»ö« | perl6.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot usage: 'perl6: say 3;' or rakudo: / pugs: / std: , or /msg p6eval perl6: ... | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by wolfe.freenode.net on 30 October 2009.
Wolfman2000 Perl 6 is not fully backwards compatible with Perl 5 for many reasons. 00:00
jnthn emma: Some people dislike the break in compatibility. Others are disillusioned because, well, people ain't good with coping with the fact that some things just take a long time to do right.
They're entitled to their opinion.
emma Maybe eventually Perl6 will be thought of as the defacto perl and all of you will chat in #perl instead. 00:01
jnthn Maybe. :-)
Tene emma: Perl has a long tradition of strict backwards compatibility, at any price. Perl 6 changes several things in incompatible ways, which is enough for some people.
And then some people odn't like those changes, and then conclude that it must not be Perl anymore. 00:02
00:03 tak11 left
Tene I've got a friend who is fanatically upset about Perl 6 because of minor changes to how it handles whitespace, for example. 00:03
He ranted to me for half an hour about it the other night.
Wolfman2000 But what can be done with Perl 6? Watch
rakudo: multi sub prefix:<√>(Num $x) { return sqrt $x; } ; say √25;
p6eval rakudo fe6dd2: 5␤
fax lol 00:04
emma Well I was planning to learn perl.
fax rakudo: multi sub prefix:<√>(Num $x) { return sqrt $x; } ; say √√25;
p6eval rakudo fe6dd2: 2.23606797749979␤
emma But now i don't know what I should learn. I was going to use this book -- www.perl.org/books/beginning-perl/
Wolfman2000 rakudo: multi sub prefix:<√>(Num $x) { return sqrt $x; } ; say √√10000;
p6eval rakudo fe6dd2: 10␤
emma but now i supose that book is worthless if you believe Perl 6 is perl.
fax emma irssi scripts would be in perl 5 (the normal perl)
jnthn emma: I'm helping Perl 6 come about, but I also still a lot of my day to day work in Perl 5. Perl 5 is still a great, and very capable, langauge. 00:05
Wolfman2000 emma: I'm also helping Perl 6 by at least giving some of what's out a shot.
I'm one of the guinea pigs that makes sure the language works as it should.
Tene emma: A large part of what you'd learn from that book would apply about the same to Perl 5 and Perl 6. As well, most people don't consider Perl 6 ready for production use yet, and wouldn't recommend it. It's still missing quite a bit, although improving rapidly. 00:06
sjohnson Wolfpig, if you will :)
Wolfman2000 I'll soon be working with masak on a feature he's working on.
sjohnson--: Not funny.
sjohnson tough crowd
Tene rakudo.org/status/
emma: That book would not be a bad place to start learning Perl, and if you want to use Perl today, then Perl 5 is your best option.
Perl 6 has been in progress for a long time, and while it's much closer today, we don't expect to have a release we can recommend for much until... April, I think? 00:07
Wolfman2000 ...woah. didn't notice rakudoperl had a twitter page. I better follow it now
Tene It's not very active.
Wolfman2000 Well, you've just gained a follower 00:08
Tene No I haven't.
My twitter account is 'tene', not 'rakudoperl'.
Wolfman2000 Tene: you get my point...I hope.
Tene Wolfman2000: Oh, I understand now. Sorry. 00:09
Wolfman2000 it's alright
00:09 muixirt left
emma Tene: okay thanks! 00:10
jnthn emma: If you want to learn a version of Perl that lets you Get Stuff Done today and that's your main objective, and you don't want to run into compiler bugs, go for Perl 5. What you learn there will prime you well to learn Perl 6 when it's ready and if you like the look of what it offers. :-)
emma well that sounds pretty good then :)
00:10 orafu left
emma This is sort of like an experimental perl channel and some day you will all come back to #perl perhaps. 00:10
Tene Hopefully. 00:11
This is mostly a separate channel for *developing* Perl 6.
00:11 orafu joined
Tene There's some segment of the Perl community that just *hates* Perl 6. I'm curious to find out how that changes over time after the big release next year. 00:12
s1n pmichaud: will you have time tomorrow to get that wiki up and running (or give me some control to do what needs to be done)? 00:14
00:16 nihiliad left 00:17 nihiliad joined
s1n Tene: i suspect none. it hasn't changed much over the course of p6, so i'm not hopeful, but forward is still progress 00:17
pmichaud s1n: I think I can arrange it, yes. 00:19
jnthn pmichaud: Didn't get any energy for ng today, 'fraid (had to take care of a deployment of some web thingy...on the upside, that's taken care of now :-)) 00:22
Next week is looking pretty clear though. :-) 00:23
pmichaud jnthn: no problem, you'll be more productive after I land the speed improvements anyway :)
jnthn pmichaud: Well, that too. :-)
I'll take 20 seconds compiles over 5 minute ones any day.
pmichaud I'll see how low I can get things :)
jnthn You're costing icanhascheezburger some hits in improving it though. 00:24
Anyway, I'm hoping that the Actions.pm won't grow too huge. 00:25
00:25 SmokeMachine left 00:26 starek joined 00:28 xabbu42 left, starek left
pmichaud it should end up less huge than before 00:31
although it might be "as big" because we duplicate a lot of small methods
jnthn Hmm 00:32
Yeah, there is that.
pmichaud anyway, we should end up with not-as-big methods, at any rate :)
testing new lineof code now
jnthn \o/
I'm glad we got to the bottom of that one.
And that I didn't have to write the fix. ;-)
pmichaud yes. and I feel better that we'll handle utf8 there also
we should see some good code improvement there. 00:33
colomon \o/
jnthn is moving a physical timezone west for much of December. 00:36
sjohnson masak++ for regex help earlier 00:37
00:37 orafu left 00:38 orafu joined 00:40 jaffa8 left 00:42 SmokeMachine joined
pmichaud ugh, my patch isn't working 00:48
src/string/api.c:1191: failed assertion '!PObj_on_free_list_TEST(src)'
oh. I think I know. 00:49
00:50 payload left
pmichaud nope, not it. 00:50
jnthn pmichaud: gah, that's nasty :-( 00:52
pmichaud: mark fail?
pmichaud probably, but I'm not sure where
want to look at the patch for me real quick?
jnthn Sure
pmichaud (you're more familiar with this stuff than I am)
gist.github.com/229000 # CodeString.pmc patch 00:53
jnthn looking 00:54
That mark is probably wrong. 00:55
moment
pmichaud I was trying to follow marks in other PMC code :-| 00:56
jnthn pmichaud: Yes, it's wrong.
You need to call SUPER()
pmichaud I couldn't find SUPER() in any other mark code, though.
I'll try it, though. 00:57
jnthn You need to call SUPER() :-)
pmichaud: Look in the string.pmc. It marks the STRING *.
pmichaud that fixed it.
jnthn Good.
pmichaud ...wonder why other mark() don't have to do it when they're in a subclass, though.
jnthn What did you look at? 00:58
I mean, Perl6MultiSub doesn't call SUPER()
But that's because it doesn't share the same layout as what it inherits from.
pmichaud oh, I found another one -- ExceptionHandler calls SUPER()
jnthn (Becase MultiSub's layout is wrong.)
Wolfman2000 Which Synopsis covers classes again? I think I need to study that once more before I attempt the Linked List again.
pmichaud filehandle.pmc doesn't call SUPER() in mark()
jnthn pmichaud: Yes, but Handle doesn't have any markables. 00:59
pmichaud okay.
jnthn Granted, that means it's assuming stuff about what it inherits from.
pmichaud that makes sense then.
jnthn Which is probably an anti-pattern.
Since if Handle was to start having a PMC reference, then FileHandle would not mark stuff it should. 01:00
oh lol guess what
/* TODO: Consider encapsulating PIOHANDLE as a PMC type, for subclassing */
pmichaud what?
jnthn ATTR PIOHANDLE os_handle; /* Low level OS descriptor */
bwaha
pmichaud anyway, I have news to report
with the lineof fix in place
my test file b.pm 01:01
jnthn So anyway, if they do that TODO, FileHandle will need to start calling Super. :-)
01:01 darln joined
pmichaud was 55 seconds 01:01
01:01 payload joined
jnthn uh-huh. 01:01
pmichaud now 24 seconds
jnthn Win.
pmichaud++
pmichaud we'll get more win when I put the transcode back in place
I'll do that while running make test
jnthn Sure, but you've no doubt solved the major bottleneck.
Wolfman2000 ah, here we go: S12 01:04
lichtkind good night
01:04 lichtkind left
darln is there any place one can get a test package of Perl 6 for testing? 01:05
pmichaud darln: rakudo.org/how-to-get-rakudo
darln: perl6.org/
darln perl6.ord gives me NXDOMAIN, I'll try out rakudo though 01:07
is rakudu really Perl 6 though?
pmichaud .org
darln ?
yeah I can't open that page
pmichaud oh, perhaps the dns server is down righ tnow
Tene darln: Yes, Rakudo is the predominant implementation of Perl 6.
pmichaud yes, rakudo really is an implementation of Perl 6 01:08
darln not getting anything with dig +trace either
pmichaud I bet the dns server was done. I know they were doing upgrades on the server earlier today.
s/done/down/
darln sorry, let me get this straight, theres implimentations of an imcomplete spec? curious 01:09
pmichaud why not?
Tene rakudo.org/status/ -- Rakudo implements this much of the spec.
pmichaud a spec doesn't have to be complete to implement it
Tene darln: The implementations drive the spec, in many cases. 01:10
darln well I'm just trying ot imagine a Perl with half the features missing :p
pmichaud ODF 1.0 was an incomplete spec, but plenty of people implemented it
in fact, on the internet, the official way to do things is to have multiple implementations before a specification is designed and approved :)
darln pmichaud: actually didn't many people have ot fill in the gaps?
pmichaud sure. but anyone who believes a spec is complete before implementation begins is still designing in the 1980s :) 01:11
darln and don't even get me started on 802.11 N
jnthn darln: The waterfall model just doesn't work out so well for language design. Sometimes things that seem like a Great Idea don't work out quite so well once people start writing code in the language, but there's no way for people to meaningfully write code if they don't have a way to run it. 01:12
Tene Perl 6 subscribes to the Whirlpool development strategy.
jnthn It's a process of convergence.
darln pmichaud: well most langs have a solid base spec before it's released in any usable fashion
pmichaud Perl 5 didn't.
nor did PHP, for that matter. :) 01:13
Tene darln: Perl 6 had a pretty good spec before anyone started working on Rakudo. We just were realistic enough to recognize that it would change in response to the implementation, so we didn't play silly naming games with it.
pmichaud I'd actually say that most languages started without a solid base spec before they were released :) 01:14
darln um, Perl 5 built on Perl 4 which built on Perl 3, .... and similar for PHP, they all had a solid BASE
s1n darln: no, most things that have a spec were either 1 RFC'd in the 70's or have the spec created while the impl is created
pugs_svn r29022 | kyle++ | [t/spec] typo fix
pmichaud the "solid specification" came long after the language.
s1n C had no solid spec for liek what, 15 years?
pmichaud darln: there may have been a solid "base" for Perl 5, but there certainly wasn't a spec that preceded an implementation
s1n and that language is far less complex
pmichaud unless you claim that an implementation is somehow a spec, in which case my point is proven :) 01:15
darln ok maybe spec isn't the word I was looking for then
well, I would consider Perl5's perldoc lib to be it's spec
01:15 nihiliad left
pmichaud heh 01:15
darln I don't think Perl ever had na officla RFC or so 01:16
s1n heh
pmichaud that's.... funny
s1n does perl5 pass it's own test suite yet?
pmichaud because perldoc explicitly says that if the docs don't match the implementation, it's the documentation that's wrong :)
darln well to me a spec is what describes all the aspects of something, and perldoc seems to be as clsoe ot that as you can get for Perl
s1n: actually yes it does, form what I've seen
s1n darln: which do you think came first, perl or perldoc?
pmichaud darln: sure. but the implementation clearly comes before the spec 01:17
darln like I said, when I said "incomplete spec" a bit back I clearly used the wrong terms/words
pmichaud darln: yes, I'm picking on you a bit. :)
s1n the only thing that speaks is code, everything else is fluff
pmichaud darln: the truth is that the language and specification co-evolve together
darln I was meaning some sort of definition or reference so one would know what is what in the lang itself 01:18
pmichaud right
the language and specification co-evolve together
darln you can't have a lang with out some clear definitions of syntax and keywords and such
or I should say a usable lang
pmichaud we have a spec, and we implement the spec, and as we do the implementation we discover places where the spec is imprecise or doesn't work, so we revise the spec based on the implementation, which drives more implementation which ...
Perl 6 is definitely a usable lang. 01:19
Tene darln: There *was* a fairly significant spec before work on Rakudo began, as I said. It's just that nobody believed that it was final, because we all knew it would change in response to the implementation.
s1n rakudo is only now catching up to ow complete pugs was (in comparison to the spec then) 01:20
pmichaud might argue it's better now with nqprx 01:21
pmichaud I think rakudo probably surpassed pugs sometime over the summer
with nqprx it's way behind...but we should be able to catch up again in a few weeks
Tene +1
s1n the point is the spec used to drive more implementation, now the implementations are catching up to the spec and they _have_ to co-evolve, just like pmichaud said 01:22
lisppaste3 wolfman2000 pasted "Nominal type check failed for parameter '$next'; expected Node but got Failure instead in Main (um, what if I WANT undef?)" at paste.lisp.org/display/89963 01:23
Wolfman2000 I thought I had this...
jnthn Wolfman2000: The type constraints are optional. :-) 01:24
Wolfman2000 jnthn: If that's supposed to be a hint to the solution, I'm dense.
pmichaud jnthn: okay, compiling b.pm was 55 seconds, now 16.5 seconds
jnthn pmichaud: omg 01:25
pmichaud (put in the transcoding optimization)
jnthn Nice.
pmichaud I'll measure Actions.pm from ng -- just a sec
jnthn Wolfman2000: method new(Str $name, Node $next)
Could be
method new(Str $name, $next)
Then you can pass what you like to $next
Wolfman2000 jnthn: Can type junctioning also work then? Say: method new(Str $name, Node|Failure $next)? 01:26
pmichaud jnthn: Actions.pm on my box is now 13 seconds to pir
jnthn pmichaud: huh, I want your box.
pmichaud you'll have it when I check in these changes :)
jnthn \o/
pmichaud and I still have quite a bit more optimization to put in place :)
jnthn Wolfman2000: No
lisppaste3 wolfman2000 annotated #89963 "As jnthn just said, no type junctioning. Oh well..." at paste.lisp.org/display/89963#1 01:27
jnthn Wolfman2000: by the way, if you don't want to have to specify the next node, why not make it an optional param?
method new(Str $name, Node $next)
Wolfman2000 jnthn: One step at a time.
jnthn gah, paste fail
method new(Str $name, Node $next?)
Then you can call it as Node.new("Wolfman2000") :-)
01:27 envi^home joined
Wolfman2000 jnthn: then what's $next set to? 01:28
jnthn An undef, iirc
jnthn thinks something might still be fishy here...
pmichaud I wonder if it should get set to the Node type object, though.
jnthn pmichaud: Yeah
pmichaud: That was what I was thinking might be fishy.
Wolfman2000 doesn't seem to set to undef. Otherwise, it would have failed.
...eprhaps it's being set to Failure automatically?
pmichaud undef kinda is a Failure in rakudo at the moment 01:29
jnthn pmichaud: If it's meant to happen, I can do it easy enough.
pmichaud: It'd be consistent with var decls.
pmichaud right
jnthn So it feels right.
pmichaud I've often thought we should do that for optional params
jnthn Also it's Really Really Easy.
Yeah.
pmichaud I haven't thought that it's Really Really Easy. At least not under the old param code :)
Wolfman2000 So...what types of junctioning are available then? 01:30
jnthn pmichaud: iirc under the new param code I think it's like, changing one line or something.
pmichaud right
Wolfman2000 If you can't junction types within a variable, what can you junction?
pmichaud jnthn: that's what I had hoped and expected
jnthn pmichaud: Heh. Maybe I did manage to get it about rightish after all. :-)
Wolfman2000: We don't handle junctions of types. It was a proposed feature of Perl 6, it may happen in the future (after 6.0.0), but it raises...issues. 01:31
Wolfman2000: I mean, you can make a variable that contains two types. 01:32
pmichaud jnthn: okay, I wonder how much the change I just made to lineno will improve the spectest speed in master :)
jnthn But you can't write a junction of types where a single type is expected.
Sorry if this is confusing...
Wolfman2000 jnthn: It is a little confusing.
jnthn I might explain it better if it wasn't 2:30am. ;-) 01:33
pmichaud in general I suspect one never needs to do |Failure, because an undefined value can be assigned to any container
jnthn pmichaud: Oh hmm. And bound to any param?
(when there's a nominal type there?)
pmichaud don't know about that... but it might be consistant
*consistent
jnthn Yeah
Actually it does bug me a little that Wolfman2000++'s code that he nopasted gave that error. 01:34
pmichaud I'd have to think about it a bit more when my brain is less fried and there's less noise in the house
jnthn Yeah, I'd have to think about it when I'm more lucid too. :-)
Wolfman2000 jnthn: ...I found a bug?
pmichaud currently there are 7 or 8 girl scouts here making dinner :) 01:35
so it's a bit noisy
jnthn Wolfman2000: I'm not entirely sure. I'm not sure if the spec calls it one way or the other (can't remember it saying so).
Wolfman2000: I think I'd like it if the spec was clarified, and Rakudo tweaked, so what you nopasted would work.
Wolfman2000 pmichaud: There is something that bugs me in the current implementation. If you try to say any variable that is set to undef or Failure, it often says "Use of uninitialized value". Is there a way to...well, not have it say that? (besides check if defined)
pmichaud Wolfman2000: at the moment, no -- eventually we'll be able to do "no warnings;" or something like that 01:36
actually, that might not be all that far off, now that I think about it
Wolfman2000 pmichaud: Let me try this approach then. Is there a way within a class to make variables with "has" as "never undef"?
jnthn pmichaud: Heh. Does that end up compiling to setting some contextual variable like $*WARNINGS? :-)
pmichaud jnthn: no, I think it's a pragma namespace 01:37
jnthn pmichaud: ah, ok
pmichaud (and namespaces are lexical now, so...)
jnthn Yeah
OK
That'll work.
pmichaud anyway, I think it just sets a lexical
Wolfman2000: at the moment the spec says that any container is allowed to have an undef in it
I'm not sure there's a way to prevent that. 01:38
oh, you could probably do it with a where clause
Wolfman2000 ...right, the new constraining thing
pmichaud I don't know if where clauses would supercede the other rule
jnthn subset DefinedNode of Node where *.defined;
has DefinedNode $!x
pmichaud well, we also have the new :D and :U syntax
has Node:D $!x
jnthn :D
eternaleye Wolfman2000: You can't junction type _constraints_. You can junction any kind of object AFAIK.
jnthn But we didn't implement those yet
:S 01:39
pmichaud (haven't read the latest spec to see what that does exactly)
lisppaste3 wolfman2000 annotated #89963 "Well, this constraint worked." at paste.lisp.org/display/89963#2
jnthn pmichaud: I think once STD starts parsing them, we probably can expect they're gonna stay around.
01:39 darln left
jnthn I don't know that it does yet. 01:39
pmichaud ...I think I saw that it does...
more precisely -- I think I saw a commit message that said it did. But I may be mis-remembering 01:40
I know I saw a recent commit to either std or the spec about them.
jnthn std: sub foo(Int:D $x) { }
p6eval std 29022: ok 00:01 109m␤
jnthn fajn, it does.
uh, fine
...oh well, I prefer the Slovak spelling anyway. 01:41
Tene std: class prefix { ... }; sub foo(prefix:D $x) { ... }
p6eval std 29022: ok 00:01 110m␤
pmichaud gotta be a little careful with prefix: :-).
jnthn std: class IHASCHEEZBURGER { }; sub omg(IHASCHEEZBURGER:D!) { }
p6eval std 29022: ===SORRY!===␤Unable to parse signature; couldn't find final ')' at /tmp/NKfRMD0fxJ line 1:␤------> EEZBURGER { }; sub omg(IHASCHEEZBURGER:D⏏!) { }␤ expecting any of:␤ circumfix␤ constraint␤ param_sep␤ parameter␤
..trait␤ type_constraint␤ whitespace␤FA…
pmichaud Also, I'm not sure if Int:D was parsed as a defined/undefined type or in the same category as token name:identifier { ... }
Tene std: class postfix { ... }; sub foo(postfix:D $x) { ... } # is this better?
p6eval std 29022: ok 00:01 110m␤
jnthn aww
01:42 s1n left
jnthn std: class IHASCHEEZBURGER { }; sub omg(IHASCHEEZBURGER:D $nom) { } 01:42
p6eval std 29022: ok 00:01 110m␤
jnthn phew
Tene pew
jnthn Anyway, yeah, so, STD implements it.
lisppaste3 wolfman2000 annotated #89963 "Hooray defined or syntax! :D" at paste.lisp.org/display/89963#3
jnthn When I work out how STD parses it, I'll implement it. :-)
eternaleye rakudo: multi postfix:<:D>( Str $item ) { say "Sitem makes me happy!"; }; "Perl 6" :D
p6eval rakudo fe6dd2: Sitem makes me happy!␤
Tene Sitem? 01:43
eternaleye grr
rakudo: multi postfix:<:D>( Str $item ) { say "$item makes me happy!"; }; "Perl 6" :D
p6eval rakudo fe6dd2: Perl 6 makes me happy!␤
jnthn I love how you can add smilies as custom operators.
Wolfman2000 jnthn: Think my latest pastebin version will be fine for now?
pmichaud at least for the moment
that one might end up being an adverb
jnthn pmichaud: aww 01:44
pmichaud but smileys should be adverbs anyway
since they modify the operation
jnthn True
Heh.
pmichaud note the difference between
"Perl 6 is dead."
jnthn The Perl 6 grammar is optimized for parsing smilies in the correct grammatical category. :-)
pmichaud and
"Perl 6 is dead." :-)
jnthn Wolfman2000: looking
pmichaud it's definitely a modifier.
yay, all tests successful in parrot trunk with my patch 01:45
jnthn yay
pmichaud now checking codetests
jnthn Wolfman2000: It's fine
Wolfman2000 jnthn++: Thanks. Maybe NOW I can make a proper linked list.
jnthn Wolfman2000: However, neater is where { .defined } since it puts the thingy it's chekcing into $_ 01:46
Wolfman2000: And you can probably get away with *.defined too
... where *.defined;
Wolfman2000 jnthn: Is there a difference between .defined and *.defined?
jnthn Wolfman2000: *.foo is equivalent to { .foo } which is equivalent to { $_.foo }
Wolfman2000: The difference is that you can drop the curly brackets with the *.defined form. 01:47
With { .defined } you need them.
