)
23:34 clkao champagne++
23:34 autrijus stevan: any time in the next 14 or so hours
23:34 autrijus theorbtwo: <aol />
23:34 stevan autrijus: sounds good to me, I will update later tonight (about lunchtime for you)
23:34 clkao quick, let's make arrayref access work in this release
23:34 gaal 2. can't use more files, it's all inlined from Prelude.pm (but i can fix that)
23:34 autrijus stevan: cool
23:34 clkao next one we fix scalar crap (30% in ast?) and hashref
23:34 autrijus clkao: just go sleep :)
23:35 autrijus I need hashref so i'll make it happen.
23:35 clkao no, i want tosee it owrks
23:35 stevan gaal: why no return?
23:35 eric256 for units it would be easy to make postfix operators that converted a number into a Unit object.  then you don't worry about it at all internaly..
23:35 clkao autrijus: make it rw, and i'll make hashref tomorrow
23:35 clkao autrijus: see the current STORE code
23:35 theorbtwo I think it should be an infix operator ` on Any and Type.
23:35 clkao in perl5.c
23:35 simcop2387 has joined #perl6
23:36 gaal autrijus, can you take a look at Run.hs:157 and see if i turned off debugging correctly? i don't understand why this is local :-)
23:36 autrijus $$self?
23:36 autrijus ah. I see.
23:36 eric256 infix / postifx... doesn't matter much. just the fact that we will have such power with creating operators means somehting like units doesn't need to be part of the actual language
23:36 ramzi has quit IRC (Connection timed out)
23:36 autrijus gaal: that's correct
23:37 clkao autrijus: also have you seen my plan for exportation in perl5/README?
23:37 eric256 as soon as coerce:<as> or infix:<+>  dispatch correctly for classes that will be easy...hint hint. ;)  /me would love to see operator overloading for classes work...
23:37 mugwump So, if I'm returning a number of days, then you'd expect that to be a Num and a Num Gregorian::Day
23:37 autrijus clkao: yeah.
23:37 clkao anyawy, i shall go sleep and expect everything to work when i get up, right, autrjus?
23:37 autrijus clkao: right.
23:38 clkao ok. good night
23:38 theorbtwo (Though I'm not sure how you spell type... possibly on Any and Unit, and then export functions with the name of units.
23:38 gaal stevan: http://colabti.de/irclogger/irclogger_log/p​erl6?date=2005-05-31,Tue&amp;sel=1302#l1954
23:38 autrijus nite!
23:38 gaal autrijus, but why is it correct? :) am i passing a copy of the env and modifying it on the fly?
23:39 autrijus gaal: you're passing a modifier function.
23:39 autrijus or rather
23:39 gaal also: woudl it be wiser to pull out the modification from initPrelude and put it in the caller, so that someone else could theoretically call initPrelude themselves?
23:39 autrijus you're passing in the env that was modified
23:39 gaal (with different debugging options)
23:40 autrijus sure, if you want that (although I don't quite see why), move the {envDebug} thing up to line 151
23:40 mugwump I want it to be a conjunctive type.  One a plain num and one a higher order type.  Num&::(Num of Date::Gregorian::Day) or something
23:40 mugwump Actually maybe auto-casting is a better way for that to work.
23:41 gaal back to my previous question, what causes the change to the env to be temporary? how come if beforehand envDebug was Just True, it would revert to that afterwards?
23:41 mugwump or maybe just being Num of Something matches the plain type Num
23:42 autrijus gaal: updates are never destructive.
23:42 autrijus env is not a IORef.
23:43 autrijus you can't change anything really.
23:43 autrijus so this
23:43 autrijus env{ envDebug = Nothing }
23:43 autrijus merely creates a clone of env
23:43 autrijus changes one field
23:43 autrijus lazily reusing all other fiends
23:43 autrijus fields, even, freudian slip
23:43 gaal ah, like OS environment, cute.
23:43 autrijus something like that, yeah.
23:44 autrijus bbiab again...
23:51 evalbot6 has quit IRC (Read error: 148 (No route to host))
23:53 elmex has quit IRC ("leaving")
23:56 eric256 has left
23:57 Limbic_Region has quit IRC ("TV Time")
23:59 zuhn has quit IRC (Read error: 110 (Connection timed out))

| Channels | #perl6 index | Today | | Search | Google Search | Plain-Text | summary

Perl 6 | Reference Documentation | Rakudo