Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2009-08-24

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:01 wknight8111 joined #perl6
00:08 japhb Is there an example somewhere of defining and using classes in NQP?
00:09 * japhb trying to avoid shaving too many yaks before the day is over
00:09 jnthn Any language's action grammar is a class.
00:09 jnthn And using - TheClassName.new()
00:10 jnthn I think it only likes the class Foo::Bar; syntax though (over the curly brace one).
00:10 japhb jnthn, AH!  That was my mistake!
00:11 japhb Hmmm, still no good.
00:12 jnthn No? Hmm.
00:12 jnthn "no good"?
00:13 japhb Actually, no, the class Foo { ... } syntax works too.  I had an unrelated bug in the same area, which I didn't see because it was valid Perl 6 but not valid NQP.
00:13 jnthn japhb: Oh no, it should be able to take a block
00:14 jnthn japhb: Ah, so you found
00:14 jnthn I was just reading the grammar.
00:14 japhb :-)
00:14 justatheory joined #perl6
00:15 japhb jnthn, NQP has no 'use' equivalent, right?  Just load_bytecode()?
00:18 jnthn japhb: correct
00:18 japhb jnthn, can you override an NQP class's new()?
00:18 jnthn japhb: Should just be able to write a method new
00:18 jnthn I can't think of any reason off ahnd that would not work.
00:19 japhb jnthn, OK.
00:19 jnthn oh, I lied about NQP having no built-ins. It appears it does...
00:19 japhb Yeah, found them.  Just a couple.
00:19 japhb I've been adding back modified versions from Rakudo of a few key builtins and vars.
00:19 jnthn Oh, I wonder if we can't see them in Rakudo because we compile the code to PIR, but don't load the NQP library or something.
00:19 jnthn That's fine.
00:20 jnthn You can probably negotiate with pmichaud over what can go in and stay in.
00:21 japhb jnthn, for now, I'm just keeping them local to the project, but I may suggest to pmichaud that some standard builtins collections be available.  For instance, the "glue" builtins, such as run(), qx(), %*ENV, %*VM, and so on.
00:21 japhb Not as core,
00:22 japhb but as a well-known PBC to load if desired.
00:22 jnthn japhb: nod
00:22 jnthn That makes sense.
00:25 japhb The biggest problem I'm having writing NQP is finding sample code that's not a language grammar.  :-)
00:26 japhb .oO( I'm writing the sample code that I needed, I guess ... )
00:27 jnthn You probably are, I'm afraid.
00:36 Tene jnthn: japhb asked for it, so NQP will have 'use' and 'eval' soon.
00:37 japhb Tene++ # wheeee
00:37 jnthn Tene++
00:38 wayland76 Tene++
00:38 jnthn karma Tene
00:38 jnthn @karma Tene
00:38 lambdabot Tene has a karma of 58
00:38 jnthn @dogma Tene
00:38 lambdabot Unknown command, try @list
00:39 jnthn aww
00:39 japhb OK, I'm getting frustrated with my grepping.  How do you define/set attributes for a class in NQP?  Alternately, since there's no 'bless' (at least, not in the NQP sources, I suppose it could come from elsewhere in PCT), what do you return from a new() method?
00:39 jnthn .oO( must not make "karma ran over dogma" joke )
00:40 jnthn japhb: Aww crap.
00:40 jnthn japhb: Well at first glance NQP appears to lack "has"
00:40 jnthn :-/
00:40 japhb jnthn, Oh, that's *very* reassuring.
00:40 japhb jnthn, yup.
00:41 jnthn like, the word doesn't appear in the grammar
00:41 Tene embedded PIR!
00:41 japhb jnthn, yup, that's what I saw
00:41 jnthn Well yes
00:41 jnthn That's probably the best bet ATM.
00:42 japhb Tene, OK, sure, but we're now beginning to get to things that really ought to be in there without going to inline PIR.
00:42 jnthn japhb: I agree.
00:42 japhb Well, at least something *minimal*.
00:43 japhb pmichaud, around yet?
00:46 Tene jnthn: you ruin my trolling with your reason.