*.methodname will make a closure for you.
It means you can do a lot of very neat things
like @products.sort(*.price) 01:48
Wolfman2000 ...wasn't there a new object method that expanded out classes and stuff into Strings?
jnthn To sort the array of, say, Product objects by their .price attribute.
Wolfman2000 I want to see my class node expanded into the string and its other node
jnthn Are you thinking of .perl ?
It's not *that* magical though.
But you can override it and make it output something nice.
oh wait 01:49
Or you can override .Str
Wolfman2000 jnthn: Unfortunately, $noth.perl only showed $noth.new()
jnthn Wolfman2000: Yeah.
Wolfman2000: That's one of those things that sorta sucks at the moment...I wish it did something better.
Wolfman2000: .perl is just a method though
You can override it.
Wolfman2000 If I can figure out a good way to call it...let's see. 01:50
I need to be able to print the $!name, and $!next if defined
jnthn method perl() { "Node.new(name => {$!name.perl}, next => {$!next.perl}" }
Basically, your .perl method generally wants to make recursive calls liek that 01:51
oh sorry
given your construcotr that's just
method perl() { "Node.new({$!name.perl}, {$!next.perl}" }
The idea of .perl is that it gives back something that you can eval to build something as close as possible to the original thing.
oh heh, I missed a paren too 01:52
method perl() { "Node.new({$!name.perl}, {$!next.perl})" } # finally
Wolfman2000 jnthn++: I think your original version also worked.
jnthn :-)
Yes, but you overrode .new, so it wouldn't ahve eval'd properly. 01:53
Wolfman2000 jnthn: Was I not supposed to override new?
jnthn Wolfman2000: btw, one problem with what I just showed you, if you want to do this Really Perfect.
Wolfman2000: No, you're allowed to override new. I'ts completely fine.
Wolfman2000: My point was that you're .perl method shuld hand back something that when eval'd would actually work. :-)
Wolfman2000: oh 01:54
Wolfman2000: It would have actually
Tene pmichaud: will ng get class stubbing so the setting can be broken apart?
Wolfman2000 jnthn: Which version works?
jnthn Wolfman2000: My first one.
Wolfman2000: Because positionals can be passed by name to anyway.
Wolfman2000 return "Node.new(name => {$!name.perl}, next => {$!next.perl})"; <-- that one?
jnthn Wolfman2000: Heh. Perl6 is too awesome for me.
Yes, that one. :-)
pmichaud Tene: class stubbing, of course (part of the spec)
breaking the setting apart-- we're not sure about that.
(yet) 01:55
jnthn pmichaud: The question was if we'll do it in ng though ;-)
pmichaud jnthn: of course we'll do it in ng, it's part of the spec :)
the setting is supposed to go into a common lexical scope -- in some ways that's easier if the setting is all in a single compilation unit
Wolfman2000 jnthn: Perl6 may be too awesome for me as well. Perhaps we should team up more often.
jnthn pmichaud: OK, that's fine, but if we're doing the whole spec, you're doing the PDL stuff in S09. :-P
lisppaste3 wolfman2000 annotated #89963 "Now with overridden perl(). Next up: what's the right way to override the submethod BUILD?" at paste.lisp.org/display/89963#4 01:56
jnthn Wolfman2000: Well, it's a new language. We're all still learning how to make the most of it. :-)
Wolfman2000 jnthn: serious question up there. I know that both new and BUILD are used for actually creating the classes.
jnthn Wolfman2000: Yeah. 01:57
Wolfman2000: Basically, they exist for different purposes.
If you want to provide a custom constructor - that is, a custom interface to the constructor - you override .new
That's what you have done.
If you just want to use whatever constructor you inheirt, but you want to do some initialization of attributes, you can instead have a submethod BUILD. 01:58
Wolfman2000 jnthn: Let me modify my file to see if I understand something...that can use both .new and .BUILD
jnthn Wolfman2000: In your example that you nopasted, you don't need to do the BUILD thing.
I mean, there's no natural use for it here. 01:59
If you want to see BUILD submethods put to decent use, see masak++'s druid.
e.g. github.com/masak/druid/blob/master/...id/Game.pm
Here, there's some complex init of the attributes to be done, so that logic is put in a BUILD method. 02:00
Wolfman2000: Don't expect to be writing a new and a BUILD in every class though.
Wolfman2000 ...but it doesn't seem to have a .new
Also, I'm mixing up what $. and $! do again. 02:01
jnthn Wolfman2000: Right. It uses the default one that is inherited from Object.
Wolfman2000: It's perfectly fine to have a class with no BUILD submethod and no new method.
They're there if you need them, they're not things Perl 6 makes you write. :-)
$!foo always refers to the storage location for the attribute. 02:02
pmichaud the default .new and .BUILD are intended to cover a lot of cases for you :)
Wolfman2000 ...and feather is going slow now. This isn't good
jnthn Right, that's a better way of putting it.
lisppaste3 wolfman2000 annotated #89963 "Broken version. Don't want to force users to initialize the time the nodes were made. Is this where BUILD comes in handy?" at paste.lisp.org/display/89963#5 02:03
jnthn I think that's a good example. 02:04
Wolfman2000 Method 'get_name' not found for invocant of class 'Str' in Main <-- that's the error message I get for running that file.
jnthn Wolfman2000: Oh. :-/ 02:05
Wolfman2000: Oh, and I know why do.
*too
:-(
Wolfman2000 jnthn: Alright, I'll establish that I did it wrong somehow. How did I do it wrong? 02:06
jnthn Wolfman2000: Your code is probably valid Perl 6. You just hit a Rakudo bug...one that I think we'll deal with in the ng branch. Basically, if you write a BUILD submethod today, it forgets about the other initializations you're doing.
Sorry. This is just one of those "Rakudo doesn't always get it right yet".
In masak's code, there was a comment to the effect. 02:07
It's a bug in the compiler, not your mistake.
Wolfman2000 jnthn: I'm guessing ng is the most popular branch. So far, it's been suggested that I hold off on using that branch. When should I use it?
colomon ng is only "used" by maybe five people yet. 02:08
jnthn Wolfman2000: It's not popular in terms of people using it.
colomon it's under heavy development.
jnthn Wolfman2000: It's popular to talk about because it's where a lot of active development/changes are happening.
So those involved in hacking on it are talking a lot about it.
colomon it's popular because we're all super-excited for what it will bring when it is ready to be merged in with the master branch.
jnthn It's certainly not usable yet.
Wolfman2000 colomon, jnthn: when will it be merged in?
I've got perfectly valid code that I don't want to have go to waste here...and the $!timecheck variable will be useful for when I eventually implement a priority Linked List 02:09
pmichaud when we're passing roughly the same number of tests as the master branch
jnthn Wolfman2000: It's hard to estimate. I really, really hope we land it before the next Rakudo release.
pmichaud and I suspect we won't "merge" the branch
colomon pmichaud: no?
pmichaud I suspect we'll rename the existing master branch to something else, and the ng branch to master
Wolfman2000 Considering that I'm using the source of rakudo, I don't think releases mean anything to me. 02:10
jnthn pmichaud: Yeah. We may have to cherry pick back some of the changes from master too.
pmichaud jnthn: there shouldn't be too many
Wolfman2000 So...when exactly do I recompile the master for the stuff that has been put in?
colomon is that a subtle hint not to do any development on master at the moment?
jnthn pmichaud: Some folks++ have done some settings tweaks.
02:10 mubot left
pmichaud colomon: I'd probably not do too much major tweakage 02:10
we could certainly pull settings in from master
and in fact, I suspect we'll do that instead of copying settings over from the existing branch
rephrasing 02:11
jnthn Wolfman2000: When it happens, it'll certainly be announced here, and if you follow any of the blogs, it'll get anounced there too.
pmichaud when we get to compiling settings in ng, we'll copy them from the master branch rather than use the ones already in ng
colomon I promise not to do anything in master outside of settings. :)
jnthn Wolfman2000: I'd like to give you a more accurate figure, but to be honest I'd be pulling it out of thin air.
02:11 checkin-it-out joined
jnthn We'll probably have a better idea in a few days time. 02:11
pmichaud yes, I think so
I plan to push to have Test.pm running tomorrow sometime 02:12
Wolfman2000 So for now, no time stamp priority linked list. Understood.
pmichaud okay, jnthn -- ng and nqp-rx now updated to new faster code gen 02:13
Wolfman2000 ...I now have to wonder. With Rakudo in the state that it's in...how am I supposed to be sure Web.pm will be alright when masak and I handle that? 02:14
colomon Wolfman2000: The current Rakudo branch certainly works better than the Rakudo Web.pm was developed under.
checkin-it-out Hi. Regarding the standard library that's going to be part of Perl 6: is the plan to keep it minimal and rely on the CPAN|CP6AN a lot, or are there plans to make Perl 6 more "batteries-included" than Perl 5?
pmichaud checkin-it-out: we plan to distinguish "compiler" from "distribution" 02:15
checkin-it-out: there will be multiple compilers, and multiple distributions
some distributions may be more batteries included than others
dalek p-rx: e8e145d | pmichaud++ | src/Regex/Cursor.pir:
Remove no-longer-used $!type attribute. (It may get replaced with
 02:16
p-rx: 95cd73d | pmichaud++ | src/cheats/hll-compiler.pir:
Provide transcoding logic to HLL::Compiler, so that it can

for faster parsing, when possible.
checkin-it-out Here perlcabal.org/syn/ I see:
dalek p-rx: 238f21b | pmichaud++ | (5 files):
Bump PARROT_REVISION, update bootstraps.
checkin-it-out "30Standard perl library(TBD)"
jnthn pmichaud: It's after 3am, but I'm gonna build and check it out anyway ;-)
02:17 NorwayGeek left
pmichaud huh 02:17
checkin-it-out I'm just curious to get an idea of how much will be specced, vs. how much will be implementation-dependent.
pmichaud checkin-it-out: our goal is to keep the core libraries as small as reasonably possible
checkin-it-out Ah. I see. Interesting. Thank you. 02:18
pmichaud and rely more on modules and distributions to provide the value-added features
02:18 NorwayGeek joined
Wolfman2000 To whoever knows this: when overriding .perl() and .Str(), should multi be used? 02:18
checkin-it-out That philosophy seems to have served Perl 5 well, even though it means for searching around cpanratings for me. :)
jnthn Wolfman2000: no 02:19
Wolfman2000: They don't take any arguments anyway (well, the invocant aside). :-)
pmichaud jnthn: 13.6sec for Actions.pm here 02:20
jnthn checkin-it-out: I think there'll be more encouragement for people to build "distributions", so there's an easy way to say "I want to do web stuff with Perl 6, so I can grab this package that gives me a compiler + a bunch of useful modules for web programming".
Wolfman2000 ...embedded comments now require backticks? What's wrong with just #?
pmichaud Wolfman2000: it was too easy to get things confused with #{...}
especially if someone was adding # to the beginnings of lines and wanted to comment out a block 02:21
jnthn checkin-it-out: Think of it a bit like Linux distros. There's the Linux kernel, but they all add other stuff.
colomon jnthn: a significant quantity of Str definitions in settings use multi...
jnthn colomon: orly? Hm.
Wolfman2000 Node.new(name => Str()<0xb5f37594>, next => Node.new(name => Str()<0xb5f50c38>, next => Failure()<0xb660aacc>)) <-- As a heads up, I think .Str looks ugly, even for what should be strings already.
checkin-it-out It's interesting to me that the Scheme community seems to be fighting the opposite battle right now: they're trying to wrangle multiple implementations together and settle on a more common spec.
Wolfman2000 rakudo: my Str $test = "Hi there."; say $test; say $test.Str; 02:22
p6eval rakudo fe6dd2: Hi there.␤Str()<0x2b4fd73ece00>␤
jnthn Wolfman2000: youch.
pmichaud checkin-it-out: we have multiple implementations of a common spec already :)
Wolfman2000 Did I do this wrong as well?
pmichaud anyway, time for me to get some dinner
checkin-it-out Yes, but Scheme has them coming out the wazoo. :)
pmichaud jnthn: post a quick note to see if things are faster for you as well (and by how much)
checkin-it-out Thanks for the info pm. 02:23
colomon jnthn: really. and I didn't write all of them. honest. :)
jnthn pmichaud: will do
pmichaud oh, I'm going to also do a profiling run and see if I see other optimizations that might bubble up now that we aren't swamped by that other problem
jnthn colomon: Curious - are they marked "is export" too?
colomon jnthn: nope
jnthn: I know I thought multi was necessary in that case... 02:24
jnthn colomon: Ah. multi IIRC implies is export. 02:25
colomon jnthn: so Str(1/3) (for instance) is expected to work? 02:26
jnthn Thing is, TypeObj($x) is a special form, not just a normal sub call.
Because Str is a type object in the namespace
I think in the case of Str($x) it's *meant* to then call $x.Str
But then I've asked aobut things like
Wolfman2000 rakudo: my Str $test = "Hi there."; say $test; say $test.Str(); 02:27
p6eval rakudo fe6dd2: Hi there.␤Str()<0x2b2bda091028>␤
Wolfman2000 no difference. of course
jnthn colomon: My::Type::Like::This($x) and what that de-sugars to.
pmichaud Str is inheriting Object's .Str, and shouldn't.
jnthn colomon: And I'm not sure I've got a clear answer on that yet.
pmichaud Str should provide its own .Str, likely.
jnthn pmichaud: Yeah, agree.
colomon pmichaud: agree
Also Num its own .Num.
jnthn Heh
pmichaud oh dear, looks like my pprof output is 945MB
that will take a while for pprof2cg to crunch 02:28
time for some dinner, then
jnthn pmichaud: Long dinner. :-)
Wolfman2000 ...did I just make you guys have to do more work on Rakudo now?
jnthn Wolfman2000: Yep. :-D
colomon rakudo: say 1.0.Num;
pmichaud Wolfman2000: no, it's the same amount of work as before. :)
p6eval rakudo fe6dd2: 1␤
pmichaud Wolfman2000: you just prioritized the work a little differently for us :)
colomon rakudo: say 1.0.Num.Num;
jnthn Wolfman2000: Though for me (don't know about others), watching people using it and find faults just motivates me to make it btter. :-)
p6eval rakudo fe6dd2: 1␤
jnthn *better
colomon rakudo: say "hello".Str.Str; 02:29
p6eval rakudo fe6dd2: Str()<0x2b0f3af649d0>␤
02:29 checkin-it-out left
Wolfman2000 I'll take a name on the Thanks To section of the pod then. 02:29
pmichaud rakudo: my $a = 'Hello'; say $a.Str; say $a.Str; say $a.Str.Str
p6eval rakudo fe6dd2: Str()<0x2b7c2cdb4410>␤Str()<0x2b7c2cdb4410>␤Str()<0x2b7c2d7e6328>␤
pmichaud yup
jnthn oh gah, Rakudo's Configure.pl screws me over again.
pmichaud sneaks away 02:30
jnthn lol
colomon any objections to my just adding Str.Str right now?
(and should it be multi or no? :) 02:31
Wolfman2000 colomon: If you're somehow able to figure it out, I don't object.
As far as multi...something's telling me yes.
jnthn chases pmichaud, yelling something about him not updating PARROT_REVISION in rakudo's ng branch
colomon: Add it. 02:32
colomon: With multi, I need to ponder that a little more.
colomon okay, I'll try to get it in before bed.
jnthn colomon: It feels a tad odd to me.
colomon: But if the rest are doing the same thing, then I guess go for consistency. 02:33
That way, they might be wrong, but they'll at least be consistently wrong. ;-)
colomon jnthn: but it's not the rest, it's like half the rest. there's no consistent pattern I could see.
jnthn Oh.
Well in that case, do what you like. :-)
02:34 synth left
jnthn My leaning is "not". 02:34
02:34 yath left
jnthn 'cus it doesn't make sense to me. 02:34
(that is, doesn't make sense why they'd want to be multi)
colomon jnthn: and I don't understand the issue enough to make a reasonable judgment. I was just sprinkling multis everywhere because I couldn't see any drawbacks.
jnthn colomon: There probably aren't any. 02:35
colomon (though as I said, I'm only responsible for a couple of them in .Str in setting.)
jnthn colomon: I'm just not sure there's any wins either. :-)
colomon Where should I add a test for Str.Str?
02:35 yath joined
jnthn t/spec/...heck knows... 02:35
S32-str somewhere maybe
colomon S32-str?
bother, none of those filenames look promising. 02:36
jnthn colomon: lol. I'm looking at them thinking the same. :-) 02:37
My second guess was gonna be S02-builtin_data_types/
But there's no str.t in there either! 02:38
Wolfman2000 I kind of have to side with S32-str myself
jnthn Yeah
colomon S32-type-casting/methods.t ? looks like there is something likely in there already...
jnthn colomon: oh, that's not at all a bad choice. 02:39
I think you meant S13
colomon yes, right.
already editing the file. :)
jnthn Do that for now.
if moritz++ has a better idea, I'm sure he'll move it.
colomon Okay, two simple tests added. 02:40
oh, actually one of the tests there already fails for this reason! 02:41
lovely.
jnthn pmichaud: oh you complete show off.
pmichaud: actions.pm builds in 10 secons for me now.
Wolfman2000 colomon: Is it failing due to design or bug?
jnthn *seconds
colomon Wolfman2000: existing failure is marked as a Rakudo bug. 02:42
jnthn pmichaud++ # 27 times faster compilation of Actions.pm!
Wolfman2000 the $stringvar.Str() thing, right
Honestly, if a variable is a string, $stringvar.Str() should just return itself
jnthn Wolfman2000: Right. :-) 02:43
colomon method Str() { self; }
Wolfman2000 ...why does it feel like we're taking that from Python?
colomon compiled... 02:44
make test passes... 02:45
jnthn Wolfman2000: 'cus Perl hapily steals from all the other languages? :-)
Wolfman2000 jnthn: wasn't aware of that history.
If anything, I thought everyone stole from Perl...or at least, their regexes
colomon My tests pass, the existing one still fails.
jnthn Wolfman2000: Well, it's all give and take.
Wolfman2000: I think Perl has kinda led the way in regexes really though, and I think Perl 6 is notching that up. 02:46
lisppaste3 colomon pasted "failing test" at paste.lisp.org/display/89966
colomon "got: undef" 02:47
Wolfman2000 ...what do we have to do to make paste.lisp.org allow colorizing as Perl anyway?
colomon looks like it's calling Str.Str instead of RT69378str.Str
jnthn colomon: That's...odd. 02:49
colomon yeah.
jnthn colomon: I'm not sure. That looks like a weird bug.
Wolfman2000 Can a class be a String?
colomon I'm going to go ahead and check in what I've got so y'all cam look at it more closely. Then I will go to bed. :)
jnthn colomon: I need to sleep now. I'll look another time. :-)
Gah, it's nearly 4am. If I make church tomorrow^Wtoday morning it'll be a miracle... 02:50
(insomnia and sleeping in until 1pm to compensate)--
Anyway, night all
colomon night 02:52
pugs_svn r29023 | colomon++ | [t/spec] Add two simple tests for calling .Str on strings.
02:53 colomon left
Wolfman2000 ...what's the difference between pugs and rakudo again? 02:55
dalek kudo: 37d480a | (Solomon Foster)++ | src/setting/Str.pm:
Add Str.Str, with a caveat.
 02:57
03:02 justatheory left
Tene Wolfman2000: Pugs: haskell, first major impl, sleeping for years. Rakudo: parrot, currently being actively developed. 03:04
quietfanatic jnth++ # overloading trait_mod:<is> works great! 03:05
03:05 dj_goku left
Wolfman2000 Tene: Then...why is it being reported on pugs_svn? 03:07
Tene Wolfman2000: why is what being reported where? 03:12
Wolfman2000 Tene: The robot, pugs_svn
Is that on pugs or rakudo?
Tene Wolfman2000: that's reporting commits to the pugs repository, which also houses the Perl 6 test suite.
which is distro-independant. 03:13
s/distro/impl/ 03:14
Wolfman2000 So Pugs has the test suite, yet Rakudo is the one being worked on
Tene the pugs *repository* is where the official test suite is currently located.
because pugs has a completely open commit policy 03:15
rakudo has a stricter policy for who is allowed to commit, but we want to allow anybody to help with the spec tests.
so it was just left in the pugs repo... no reason to move it.
there are a few other Perl 6 projects being worked on in the pugs repo. 03:16
there's an evalbot in there, I think the perl6.org website is in there, a couple other smaller compiler projects... 03:17
Wolfman2000 Tene: I know feather.perl6.nl is nosted on there
hosted*
I commited some fixes to index.html
03:22 tak11 joined
quietfanatic I'm getting a "Internal Error: Rakudo_binding_bind_signature passed invalid signature" 03:23
03:26 pointme left, pointme joined, mubot joined, mubot left, zaslon left, zaslon joined 03:27 mubot joined, mubot left, seanstickle joined 03:28 mubot joined, mubot left
carlin o_o that's three times in a row mubot has segfaulted trying to start up 03:28
03:30 mubot joined, mubot left 03:32 ted__ left
quietfanatic I think this error has to do with my overloaded trait_mod:<is> interfering with "is export"--but it shouldn't. 03:33
eternaleye We need a paste.perl6.org 03:37
Preferably with P6Regex-based highlighting 03:38
Wolfman2000 I may be able to make a database for planning a pastebin, but I would need help with some of the other features such as the regex highlighting and nickname remembering. 03:39
Perhaps my web development skills will come in handy 03:40
...no psql. Crap. Alright, mysql will have to do.
Can't bother Juerd with everything.
quietfanatic Erk this is a (different) weird bug... 03:43
rakudo: BEGIN { .signature.perl.say for &trait_mod:<is>.candidates }; say &trait_mod:<is>.defined
p6eval rakudo 37d480: :(Object $child where all(), Object $parent)␤:(Object $child where all(), Any :hidden($hidden)!)␤:(Code $block, Any $arg?, Any :export($export)!)␤:(Code $block, Any :default($default)!)␤:(ContainerDeclarand $c, Any :rw($rw)!)␤1␤
quietfanatic This is fine (note the 1 at the end)
rakudo: module M; BEGIN { .signature.perl.say for &trait_mod:<is>.candidates }; say &trait_mod:<is>.defined 03:44
p6eval rakudo 37d480: :(Object $child where all(), Object $parent)␤:(Object $child where all(), Any :hidden($hidden)!)␤:(Code $block, Any $arg?, Any :export($export)!)␤:(Code $block, Any :default($default)!)␤:(ContainerDeclarand $c, Any :rw($rw)!)␤0␤
quietfanatic See the 0?
Wolfman2000 quietfanatic: I do.
quietfanatic Actually that BEGIN is quite unnecessary to reproduce.
rakudo: say &trait_mod:<is>.defined; module M {say &trait_mod:<is>.defined>} 03:45
p6eval rakudo 37d480: Confused at line 2, near ">}"␤in Main (file <unknown>, line <unknown>)␤
quietfanatic rakudo: say &trait_mod:<is>.defined; module M {say &trait_mod:<is>.defined}
p6eval rakudo 37d480: 1␤0␤
quietfanatic Not defined in a module :(
lisppaste3 wolfman2000 pasted "eternaleye: Rough database plan for new pastebin. Unsure of pword file, but I don't want anyone to claim a name that others use." at paste.lisp.org/display/89971
quietfanatic rakudo: say &trait_mod:<is>.defined; module M {say &CORE::trait_mod:<is>.defined} 03:46
p6eval rakudo 37d480: Confused at line 2, near ":<is>.defi"␤in Main (file <unknown>, line <unknown>)␤
quietfanatic Is &trait_mod:<is> in the module referring to M::<&trait_mod:\<is\>> instead? 03:47
(a similarly-named subroutine but in the M:: namespace)
03:47 synth joined
Wolfman2000 eternaleye: your thoughts on my paste? 03:53
eternaleye Wolfman2000: looking at it now 03:54
Wolfman2000 *nods*
eternaleye Wolfman2000: I'm not sure I understand it 03:55
Wolfman2000 eternaleye: It's a database plan
to organize the pastes
eternaleye Wolfman2000: Yes, but I'm not quite sure what you mean by that. Are these tables in an SQL database? Are these directories? It's very light on any form of explanation - it's like giving the datatypes of the elements in a struct, without giving any information on how the members (or the struct) will be accessed/used/etc. 03:58
Wolfman2000 eternaleye: my apologies on that. They are tables in a SQL database. At least, that's the plan.
eternaleye It's data without an algorithm 03:59
pmichaud hello again, back from dinner 04:00
02:31 * jnthn chases pmichaud, yelling something about him not updating PARROT_REVISION in rakudo's ng branch
Wolfman2000 evening pmichaud
04:00 mubot joined, mubot left
pmichaud I did update PARROT_REVISION! 04:00
...I just forgot to commit+push... :-( 04:01
eternaleye Well, there are a few things I'd recommend. One - How do you plan to handle passwords? I see the pword field, but how do you plan to store them? Hashed? Hashed and salted? For an example of a high-profile site that didn't properly secure its passwords, look at the semirecent PerlMonks snafu.
Wolfman2000 eternaleye: This was meant to be a base plan. I'm well aware of the implementation that needs to take place. I do have other websites to my name after all. ;)
eternaleye Cool. 04:02
Wolfman2000 I wanted to know if the data part was fine. I know of the algorithms to implement for the database part.
Thank you for your concern though.
04:03 seanstickle left
eternaleye Also, I recommend looking at ix.io for a _really_ simple pastebin implementation that actually does really well. Its commandline client in particular is very well-done 04:03
Wolfman2000 ...alright, time to play the ignorant card. Why should I worry about having a commandline client? 04:04
eternaleye The only things I can think of that it _doesn't_ do and which are desirable are a.) the channel notification thing that paste.lisp.org does and b.) soem form of 'paste channels' so that you can group pastes of a certain type (like how lisppaste has a perl6 section) 04:05
Wolfman2000: As far as a commandline client goes, there have been a _lot_ of times when I've needed to get input on an issue when I don't have a browser. 04:06
04:06 NorwayGeek left
Wolfman2000 eternaleye: I'll admit that I could use some help with the channel notification part. As far as "paste channels" goes...hmm...is there perhaps a reason to allow a paste to go to multiple channels at the same time? 04:06
eternaleye Almost everyone has irssi+wget/curl/whatever. MFar fewer have links/elinks/lynx 04:07
s/M//
Wolfman2000 eternaleye: you mean a text browser? Ah.
I wish you said that in the first place. I can code for a text browser easily.
...well...almost easily.
Perl 6 will involve more UTF-8 than before. 04:08
eternaleye Wolfman2000: Possibly (think parrot and rakudo both being involved in a bug), but it would complicate implementation greatly
Wolfman2000: Also, pasting things is _very_ difficult without X
Wolfman2000 eternaleye: You notice that I used groups and pastes separately, right? My thinking here is as follows: 04:09
eternaleye I may want to paste a file. It's much easier to go to a terminal and type 'ix $filename' than it is to open it in an editor, copy the text, go to a webpage, and paste it
Wolfman2000 ...holding my thought while I read that
eternaleye Plus, stdin. <command> | ix gives instant pasting of logs 04:10
Wolfman2000 ...so that's what you mean by a command line client.
eternaleye Wolfman2000: Yes, The ix client can also fetch pastes given the pasteid, though
Wolfman2000 I'll need a partner for this one then. I'm better with the web development part than the command line part.
eternaleye Wolfman2000: If you implement the same POST API as ix.io, or very similar, we can basically steal their client ;D 04:11
GPL++
04:11 meppel joined
Wolfman2000 ...the GPL may be useful here, but...I admit I have my own ideas on URL implementation. 04:12
And it's that that I want to discuss here.
04:12 mubot joined
Wolfman2000 Some basics first: I'm assuming Catalyst::Perl: I have some experience with it, and this is a perl channel after all. Second: assume URL is where it's located for now. 04:13
04:13 mubot left
eternaleye Well, the root page of ix.io gives a rather complete description - sufficient to interact with the server using only curl or wget. THat means it should be trivial to adapt it to fit your ideals. 04:13
04:13 mubot joined
eternaleye Okay 04:13
Wolfman2000 I have Pastes and Groups separately for the purpose of annotations. I would say URL/GroupID/GroupNum to load the specific paste, or GroupID on its own to load all of the pastes in that group. 04:14
If one wanted to see all of the pastes of the user, then we could go into URL/user/UserID or something along those lines.
eternaleye Okay 04:15
Might I suggest URL/user/$username and URL/group/$groupid, so that people can look up all posts by as pecific user as well? That also makes it more extensible
Wolfman2000 What I'm right now torn about...do we allow multiple people to have the same username? Should we try to lock a screen name to one person only?
eternaleye: That's fair.
As long as we enforce that usernames have to have at least one letter. 04:16
eternaleye How about this: If a username has been 'registered' and given a password, only accept pastes with that password as belonging to that username. Otherwise, free-for-all (the IRC system, basically) 04:17
Wolfman2000 so the user's table should include an isRegistered field?
eternaleye Wolfman2000: I mean the actual text 'user' to denote that after the next slash is a username, and 'group' to denote that after the next slash is a groupid. No restrictions on usernames needed then.
Wolfman2000: Simpler. Only add to the table when registering. If the username isn't in the table, short-circuit. 04:18
Wolfman2000 Otherwise, have all non registered posts point to an anonymous user. Fair enough. 04:19
We may want to restrict viewing the anonymous user's posts publicly though.
eternaleye Ack, I think I made a mistake in my thinking. One sec.
Wolfman2000 take your time 04:20
eternaleye How are you planning to group posts by user? As a list of ids in a field of the user table, or by selecting from pastes where user = $whatever ?
I was assuming the latter 04:21
Wolfman2000 the second one
the first one doesn't exactly allow for good table design if I'm interpreting your statement right
eternaleye Okay. Then you don't need to map them onto an anonymous user. You've already validated that the username they want isn't registered; therefore, just give the paste the username they want! 04:22
04:22 bpetering joined
bpetering hai everyone :) 04:22
Wolfman2000 ...okay, I need to back up
eternaleye Simplest thing that could possibly work etc. 04:23
hi bpetering!
bpetering howdy eternaleye, Wolfman2000 :)
Wolfman2000 I think there should be a table of registered users, and the pastes table should point to a registered user. If the paste is NOT of a registered user, should we still include a username field in the paste? Or do we put all of the names in the users table, and just have a boolean to see if a user is registered?
eternaleye have them type in a username when they submit a paste. Blank gets anonymous user, anything registered prompts for password, unregistered gets accepted w/o question
Wolfman2000: I was assuming 'username' field was a string 04:24
Wolfman2000 eternaleye: If the person is unregistered, how do we get them to register?
04:24 gfx joined
eternaleye URL/register 04:24
Separate form
Wolfman2000 alright, fair enough on that 04:25
carlin What happens if someone registers a username that has been used anonymously in the past?
Wolfman2000 carlin: ...good question.
eternaleye Hm, that is a problem. 04:26
Wolfman2000 That's partly why I think unregistered people should just be sent to anonymous 04:27
Either that, or we do away with the login stuff entirely.
eternaleye I'd say we can explicitly state that there are no guarantees about paste lifetimes made with unregistered usernames, and just wipe them on registration
All's fair if you predeclare
04:27 meppl left
Wolfman2000 ...tough love, but fair enough 04:28
eternaleye no-username pastes will be guaranteed to live for the specified lifetime - just give people the choice between vanity + security, vanity - security, and security - vanity
Wolfman2000 ...what's the difference between vanity - security and security - vanity? They seem similar sounding 04:30
eternaleye Vanity without security = unregistered username (no lifetime guarantees, but it has your name). Security without vaity = no username (Duration guarantee, but no associated name) 04:31
*vanity
subtraction is not commutative ;D 04:32
Wolfman2000 eternaleye: thanks for that reminder
Alright...next questions. 1) Should we host this on feather, even though this means whatever password we put on the database won't be super secure? 2) Does feather allow us to cpan and install perl 5 modules? 04:33
04:33 synth left
eternaleye I'd say security is probably not a big deal _yet_ - it will be mainly used for sharing snippets of p6 code for the forseeable future. When necessary, it can be moved and the password changed. 04:35
Also makes it easy for it to be on the perl6.org domain, which is already hosted on feather
Wolfman2000 That answers my 1st question, and my unasked 3rd question: whether I'd have to buy a domain name. 04:36
eternaleye and IIRC Juerd admins it, which means at most you'd have to ask him - and he's in-channel
04:37 c9s joined
Wolfman2000 ...alright, I'll find out the 2nd one 04:38
lisppaste3 wolfman2000 pasted "Umm...guessing we shouldn't mess with cpan yet." at paste.lisp.org/display/89973 04:39
dalek p-rx: 57d2a44 | pmichaud++ | src/NQP/ (2 files):
Add <block> subrule.
p-rx: 5559916 | pmichaud++ | (3 files):
Add closure interpolation.
04:40 xp_prg joined
eternaleye Wolfman2000: What are you thinking as far as the POST API goes? 04:42
Wolfman2000 eternaleye: browser based or command line based?
eternaleye CLI 04:43
Wolfman2000 ...I'm primarily thinking of the browser end, I must admit.
The big issue I'm having is the URL format suggested by ix.io
I don't like the idea of including so much in the URL. 04:44
Plus, I'm unsure about pasting multiple things in one go.
eternaleye Wolfman2000: The thing is, the ix.io client is easily adapted to anything matching /URL\/<string.?\?[<key>'='<value>]*/ 04:45
Wolfman2000 Sorry if that's not what you wanted to hear
eternaleye And multipasting is not particularly important.
Wolfman2000 eternaleye: If we use Catalyst::Perl, I'd rather NOT use key=value. I'd rather use /value/value2
And Route the URL as required
eternaleye Although it might be best to adapt their client to Perl (5|6) (it's in python) 04:46
Wolfman2000: Yes, but what about optional keys? What if I want to specify key1 and key3, but leave key2 undefined?
Wolfman2000 I caught the python comments. One of my websites, www.pumpproedits.com/, is in Pylons (Python framework) (hopefully that page still works). I suggest we adapt to Catalyst::Perl, which uses Perl 5.
eternaleye: Let's go over the contents then. 04:47
f:N: urlencoded contents. I assume that's the paste stuff. That should NEVER be in the URL.
eternaleye Wolfman2000: I'm not suggesting using their server; just taking their client as a model of what a 'good CLI user interface' should be
Wolfman2000: Agreed 04:48
Wolfman2000 name:N: filename...ah, I see what you mean by command line. Rip the file contents, make those the paste.
filename: should be required honestly. If more than one file is sent in, more power to you.
ext:N: filetype: irrelevant. 04:49
eternaleye Wolfman2000: I totally agree the file contents should be POSTed, not uploaded via a GET URL
Wolfman2000: Stdin.
No filename in that case.
We could default that to 'Standard Input' though
Wolfman2000 eternaleye: possibly. I don't know if we need an ext if we're using the filename though.
login: "quote" the name. 04:50
eternaleye Wolfman2000: The way ix does it is it submits non-extension filename as filename, and extension as ext. It splits off the extension when given a file, so the server doesn't need to paste the filename. 04:51
Wolfman2000 token: instead of including the token/password here, have CLI query the database, see if the name exists, and then prompt for a password.
04:51 keeth joined
eternaleye *parse 04:51
Wolfman2000 rm: not allowing people to delete that easily
id:N: no replacing easily either.
eternaleye Wolfman2000: The thing is, if it's being used at the end of the pipe, how will it read a password?
Wolfman2000 eternaleye: end of a pipe? Clarify that. 04:52
eternaleye <command> | client
Unix pipes
It won't be able to read a password from standard input, since that';s where the data is coming from
Wolfman2000 ...honestly, do you like the idea of people putting in a plain text password in the command line? Or am I missing something? 04:53
eternaleye 1.) Only root can read their files (.bash_history) unless they're idiots. 2.) THe client isn't long-running, so the chance of someone seeing it in the process table with something like 'ps aux' is very low 04:54
Wolfman2000 ...so we need to define the command line process. 04:55
along with how to pass files along.
eternaleye Yeah 04:56
bpetering Wolfman2000: what if you get the wrong file by (honest) mistake?
Wolfman2000 bpetering: then the file pastes
Tene see also, tools/dev/nopaste.pl in the parrot repo
eternaleye bpetering: I'm thinking we should only allow deletions whn authenticated
bpetering: The choice is between allowing people to redact their stuff when authenticated, and preventing malicious users from deleting unauthenticated pastes 04:57
bpetering eternaleye: sounds good
eternaleye s/ authenticated/unauthenticated/ 04:58
Wolfman2000 eternaleye: Permission to PM you?
eternaleye Granted 04:59
Tene eternaleye: Permission to steal your wallet?
bpetering Tene: that's rather... malicious :)
Tene s/: that// 05:00
eternaleye Tene: Denied
Tene dammit
Talk about malicious...
bpetering Tene: me? think i'm getting confused with someone else... :) 05:02
Tene No, I was saying that it was malicious for eternaleye to deny me.
Wolfman2000 Tene: no wallet stealing 05:03
bpetering Tene: oh, sorry... my mistake
Wolfman2000 Alright...I can make the database structure that I'm thinking of tomorrow. I hope to get to bed within an hour. 05:04
anything else that needs discussing of this subject? 05:05
05:09 dj_goku joined
bpetering Wolfman2000: i don't think so :) 05:09
eternaleye Tene: Besides, KDE encrypts my password wallet with AES
Wolfman2000: Maybe we can call the client 6p (perl6 Paste), so it's nice and short like is :D
AGHHHHH I'm lagging so ridiculously badly. I should not have tried testing BtrFS' online resize with a 250g delta. Now it's in kernelspace and I can't even kill it, despite it eating all my disk bandwidth.
Wolfman2000 We'll argue over the name later. There are still some things I need to take care of before bed. 05:11
1) finish up a homework assignment. 2) actually take care of that linked list implementation I'm behind on doing.
bpetering Wolfman2000: good luck with the homework assignment :) 05:12
Wolfman2000 bpetering: the homework is simple
just write a report based on my teacher observation experience this past Wednesday
bpetering Wolfman2000: did you learn anything interesting? :) 05:14
Wolfman2000 bpetering: last wednesday, or in general? 05:15
bpetering Wolfman2000: in general
Wolfman2000 Wilmington, NC does not offer programming classes in high schools. I'll have to just accept Math...and hope I don't get forced to teach History 05:16
I'll say more once I make my post 05:17
quietfanatic phenny: tell jnthn Unfortunately it looks like overloading trait_mod:<is> doesn't make a versatile subclassing hook, because the left argument given to it is an unnamed and uninitialized ClassHOW object. 05:23
phenny quietfanatic: I'll pass that on when jnthn is around.
Wolfman2000 right, phenny 05:24
diakopter quietfanatic: but you could grab it from the .source it throws by throwing and catching an exception :D
Wolfman2000 phenny: tell Juerd Planning a Perl 6 Pastebin: any assistance you can provide would be appreciated. Contact myself or eternaleye for more information.
phenny Wolfman2000: yeah, yeah
quietfanatic .source?
diakopter the error message, I mean..
Wolfman2000 ...did phenny pass the info on, or is phenny going to ignore me on that
diakopter Juerd: hilite 05:25
carlin Wolfman2000: phenny does that from time to time, IIRC in its source it has a 1/1000 chance to say that
diakopter nice
Wolfman2000 carlin: So where's my money for hitting a mini jackpot?
quietfanatic diakopter: I don't think that'll work because it's just a bare ClassHOW object that doesn't have any metadata. It only expects me to add a parent class before it initializes it properly.
diakopter oh 05:26
bpetering Wolfman2000: your odds are better with hugme, it'll blush 1 time in 100 :)
carlin inamidst.com/phenny/modules/tell.py # CTRL+F for "yeah, yeah"
Wolfman2000 okay, assignment submitted
quietfanatic I don't think there's anything I can throw and catch that contains the name of the class that hasn't been defined yet :)
Wolfman2000 I got the super rare one. :D 05:27
quietfanatic But surely the name must exist somewhere higher in the stack, I just can't reach it. :( 05:28
Wolfman2000 ...alright, this may be a silly question here, but...why do we have a python bot in a perl room?
bpetering Wolfman2000: thievery is a form of laziness? :) 05:29
Wolfman2000 *shrugs* 05:30
Now, to my linked list...hopefully
bpetering Wolfman2000: what're you blocking on? 05:31
Wolfman2000 bpetering: blocking on? that's a new one on me
bpetering
.oO( we need a definition bot... )
carlin I could port Purl to Perl 6 05:32
bpetering ... but i'll be happy to help: "X is preventing Y from being accomplished" === "Y is blocking on X" 05:33
Wolfman2000 bpetering: Starting fresh for this one.
This time, I built the Node class separately, and got that tested. Check the IRC logs for that one.
bpetering carlin: purl seems cool. :) 05:34
carlin bpetering: you obviously haven't seen jnthn and masak complaining about how annoying it is? :p 05:35
05:36 frew joined
bpetering Wolfman2000: #89963? 05:37
Wolfman2000 bpetering: I don't recall which paste it was, sorry.
eternaleye carlin: Maybe make one change to purl, and require a leading '?' (ignoring lines that are _only_ '?')
bpetering carlin: can't be as annoying as lambdabot... unless you made it trigger on vowels, or something :| 05:38
even lambdabot is useful though. just a bit trigger-happy :) 05:42
lisppaste3 wolfman2000 pasted "Null PMC access in type() in Main <-- Umm...what does this mean?" at paste.lisp.org/display/89974
eternaleye bpetering: lambdabot's only really trigger-happy in perl-based channels, because arrays lead to leading @-signs 05:44
carlin and rakudo's REPL leads to leading > signs 05:45
Wolfman2000 ...is this going to be one of those questions where I have to wait for the morning crew? 05:47
bpetering well, i couldn't have written lambdabot. i guess i'll stop complaining until i write my own. :)
carlin Wolfman2000: a Null PMC access is sort of like a rakudo segfault 05:48
Wolfman2000 ...just once, I'd like to be able to make code that just WORKS 05:49
carlin if you had a working version of that and you know what you changed to caused it would help 05:50
Wolfman2000 And I don't mean just simple square root problems.
carlin: This is my only version. Well...my feather box has node_good.pl, but that's before I implemented the linked list stuff. The Node class by itself worked. 05:51
...well, except for when you tried to Str() a Str, but that's a rakudobug
lisppaste3 wolfman2000 annotated #89974 "Rakudo must not like inner classes. Still an error though: invoke() not implemented in class 'Node'" at paste.lisp.org/display/89974#1 05:53
Wolfman2000 ...and when I google the error message, paste 89892 is the first result. Great, I'm famous. 05:54
carlin Try changing new Node(""); to Node.new 05:56
Wolfman2000 carlin: ...did I make the same mistake again?
...BAH 05:57
at least I have a new error now 05:58
bpetering Wolfman2000: which error? 05:59
lisppaste3 wolfman2000 annotated #89974 "set_integer_keyed() not implemented in class 'Sub' in Main <-- Umm...my classes are LinkedList and Node." at paste.lisp.org/display/89974#2
Wolfman2000 bpetering: ^
06:00 araujo left 06:01 c9s left, c9s joined, c9s left, c9s joined
bpetering Wolfman2000: some things just aren't ready yet... eventually they'll "just work" :) 06:03
Wolfman2000 ...so even though I've designed the basics of the linkedlist properly this time, I still can't make it work. OY 06:04
bpetering Wolfman2000: frustrating, isn't it?
Wolfman2000 bpetering: Yes. :( 06:06
Looks like I'll have to wait for whatever's going on in the ng branch to be merged to the main branch, then rebuild rakudo
Unless you have another idea. 06:07
bpetering sometimes leading by example works wonders... rather than waiting, i mean... 06:10
Wolfman2000 you mean I should use the ng branch? 06:11
06:13 hcchien left 06:16 keeth left 06:18 hcchien joined 06:20 synth joined
bpetering Wolfman2000: sorry, was afk - needed a drink :) 06:29
dalek p-rx: ee6743c | pmichaud++ | src/NQP/ (2 files):
Convert package_def to use <block> instead of <pblock>.
 06:30
p-rx: 1e2187a | pmichaud++ | t/nqp/38-quotes.t:
Fix bug in interpolated closure test.
p-rx: 457fa04 | pmichaud++ | src/NQP/Actions.pm:
Update handling of block arity.
p-rx: 992b220 | pmichaud++ | (3 files):
Enable pointy blocks.