00:48 wayland76 We have to remember the reasonable defence against trolling next time we get trolls
00:48 wayland76 But OTOH, it was a successful troll -- it got a response :)
00:49 japhb Rrolls rarely enjoy they response that involves getting cleaved.  :-)
00:49 japhb er
00:49 japhb Trolls
00:50 jnthn lol you fail it you can't even spell troll lol
00:50 wayland76 cleave together, or cleave apart?
00:50 jnthn ;-)
00:50 seeken joined #perl6
00:52 japhb The silly thing was my typo was caused by fat-fingering a minor cleanup of an otherwise grammatically correct sentence.  Sigh.
01:14 japhb If I say create a '$OS' variable in PIR using set_hll_global, I still have to say 'our $OS' in NQP before trying to use $OS, or the parser complains of no declaration.  Is there some magic way to poke the parser so it knows about the "builtin" global?
01:15 jnthn japhb: afaik no
01:15 jnthn japhb: let me glance the source
01:18 jnthn japhb: Ah
01:18 jnthn If you can tweak the outermost block between the past and the post stages, you can probably call .symbol('$OS', 'scope'=>'package')
01:19 jnthn Oh wait...that's a parser tweak.
01:20 jnthn Hmm. I can't think of a good way. Sorry.
01:26 leedo joined #perl6
01:27 jnthn japhb: I'll sleep in a moment (very late here...) Any final questions before I go?
01:33 jnthn Timed out
01:33 * jnthn -> sleep, night all
01:35 wayland76 'night :)
01:39 japhb Hope he sleeps well ... and jnthn++ for all the help today
01:57 alester joined #perl6
02:32 cotto joined #perl6
02:45 carlin joined #perl6
02:53 ihrd rakudo: my $code = {"Foo"}; my %h = :bar; $code(| %h);
02:56 jaldhar joined #perl6
03:07 molaf joined #perl6
03:09 carlin heh, it killed the bot
03:10 ihrd :(
03:12 zamolxes joined #perl6
03:26 SmokeMachine joined #perl6
03:40 PerlJam @tell alester Several pages on rakudo.org go to a blank page (/documentation, /community, /developers-guide, /status).  Same thing happens when I attempt to login.
03:40 lambdabot Consider it noted.
03:56 quietfanatic joined #perl6
03:57 leedo joined #perl6
04:09 wayland76 joined #perl6
04:23 kyle_l5l joined #perl6
04:23 japhb http://gitorious.org/parrot-plumage/parrot-plumage
04:24 japhb jnthn, pmichaud: Note especially the NQP section of http://gitorious.org/parrot-plumage/parrot-plumage/blobs/master/TODO
04:33 Zloyrusskiy joined #perl6
04:51 _jaldhar joined #perl6
05:13 leedo joined #perl6
05:14 s1n jnthn: i created a really nice work-around for the interim, it took several hours of thinking and 2 minutes to implement, it's _far_ from perfect, but i'm going to use it for now
05:15 s1n jnthn: i'll send it to you tomorrow
05:16 sri_kraih_ joined #perl6
05:36 adhoc left #perl6
05:41 orafu joined #perl6
06:00 justatheory joined #perl6
06:04 moritz_ good morning
06:06 finanalyst joined #perl6
06:07 japhb Good morning, moritz_
06:28 mj41 joined #perl6
06:28 rfordinal joined #perl6
06:30 japhb http://use.perl.org/~geoffrey/journal/39516
06:33 moritz_ japhb++
06:33 japhb moritz_, thx!
06:47 szabgab joined #perl6
06:51 polettix joined #perl6
07:00 Su-Shee joined #perl6
07:00 Su-Shee heyho :)
07:00 moritz_ ho yo!
07:01 sjohnson hi!
07:02 Su-Shee *hihi* perl on android. :)
07:08 carlin perl on android?
07:10 Su-Shee google phone. http://code.google.com/p/android-scripting/source/browse/perl/ase/scripts/hello_world.pl
07:11 ihrd japhb++
07:12 rfordinal joined #perl6
07:13 japhb ihrd, thx!
07:21 Kentrak joined #perl6
07:22 Kentrak Anyone here know the state of proto/multi in objects?
07:23 Kentrak Trying to define another .new and finding that it makes the original .new no longer accessible.  Not sure if it's a bug in the object code and .new isn't defined as proto, or if it's a multiple dispatch bug.