Wolfman2000 bpetering: that's fine. apparently I never remembered to turn off /away, so I was apparently away the hwole day. :P
bpetering Wolfman2000: :P 06:32
Wolfman2000 but I've stayed up too late, so I'm off to bed for real. 06:33
We can talk more later.
bpetering Wolfman2000: sounds good. 06:35
06:36 tak_ joined 06:38 tak11 left, tak_ is now known as tak11 07:04 synth left 07:05 synth joined 07:14 bpetering left 07:16 tak11 left 07:34 xenoterracide left 07:36 xenoterracide joined 07:40 SmokeMachine left 07:51 ihrd joined, ihrd left 07:52 frew left 07:54 [synth] joined 07:55 SmokeMachine joined, synth left 08:05 fax left 08:11 ihrd joined 08:13 ihrd left 08:17 gbacon left 08:22 gfx left 08:26 jrockway left, [synth] left, akl left, SmokeMachine left, dj_goku left, zaslon left, emma left, xinming left, markmont left, c9s left, envi^home left, pnate2 left, IllvilJa left, nbrown left, rgrau left, shachaf left, reid05 left, payload left, p6eval left, cotto_work left, tylerni7 left, ssm left, daemon left, yath left, lmc left, hicx174 left, _jaldhar left, _eMaX_ left, japhb left, simcop2387 left, hugme left, jnthn left, Juerd left, yves left, cosimo left, athomason left, Maddingue left, omega_ left, lambdabot left, parduncia left, constant left, krunen left, Infinoid left, pugs_svn left, dukeleto left, hatseflats left, jan_ left, szabgab left, jiing_ left, kst left, lisppaste3 left, spinclad left, Woodi left, cls_bsd left, crazed left, sri__ left, phenny left, Patterner left, Wolfman2000 left, pmichaud left, dukelet0 left, KatrinaTheLamia left, silug left, xenoterracide left, hcchien left, xp_prg left, pointme left, hsb left, alester left, allbery_b left, Helios- left, drbean left, charsbar left, obra_ left, leedo left, revdiablo left, rafl left, moritz_ left, awwaiid left, r0bby_ left, sjohnson left, gfldex left, baest left, renormalist left, Grrrr left, cotto left, mtve left, wolverian left, nothingmuch left, literal left, christine left, Trey left, TimToady left, diakopter left, Bucciarati left, stepnem left 08:27 pmichaud joined, SmokeMachine joined, [synth] joined, xenoterracide joined, c9s joined, dj_goku joined, xp_prg joined, zaslon joined, pointme joined, yath joined, envi^home joined, payload joined, emma joined, xinming joined, pnate2 joined, IllvilJa joined, lmc joined, Patterner joined, nbrown joined, rgrau joined, jrockway joined, markmont joined, shachaf joined, reid05 joined, hicx174 joined, _jaldhar joined, dukeleto joined, _eMaX_ joined, Wolfman2000 joined, p6eval joined, cotto_work joined, tylerni7 joined, akl joined, krunen joined, hsb joined, japhb joined, alester joined, allbery_b joined, simcop2387 joined, dukelet0 joined, KatrinaTheLamia joined, hugme joined, Infinoid joined, pugs_svn joined, jnthn joined, irc.freenode.net sets mode: +ovvo pmichaud hugme pugs_svn jnthn, Juerd joined, hatseflats joined, jan_ joined, sjohnson joined, szabgab joined, yves joined, gfldex joined, jiing_ joined, kst joined, Helios- joined, drbean joined, charsbar joined, lisppaste3 joined, spinclad joined, lambdabot joined, cls_bsd joined, constant joined, cosimo joined, crazed joined, athomason joined, daemon joined, ssm joined, Woodi joined, parduncia joined, sri__ joined, omega_ joined, Maddingue joined, phenny joined, r0bby_ joined, wolverian joined, cotto joined, Grrrr joined, renormalist joined, baest joined, revdiablo joined, awwaiid joined, christine joined, Bucciarati joined, silug joined, stepnem joined, leedo joined, literal joined, rafl joined, Trey joined, obra_ joined, moritz_ joined, irc.freenode.net sets mode: +vvvo lisppaste3 lambdabot phenny moritz_, nothingmuch joined, mtve joined, TimToady joined, diakopter joined, irc.freenode.net sets mode: +oo TimToady diakopter 08:32 hcchien joined 08:33 hcchien left 08:34 ihrd joined 08:37 hcchien joined 08:42 ihrd left, ihrd joined 08:46 jrockway left, hcchien left, [synth] left, akl left, markmont left, zaslon left, dj_goku left, SmokeMachine left, xinming left, emma left, nbrown left, IllvilJa left, pnate2 left, envi^home left, reid05 left, shachaf left, c9s left, rgrau left, ihrd left, cotto_work left, payload left, p6eval left, tylerni7 left, ssm left, daemon left, athomason left, cosimo left, lmc left, _eMaX_ left, yath left, Maddingue left, hugme left, japhb left, hicx174 left, _jaldhar left, yves left, Juerd left, jnthn left, lambdabot left, omega_ left, simcop2387 left, parduncia left, constant left, sri__ left, Woodi left, crazed left, spinclad left, jiing_ left, pugs_svn left, Infinoid left, krunen left, kst left, jan_ left, szabgab left, dukeleto left, hatseflats left, cls_bsd left, phenny left, lisppaste3 left, dukelet0 left, Wolfman2000 left, Patterner left, KatrinaTheLamia left, silug left, pmichaud left, obra_ left, rafl left, leedo left, revdiablo left, Helios- left, hsb left, pointme left, xenoterracide left, xp_prg left, charsbar left, moritz_ left, drbean left, awwaiid left, alester left, r0bby_ left, allbery_b left, nothingmuch left, christine left, baest left, renormalist left, Grrrr left, gfldex left, sjohnson left, cotto left, wolverian left, mtve left, literal left, TimToady left, Trey left, diakopter left, Bucciarati left, stepnem left, orafu left, eternaleye left, mj41 left, estrai left, PZt left, base_16 left, eiro left, jjore left, mubot left, PacoLinux left, Caelum left, slavik left, tomaw_ left, quietfanatic left, cognominal left, Woody2143 left, BinGOs left, jantore left, Gothmog_ left, presh_ left, felipe left, bloonix left, kolibrie left, elmex left, he_ left, rhr left, viklund left, mathw left, frettled left, huf left, pjcj left, Khisanth left, nsh left, sunnavy left, avuserow left, ilbot2 left, sbp left, IRSeekBot left, lestrrat left 09:00 yahooooo left 09:35 slavik joined, PacoLinux joined, mubot joined, Jedai joined, he_ joined, sunnavy joined, Khisanth joined, ilbot2 joined, pjcj joined, rhr joined, sbp joined, nsh joined, IRSeekBot joined, huf joined, avuserow joined, viklund joined, lestrrat joined, mathw joined, frettled joined, elmex joined, bloonix joined, kolibrie joined, PZt joined, estrai joined, mj41 joined, eternaleye joined, orafu joined, BinGOs joined, Woody2143 joined, cognominal joined, Gothmog_ joined, jantore joined, presh_ joined, felipe joined, jjore joined, eiro joined, base_16 joined, yahooooo joined, gbacon joined, quietfanatic joined, irc.freenode.net sets mode: +vv ilbot2 IRSeekBot, tomaw_ joined, pmichaud joined, Woodi joined, reid05 joined, iblechbot joined, hcchien joined, SmokeMachine joined, xenoterracide joined, c9s joined, dj_goku joined, xp_prg joined, zaslon joined, pointme joined, yath joined, envi^home joined, payload joined, emma joined, xinming joined, pnate2 joined, IllvilJa joined, lmc joined, Patterner joined, nbrown joined, rgrau joined, jrockway joined, markmont joined, shachaf joined, hicx174 joined, _jaldhar joined, dukeleto joined, _eMaX_ joined, Wolfman2000 joined, p6eval joined, cotto_work joined, tylerni7 joined, akl joined, krunen joined, hsb joined, japhb joined, alester joined, allbery_b joined, simcop2387 joined, dukelet0 joined, KatrinaTheLamia joined, hugme joined, Infinoid joined, pugs_svn joined, irc.freenode.net sets mode: +ovv pmichaud hugme pugs_svn, jnthn joined, Juerd joined, hatseflats joined, jan_ joined, sjohnson joined, szabgab joined, yves joined, gfldex joined, jiing_ joined, kst joined, Helios- joined, drbean joined, charsbar joined, lisppaste3 joined, spinclad joined, lambdabot joined, cls_bsd joined, constant joined, cosimo joined, crazed joined, athomason joined, daemon joined, ssm joined, parduncia joined, sri__ joined, omega_ joined, Maddingue joined, phenny joined, irc.freenode.net sets mode: +ovvv jnthn lisppaste3 lambdabot phenny, r0bby_ joined, wolverian joined, cotto joined, Grrrr joined, renormalist joined, baest joined, revdiablo joined, awwaiid joined, christine joined, Bucciarati joined, silug joined, stepnem joined, leedo joined, literal joined, rafl joined, Trey joined, obra_ joined, moritz_ joined, nothingmuch joined, mtve joined, TimToady joined, irc.freenode.net sets mode: +oo moritz_ TimToady, diakopter joined, irc.freenode.net sets mode: +o diakopter 10:02 rgrau` joined
moritz_ good morning 10:04
dalek kudo: 97ced10 | moritz++ | t/spectest.data:
[t/spectest.data] track file rename
 10:11
10:11 fax joined
Woodi good morning 10:15
10:15 NorwayGeek_ joined 10:16 NorwayGeek|Away joined
Woodi i got impression that this # is overloaded like Lisp brackets :) 10:17
10:18 rgrau left
Woodi kind a virtual room would be good for it :) 10:18
with nice graohics and 3d space :)
10:23 NorwayGeek joined 10:35 NorwayGeek_ left 10:36 NorwayGeek|Away left
moritz_ the rakudo ng branch builds in 28s (with three jobs parallel on 2 cores) 10:40
it even passes a few subtests of the spectests ;-) 10:43
Woodi woow :)
any prediction on master merge time ? :)
moritz_ this month 10:44
10:44 kaare joined
Woodi ng branch is that refactoring to nqp ? 10:45
moritz_ not only that; it is also a rewrite of most of rakudo :) 10:46
at least the parser and action methods
10:46 arthur-_ joined 10:47 envi^home left
Woodi as a user i mainly care about tests passed and build proces :) 10:47
and stories about development proces :>
love stories :)
moritz_ then read jnthn's and pmichaud's blogs ;-) 10:48
Woodi moritz_: i hear onece something like "monadic revolution"... i am specially interested how it grow in perl community... 10:49
begin was sponsored by Audrey..
and what happens next ?
moritz_ only know nomads 10:50
10:51 envi^home joined
Woodi :) 10:51
nvm. but i am realy concern about this story... it is big paradigm shift from structural-objects to something more :) 10:52
moritz_ I always viewed monads as a concept to work around statelessness in haskell, and thus superfluous in Perl
but maybe I'm on the totally wrong track here 10:53
or maybe just nobody explained to properly why we might want/need explicit monads in Perl
Woodi i can be wrong... for me monads are like sandboxes with cleaners or just classes... but i thinked perl6 want to be functional language too... 10:54
or maybe extracting lazines component from functional languages is everything what we need ? :) 10:57
10:59 rgrau`` joined
moritz_ we do have other features from functional languages 11:00
like multi sub, easy currying, closures 11:01
11:03 rgrau` left
Woodi i am new on that names... - no idea when to use that features... 11:06
closures looks just like static fields... 11:07
btw. do rakudo have threads atm ?
11:08 fax left
moritz_ no 11:08
Woodi but planned ? 11:09
moritz_ planned, yes
afk, lunch 11:10
Woodi smacznego :)
11:11 pkkm joined
pkkm can I ask about perl regular expressions here? 11:11
Woodi yes 11:12
but you should not ask if you can ask - info in topic usually:)
pkkm: polak ? 11:14
pkkm I want my pcre to catch all google ads. I don't know what I did wrong. Screenshot from online regex tester: img32.imageshack.us/img32/3122/zrzutekranu1y.png
Woodi: tal
tak*
In that case pcre will be the same as perl regex. 11:15
Woodi mainly should :) perl5 like
here is perl6 # so a lot of new syntax ... 11:16
pkkm: you want match everything in one step... 11:20
11:21 kidd joined
Woodi pkkm: maybe first catch right <script> ... </scrip> then extract what you need ? 11:21
pkkm Woodi, will it change anything?
anyway, I need this pcre in one line 11:22
Woodi pkkm: iw will make more clear regex for example 11:23
pkkm Woodi: thany You.
Woodi and error on top of image can mean that you put \ before some letter which isn't a modifier 11:24
at least not in pcre
pkkm I put it before * in (?:/\*.*\*/)? 11:25
but how could it be done without it?
\s* - whitespaces 11:26
Woodi: Do you want me to send You this pcre in plaintext?
11:27 am0c joined
Woodi i think this regex is overcomplicated... for example you match two <scripts> at once and second looks like constant... 11:27
why put second in regexp ? 11:28
11:29 fax joined
pkkm I want to catch whole block of code responsible for google ads 11:32
I need it
It will be ettercap filter
look at this: paste.ubuntu.com/313194/
Woodi if you need blocks then catch blocks and throw away details from inside 11:33
pkkm It's intended to catch ALL google ads, even with comments and additional/removed whitespaces. 11:34
Woodi: I need to catch whole google ads code. 11:35
Woodi: And I need it in the exact form it's on website 11:36
Woodi i think that for catching blocks there is not need to describe everything inside blocks
11:37 payload1 joined 11:38 payload left
Woodi just match <script> google_ad_client ap to FIRST </script> and then next <script with syndication 11:38
you describe everything inside 11:39
if you want to extract more details from script blocks then use catching () 11:40
pkkm Woodi: and what if someone has broken website like: paste.ubuntu.com/313202/ 11:41
my pcre needs to be "bullet proof" 11:42
Woodi first script with google sindication is not ads 11:44
so you need to match only scripts with google_ad_client 11:45
pkkm <script type="text/javascript" src="pagead2.googlesyndication.com/pagea...ds.js"> </script>
it's responsible for showing ads 11:46
Woodi eg: /<!-- .+ google_ad_client = .+ -->/
pkkm I want this regex to catch WHOLE code responsible for ads
Woodi ah, right :)
look
html allows to send scripts not in order 11:47
so you cannot expect sindication will be just after google_ad_client
pkkm this regex is intended to be multipurpose, from counting ads, through marking them, to commenting them out 11:48
pcre*
Woodi so you need two regexes
it is not important yet 11:49
just make regex match what you want then use it for substitution 11:50
pkkm how can I make some strings to be matched, independent on order, with repetitions possible? 11:52
Woodi use separated regezes
keepp it simple
pkkm look at (?:(?#CAW - Comments And Whitespaces)(?:/\*.*\*/)?\s*.{0,5}(?:/\*.*\*/)?(?#END-CAW)) 11:53
Woodi terrible :)
pkkm (?:(?#CAW - Comments And Whitespaces)(?:/\*.*\*/)?\s*.{0,5}(?:/\*.*\*/)?(?#END-CAW))
if I could do something i told about before, I wouldn't have to repeat this 11:54
11:55 masak joined
masak oh hai, #perl6. 11:55
Woodi hi masak++
masak I find the backlog's discussion with emma very interesting.
pkkm it would look like noorder("(?:/\*.*\*/)?","\s*",".{0,5}")
masak emma: welcome, and thanks for posing interesting questions. 11:56
pkkm Woodi 2^
masak sjohnson: (re regex help) no problem :)
pkkm Woodi, it would be much simpler 11:57
masak sjohnson: (calling Wolfman2000 'wolfpig', and receiving a negative reaction) that's not a 'tough crowd', that's you being rude, unfortunately. 11:58
Woodi pkkm: use (something|something_other)
masak or sometimes [something|something_other] 11:59
Woodi masak: in perl5 or pcre ?
masak in Perl 6 regexes.
Woodi in p6 [ ] is still class ? 12:00
masak sorry, were you talking about Perl 5 regexes? I don't have the entire context? :)
Woodi: no, it's a non-capturing group.
Woodi ok
pkkm masak: about pcre
masak kind of like (?:something|something_else) in Perl 5.
pkkm masak: ettercap pcre :) 12:01
look at context
<pkkm> how can I make some strings to be matched, independent on order, with repetitions possible?
Woodi pcre is only one :)
pkkm <pkkm> (?:(?#CAW - Comments And Whitespaces)(?:/\*.*\*/)?\s*.{0,5}(?:/\*.*\*/)?(?#END-CAW))
<pkkm> it would look like noorder("(?:/\*.*\*/)?","\s*",".{0,5}")
Woodi it's c library :)
pkkm masak: so (sth|sth2) is not able to do that 12:02
12:02 jaffa8 joined
masak pmichaud: yes, but it captures. 12:02
pkkm Woodi, so how can I make it to be independent on order? 12:04
masak er, pkkm, not pmichaud. :/
12:04 ihrd joined
Woodi pkkm: use alternatives: | 12:05
pkkm: (<script google_ad_client|<script syndication) 12:06
pkkm: btw do you are allowed to sniff that stuff ? 12:07
pkkm Woodi: so will (?:(?:/\*.*\*/)?|\s*|.{0,5})* capture /*sth*/saf/*sth*/ /*comment*/
Woodi private or inside-company ?
pkkm Woodi: my home network
so I think I am
Woodi looks it will match *sth* and match saf and match last one separatelly 12:10
pkkm and will (?:(?:/\*.*\*/)?|\s*|.{0,5})* capture /*sth*/saf/*sth*/ /*comment*//*sth*/saf/*sth*/ /*sth*/saf/*sth*/ /*comment*/ /*comment*/ 12:11
?
what do You mean 'separately'?
whole pcre returns true/false 12:12
so how can something be separate?
Woodi ehmmi think /\*.*\*/ will match from start to end 12:13
masak pkkm, Woodi: so, how did you come to be discussing this on #perl6? 12:14
Woodi just anseringbot i am... 12:15
sory :)
masak no problem. just mind the signal-to-noise ratio.
pkkm Woodi: oh yes :D
Woodi is answering&asking addicted :) 12:16
pkkm can (sth|sth2) catch more than one time? 12:17
Woodi pkkm: yes, default * and + are gready
masak greedy.
Woodi right :) and both in p5 and p6 12:18
12:18 NorwayGeek left, hcchien left 12:19 Whiteknight joined
pkkm so (?:sth1|sth2|sth3|sth4)* would do? 12:20
moritz_ std: / (?:sth1|sth2|sth3|sth4)* / 12:21
Woodi will match one of sth, as far as it can
p6eval std 29023: ===SORRY!===␤quantifier quantifies nothing at /tmp/Q6fH9m0Qzs line 1:␤------> / (?:⏏sth1|sth2|sth3|sth4)* /␤ expecting quantmod␤FAILED 00:01 104m␤
Woodi std: / (sth1|sth2|sth3|sth4)* / 12:22
p6eval std 29023: ok 00:01 103m␤
Woodi std: "sth" ~~ / (sth1|sth2|sth|sth4)* / 12:23
p6eval std 29023: ok 00:02 105m␤
12:23 ihrd left, hcchien joined
pkkm oh I didn't konw we've got testing bot here :) 12:25
moritz_ rakudo: say 'foo sth3' ~~ / [ sth <[1234]> ]* /; 12:27
p6eval rakudo 97ced1: ␤
moritz_ rakudo: say 'foo sth3' ~~ / [ sth <[1234]> ]+ /;
p6eval rakudo 97ced1: sth3␤
Woodi btw. constant installing parrot and rakudo makes i write Makefile for that :) 12:32
moritz_ git pull && perl Configure --gen-parrot && make -j 3 install
no rocket science ;-)
Woodi maybe it will be usefull for someone... located at perl.lunski.pl/perl6/spi 12:33
moritz_: that is a problem
pkkm std: /(?:[/\*.*\*/|\s*|.{,5}]*)/ 12:34
Woodi what if you want it in /usr/local ?
p6eval std 29023: ===SORRY!===␤quantifier quantifies nothing at /tmp/mMMNGyv0Jm line 1:␤------> /(?:⏏[/\*.*\*/|\s*|.{,5}]*)/␤ expecting quantmod␤FAILED 00:02 103m␤
moritz_ pkkm: (?: is not Perl 6 regex syntax
pkkm moritz_: but it's working in pcres
std: /([/\*.*\*/|\s*|.{,5}]*)/
p6eval std 29023: ===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/BJ86v7WoW9 line 1:␤------> /([⏏/\*.*\*/|\s*|.{,5}]*)/␤ expecting any of:␤ regex atom␤ regex_infix␤ termish␤ ws␤FAILED 00:02 106m␤
moritz_ pkkm: maybe, but this is a Perl 6 channel 12:35
Woodi: then you need something more elaborate, that's true
Woodi pkkm: pcre was made for compability with superb-perl5 regexes
pkkm std: /[/\*.*\*/|\s*|.{,5}]*/
moritz_ but it's still not compatible :/
p6eval std 29023: ===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/GGL8dgX0Tx line 1:␤------> /[⏏/\*.*\*/|\s*|.{,5}]*/␤ expecting any of:␤ regex atom␤ regex_infix␤ termish␤ ws␤FAILED 00:01 103m␤
Woodi moritz_: and what if you must install parrot then rakudo, make mackages to place system-wide, then remove becouse new wersion was developed by someone ? :) 12:36
and even proto ask me to install it outside so i did :) 12:37
coffee & 12:38
12:39 araujo joined, hcchien left
pkkm std: /[/\*.*\*/|\s*|.{,5}]*/ 12:39
p6eval std 29023: ===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/Ooz6ZyDSlX line 1:␤------> /[⏏/\*.*\*/|\s*|.{,5}]*/␤ expecting any of:␤ regex atom␤ regex_infix␤ termish␤ ws␤FAILED 00:01 103m␤
pkkm std: / [/\*.*\*/|\s*|.{,5}]* /
p6eval std 29023: ===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/A6cj6Ct2YG line 1:␤------> / [⏏/\*.*\*/|\s*|.{,5}]* /␤ expecting any of:␤ regex atom␤ regex_infix␤ termish␤ ws␤FAILED 00:01 103m␤
12:39 hcchien joined
moritz_ it's complaining about / not being escaped 12:40
pkkm std: / [\/\*.*\*\/|\s*|.{,5}]* / 12:41
p6eval std 29023: ===SORRY!===␤Preceding context expects a term, but found infix , instead at /tmp/b2YOAUwAy1 line 1:␤------> / [\/\*.*\*\/|\s*|.{,⏏5}]* /␤FAILED 00:01 106m␤
12:42 fax left
moritz_ and in Perl 6 the generig quantifier is **0..5, not .{,5} 12:42
s/generig/generic/
pkkm std: / [\/\*.*\*\/|\s*|.**0..5}]* / 12:43
p6eval std 29023: ===SORRY!===␤Unable to parse regex; couldn't find right bracket at /tmp/w46Sj78XRm line 1:␤------> / [\/\*.*\*\/|\s*|.**0..5⏏}]* /␤ expecting any of:␤ regex atom␤ regex_infix␤ standard stopper␤ terminator␤ ws␤FAILED 00:01 104m␤
pkkm ...
std: / [\/\*.*\*\/|\s*|**0..5}]* /
p6eval std 29023: ===SORRY!===␤quantifier quantifies nothing at /tmp/sVglybw7E9 line 1:␤------> / [\/\*.*\*\/|\s*|**0..5⏏}]* /␤FAILED 00:01 103m␤
pkkm std: / [\/\*.*\*\/|\s*|.**0..5}]* / 12:44
p6eval std 29023: ===SORRY!===␤Unable to parse regex; couldn't find right bracket at /tmp/91c7SkqoYK line 1:␤------> / [\/\*.*\*\/|\s*|.**0..5⏏}]* /␤ expecting any of:␤ regex atom␤ regex_infix␤ standard stopper␤ terminator␤ ws␤FAILED 00:01 103m␤
pkkm std: / [\/\*.*\*\/|\s*|.**0..5]* /
p6eval std 29023: ok 00:01 103m␤
pkkm :)
moritz_ std: / [ '/*' .*? '*/' | \s+ | .**1..5 ]* / 12:45
p6eval std 29023: ok 00:01 105m␤
moritz_ pkkm: since the group is quantified with a * which allows zero matches, it doesn't make much sense to allow \s* inside, which also matches an empty string
pkkm Oh! 12:47
it will match any string!
moritz_ right
pkkm lol
moritz_ that's why it's important to always test for both matches and non-matches 12:48
too permissive matches are quite common 12:49
pkkm std: / (\s*(/\*.*\*/)?\s*) / 12:50
p6eval std 29023: ===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/6Fikrpb5CG line 1:␤------> / (\s*(⏏/\*.*\*/)?\s*) /␤ expecting any of:␤ regex atom␤ regex_infix␤ termish␤ ws␤FAILED 00:02 107m␤
pkkm std: / \s*(/\*.*\*\/)?\s* /
p6eval std 29023: ===SORRY!===␤Unrecognized regex metacharacter (must be quoted to match literally) at /tmp/jfIA2iI7XX line 1:␤------> / \s*(⏏/\*.*\*\/)?\s* /␤ expecting any of:␤ regex atom␤ regex_infix␤ termish␤ ws␤FAILED 00:01 104m␤
12:50 rafl left
pkkm std: / \s*(\/\*.*\*\/)?\s* / 12:51
moritz_ pkkm: anyway, if you want to test PCRE regexes, a Perl 6 regex checker is surely not your best option ;-)
p6eval std 29023: ok 00:01 103m␤
pkkm brb dinner :) 12:56
12:58 masak left
pkkm I have to go. 13:28
see You at 17 13:29
moritz_ whenever that might be outside your timezone ;-)
13:30 payload1 left
Woodi moritz_: maybe you know something about including Actor model in p6 ? 13:32
moritz_ Woodi: I have no idea what the Actor model ist
I just know that most models suck at acting ;-)
Woodi model of concurent programming... 13:33
13:33 pkkm left
Woodi allow no low-level races 13:33
k, thanx
moritz_ rakudo: sub f(*@_, *%_) { say @_.perl, %_.perl }; my $x = 'y'; f( $x => 5 ) 13:36
p6eval rakudo 97ced1: ["y" => 5]{}␤
13:42 envi^home left 13:50 lichtkind joined 13:51 SmokeMachine left 13:52 SmokeMachine joined
Juerd Wolfman2000: What kind of help would you need, for building that pastething? 14:00
phenny Juerd: 05:24Z <Wolfman2000> tell Juerd Planning a Perl 6 Pastebin: any assistance you can provide would be appreciated. Contact myself or eternaleye for more information.
Wolfman2000 *yawn* morning. Wasn't expecting phenny to wake me up 14:01
Juerd Moin 14:02
14:02 yjh joined
Wolfman2000 Juerd: We've figured out most of the basics. Most likely we'll lose Feather for hosting right now until we end up with a paste.perl6.org or something. We've got a database plan, login plan...the big issue the command line interface, but I think eternaleye has msot of that figured. 14:03
...oh yeah. We need some magical ability to turn perl code into colored syntax. a P6Regex matching thing
Juerd I can help you with hosting on feather but not much else :0 14:04
:)
Wolfman2000 Then the other questions...can we use mysql and cpan without issue? I tried to start cpan, but it's apparently meant for sudo users only. We each don't have our own. I wonder if mysql is similar. 14:05
Juerd See 'How do I keep my own module/library directory?' in perlfaq8 or ask me to install some modules
And ask me if you need mysql databases 14:06
Wolfman2000 Juerd: mysql I saw was available. I didn't know if I was allowed to run that on my own.
As far as modules to install...well, since this is going to be built on Perl 5 right now...Catalyst::Perl and related packages. If there is a P6Regex style package for Perl 5, that'd be great too. 14:07
Juerd You're allowed to run your own but it's easier to just get a database in the system-wide mysql
For Perl 6 projects, you're allowed to run *anything* as long as you play nice with resources, and don't do anything illegal. 14:08
Wolfman2000 We plan on moving to Perl 6 once it's more stable.
But even you have to admit that the current pastebin with its lack of Perl syntaxing is...kind of sad 14:09
Juerd It's suboptimal
Wolfman2000 ...ah. I can use cpan. I just have to set o confs
14:09 NorwayGeek joined
Wolfman2000 ...and adjust my own @INC path. Is that even possible? 14:11
moritz_ sure, with PERL5LIB 14:12
14:12 envi^home joined
Juerd use lib 'some path'; or set PERL5LIB 14:12
Or BEGIN { add to @INC here }
Wolfman2000 Juerd: I think an environment variable would work best here.
moritz_ or perl -Ipath script.pl
perldoc perlrun # more alternatives
Juerd Wolfman2000: Catalyst is being installed right now.
Done 14:14
Wolfman2000: What's the preferred name for a database and mysql login?
Woodi there is ncurses in parrot, can i use it in rakudo ?
Wolfman2000 I forgot if mysql databases can have numbers in them. p6paste for the database, use jafelds for login. password...guess I set that. 14:15
Key thing about the database: MUST BE UTF-8
I recall postgresql really prefers the encoding set when the database is made.
I don't recall if mysql is similar. 14:16
dalek ok: 5db9456 | moritz++ | src/grammars.pod:
[grammars] describe more of the action methods, and add some RAKUDO comments
Juerd Wolfman2000: Databases don't have encodings, tables do. You're creating the tables yourself :) 14:17
Wolfman2000 Juerd: Right, mysql difference. I admit I'm more of a fan of postgresql, but I can adapt.
14:18 payload joined 14:24 icwiener joined
Wolfman2000 Note to self: careful when you use your own CPAN. You get to see all of the packages installed. 14:24
14:29 _eMaX_1 joined, _eMaX_ left 14:39 colomon joined
colomon rakudo: say 1.0.Str.Str; 14:39
p6eval rakudo 97ced1: 1␤
Wolfman2000 ...what's 1.0 supposed to be? 14:40
rakudo: say 1.0;
p6eval rakudo 97ced1: 1␤
Wolfman2000 ...right, a number
moritz_ rakudo: say 1.0 * 2.0
p6eval rakudo 97ced1: 2␤
colomon 1.0 is supposed to be a Num.
Wolfman2000 So a num to a String prints the integer string
pmichaud actually, a Rat
Wolfman2000 ...but wait. It's actually printing...the number.
And not a hex code
14:41 masak joined
colomon rakudo: say 1.0.Str.Str.WHAT 14:41
p6eval rakudo 97ced1: Str()␤
colomon It's printing the string "1"
Wolfman2000 colomon++: you fixed it! 14:42
colomon Wolfman2000: It was very, very easy. :)
Wolfman2000 Well, how soon will it show up in the master branch? 14:43
masak Wolfman2000: I'm finally putting together the clues: you're building a pastebin in Perl 6.
Wolfman2000: incidentally, that's something Tene++ and I have been toying with a bit.
Wolfman2000 masak: That's...not exactly how it ended up.
colomon Wolfman2000: I was just checking to see if the fix was incorporated in p6eval yet.
masak Wolfman2000: no? how did it end up?
moritz_ colomon: what fix? 14:44
not saying OUTPUT every time?
Wolfman2000 masak: I wanted to transform one of my own websites to Perl 6. However, it was agreed by the people in here that we should have our own pastebin. We'll build it in Perl 5 first, and then when Web.pm is more stable, we transfer it into Perl 6.
colomon moritz_: Str.Str exists and works now.
moritz_ oh, I thought it was a fix to p6eval itself ;-)
Woodi masak: can you tell me how to use proto with parrot/rakudo installed system-wide ?
masak Wolfman2000: that sounds like an excellent plan.
Woodi: I'm not 100% sure proto handles that. 14:45
colomon moritz_: There's still one really funky test existing test to sort out.
masak Woodi: proto needs a lot of love right now. but most of it is spent on the new branch, installed-modules.
Woodi ok, thanx
colomon moritz_: paste.lisp.org/display/89966
Wolfman2000 masak: glad you approve. I already have a database set up with one table on feather until we're ready for the big move off 14:46
Woodi masak: i am probably to stupid to uncipher how it work...
colomon moritz_: The test violates the LSP, but should probably still work, I reckon.
masak Wolfman2000: we're currently in the paradoxical situation wherein Perl 6 syntax highlighting is done most easily in Perl 5.
colomon masak: Will that change when ng lands? 14:47
Wolfman2000 masak: So I've heard. Unfortunately, I'm not...too familiar with Perl 6 syntax highlighting. I'm hoping there is a CPAN library for that.
masak Woodi: no, that's not a great way to approach a problem. :) you should use curiosity, not feelings of unworthiness. :)
colomon: perhaps.
Wolfman2000: well, you should definitely speak to azawawi about it. 14:48
Woodi yea, positive thinking is better than self-destruction :)
colomon wasn't there a perl 6 syntax highlighting module on CPAN? (In perl 5, of course.)
Wolfman2000 colomon: that's what I asked. If there as 14:49
was*
masak colomon: there certainly is one for Padre.
moritz_ colomon: uhm, I don't think that test is valid
Wolfman2000 masak: Unfortunately, I borked my local Perl enough so that I can't run Padre.
moritz_ colomon: since it's already a Str, there's no need why stringification should have to call .Str on it
Wolfman2000 And I'm not in a position to wipe everything off the computer and start again.
colomon masak: There was a blog post six months ago describing ripping the padre highlighting out into a module.
masak Wolfman2000: sorry, can't help you there. :)
colomon: URL? 14:50
colomon moritz_: ???? That was the entire point of writing Str.Str -- you should be able to call .Str on a Str, for sure.
masak: no idea, that was an ironman post from the first half of the year, I think. 14:51
moritz_ colomon: sorry, I misread the test
szabgab try looking for the blog post here padre.perlide.org/about.html 14:52
moritz_ I thought you were using ~$thing instead of $thing.Str
masak here's the module, anyway: cpan.uwinnipeg.ca/dist/Syntax-Highlight-Perl6
colomon masak++ 14:53
Wolfman2000 I won't give the ++ yet until I see how the HTML output format.
masak well, azawawi++, surely.
Wolfman2000 As a web developer, I do care about the output.
masak ah! an idealist! :)
colomon Wolfman2000: I doubt it has HTML output yet.... 14:54
Wolfman2000 Syntax-Highlight-Perl6? It does.
moritz_ it has
colomon it does! crazy. 14:55
I thought it was supposed to plug into VIM or something like that.
Wolfman2000 colomon: I think that may be an option too.
The point is...we've got possibilities.
Granted, modifications will need to be done, but...we can probably yank out the needed text.
moritz_ colomon: that's what Text::VimColor does 14:57
Wolfman2000 Basically, we can use a modification of Syntax::Highlight::Perl6->simple_html. If we can overload it and then just rip out the code itself, even better.
moritz_ colomon: which is what I use for my websites
colomon ah. 14:58
Wolfman2000 enjoys being able to use CPAN without worrying about permissions issues. 14:59
Alright, I have Syntax-Highlight-Perl6 installed via CPAN.
I know Catalyst is also installed, so I can soon use the custom apache I have.
15:00 Whiteknight left
Juerd automysqlbackup++ debian++ 15:00
# apt-get install automysqlbackup # no config, works well out of the box.
15:01 Whiteknight joined
Juerd It simply makes daily db dumps and puts them in /var/lib/automysqlbackup 15:01
Keeping weeklies and monthlies too.
Wolfman2000 Juerd++ masak++ good. 15:03
masak is currently blogging about the midnight "Perl 6 isn't Perl" discussion 15:04
Wolfman2000 which one is this?
colomon moritz_: On the paste.lisp.org/display/89966 failure, I've just added debugging says, and it is the RT69378str.Str that is being called. 15:05
masak Wolfman2000: irclog.perlgeek.de/perl6/2009-11-07#i_1699890
Wolfman2000 ...and of course the log ends with me being silly.
masak I find it comforting to see that emma still lurks on this channel, but is no longer in #perl. :) 15:06
that makes me hypothesise that we came off as the saner crowd. 15:07
Wolfman2000 Umm...guys? How do we install STD for Perl 5? I can't just install STD on CPAN
moritz_ 4
masak Wolfman2000: most of us just build it ourselves.
Wolfman2000: it's in the Pugs repo.
Wolfman2000 masak: Well, while you're here...what is $PAGER supposed to be? 15:08
masak moritz_: 4? is that your final answer?
Wolfman2000 I didn't understand that environment variable
masak Wolfman2000: something like 'less'.
Wolfman2000: i.e. a program that serves you text page-wise.
Wolfman2000 ...are we supposed to build via traditional means or Cabal? 15:09
masak Wolfman2000: you don't need to isntall Pugs. 15:10
it's just in the same repo.
15:10 Whiteknight left
Wolfman2000 masak: may as well 15:10
...or I would, if it worked.
lisppaste3 wolfman2000 pasted "Umm..." at paste.lisp.org/display/89985 15:11
masak Wolfman2000: Pugs has nothing to do with STD.pm, besides being hosted in the same repo.
Wolfman2000: I'm not one to decide where you spend your time, but if it's STD.pm you're after... 15:12
Wolfman2000 masak: I will require STD.pm for Syntax::Highlight::Perl6
15:12 lichtkind left
masak Wolfman2000: yes. 15:12
Wolfman2000: but for that, you don't require Pugs. 15:13
just the repo. I know that's a bit confusing. sorry :/
Wolfman2000 Slightly.
masak it's an accident of gistory.
s/g/h/
Wolfman2000 masak: pugs/src/perl6/STD.pm? 15:14
masak that's the one.
there's a README with build instructions, if memory serves me.
Wolfman2000 I'll expand PERL5LIB then
...oh, can't use just yet?
/bin/sh: ./gimme5: /usr/local/bin/perl: bad interpreter: No such file or directory <-- :( 15:16
masak Wolfman2000: where do you keep Perl 5.10? 15:17
Wolfman2000 masak: I don't believe I custom built Perl 5.10 on Feather
Juerd feather has 5.10.1 as the main perl
Wolfman2000 ...I'm using the default perl, located on /usr/bin/perl
ah: gimme5 has the shebang set to /usr/local/bin/perl 15:18
I wonder if that's intentional.
Wolfman2000 gets to making now 15:19
Juerd Hm, let's re-symlink that
15:19 s1n joined
Wolfman2000 Tell me when please: there are probably too many places I'd ahve to change the path. 15:19
Juerd That used to be a separate from-source install of Perl 5.10, when Debian still clinged to 5.8
Wolfman2000: Done 15:20
Wolfman2000 Yay for finding bugs...I think.
If nothing else, this whole experience will be a resume builder. 15:21
Not every day one gets to say they helped with the next great language.
masak: make finished. there is no rule for make install. Is it safe to use STD.pm now, and will I still have to adjust PERL5LIB to point to the pugs directory? 15:22
masak Wolfman2000: it's safe to use STD.pm now. 15:23
15:23 Psyche^ joined
masak Wolfman2000: I suggest you try it out in that directory until you feel comfortable with what it does. but after that, setting PERL5LIB should be enough. 15:23
Wolfman2000 Juerd++, masak++: works without a problem. 15:24
15:26 Whiteknight joined
masak does #perl have logs? 15:26
Wolfman2000 ...I just realized. simple_html automatically generates <link href="p6_style.css" rel="stylesheet" type="text/css">. I don't know if such a file already exists. 15:28
...and it does. azawawi++, thank you for already making said file. 15:29
15:30 Patterner left, Psyche^ is now known as Patterner
pmichaud phenny: tell jnthn I think that creation of Code/Block/Routine/Sub/Method wrappers needs to happen at compile (BEGIN) time instead of at runtime. 15:30
phenny pmichaud: I'll pass that on when jnthn is around.
Wolfman2000 Juerd++: Thanks for perhaps the one advantage of having everyone able to view each other's directories. 15:31
diakopter masak: for brains, yeah 15:32
Juerd Wolfman2000: Actually not my idea but I forgot who convinced me.
Wolfman2000 brb for quick breakfast...then, the Catalyst begins.
masak diakopter: I will attempt to frequent #perl more in the future. I want to be part of Perl 6 discussions when they happen there. 15:34
diakopter heh 15:35
colomon what you need is a bot that notifies you when perl 6 discussions are started there.... ;) 15:36
diakopter throw ilbot at it... 15:37
pugs_svn r29024 | colomon++ | [t/spec] Add a new test which highlights why one of the existing tests fails.
diakopter masak: you're a firefox addict, right? use ReloadEvery and GreaseMonkey to get notified...
Juerd colomon: Why not just a highlight? :) 15:38
I've had a highlight on "unicode" and "utf-?8" for a long time.
Just to jump into discussions on several #perl's 15:39
masak diakopter: hm, sounds interesting. will investigate.
Juerd: ah, so that's how to clandestinely summon you... :P 15:40
diakopter masak: I mean, once moritz_ throws ilbot at it
Juerd masak: I have a highlight on feather too.
I've removed the one on UTF-8 by the way
15:40 diakopter sets mode: +oooo masak Juerd colomon Wolfman2000
Juerd It was triggered too often by Google URLs. 15:40
huf most of the perl6 discussion on #perl is along the lines of "you abandoned us you bastards!" ;)
masak huf: really? 15:41
huf that was my impression
some people think it's bloody stupid but they can at least mine it for ideas in "the real perl"
lots of them cant deal with the fact that it's a long time coming
which is a strange reason to hate perl6, but meh 15:42
Juerd People who love Perl 6 usually keep quiet, is my experience.
masak huf: I know mst's stance on it, and I respect that.
diakopter Wolfman2000: for highlighting of Perl 6 code, STD has a script for it, but it may be a bit bitrotted
colomon Juerd: because I had no idea you could set a highlight like that. sweet!
pmichaud I think for many people it's like a jilted lover, or a promise unkept
masak huf: he doesn't think it'll make any large impact ever, but that Perl 5 will soak up the good parts and move on.
huf yeah
Juerd colomon: If you use irssi, I recommend -word -color %Y -actcolor %Y 15:43
Or some other color, but include -word so the actual match is highlighted rather than the name of the person who said it.
masak if someone wants to read and comment on my blog post about "Perl 6 is not Perl", here it is: gist.github.com/raw/229327/7d58b2cd...st%20draft
Juerd Makes finding the highlights much easier
pmichaud many people expected Perl 6 to be available in just a year or two, and that was a reasonable expectation given the things being said about it at the time
diakopter .. prophecy unfulfilled
Juerd juerd.nl/i/b454ff6aaf3aeb462e68f32c75ee1485.png # example of a highlight on "feather" 15:44
s1n i've heard a handful of comments to the effect that people think perl6 is fun thought experiment of features that might be useful to Perl (not perl5 or perl6, because perl6 doesn't exist)
pmichaud when it didn't appear, and then took longer and longer... well, people get tired of having their hopes dashed. :)
s1n: yes, many people who started to be opposed to perl6 begain calling it a "research project" as opposed to a real project with real deliverables
huf that's why you dont put much hope in something that's not here yet. you just wait and see.
pmichaud *began
s1n that is not the open source mentality 15:45
if you want it, help make it so
Juerd masak: gist.github.com/raw/229327/7d58b2cd...st%20draft is formatted in a terrible way :)
masak Juerd: yeah, sorry about that. :/
huf true, i'm just saying that because i dont understand how perl6 (wether it ever gets done or not) hurts perl5
and yet that's the impression i get from a lot of people
pmichaud huf: some people think that the focus on perl 6 stole resources from perl 5 15:46
huf that somehow the existence of perl6 in its current state is harmful
s1n perl6 doesn't, impatient devs who don't want to help with perl6 hurts
huf pmichaud: so that's why perl5 development seems to have speeded up in the recent past?
pmichaud huf: i.e., that people who were working on perl6 would be working on perl5 instead
huf: I'm not saying those people are right... I'm just saying why they think that
huf ah ;)
well yeah, that's the bit i *really* dont get. like it? help or at least shutup. dont like it? it doesnt hurt you so shutup. :D 15:47
pmichaud huf: many things have been (are being) said about perl6 that are at quite a distance from reality
Juerd masak: At the amsterdam.pm meetings most people say that Perl 6 isn't Perl, but they don't mind presentations about Perl 6, so they can't be THAT convinced.
Wolfman2000 ...I'm soon going to have to do port forwarding for this one
jaffa8 is there profiling for perl 6?
s1n huf: keep in mind the people who think perl6 is stealing resources have flatly refused to join the community
masak hm. maybe I should add a paragraph about why it's the right time to be excited about Perl 6.
Juerd masak: My opinion is still that using the name Perl with a version number is a bad idea for something that isn't just an incremental update. It is confusing. 15:48
masak jaffa8: there are things in the works for Parrot, yes.
jaffa8 "in the works"
Wolfman2000 Juerd: Not all of the needed catalyst stuff was installed. I'm taking care of that on my own CPAN now.
masak Juerd: yes, but that's what I'd call a "balanced view".
jaffa8 masak, what is that?