07:24 moritz_ good question
07:24 moritz_ rakudo: class A { multi new ($x) { } }; A.new()
07:24 moritz_ bah, p6eval is gone (server down)
07:25 lisppaste3 kentrak pasted "untitled" at http://paste.lisp.org/display/85940
07:26 Kentrak http://paste.lisp.org/display/85940 should show the problem somewhat succinctly...  although I just realized I should have included the output of the latest rakudo build.
07:27 carlin Su-Shee: nice, I knew ASE had accepted the suggestion to add perl but I didn't think it was close to being released
07:27 lisppaste3 kentrak annotated #85940 "untitled" at http://paste.lisp.org/display/85940#1
07:28 moritz_ Kentrak: this is a bug in rakudo. If your custom new method would have hidden the default constructure, you'd have gotten a dispatch error
07:28 moritz_ Kentrak: could you please submit a bug report to rakudobug@perl.org?
07:28 Kentrak yep
07:28 Kentrak Just wanted to check before submitting. :)
07:29 moritz_ great
07:32 Kentrak Also, is there a defined object method for determining the available mthods/submethods an object has?
07:33 moritz_ you can loook at Class.^methods(:local) for example
07:33 moritz_ or ask about specific methods with .can or .^can
07:33 moritz_ don't know how any of those deal with submethods though
07:33 Kentrak moritz_: ah.  Somehow missed the introspection part of S12.  Thanks
07:34 moritz_ Kentrak: if the specs are confusing (which wouldn't surprise me) you might also look at the tests in t/spec/S12-introspection/
07:35 moritz_ rakudo passes all of them right now
07:35 nbrown joined #perl6
07:36 Matt-W Morning
07:37 Kentrak moritz_: thanks, on the right track now.
07:40 Kentrak left #perl6
08:03 agentzh joined #perl6
08:08 jaffa8 joined #perl6
08:09 jaffa8 hi
08:09 moritz_ hi japhb
08:09 moritz_ erm, hi jaffa8
08:09 moritz_ sorry, tab fail :-)
08:09 jaffa8 has anyone tried to compile std.pm?
08:09 jaffa8 Because I cannot
08:09 moritz_ worked for me last I tried
08:10 jaffa8 in what version?
08:10 jaffa8 I get an error
08:10 jaffa8 Statement not terminated properly at line 171, near "= %::($var"
08:10 moritz_ what exactly did you do?
08:10 jaffa8 I wrote a small program
08:10 jaffa8 use STD;
08:10 jaffa8 my $e=1;
08:11 jaffa8 and ran perl6.exe on it.
08:11 moritz_ that's not how it works
08:11 moritz_ rakudo can't run it yet
08:11 moritz_ cd pugs/src/perl6
08:11 moritz_ make
08:11 moritz_ # and hope
08:11 moritz_ I don't know if anybody tried it on windows though
08:16 jaffa8 the lesson is
08:16 jaffa8 never base a project on one person.
08:17 moritz_ jaffa8: you're very welcome to increase our bus number
08:17 jaffa8 what is bus?
08:18 moritz_ http://www.c2.com/cgi/wiki?BusNumber
08:19 jaffa8 IS that a joke?
08:19 dukeleto joined #perl6
08:19 moritz_ no, it's serious
08:20 moritz_ just that people don't have to be hit by a bus to drop out of a project
08:20 moritz_ they can just have a shift of interest, illness, a new job with no time left, whatever
08:20 jaffa8 exactly
08:20 jaffa8 the term is oversimplifying
08:21 moritz_ yes, but it's catchy :-)
08:21 moritz_ it has a nice second conotation
08:21 moritz_ a bus is a communication medium
08:22 moritz_ so if you have bus number of 1, only one person can efficiently program on that project
08:22 jaffa8 so what is the bus number of rakudo?
08:23 moritz_ 2, I'd say
08:24 jaffa8 Specifically, who are they?
08:24 moritz_ jnthn and pmichaud
08:29 ihrd ah, good link
08:30 ihrd moritz_++
08:33 jaffa8 moritz_, why is catch in uppercase?
08:33 jaffa8 it does not look logical to me...