15:48 NorwayGeek left
masak jaffa8: "in progress". I don't know much more about it. people on #parrot might. 15:48
15:49 NorwayGeek joined
Juerd masak: Perl 6 really isn't what I had always thought Perl to be. 15:49
emma In the book Im reading to learn perl it mentions a project called Topaz, is this really Topaz?
masak Juerd: oh?
s1n in the end, it won't matter, perl5 is not a way forward, the people who are currently hostile towards perl6 will have a difficult choice in the future, like it or not
jaffa8 Juerd, that is like life.
huf it's rather perlish still, so i kinda get why it's called perl6 ;)
diakopter emma: I think Topaz was Perl compiler/vm in C++
s1n diakopter: it was just an attempt to rewrite the guts in c++ 15:50
masak emma: Topaz got stalled around 2000-2001.
s1n it was still ugly c++
Wolfman2000 Juerd: Even with my changes to the CPAN libs, two packages didn't install. I need File::ChangeNotify and MooseX::SemiAffordanceAccessor installed please.
Juerd masak: To me, "Perl" is Perl 1..5 and "Perl 6" is Perl 6.
huf isnt that the intention tho?
markmont masak: add link to perl 6 faq to remind people, and maybe the anti-FUD page at www.perlfoundation.org/perl6/index.cgi?fud ?
Juerd masak: But I'd prefer Perl 6 to have a different name instead.
emma what if the people who make perl want to make a new edition of their perl? What will they have to call it now? 15:51
masak Juerd: yes, but explain what you mean by 'what I had always thought Perl to be'.
markmont: good idea.
Juerd markmont: There are also my old talks "Perl 6 Myths" and the "Perl 6 fears" project.
s1n Juerd: why? Perl is what larry defines it to be, period
colomon jaffa8: there is a parrot profiler that works now, search the #perl6 backlogs (gotta run, alas)
pmichaud (parrot profiler) for some values of "works"
Juerd s1n: Yes, that is reality. However, thoughts don't always match reality :)
s1n then realign your thoughts to reality
emma They might jump ahead of you guys and call theirs Perl 7, so that now you are stuck with Perl 8. I think maybe you guys should be Perl (2n) and they could be Perl (2n-1) where n is an integer. 15:52
s1n i think it should be Perl and Perl5
i.e. future Perl, dead Perl
huf emma: they are free to call the next version Perl 6.0 ;) wont affect Perl6 1.0 in any way :)
Juerd Oh 15:53
It appears the Perl6::Doc::FUD document is based on the fears project.
huf btw have you decided to call it Perl 6.0 or Perl6 1.0 yet? :)
pmichaud
.oO( I wonder when Linux kernel version 3 will come out. )
Juerd Perl6::FAQ::FUD,sorry.
s1n huf: it will be Perl :)
emma If they wanted to sort of give you a taste of your own medicine they might call theirs Perl6 1.0 2
huf ;)
s1n emma: that sort of thing is childish and a huge waste of effort
huf emma: you still seem to be under the impression that there's a problem somewhere
pmichaud I think I would say it this way 15:54
masak emma: one important aspect of this, I think, is that Larry Wall, the author of Perl 1..5, is very much involved in Perl 6 development.
pmichaud there's a portion of the perl community that believes that Perl 6 represents an abandonment and "fork" of Perl
in reality, Perl 6 is the mainline development.
Juerd I wrote parts of that document but wasn't included in the README :/ 15:55
masak emma: I have the feeling that those who actually understand what it is we do also have some amount of understanding for it.
emma huf: i've seen people in both channels feeling like there's a problem. To me it is worrisome. I think schisms are bad news. I mean bad news for everyone who likes perl of any kind. You should all have a lot of forboding. It's not good when a community starts talking about itself more than anything else. It is a prelude to the end, if people don't restore harmony.
s1n pmichaud: yeah, i explained that to mst before and he laughed
pmichaud it may take a very long time for us to see that Perl 6 is the mainline, but ultimately that's the case.
s1n i told him there is Perl and Perl5 and he sorta refused to accept that
pmichaud s1n: mst is actually quite reasonable here, from his perspective
masak emma: that's why I'm writing that post. did you read it?
markmont I wonder if people felt the same way about FORTRAN {90,85,2003} being an abandonment or fork of FOTRAN 77? There were similar sized changes to what Perl 6 is making, starting with free-form source input (no more character columns!)
pmichaud if someone believes that Perl 6 is an evolutionary dead end, then they'd be perfectly correct that Perl 6 is a branch and that Perl 5 represents the mainline 15:56
huf i dont see why all members of a community should believe the same thing
jaffa8 perl 6 is a step ahead concerning functionality
huf and this is especially true for the Perl community :)
masak markmont: I'm not sure FORTRAN ever had the same flavor of community as does Perl.
pmichaud huf: precisely, we don't all need to believe the same thing
huf ;)
s1n in the end, none of it matters, Perl v6 is great, if you prefer v5 don't upgrade
emma If you are all not careful about what you are doing, the worst of all possible outcomes will happen -- People will no longer consider perl 5 relevant, and people will not feel compelled to bother with perl 6 since there are many other options and nothing is written in perl 6. All of this could mark the end of perl history. 15:57
s1n just like python3000, it's great, needs work, and is the future
huf 5 will be around for at least a decade so... it's not yet time to decide anything
emma: ask around. most people already consider perl to be a number of stupid fud arguments. one more wont change a thing
markmont masak: very true
15:57 am0c left
emma The reason that bothers me is that I'm reading a book to learn perl so I don't want to finish it just at the end of perl history. 15:57
s1n emma: alas, all of cpan will work under perl v6
Wolfman2000 ...and I'm an idiot. I need a way to kill a socket I made 15:58
pmichaud emma: I wholeheartedly disagree. People who say that Perl 6 has foreclosed development of Perl 5 just need to broaden their thinking a slight bit
masak emma: Perl 5 will be relevant for years to come.
pmichaud If the perl 5 developers wanted to start a new version of perl, they could certainly come up with all sorts of names besides "Perl 6" or "Perl 7"
masak emma: but Perl 6 will increasingly gain mindshare too. 15:59
pmichaud there are many ways to denote projects besides integers.
Wolfman2000 Does anyone know if there is a way to kill processes based on the name of the process ran? I need to kill script/p6paste_server.pl
huf killall?
s1n Wolfman2000: in unix, killall
emma Yes perhaps it will. But it's a very bold thing you are trying to do. And a pyrrhic victory for perl 6 will still be the de facto end of Perl.
huf or pkill or what was that one that can do regex matching
Wolfman2000 no process found. :(
pmichaud emma: we disagree on that point, but that's fine. :)
Wolfman2000 Couldn't create daemon: IO::Socket::INET: Address already in use at /usr/share/perl5/Catalyst/Engine/HTTP.pm line 211. <-- I forgot to close this before I temporarily disconnected from feather. 16:00
emma Something must have gone really wrong. This isn't a normal way to advance a language.
s1n emma: it will _not_ end Perl, it will continue Perl
pmichaud emma: we're willing to let history tell the tail. At the time that Perl 6 was being designed, it was widely agreed that Perl 5 was going to die anyway.
emma I'm saying a pyrrhic victory I might not be spelling that correctly though.
pmichaud or at least that Perl 5 was on the decline
there's a very good argument to be made that Perl 6 actually freed Perl 5 to continue developing
huf one good thing that came out of perl6 is a lot of activity to save perl5
s1n keep in mind, perl5 was mostly on a downward slide to death until perl6 and all that moose stuff was added to revive perl5 16:01
Woodi Wolfman2000: port you use is still in use by os. try use other port for now
markmont emma: I'd rather focus on what I can do with a tool than what it is called. "victory" is getting useful stuff done, and Perl 6 is shaping up to be *very* useful for things that are difficult in Perl 5. 16:02
Wolfman2000 Woodi: I'm trying to free the port
Woodi Wolfman2000: and there is option to socket crating which prevents this behaviour...
Wolfman2000 okay, I got it.
Juerd++: thanks for htop
pmichaud more to the point, even though many in the Perl community envision some sort of ongoing battle between perl 5 and perl 6, the people focused on perl 6 don't view it that way.
I'm going to disappear for a while to think about modules. :) 16:03
afk
Wolfman2000 hooray port forwarding!
16:08 rgrau`` is now known as kidd`
Juerd emailed lichtkind about acknowledging contributors to the FUD page (originally: pugs/docs/quickref/fears) 16:08
zaslon lolmasakhazblogged! masak++ 'The "Perl 6 is not Perl" meme': use.perl.org/~masak/journal/39861?from=rss 16:12
Wolfman2000 ...isn't zaslon another bot? 16:13
16:14 diakopter sets mode: +v zaslon
Wolfman2000 diakopter: Thanks for confirming. 16:15
diakopter Wolfman2000: did you see what I said about std syntax highlighting Perl 6
Wolfman2000 outdated
16:15 [particle] joined
diakopter outdated? 16:15
Wolfman2000 At least, if that's what bitrotting means
diakopter I said it _may_ be a bit bitrotted... but only the presentation mechanics 16:16
the Perl 6 syntax is obviously definitive/latest.
Wolfman2000 That, I can't argue with.
16:17 astrojp joined
masak ah. it was the freenode #perl that insisted that Perl 6 isn't Perl, not the perl.org #perl. 16:17
Wolfman2000 masak: ...what? 16:18
diakopter s1n: a publicity/marketing death, perhaps
masak for some reason I assumed the latter.
Wolfman2000 Freenode's #perl hates Perl 6?
...why am I in there right now?
masak Wolfman2000: there are several #perl channels.
Wolfman2000: I don't know what they said. maybe I never will, because there are no public logs. 16:19
16:19 nihiliad joined 16:22 am0c joined
masak is now under heavy fire in #perl on irc.perl.org :) 16:25
diakopter goes to watch the fireworks 16:27
16:27 IllvilJa left 16:29 _eMaX_ joined
Woodi rakudo: say %*VM<config><build_dir>; 16:30
p6eval rakudo 97ced1: /home/p6eval/rakudo/parrot␤
16:31 _eMaX_1 left
diakopter what a joke... an intelligent discussion cannot be had at that message arrival rate 16:32
masak diakopter: it's tiring... :/ 16:33
diakopter if only people would respect each other's comments enough to slow down about 5x 16:34
masak thank you s1n and Juerd for contributing.
diakopter otherwise it just becomes a war of who can type the fastest 16:35
masak exhales
it slowed down somewhat now. :)
s1n np, this perl5 forever crap is really irritating
diakopter I suspect that's tongue-in-cheek
Woodi do http-daemon works only with rakudo configured with --gen-parrot
Wolfman2000 rakudo: say time(); 16:36
p6eval rakudo 97ced1: 1257697683.48797␤
Wolfman2000 ...how exactly do we format time to a traditional YYYY-MM-DD format anyway?
BinGOs if it was perl5 I'd suggest POSIX::strftime 16:37
Wolfman2000 ...unfortunately, I'm making a Perl 6 pastebin in Perl 5. :/ 16:38
masak http-daemon works regardless of the way you configure rakudo/parrot.
jnthn Woodi: --gen-parrot just makes sure you're building a revision of Parrot that Rakudo is known to work with, and then builds Rakudo against that.
phenny jnthn: 05:23Z <quietfanatic> tell jnthn Unfortunately it looks like overloading trait_mod:<is> doesn't make a versatile subclassing hook, because the left argument given to it is an unnamed and uninitialized ClassHOW object.
jnthn: 15:30Z <pmichaud> tell jnthn I think that creation of Code/Block/Routine/Sub/Method wrappers needs to happen at compile (BEGIN) time instead of at runtime.
BinGOs Oh great another pastebin
Wolfman2000 BinGOs: the one we use right now is suboptimal.
BinGOs and that is / 16:39
?
Wolfman2000 ...we can do better.
Woodi jnthn: i have p/r build and installed atm.
BinGOs I meant which one do you use now ?
jnthn pmichaud: Most probably, yes.
Wolfman2000 BinGOs: the topic is what's used now. It does't support Perl 6 or a concept of registered users.
We're going to fix that. 16:40
jnthn pmichaud: Now you've fixed BEGIN maybe that'll be easier.
pmichaud: If you want to just go ahead and put it in how you think it should be, I've no objections whatsoever.
BinGOs o.O CAPTCHA 16:41
Wolfman2000 BinGOs: that can be looked into.
BinGOs Nice idea, I must add it to Bot-Pastebot 16:42
jnthn quietfanatic: Things may work out better in the ng branch, when trait_mod:<is> really will just be a call on the metaclass. I'm not sure why it wouldn't work now though...
quietfanatic: I meant, will jsut do a call on...
16:43 carlin__ joined, envi^home left
jnthn quietfanatic: Thing is, since we do things in declarative order, in class Foo is Bar { has $!x; method wob { } } you'll be having the trait_mod called before the attr and method are known about by the metaclass. 16:43
Anyway, we may need a better hooker for these kindsa things. 16:44
16:45 carlin__ left
jnthn masak++ # blogging and trying to improve 5-6 community relations 16:45
masak jnthn: I'm quickly losing hope, though. 16:46
jnthn: the people over at #perl (irc.perl.org) are, for lack of a better word, thick.
huf they're being deliberately thick
because you've somehow hurt their feelings ;)
Wolfman2000 masak: Reminder needed. Larry...Wall I believe, was the creater of Perl. He's also going for Perl 6. Am I right so far? 16:47
huf resolving this is about as easy as resolving the "perl is linenoise lol" misconceptions ;)
masak Wolfman2000: yes.
16:47 meppel left
Wolfman2000 ...then why does it sound like everyone is going against the owner? 16:48
s1n i can't seem to get across to aperion
masak Wolfman2000: they don't believe in the effort.
s1n he has seriously flawed logic that he refuses to accept as flawed
Woodi Wolfman2000: language syntax CHANGED, what was working will not work now (without tricks) 16:49
masak s1n: outside of #perl6, people don't always run on 'logic' :)
s1n aparently not
he seems incapable of reasoning
Wolfman2000 ...and mysql knows how to be annoying at times. 16:50
| Warning | 1628 | Comment for table 'groups' is too long (max = 60) | <-- how come that's never talked about? 16:51
jnthn No...surely...you can have a comment for a table longer than 60 chars?!
Ouch!
16:51 carlin__ joined
Woodi s1n: sometimes pure-practical point of view is very strong point 16:52
s1n Woodi: few people in there did a "use logic" this morning 16:53
pmichaud back again
Woodi s1n: try to talk with management ppls about making investment in new hardware :)
jnthn yayitspm!
pmichaud jnthn: I haven't quite figured out all of the pieces to Code yet
jnthn pmichaud: Me either. :-) 16:54
s1n Woodi: i do all the time, just upgraded several hundreds of thousands of development machines :)
pmichaud actually, the biggest thing I think we can use at this point is traits on subs (particularly "is export")
jnthn pmichaud: The blocker for trait mods now is afaik solely "our" routines.
pmichaud hmmm 16:55
jnthn Well, our multis.
pmichaud not sure why that's a blocker
markmont If anyone wants to give feedback before I submit (email) it, I have a patch up that makes $! contextual as jnthn++ suggested. It also fixes the order in which find_contextual searches packages, per S02/Names. github.com/markmont/rakudo/commit/8...edd61c5cc8
Woodi s1n: ask now for more :) in some environment you will not get new hw until loadaverage will be visible for external users
pure practical objections...
jnthn pmichaud: How so? 16:56
16:56 carlin__ left
pmichaud markmont: that kinda doesn't look right to me 16:56
$! is a lexica
jnthn pmichaud: By blocker I don't mean "insurmountable task", just "thing that hasn't been done yet that needs doing" :-)
pmichaud $1 is a lexical
er
$! is a lexical
jnthn Third time's a charm.
Wolfman2000 okay, I have a database dump file on my feather home if anyone wants it. 16:57
Juerd I'm sorry to inform you that Perl 6 is dead.
jnthn pmichaud: $! is a lexical but also "is context".
Juerd Stop working on it!
jnthn Juerd: omg nooooo!
Juerd Really, it's dead. #perl on magnet says so.
pmichaud jnthn: contextuals are just lexicals with a * twigil
there's no special "store_contextual" required, afaik
Juerd You, my friends, are working on a dead language.
jnthn pmichaud: Well, now, yes
16:57 justatheory joined
Juerd :) 16:57
markmont pmichaud: S02/Names_and_variables: "$_, $! and $/ are always contextual, as are any variables whose declared names contain a * twigil." 16:58
Wolfman2000 Juerd: Your joking is...not exactly funny here.
jnthn pmichaud: Back before that happened I'm quite sure I remember a discusison along the lines of "$! is contextual"
Juerd Wolfman2000: I'm not joking. I'm just relaying the message.
pmichaud markmont: yes, although TimToady and I had a slight clarification on here about what "contextual" meant in that phrase.
let me see if I can find it
Wolfman2000 Who is the original person that layed the message to you?
masak we're the ones with hugme. #perl on magnet are the ones who'd really need it. 16:59
jnthn Juerd: Oh no! What to do?
pmichaud Juerd: message received, thanks.
masak really, really need it.
markmont pmichaud: What started this was jnthn said that setglobal '$!' in die() looked wrong. Especially as $! is lexical everywhere else.
Wolfman2000 masak, Juerd: you guys both realize I don't think Perl 6 is dead, right?
pmichaud markmont: sure. what should happen is that we look up $! as a contextual and then assign or rebind it 17:00
masak Wolfman2000: what, really? :P
Juerd Wolfman2000: I don't think many people in this channel think it is dead.
Wolfman2000 ...then who thinks Perl 6 is dead?
jnthn Wolfman2000: I think there's an element of satire. :-)
Wolfman2000 is not known for satire.
jnthn Wolfman2000: On this channel, that is. :-)
Juerd In fact, I believe that it's more alive than many other projects, including Perl 5.
jnthn On #perl it sounds like it's dillusion instead...
masak nod.
jnthn :-)
pmichaud TimToady moritz_: yes, $_ is contextual, but every block gets its own, so you never get the caller's $_ unless you say CALLER::<$_> 17:01
the same is largely true for $!
jnthn pmichaud: I thought the point was that things could set $! by looking it up contextually?
pmichaud jnthn: sure, but that's no different from any other contextual var
look up the variable, assign to it 17:02
moritz_ 's head jerks up
jnthn pmichaud: Right, I'm not disagreeing with that. :-)
Wolfman2000 ...and so with these...ops that I somehow gained...I guess I'm now in a holy war of Perl 6 vs Perl 5. I don't know how I feel about that.
jnthn pmichaud: Maybe I shoulda read the patch. :-)
Wolfman2000: Erm. I haz op and I'm surely not in such a thing.
moritz_ Wolfman2000: duck your head low and hope to survive the storm ;-)
jnthn pmichaud: Yeah, it should all be lookup and assign, I agree. 17:03
s1n i'm poking the bear in there
pointing out how C was a bad fork of asm heh
pmichaud jnthn: anyway, the three things that would be of most help to Test.pm at this point: augment, is export, compile-time blessing into Code 17:04
Wolfman2000 Is anyone else here familiar with Catalyst by any chance? The only catalyst website I used never had the Model.
jnthn pmichaud: I kinda was trying to avoid "blessing into" as such.
pmichaud: Just creating an instance of Code that contained the Parrot sub.
pmichaud jnthn: I think it's not avoidable. Otherwise when we do various lookups we get the wrong type of object 17:05
jnthn pmichaud: Which lookups?
pmichaud package-scoped lookups, for one
method lookups would be another
jnthn pmichaud: No, no
For method lookups we're putting them all in through the metaclass
pmichaud: So we just pass in the wrapped up code object. Done.
pmichaud right, but that's compile-time. 17:06
jnthn pmichaud: For package scoped we can wrap it and the install it in the namespace, at BEGIN time if you like.
pmichaud (where "compile-time" means "before INIT")
markmont For $! can someone clarify what is meant by "lookup and assign"? Find the callers's lexpad or set a global if there is none? Or something else?
pmichaud markmont: the code will end up looking like
jnthn markmont: Just use the fetch, and then call '&infix:<=>'(thingy_you_looked_up, thing_to_assign_to_it) 17:07
pmichaud $P0 = '!find_contextual'('$!')
'&infix:<=>'($P0, ... )
what jnthn++ said
more to the point, it will fall out naturally
17:07 NorwayGeek left
pmichaud $! = ...value... will naturally dtrt 17:07
17:08 NorwayGeek joined
markmont pmichaud, jnthn: Many thanks! 17:08
pmichaud we just have to change the compiler so that it treats $! as a contextual lookup instead of a lexical lookup
jnthn: (package scoped sub) okay, wfmfn 17:10
markmont pmichaud, jnthn: I'll re-do things and put up another proposed patch for feedback later today. 17:11
pmichaud markmont: thanks. keep in mind that a lot of this changes in the ng branch, though
although !find_contextual is likely to continue to exist
jnthn pmichaud: Once that is done, sub trait code gen should take me very little time. 17:12
markmont pmichaud: Yep, but it looks like a choice of not continuing to learn until ng lands, using ng now (looks beyond my skill), or doing stuff now and re-doing it after ng lands.
pmichaud markmont: redoing is okay, just wanted to make sure you were ready for it :) 17:13
markmont: didn't want you to suffer disappointement in a week or two (or possibly less)
markmont: if you're doing this to get more familiar with the code and structure, you're on absolutely the right track :)
jnthn pmichaud: If you want to tackle "our" scoped subs now, feel free. I'm about to make dinner and really need to spend a bit of time on other bits today.
pmichaud (and you'll "hit the ground running" when ng lands) 17:14
17:14 lichtkind joined
markmont pmichaud: I appreciate that. Yes, that's why I'm doing this, plus to help out where possible. 17:14
pmichaud jnthn: what other bits?
jnthn pmichaud: As in, things besides Rakudo.
pmichaud ah. okay.
jnthn pmichaud: Like, emails I shoulda replied to a week ago and the like. :-)
pmichaud np
yes, I'll have to work on them then if I'm to get Test.pm going the way I'd prefer 17:15
whichever one of us gets to them first, great
right now I'm working on need/import/use
jnthn pmichaud: OK. I can probably sneak a moment to throw in the sub traits application later on too.
pmichaud and getting the GLOBAL namespace set up properly
jnthn: actually, I think augment and sub traits are about equal
jnthn pmichaud: equally important? 17:16
pmichaud yes
jnthn pmichaud: OK.
pmichaud with augment we can start putting methods into the core setting
augment _should_ be pretty easy, it's just a different scope on the Package object
jnthn pmichaud: I can work a first cut of that up for us fairly quickly, I expect.
Yes, it should.
pmichaud and then we need the metaclass instance to know that it's okay to add methods to already-composed classes 17:17
jnthn Right now, it just won't complain if Parrot ain't going to complain. :-)
pmichaud: The only thing is that we probably can't re-call .compose
pmichaud I think we should be able to re-call .compose and it figures out "oh, I already composed this class, I just need to do some extra work here"
in particular, I think that if $*SCOPE eq 'augment', then "create a new metaclass" is "look up the existing metaclass object" 17:18
jnthn pmichaud: Right, I agree on that part.
pmichaud and "compose the metaclass" is "update it with whatever got registered"
jnthn Well
Yeah, it's just knowing what got added.
pmichaud it's okay to me if "add_method" is "immediate" in the case of an already-composed class
then compose is a no-op if the class was already composed 17:19
masak after spending some time on #perl on magnet, I just want to say: I really like you, #perl6.
pmichaud masak: there's a reason we talk about changing the culture as well as the language. 1/2 :-)
jnthn pmichaud: I'll go for that as my first cut, I think.
masak pmichaud: oh yes.
s1n agreed
there are a few poisonous people in the perl5 camp 17:20
masak also, it's in the atmosphere in there.
it's much less... respectful than the one here.
Juerd 18:19 <@masak> after spending some time on #perl on magnet, I just want to say: I really like you, #perl6.
+1
pmichaud #perl has unfortunately had that tradition 17:21
and that somewhat pre-dates Perl 6
Juerd By a decade at least.
masak in there, it's "you're an asshole!". in here, it's more like "after looking at your arguments, and pondering a while, I'd just like to say that you, sir, are an asshole". 17:22
Juerd Hm
markmont I just read the comp.lang.perl archives for July - September 1993 to see if there were people freaking out over the changes in Perl 5 alpha 1, and there weren't. People were focused on getting useful things done, and it was a respectful forum.
masak and even that is very uncommon.
Juerd masak: Uncommon in either channel.
pmichaud so if people think we're trying to break from Perl 5 -- then yes, there are some areas where we are trying to make a definitive break.
masak Juerd: just a fictional example. 17:23
Juerd It was one of my permanent channels for years
masak Juerd: I guess my real point is that people generally tend to avoid being deliberately hurtful in here.
Juerd I haven't seen much name calling.
17:23 cognominal left
Wolfman2000 ...well, a miracle can happen. I think I've finally got a database hooked up in Catalyst. We're partway there to having a pastebin 17:26
17:27 und3f joined
pugs_svn r29025 | pmichaud++ | [pm.txt]: Another Pm-posed question... 17:28
r29025 |
r29025 | Pm-11: S11:257 says "Without an import list, C<import> imports
r29025 | the C<:DEFAULT> imports." How does one import one of the
17:28 mubot left
pugs_svn r29025 | other tagsets? (I think I'm missing something obvious here.) 17:28
17:28 und3f left
jnthn pmichaud: iirc use Foo :TAG; 17:29
pmichaud jnthn: that seems to conflict with
use Sense :EXPORT; # import and re-export the defaults
jnthn pmichaud: I seem to remember thinking that too. 17:30
pmichaud it could be possible that :EXPORT, :MY, :OUR, etc are just "special"
jnthn pmichaud: I think that's possible.
pmichaud and that any colonpair that isn't special is a taget
moritz_ I also think use Foo :TAG; and either :EXPORT is wrong, or special
jnthn I think it's the uppercase/lowercase convention.
pmichaud it's also the case that I'm referring to C<import> instead of C<use> 17:31
s1n just to round out my opinion, it seems clear that some people are bitter that perl6 is still on-going, they see it as we abandoned perl5 and that perl6 is unnecessarily complex
none of the reasoning holds up, but then again, everyone who hates perl6 seems to forget to do a 'use logic;'
huf the root of some of that is their mistaken impression that if a project takes a long time, then it's evil.
pmichaud there's also
require Sense :OUR<ALL> # but this works
s1n huf: yes that too 17:32
pmichaud which presumably imports the :ALL tagset
s1n but the perl community is _very_ different than c++ and python
which has had the exact same problem
(c++0x and python3000)
yet no one in those communities were disdainful of the pending change
Wolfman2000 c++0x? That sounds like a mouthful. 17:33
s1n they act as if perl6 ran over their dog or spit in their food
Wolfman2000 python3000 at least sounds nice.
s1n Wolfman2000: they wanted the spec done by 2009, that obviously didn't happen (it's been about 11 years)
17:33 cognominal joined, cognominal left
pmichaud 11? 17:33
11?!
s1n yeah, they were talking about it right after c++98 17:34
pmichaud subtracts November 2009 from April 2000 and gets..... 9.5 years
s1n yay, a phd and he can do math :)
yeah, c++0x is gonna be awesome, and everyone is welcoming it with open arms
yet it's been in work _forever_
17:35 Whiteknight left
Juerd s1n: In a way it is comforting to see that people are passionate about Perl 6 17:35
Even if passionate in the wrong way.
pmichaud Juerd: +1
Juerd They wouldn't have spent all the time debating the issue if they really didn't care. 17:36
pmichaud If Perl 6 were truly irrelevant, then it would be far more ignored than it is :)
Juerd Yea.
(Don't tell them that! ;))
pmichaud considers tweeting that.
Juerd Hehe
You'd be giving them ideas 17:37
pmichaud and here I thought ideas were a good thing.
Wolfman2000 ...considering how we seem to be communicating...decently...
Juerd pmichaud: Touché
Wolfman2000 I may as well ask. What are your thoughts on using Skype for chatting with each other?
Juerd I don't use Skype and I find it much easier to type English than to speak it.
pmichaud when I want to chat voice, I'd consider skype. :) 17:38
huf skype is the devil ;)
Juerd Many of my clients want me to use skype
pmichaud but trying to illustrate a line of code over the telephone is... painful :)
huf or at least not software.
jnthn is a native English speaker, but doesn't like voice calls still
Juerd But those conversations always turn out to be a huge waste of time.
So I don't do that anymore.
Wolfman2000 pmichaud: screen sharing is starting to make its way into skype
already on the Mac version
Juerd What use is screen sharing? 17:39
pmichaud still, with skype it'd be harder to have multiple conversations going on simultaneously, as we often do here
Wolfman2000 Juerd: to illustrate the line of code
Juerd Let me put it differently - how is screen sharing more useful than IRC?
Wolfman2000 pmichaud: I'm not suggesting we chuck IRC
Juerd: ...good question. I'll get back to you. 17:40
pmichaud Wolfman2000: wasn't necessarily taking it that way :)
17:40 cognominal joined
Wolfman2000 is going to need to have to change his mysql password at this rate... 17:42
17:44 payload1 joined, payload left
Wolfman2000 Hmm...what IS my feather email address anyway? 17:44
Juerd [email@hidden.address]
Hmm 17:45
Wolfman2000 ...right, should have realized
Juerd Note .nl
17:46 icwiener left
jnthn should try and make it to .nl next year. 17:48
jnthn misses stroopwaffels. 17:49
Juerd jnthn: If you come to the Perl Workshop I'll get you a box of stroopwafels to bring home :)
jnthn I'm SO coming! 17:50
Juerd perlworkshop.nl
jnthn oooh, is there a date yet?
Juerd We have a date already
jnthn oh nice
Juerd Yes. And it's very likely that it'll be held at the StayOkay in Arnhem again.
jnthn Ahh...in Arnhem again. :-)
Juerd We love that place.
jnthn The food there is win. 17:51
And it's just an all round nice place.
Wolfman2000 ...alright, more a perl 5 question, but... should use utf8; come before or after the package declaration?
Juerd Wolfman2000: Before any non-ascii code. 17:52
jnthn Juerd: OK, I'll block that date out in my calendar. :-)
Juerd Wolfman2000: Apart from that, it doesn't matter much.
Wolfman2000 Juerd++: Noted.
Juerd jnthn: Great; I'm looking forward to meeting you again.
jnthn: Can I notify the organization that you'll be there?
Wolfman2000 Still working on the website right now...this may take awhile. Not used to using the model here, and I'll DEFINITELY need more than one Controller unlike my other page. 17:53
17:57 payload1 left, payload joined
pmichaud std: class A { 1 }; class B; say('hmmm....'); 17:57
p6eval std 29025: ok 00:01 108m␤
jnthn Juerd: Yes. :-)
Juerd jnthn++ 17:58
jnthn Juerd: Hopefully this year the Ukrainian workshop doesn't have a date conflict, but either way, you asked first, so if that happens you get me. :-)
markmont There's a problem with setting $! in die by doing find_contextuals followed by infix:= ... infix:= can call die if the variable is read-only, and that would be bad. I'd just copy a little code, but the actual work is done by Object::!STORE which not only can call die itself, but calls a bunch of other stuff that could, in theory, wind up calling die. Any thoughts? 17:59
jnthn I hope not that, 'cus I like my little Ukrainian adventure each year. :-)
Juerd :) 18:00
pmichaud markmont: I'd just use store_dynamic_lex and be done with it :) 18:01
at least for now :)
eventually I somewhat expect it'll be bind instead of assign anyway
Wolfman2000 Juerd: I recall you said you wanted to do a kernel update or something today. Do you know roughly how much time I have before I'll be disconnected?
Juerd Wolfman2000: Yeah... about that... I forgot. 18:02
So it's not going through.
(today)
Sorry
Wolfman2000 Juerd: That's fine. Tomorrow's a better day for you to do it for me anyway due to college classes.
...I'm still finding it funny how I'm going to become a teacher, yet I can't get rid of my Bachelors.
18:03 icwiener joined
masak it's much easier to think of the #perl denizens as cleverly programmed bots badmouthing Perl 6. it also explains why arguing with them yields little result. 18:03
markmont pmichaud: Will do :) 18:04
huf masak: i still think the basis of the problem isnt something you can get at with arguments or logic
because it's mostly emotional
masak nod.
pmichaud ...but we can work on the fringes a bit 18:05
masak yes, that's probably what's within range.
pmichaud every once in a while we confront someone with facts and they say "oh...."
masak :)
pmichaud I'm not sure I'd march into the lion's den, but I'll hold an occasional chat/post/tweet with a lion 18:06
s1n holy crap, #perl is chock full of people who haven't done 'use logic;' yet 18:08
moritz_ the freenode #perl, or the irc.perl.org #perl?
masak the latter. 18:09
we've been talking there all day with the fine folks who... have different views about things.
kidd` btw, is twitter.com/rakudoperl deliberately abandoned? last tweet is from September 17th. 18:11
pmichaud not deliberately, no
I've been super-busy on nqp-rx
I need to tweet some rakudoperl stuff
moritz_ just low on bus number ;-)
pmichaud I should just make it a point to tweet twice a day to that
kidd` maybe it could be a good way to comunicate things little by little, without having to fight with lions
pmichaud kidd`: good suggestion, but my experience has been that you need a bit of sustained effort with the lions 18:12
something more than 140-chars :)
maybe others would have more luck at it than I
besides, I spent much of my week battling in a different arena with different lions :) 18:13
jnthn ;)
moritz_ pmichaud: I could write an IRC bot which would allow more people to use the rakudoperl twitter account ;-)
pmichaud moritz_: I'd accept that
jnthn is SO glad that he's taking the time out to cook something decent today, rather than the quickest thing he can throw together. 18:14
18:14 nbrown_ joined 18:15 frew joined
Wolfman2000 If it helps, the #catalyst people are at least being friendly towards me...then again, I'm not asking when Catalyst will use Perl 6 or anything. 18:16
18:16 frew left
masak Wolfman2000: yes, I've also had good experiences on #catalyst. they're pretty helpful. 18:18
kidd` moritz_: the twitter-IRC bot is a good idea IMHO
18:18 xenoterracide left
Wolfman2000 moritz_: The only concern I have with the twitter bot...how would you decide who can use it? 18:18
moritz_ I hope I can make hugme do that
Wolfman2000: by nick whitelist
Wolfman2000 moritz_: Should have guessed. 18:19
18:19 xenoterracide joined
moritz_ hugme already verifies if people are registered to freenode 18:20
Wolfman2000 hugme: help
hugme Wolfman2000: (add $who to $project | list projects | show $project | hug $nickname)
18:20 frew joined
Wolfman2000 can't help but feel I'm missing some options 18:20
moritz_ hugme: cuddle Wolfman2000
hugme cuddles Wolfman2000; moritz_++
s1n masak: you should know that i'm provoking them
moritz_ s1n: that's not always a good basis for a dialog 18:21
Wolfman2000: ok, cuddling is not documented
masak s1n: I see. I'm not sure that's the way to go. but it's your call.
s1n conversing with them is not a dialog
masak s1n: no, but there are people listening in.
s1n: keeping up dignity is important on its own.
s1n i am being serious from time to time
masak I'm being serious all the time. 18:22
I don't have time to joke around.
it's draining enough as it is staying in that channel. I feel as if I've sprung a leak.
diakopter he 18:24
hee
Wolfman2000 ...and now due to what may be a misunderstanding, Catalyst is apparently giving me access to update the PODs as needed. 18:25
masak :)
#catalyst++
18:26 cognominal left
Wolfman2000 All I wanted to know was the proper way to have all controllers be able to use a common sub. Apparently now I have to create that either using a base controller or a role...whatever a role is in Perl 5. 18:26
masak probably a Moose role. 18:27
18:29 nbrown left
zaslon lolmasakhazblogged! masak++ 'Week 19 of Web.pm -- building a tutorial': use.perl.org/~masak/journal/39863?from=rss 18:34
masak many of the discussions on #perl seem to end up with 'yawn. let me know when Perl 6.0.0 is out.' as the conclusion. 18:35
moritz_ oh, they come to a conclusion?
moritz_ is genuienly surprised
masak they have no trouble coming to conclusions, trust me. :) 18:36
Wolfman2000 Building a tutorial...this should be a good read. 18:38
jnthn wow masak++ blogs twice today!
masak and a third one (by necessity) coming up :)
and I'm already pretty tired. :)
Wolfman2000 ...you had to use my own words.
masak I shall sleep well tonight.
Wolfman2000: they were very fine words. 18:39
they warmed my heart.
masak edits the post to say 'Wolfman2000++'
Wolfman2000 *chuckles*
Astaire? Who comes up with these names? 18:40
masak Wolfman2000: that program is a port of Ruby's Sinatra. I hope that explains it. :) 18:41
Wolfman2000 Frank Sinatra...guessing Fred Astaire. Good enough.
masak Wolfman2000: I'm guilty of the naming. suggestions welcome.
18:41 PacoLinux left
Wolfman2000 masak: I thank you for at least not using Spears 18:41
moritz_ lol 18:42
masak does she dance?
Wolfman2000 ...I wouldn't call it dancing
Tene hi masak
18:42 PacoLinux joined, payload left
masak Wolfman2000: bet she doesn't do this: www.youtube.com/watch?v=Y8n7WQIXQDs 18:42
Tene: oh hai
moritz_ she mostly offends the audience by not singing, but using playback
markmont Improved attempt (pmichaud++, jnthn++) at getting rid of that set_global '$!' in die: github.com/markmont/rakudo/commit/1...d5ea4db880 18:46
Tene It kinda seems to me that $! should be a context var. 18:47
... and that's what your patch does.
:)
moritz_ Tene: then apply it ;-)
Wolfman2000 masak: 4 words for that scene. 18:48
Big Lipped Alligator Moment.
masak :) 18:49
Wolfman2000: I kinda like it.
it might not take the plot forward, but hey, he's dancing on the roof.
that's a pretty clear metaphor.
18:50 oZ1 joined, oZ1 left, oZ] joined 18:53 Chillance joined
s1n pmichaud: have time this evening? 18:53
18:54 nbrown_ is now known as nbrown 18:55 nbrown is now known as nbrown_, Paks left, nbrown_ is now known as nbrown 18:59 NorwayGeek left, NorwayGeek joined
s1n masak: i'm not really trying in #perl anymore 19:00
19:00 NorwayGeek left
huf s1n: since at least 20 minutes ago :) 19:01
19:01 NorwayGeek joined
s1n they're just being fecitious (spelling?) 19:01
i think the root of the problem is 2 fold: 1) it's taken a long time to get perl6 going and 2) they would have to learn something new 19:02
maybe we can take something away from that
19:02 NorwayGeek left
huf you're still looking for rational reasons 19:03
stop it ;)
19:03 NorwayGeek joined
s1n no, i found those reasons a long time ago 19:03
huf ah. keep it up then
s1n if there had been a not too radically different perl6 in 2002, there wouldn't be this problem
i.e. if there had been a few things, like classes, junctions, and the grammar engine 19:04
but that ship has sailed, so keep up the good work #perl6
moritz_ then it would have been perl 5.somthing, not Perl 6
19:05 frew left
s1n moritz_: yeah, but even if you had called it perl6, there would have been little if any resistance 19:05
i think people just dont like the unknown, and perl6 is a big unknown to perl5 people
19:07 pkkm joined, NorwayGeek left
s1n i think the only thing that will bridge the gap is the next BigThing written in perl6 19:07
19:07 NorwayGeek joined
s1n CPAN galvanized perl5, perl6 needs something epically proportioned 19:08
colomon Philosophical Question I hit last night with my rakudopatch: 19:09
Should .Str be guaranteed to return Str, or is a class derived from Str okay? 19:10
moritz_ anything conforming to Str should be fine 19:11
19:11 zloyrusskiy joined
s1n has to go run 9 miles now 19:11
19:12 zloyrusskiy left 19:16 NorwayGeek left, NorwayGeek joined 19:19 NorwayGeek left, NorwayGeek joined
colomon moritz_: then I suppose I got it right last night. 19:23
and I should add a test or two to verify / "document" that. 19:24
I'll be away from the net for the next few hours, but I'll try to get the tests pounded out and then check them in when I have access again.
moritz_ have the appropriate amount of fun ;-) 19:25
19:26 colomon left 19:28 NorwayGeek left, NorwayGeek joined 19:30 cxreg joined 19:35 athaba joined 19:36 cognominal joined 19:41 cognominal left
zaslon lolmasakhazblogged! masak++ 'November 8 2009 -- people may call you a troll, but...': use.perl.org/~masak/journal/39864?from=rss 19:43
19:44 NorwayGeek_ joined
jnthn masak: lol :-) 19:45
Nice subject. 19:46
masak I feel exhausted.
I need to go lie down somewhere.
moritz_ suggests $Home::bed 19:47
19:47 nihiliad left
masak good idea. 19:47
Woodi masak: still here ? 19:48
Tene Hmm... getting some weird behavior from NQP.
masak Woodi: yes, but not for long, I think.
Tene -> #parrot
Woodi about last blog... look into html-template Makefile.PL and you will be know answer..
anyway, i tried 5 or so p6 modules and noone installs... 19:49
mainly build errors...
masak Woodi: oh, you're right! 19:50
Woodi: so it must be that proto sets PARROT_DIR, and that's why it gets further. yes that makes sense.
Woodi .pbc was required ?
masak Woodi++
Woodi: I think that Makefile.PL's seriously old
19:50 jan_ left
Tene dates back to pre-install days? 19:50
masak oh yes. 19:51
pre-most anything.
jnthn Prehistoric makefiles.
masak anyway, I'm in no condition to fix that today. will do it tomorrow. :)
jnthn masak: Sleep well. :-)
Woodi other modules have other bugs... not in code just in build... 19:52
masak thanks. :)
moritz_ never had problems installing SVG or SVG::Plot 19:53
Woodi what is: RAKUDO_DIR=$PARROT_DIR/languages/rakudo
masak moritz_: this is in the installed-modules branch, though.
Woodi for me rakudo installs to /usr/bin as perl6 binary...
masak Woodi: it's something that made sense when Rakudo was always in that directory in Parrot. 19:54
which it was, a year or so ago.
19:54 jaffa8 is now known as jaffa4
Woodi many hardcoded things things.. 19:54
19:54 scottt joined
Woodi i will try SVG :) 19:55
moritz_ masak: that's what i'm talking about, yes
Wolfman2000 ...I am really questioning if roles are really needed for Catalyst.
Woodi err... how to install svg ? no Makefile.*... 19:56
moritz_ ./proto install SVG
worked in the installed-modules branch last I tried 19:57
Woodi no idea how to install proto :)
jaffa4 what is perl 6?
19:57 lmc left
Woodi k, enought for me for today... gn good ppls 19:58
moritz_ good night Woodi
masak 'night Woodi
19:58 masak left
moritz_ jaffa4: you're not here for the first time - what is this, a quiz? 19:59
20:00 NorwayGeek left
jaffa4 just a test 20:01
moritz_ recommends writing tests for official test suite instead
jaffa4 really?
do not you have enough tests? 20:02
What kind of tests?
20:03 fax joined
moritz_ t/spec/TODO contains an (incomplete) list of tests that need to be written 20:05
Juerd Maybe that should be put in a chart :) 20:06
(j/k) 20:07
20:07 nihiliad joined
Wolfman2000 And...roles may not be needed after all. 20:09
moritz_ fourier-transforms Juerd and puts him into a chart
Juerd \o/ 20:13
jaffa4 ? t/spec/TODO 20:14
moritz_ svn.pugscode.org/pugs/t/spec/TODO 20:15
! 20:16
jaffa4 pugscode 20:18
how misleading!
mathw history provides the reason
Juerd The test suite started in the pugs repository 20:21
And... still lives there :)
It likes being there I think.
moritz_ for now, yes
Juerd Many things like being there. It's gezellig. 20:22
jaffa4 rakudo: print glob "*"
p6eval rakudo 97ced1: Could not find non-existent sub glob␤in Main (file src/gen_setting.pm, line 324)␤
20:22 pkkm left
jaffa4 rakudo: print (glob "*") 20:23
p6eval rakudo 97ced1: Could not find non-existent sub glob␤in Main (file src/gen_setting.pm, line 324)␤
Juerd rakudo: sub glob { `ls -1 @_` }; print glob "*"
p6eval rakudo 97ced1: Unable to parse block; couldn't find final '}' at line 2, near "`ls -1 @_`"␤in Main (file <unknown>, line <unknown>)␤
moritz_ it's qx/.../ in Perl 6, and disabled in p6eval
Juerd Oh
eternaleye Wolfman2000: Re syntax highlighting of perl6, see Syntax::Hilight::Perl6 - it was written by a guy who used the nick azawawi, and it has multiple varieties of html output. One of them even has code folding ;D 20:24
Wolfman2000 eternaleye: I don't wish to get into cold folding 20:25
eternaleye (and yes, it's a P5 module)
Wolfman2000 eternaleye: aware it's P5 20:26
eternaleye (heh, scrolling down shows someone already pointed it out) 20:27
pmichaud (pugscode misleading) two thoughts: (1) we still have the option of moving some things into the perl6 account on github, (2) since perl6.org now points to feather, perhaps we could now do something under the perl6.org domain 20:28
afk, soccer game
eternaleye (FYI, I'm backlogging)
Wolfman2000 ah
pmichaud: back up a second. I thought feather was originally...different. Now...we have control over all of perl6.org? 20:29
moritz_ we have 20:31
eternaleye TimToady: Is there any chance that gimme5's shebang could be set to '#!/usr/bin/env perl'? /usr/local/bin/perl isn't particularly portable. 20:32
Wolfman2000 eternaleye: symlink already made
moritz_ Daniel White kindly agreed to make perl6.org available for us
Wolfman2000 "Daniel White"++
Dirty HTML code...watch out. Your days are numbered.