08:33 masak joined #perl6
08:34 moritz_ jaffa8: all special blocks are, to make them stand out
08:34 moritz_ BEGIN, END, CATCH, CONTROL, PRE, POST, CHECK, INIT - you name it
08:34 moritz_ it's not only that they stand out - they might also be very useful names for the programmer to use for own stuff
08:35 jaffa8 it does look a big deal to me
08:35 jaffa8 it does not look a big deal to me
08:35 masak greetings, perl6ers.
08:35 masak jaffa8: you are so full of paradox!
08:35 moritz_ oh hai masak
08:36 ihrd moritz_: I find out why I had the strange error while 'make' tests
08:36 moritz_ ihrd: what was it?
08:36 ihrd moritz_: It`s just another one Test.pir in my system
08:37 moritz_ oh. was it in PERL6LIB or something?
08:37 ihrd yes, I just rm all of them, remake spectest and it works for now
08:38 masak that's a weak definition of 'finding out' :)
08:38 masak oh, another Test.pir. sorry. :/
08:39 masak yesterday in a blog comment I argued against having several Test.pm (and, by extension, Test.pir) copies spread out over different repos.
08:39 masak I don't think it's the responsibility of an average Perl 6 project to hold or copy in Test.pm.
08:40 masak I think it only leads to chaos and bitrot.
08:40 ihrd masak: yes, we should use one
08:41 ihrd masak: and send patches instead of tweaking our own Test.pm :)
08:41 masak maybe this is the right moment to finally codify the first practice in wtop...
08:41 masak ihrd: oh, absolutely. :)
08:41 jaffa8 masak: what makes you think that?
08:42 jaffa8 What is the paradox?
08:42 masak <jaffa8> it does look a big deal to me
08:42 masak <jaffa8> it does not look a big deal to me
08:43 jaffa8 it was a typo
08:43 jaffa8 the main problems with perl 6 now
08:43 jaffa8 compiler is slow
08:43 jaffa8 the design is not implemented
08:43 jaffa8 the error messages are not the best
08:44 jaffa8 std.pm cannot be compiled.
08:44 jaffa8 not a big deal
08:44 ihrd masak: sorry, I do`t catch, what is wtop?
08:45 masak ihrd: it's a project mberends and I thought up: "Well-Thought-Out Practices". kind of like Perl::Critic for Perl 6 projects.
08:45 ihrd masak: ah, thank you.
08:46 jaffa8 masak: I was responsing on moritz_'s response.
08:47 jaffa8 he wrote that INIT is an uppercase to allow the programmer to use init for his own use.
08:47 masak jaffa8: that's one of the reasons, yes.
08:47 jaffa8 it would not bother me if INIT was in lower.
08:47 jaffa8 I would use init_program
08:47 jaffa8 or init_parsing
08:48 jaffa8 or whatever if it were a keyword
08:48 masak jaffa8: another reason is that INIT is special in that it runs automatically.
08:48 masak jaffa8: I'm glad it would hypothetically not bother you.
08:48 Matt-W moritz_: great blog post on object creation btw. Liked it a lot.
08:48 Matt-W moritz_++
08:49 masak jaffa8: I'm also happy that Perl 6 gets out of the way in the way it does for common keywords.
08:49 moritz_ Matt-W: thanks, and glad that I could be of help
08:49 masak the sigils help a lot with that, of course.
08:49 ihrd ah, yes, join one's voice to that
08:49 ihrd I post link in my own blog :)
08:49 Matt-W the all-caps stuff is a nice hint that Something Special Is Happening
08:50 Matt-W a bit like .WHAT and .HOW and friends
08:51 masak indeed.
08:52 moritz_ just like BUILD
08:55 sjohnson Perl 6 Is A Big Deal
08:55 sjohnson :)
08:55 masak sjohnson: :)
08:59 kidd joined #perl6
09:00 al joined #perl6
09:00 moritz_ btw the server on which p6eval is hosted is down
09:00 moritz_ I apologize for the inconenience
09:00 moritz_ $ ssh moritz@timtowtdi.org
09:00 moritz_ ssh_exchange_identification: Connection closed by remote host
09:01 masak moritz_: no problem. thanks for a generally very consistent service.
09:01 moritz_ you're very welcome.
09:14 braceta joined #perl6
09:15 jaffa8 moritz_, what is your role in Perl 6?