20:40 ash__ joined
ash__ hi 20:41
moritz_ hi ash__
ash__ i am a perl developer for last 5 years
is perl 6 ready for production use? 20:42
moritz_ not yet
Juerd It will be, before another 5 years have passed :) 20:43
ash__ this is a childish question
mathw Juerd: It's so good to be able to say that :)
ash__ is there any value for using perl6
moritz_ sure
Juerd mathw: I've been saying it for at least 5 years now ;)
mathw ash__: definitely 20:44
Juerd As a new product, Perl 6 needs a lot of testing.
ash__ every now and then i check for perl6 but things are not geting anywhere, 20:45
mathw I beg to differ
Progress is being made very steadily now
Juerd And knowing Perl 6 early might help in the future. I hope to guide businesses through the transition from 5 to 6.
moritz_ rakudo.org/status <-- doesn't look like "going nowhere" to me
Juerd ash__: It's certainly getting somewhere. 20:46
ash__ could you please tell me why shall we use perl6
mathw Because it's better
Juerd ash__: Because you like it, love it, or because it's the best tool for the job.
ash__ in what sense
mathw That's the overly simple answer of course
oZ] Everyone: This is my first day in #perl6, and it seems there's already been three queries on whether perl6 is real and relevant or not. Is this the usual, or does today just suck?
mathw Plenty of better languages have fallen over the years
moritz_ because Perl 5 has not much space left for evulation
Juerd ash__: But if you don't like it, don't love it, or if it isn't the best tool for the job, there are still many other programming languages that may suit you better.
moritz_ oZ]: today is particularly bad
mathw oZ]: It's high, but a lot of people still ask. The thing with the people who ask in here is that they can get good answers. 20:47
Juerd oZ]: That's today.
oZ] Gotcha. I've noticed you guys are 20x more pleasant than #perl, so we're off to a good start. Forgive the lurking.
moritz_ oZ]: and welcome to #perl6 ;-)
ash__ i know perl, php, java, c#, perl is the one provide me living
Juerd Lurkers are great... just a bit scary ;)
oZ] Thanks. :)
mathw ash__: For me, the key things in Perl 6 are the object-oriented programming model, the grammar system and lazy evaluation (as well as the cleaned up syntax, increased consistency and junctions)
Having the ability to use Perl 5 modules will help people make the transition 20:48
ash__ from my experience its not only the language other thing like good IDE etc for example ruby on rails
Juerd ash__: Perl 6 is not yet interesting as a tool to use in your job. For now, it's being used by people who do it for fun and/or because they care.
mathw Hopefully the Rakudo Star release will get Perl 6 into a lot of hands so people can try it for themselves
There's a lot of stuff in my job I'd do with Perl 6 now, if it was up to speed 20:49
ash__ without having supprting ide and framework it will be very difficlt to become succesful
Juerd ash__: Ruby on Rails makes people sad in the long run.
eternaleye (backlogging) jnthn: dillusion (n.): You think you see a pickle, but it isn't really there. 20:50
mathw yes, and Perl 5 has huge momentum with CPAN and the installed base
oZ] With any luck, Padre may be actually useful and usable some day, and the IDE issue will imrpove.
Tene ash__: there are MANY people who successfully use Perl with no IDEs.
mathw it should be more possible to write good IDEs for Perl 6
Tene Me, for example.
mathw But I'm one of those programmers who don't get on with them very well
Juerd never used an IDE with Perl. 20:51
Linux is my IDE :)
mathw With the noteable exception of C#, I do tend to use an IDE for that
Juerd I never understood the point of IDEs, to be honest.
oZ] Tene: That may be true, but IDEs are useful for many people, myself included. I may currently use TextMate, but I'd be a Komodo user if it wasn't so slow.
szabgab and you are in the top 0.001% of the programmer community
huf shells and programmer's editors are complex enough to do anything an ide can
Juerd szabgab: Ordered by what? :)
ash__ well personaly i use notepad
szabgab amount of perl usage :-)
Wolfman2000 ash__: Get thee to notepad++ right now. 20:52
mathw Juerd: a really good one is helpful if you're unfamiliar with the libraries, or for exploring an unfamiliar codebase, but it has to be a good one. C# gives the advantage of the reflection systems
ash__ i have three developer work underme
Juerd n...notepad. For real?
mathw wow, notepad
ash__ they dont want to use notepad , they need ide
oZ] I think he meant 'Notepad++'.
mathw man
huf i hope so
ash__ yes notepad++
mathw ah
okay
moritz_ ash__: there's an IDE with Perl 6 support already, and good syntax hilighting for vim
mathw that's a completely different program then
szabgab karma notepad
Juerd ash__: You need a programmer's editor; IDE's have that. But you can use a good editor without the rest of the environment. 20:53
Wolfman2000 @karma notepad
lambdabot notepad has a karma of 38
moritz_ @karma notepad
lambdabot notepad has a karma of 38
szabgab ty
Wolfman2000 ...we're giving notepad too much karma when we suggest the ++ version
We need to fix that.
Tene ash__: moritz is right, if you want a good Perl (5,6) IDE, you should take a look at padre.
Juerd @karma c
lambdabot c has a karma of 0
Juerd Huh, what?
huf :))))
moritz_ @karma C
Juerd C is supposed to top the karma ranking.
lambdabot C has a karma of 0
Juerd In EVERY channel.
Wolfman2000 Apparently not.
moritz_ Juerd: unless people like C-- ;-) 20:54
mathw I think it's special-cased
szabgab overflow ?
moritz_ @karma C
lambdabot C has a karma of 0
moritz_ yes, special cased
szabgab C reached karma overflow
moritz_ I just --'ed
Juerd (special case)--
szabgab karma++
@karma
lambdabot You have a karma of 75
szabgab @karma karma
lambdabot karma has a karma of 1
huf @karma a
lambdabot a has a karma of 5
Juerd @karma
lambdabot You have a karma of 6
Juerd Yay
I can haz 6
ash__ i used debian, but with ubuntu i am more comfortable 20:55
Wolfman2000 I know this isn't the right room, but...any reason why my database connecting isn't working here? (This is for work on the new website.) scsys.co.uk:8001/35949
ash__ with a shell script(from net) I been able to install whole contelpanel
oZ] ->search()->count();
ash__ and hosting 200 website so far so good
oZ] er, without the parens for your code. :)
Wolfman2000 oZ]: I'm not searching for anything. I want all.
Juerd Wolfman2000: That's searching with no conditions. 20:56
ash__ i love the power of commandline but still ease of use is a key factor
oZ] What Juerd said.
You have to have a query to get the count.
Wolfman2000 ...and now I have an exception.
due to lack of >
oZ]: placed the empty search() in, didn't change the query, still have the original error. 20:57
oZ]: If it matters, it seems that my Model folder only contains DBIC.pm, while my tables are in Schema/Result/TableName.pm 20:59
21:00 fax left 21:01 fax joined
oZ] Wolfman2000: That should be the proper placement, and then your model() line is checking the DBIC model for a schema called Users. Does your Users class' defined columns match the database? I'm still checking. 21:01
Wolfman2000 oZ]: Not calling any columns here: just the table names.
ash__ ok bye for today 21:02
oZ] Wolfman2000: Could you paste your connect_info, whether it's inside Model/DBIC.pm or the yml file? 21:03
Wolfman2000 oZ]: hang on
21:03 ash__ left
lisppaste3 wolfman2000 pasted "Connect info" at paste.lisp.org/display/90004 21:04
Wolfman2000 oZ]: that's from Model/DBIC.pm
Everything below that connect batch is POD
21:05 [particle] left
eternaleye Wolfman2000: (re Symlink already made) Yes, on feather. But my box doesn't have a /usr/local/bin/perl either. `env perl` just calls the first one in $PATH. 21:05
Wolfman2000 eternaleye: can't help you there 21:06
Juerd env is not portable.
oZ] Wolfman2000: Sorry for the delay. Get rid of that last hashref, and just supply mysql_enable_utf8 as a fourth key to the connect_info hashref. 21:08
Wolfman2000 oZ]: No error...but at this point it's me not using a template correctly. 21:09
oZ] Heh.
21:09 nihiliad left
oZ] Well, it's progress. :) 21:09
21:09 mubot joined
Wolfman2000 Okay, I stashed it in $c->stash->{users}. The templates should be able to call c.users 21:09
21:09 pointme left 21:10 zaslon left, pointme joined, zaslon joined, carlin left
eternaleye Wolfman2000: My point was that /usr/local/bin/perl breaks for a lot of people. `env perl` falls under the heading of 'sensible defaults', because almost all of the time it will DTRT 21:11
oZ] Wolfman2000: Stash gives keys to root, so it's just [% users %]
Wolfman2000 oZ]++: We have progress!
oZ] Woo! 21:12
moritz_ might I remind you that this is a Perl 6 channel?
Wolfman2000 moritz_: Might I remind you that this is for a Perl 6 website?
21:13 carlin joined
moritz_ Wolfman2000: it's still Perl 5 21:13
Wolfman2000 Besides: normal support channels went dead awhile ago.
oZ] Wolfman2000: Feel free to message me privately.
eternaleye moritz_: He's working on a pastebin for Perl 6
moritz_ eternaleye: I know
21:13 IllvilJa joined 21:16 jan_ joined, nihiliad joined 21:22 pdcawley joined 21:27 jaffa4 left 21:31 _eMaX_ left 21:33 icwiener left, icwiener joined 21:44 mj41_ joined
Wolfman2000 i34.tinypic.com/4j429y.png The colors and layout can change. The point is: we have a basic front end. 21:54
eternaleye \o/ 21:55
Wolfman2000 I figure the recent pastes will go on the left, and any other needed navigation can go in the footer. 21:56
22:00 xabbu42 joined 22:02 mj41 left 22:15 hcchien left 22:20 hcchien joined 22:21 NorwayGeek_ left, NorwayGeek joined 22:23 kaare left 22:24 NorwayGeek left, NorwayGeek joined 22:27 eternaleye left 22:29 icwiener left 22:36 athaba left 22:37 nihiliad left, _eMaX_ joined 22:38 pmurias joined
pugs_svn r29026 | colomon++ | [t/spec] Four tests showing that DerivedFromStr.Str returns an object of type DerivedFromStr (when inheriting Str.Str's implementation). 22:40
22:45 colomon joined
carlin finally got Apache to exec perl6 22:48
22:48 am0c left, iblechbot left
Wolfman2000 carlin: Using Web.pm for this, or something else? 22:48
carlin Wolfman2000: just a basic Hello world script at the moment, was going to have a poke around at Web.pm later though 22:49
Wolfman2000 carlin: You may get to poke at Web.pm first: I've been charged to take care of a new pastebin for us. 22:50
22:51 kst left 22:52 kst joined 22:56 hugme left, hugme joined
moritz_ hugme: twit hugme_test this line now appears on twitter, if everything worked 22:58
hugme hugs moritz_; tweet delivered
Wolfman2000 moritz_: you use rakudoperl, right? 22:59
that is, for twitter
moritz_ Wolfman2000: no, that's pmichaud's account 23:00
Wolfman2000: but I hope to use that too, soon ;-)
Wolfman2000 which one is yours?
dinner: afk 23:01
moritz_ Wolfman2000: currently I have only hugme_test for testing hugme 23:02
frettled Hey, guys, what should I blog about tonight? :D 23:03
Sorry, didn't mean to scare everyone else into fetal positions beneath their desks! 23:06
moritz_ frettled: blog about the bots in #perl6 ;-) 23:07
frettled Hmm!
And those are the ones that have a Voice instead of op, right?
:D
23:08 moritz_ sets mode: +v hugme
moritz_ right ;-) 23:08
frettled How do I classify the bug-submitting bot, AKA masak?
moritz_ your choice ;-)
frettled muahaha. 23:09
Let's see. I've forgotten what dalek does. I genuinely don't know what ilbot2 and ilogger2 do, except I suspect the latter is responsible for the web updates.
Add IRSeekBot to that list.
moritz_ dalek reports commmits to rakudo, nqp-rx, the book and so on 23:10
Juerd dalek announces commits
"the book" ;0
s/;0/:)/
moritz_ ilbot2 is the logging bot corresponding to irclog.perlgeek.de
ilogger2 for another one; don't know about IRSeekBot
frettled lambdabot takes care of karma for us, lisppaste3 handles the paste.lisp.org bits, phenny is a secretary, and pugs_svn tracks pugs commits?
Juerd Pugs repository commits 23:11
frettled p6eval handles various online Perl 6 code evaluation
Juerd Which holds more than just pugs
jnthn Mainly the test suite and spec are what get reported by pugs_svn
frettled yup, like the spec
mm
moritz_ and masak usually doesn't have voice, but handles RT submissions nonetheless ;-) 23:12
frettled :D
moritz_ transfixedbutnotdead.com/2009/11/08...in-threes/ lovely 23:13
23:13 Juerd sets mode: +v moritz_
frettled haha :) 23:13
23:13 pmurias_ joined
moritz_ zaslon reports blog posts 23:14
23:14 moritz_ sets mode: +v zaslon
frettled ah, yes, our newish bot member 23:15
jnthn *and* zasolan is written in Perl 6 \o/ 23:16
carlin pointme provides links to projects tracked by proto
and mubot dies every 12 hours or so but also tracks karma 23:17
Juerd Where's zasolan's source? :)
s/zasolan/zaslon/
github.com/moritz/rssbot 23:18
carlin pointme: rssbot
pointme carlins's rssbot is at github.com/carlins/rssbot/
Juerd Correct spelling helps in google too.
frettled Are the numbers for ilbot2, ilogger2 and lisppaste3 fixed, or just chosen because the names without numbers were accidentally taken when the bots logged back on the last time? 23:19
moritz_ frettled: ilbot2 is the successer of ilbot, which sucked :-) 23:20
23:20 lichtkind left
frettled moritz_: :D 23:20
23:21 meppl joined
frettled perl6: my $a; say $a+$a; 23:23
p6eval pugs: 0␤
..rakudo 97ced1: Use of uninitialized value␤Use of uninitialized value␤0␤
..elf 29026: Use of uninitialized value $a in addition (+) at (eval 122) line 4.␤Use of uninitialized value $a in addition (+) at (eval 122) line 4.␤0␤
frettled hah
std: my $a; say $a+$a;
p6eval std 29026: ok 00:02 109m␤ 23:24
23:24 pmurias left 23:26 NorwayGeek left, eternaleye joined 23:27 markmont left 23:28 markmont joined 23:30 sunnavy left, sunnavy joined 23:31 nsh left, PacoLinux left 23:32 baest left, sbp left, frederico joined, arnsholt left 23:33 sbp joined, arnsholt joined, Khisanth left 23:34 frederico left 23:35 nsh joined
frettled There, I think I got it all down now, let's see. 23:35
hrm, paste.lisp.org is irresponsive.
No, it's my ISP's international provider's link across the Atlantic that's buggy. GRR. 23:37
23:38 oZ]1 joined 23:39 Whiteknight joined
frettled whoa. that killed Firefox, too. 23:39
good thing I saved my blog entry first, haha 23:40
pastebin.com/d473612bc 23:41
Juerd, jnthn, moritz_ - any corrections/comments?
23:42 starek joined, starek left
moritz_ frettled++ 23:42
frettled: mabe you could mention that mubot also handles aliases for karma tracking 23:43
frettled What kind of aliases are those?
As in moritz_ and moritz being the same?
moritz_ for example kyle is know as KyleHa here on the channel, but as kyle on pugs, so he has two karma accounts on lambdabot
right
jnthn frettled: If you're noting that zaslon is written in Perl 6, it may be wroth noting that mubot also is. :-)
Our own dog food. We're nomming it. 23:44
Wolfman2000 back from dinner. I have to say...I'm not a big fan of steak.
moritz_ jnthn: s/dog food/bot food/ ;-)
jnthn ;-)
frettled jnthn: ah, brilliant, thanks!
jnthn Wolfman2000: Next time, I'll happily eat it for you...
jnthn <3 steak
Wolfman2000 jnthn: Fine a way to allow teleportation, and perhaps we'll have something. 23:45
s1n heh, now if the irc bots saved all the logs and the log browser were hosted up off Web.pm (or some such), that would be dogfooding it :)
Wolfman2000 s1n: I think that's one of the goals.
carlin frettled: pointme is also written in Perl 6 23:46
frettled whoa, neat!
Wolfman2000 eternaleye: You here? Got a question for you regarding the new website.
quietfanatic Hi #perl6
moritz_ hola
23:46 kst left
frettled okie-dokie 23:47
quietfanatic jnthn: When executing "Foo is Bar {...}" the call to trait_mod:<is> has no knowledge of the name "Foo". That's what I meant.
23:47 Khisanth joined, kst joined
Wolfman2000 Evening Khisanth. Surprised to see you in here. 23:47
frettled Now all I need to do is to add my blog to zaslon, and, ehrm.
jnthn quietfanatic: Ah, ok 23:48
quietfanatic: That may well be fixable.
carlin zaslon: help
zaslon usage: add <name> <link to rss feed> | remove <name> | link <name>
frettled hrm, zaslon doesn't answer privmsg with help. *clickaurl*
moritz_ frettled: or play zaslon yourself ;-)
frettled ah, there it happened anyway.
23:48 oZ] left
s1n what is zaslon? #feels out of the loop 23:48
jnthn quietfanatic: I mean, we parsed the name already...
frettled zaslon: add frettled howcaniexplainthis.blogspot.com/fee...ts/default
zaslon I am now following frettled's blog
jnthn quietfanatic: I'll see what I can do. :-)
23:49 reid05 left, reid05 joined
carlin lets see if it worked 23:49
zaslon: link frettled
quietfanatic jnthn++
zaslon frettled's blog is at Failure()<0xb71e5584>
frettled heh
s1n lol
carlin aww :(
frettled carlin: I suspect the problem is that blogspot uses Atom, not RSS.
moritz_ astrojp: add frettled howcaniexplainthis.blogspot.com/fee...lt?alt=rss 23:50
zaslon: add frettled howcaniexplainthis.blogspot.com/fee...lt?alt=rss
zaslon Sorry, I am already following frettled's blog
moritz_ zaslon: remove frettled
zaslon I am no longer following frettled's blog
moritz_ zaslon: add frettled howcaniexplainthis.blogspot.com/fee...lt?alt=rss
zaslon I am now following frettled's blog
moritz_ zaslon: link frettled
zaslon frettled's blog is at howcaniexplainthis.blogspot.com/
carlin \o/ 23:51
moritz_ astrojp: sorry for the mis-hilight
s1n what's the general purpose of zaslon? linking perl6 blogs in irc?
frettled yup
carlin It won't announce the most recent post though
frettled s1n: and what moritz_ suggested I blog about tonight, is what the bots here do.
s1n frettled: okay, i'll have to read it after my system upgrade 23:52
moritz_ frettled++ nice post
maybe we should link that on perl6.org or so
jnthn It'd probably make the channel a little more understandable for those who are new to it. :-) 23:53
eternaleye Wolfman2000: What's the question?
Wolfman2000 eternaleye: We didn't discuss all URL formats. When it comes to registering...should we attempt to keep the same URL, or can we allow a different URL to take place?
23:54 pnate joined
s1n pmichaud: ping 23:55
eternaleye Wolfman2000: Not sure what you mean by 'when it comes to registering'. Are you asking whether we should use the same URL scheme for retrieving a user's posts regardless of whether it's registered vs. different based on registration status?
pugs_svn r29027 | moritz++ | [perl6.org] link to bot explanations
frettled moritz_: sure, but perhaps better, copy the contents into a more permanent location than a blog.
moritz_: that makes it easier for other people to fix it when it gets out of date next week, too :D 23:56
Wolfman2000 eternaleye: ...no. The registration page will have a URL. Should that same URL be used for POST/submitting?
moritz_ frettled: good idea, will look into that tomorrow
eternaleye Wolfman2000: Ah. I'd say no, since they do different things. Maybe URL/register and URL/submit ?
astrojp moritz_: no problem. :) 23:57
moritz_ frettled: community/bots.html might be good
Wolfman2000 eternaleye: Basic no was all I needed. I feel awkward about using submit on its own.
moritz_ frettled: if I'm allowed to rip off your blog post for that ;-)
frettled moritz_: sure, consider it some sort of public domain thingy text, copyright transferred to the Perl 6 community™
eternaleye Wolfman2000: Well, if we use /submit, we can have that be the webinterface page for pasting something as well
moritz_ frettled++ again 23:58
Wolfman2000 eternaleye: Won't be that easy to work with on Catalyst from what I'm seeing. Let me get more done if possible.
23:58 payload joined
moritz_ bed & 23:58
Wolfman2000 You can check out what I'm doing on my feather box if you want.
frettled I doubt that the text would have what we call «verkshøyde» in Norwegian, that is, «a work worthy of (copyright) protection» :)
eternaleye Wolfman2000: /submit for webinterface and /submit/auto for POST w/ client?
frettled moritz_: goodnight!
frettled hits the sack as well.
moritz_ frettled: it's nice to ask anyway ;-)
frettled Thanks for the inspiration, guys! 23:59
moritz_ sleep well too
frettled moritz_: absolutely, that's my position as well. Thanks for the assistance and idea!
23:59 nihiliad joined
Wolfman2000 eternaleye: I'll see what I can do. No guarantee. 23:59
frettled moritz_++ of course

Powered by Cro and the Raku® Programming Language.

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

Thank you!