09:16 masak moritz_ is our teddy bear. we hug him when we need to hug someone.
09:16 masak he's also excellent at writing spectests and blog posts.
09:16 moritz_ lol
09:16 moritz_ jaffa8: according to masak <http://use.perl.org/~masak/journal/39445> I'm a Priest :-)
09:17 masak oh, right. Priest.
09:17 moritz_ teddy priest?
09:17 masak I get those two mixed up sometimes.
09:17 moritz_ jaffa8: we don't have fixed roles. I do what's necessary and within my scope
09:17 masak moritz_: now that creates interesting mental images.
09:17 masak moritz_++
09:17 moritz_ and I try to have fun along the way, which mostly works very well
09:18 carlin http://www.bearfamilygifts.com/bear/priest-bear-hat-glasses-lisar.jpg
09:19 * moritz_ can somewhat identify with that picture
09:20 masak carlin: wow.
09:20 masak that's uncanny.
09:24 * Matt-W seems to recall classifying himself as a Warrior/Priest/Archer with Mage ambitions
09:25 ihrd left #perl6
09:25 masak something like that here too.
09:26 Matt-W Most of my Perl 6 time is Priest though
09:30 masak most of mine is Warrior, closely followed by Archer.
09:31 jaffa8 moritz_,your role seems to be pretty fixed to me.
09:31 moritz_ jaffa8: how so?
09:32 jaffa8 you are writing blogs.
09:32 jaffa8 Is that all ,right?
09:32 masak certainly not.
09:32 masak moritz_ is all over the place.
09:32 moritz_ lunch&
09:32 masak even in the canteen!
09:32 jaffa8 evangelist?
09:33 * moritz_ runs IRC logs, evalbot, the test suite, several Perl 6 apps, rakudo patches, ...
09:33 moritz_ really lunch&
09:33 masak jaffa8: well, no need to bring religion into it, but sure.
09:33 Matt-W moritz_ is part of the glue which holds the project together
09:33 jaffa8 masak, I just read it
09:33 M_o_C joined #perl6
09:33 jaffa8 evangelist
09:34 jaffa8 it is used in programming field
09:34 masak moritz_ is a very important cog in the machinery that keeps moving Perl 6 forward day by day.
09:34 masak jaffa8: yes, because people keep applying religious metaphors to programming.
09:34 masak jaffa8: like the Emacs/vim "holy wars".
09:35 masak I've never truly grokked such exclusionist views of things.
09:35 masak but maybe that's not what you meant.
09:36 Gothmog_ joined #perl6
09:37 jaffa8 where is the exclusism in being an evangelist?
09:38 jaffa8 exclusionism
09:38 guest_007 joined #perl6
09:39 masak I was talking about the danger of exclusionism in applying the religion metaphor to the computer world.
09:40 masak being an evangelist is fine, it's just not a word I tend to use to describe the concept.
09:40 masak also, much of the promotion for Perl 6 that I see favourably brings in other languages, comparing and contrasting.
09:41 masak Perl tends to do that a lot, given its very horizontal heritage.
09:41 masak jaffa8: but you shouldn't pidgeon-hole moritz_ as only a PR man. he does a lot more around here.
09:45 wayland76 joined #perl6
09:46 hanekomu joined #perl6
09:48 masak seems HTML 5 people have to be very clear about when HTML 5 will be 'ready': http://html5doctor.com/2022-or-when-will-html-5-be-ready/
09:48 Matt-W o/ wayland76
09:49 wayland76 hi :).  I'll likely be in and out, because I'm trying to beat xorg into submission at the moment, and I'm using an X-based IRC client
09:49 hanekomu joined #perl6
09:50 jaffa8 masak: Are you taking this thing a way too seriously?
09:50 wayland76 My answer is who cares, because XHTML2 is better than HTML 5 anyway :)
09:50 masak wayland76: :)
09:50 * sri_kraih falls off his chair laughing
09:50 masak wayland76: wasn't XHTML2 scrapped?
09:51 wayland76 No, just some people got in a snit and developed HTML 5 as a competitor :)
09:51 masak jaffa8: with 'this thing', do you mean computers? religion? life?
09:52 masak I tend to take all those a little too seriously, I think.
09:52 carlin wayland76: http://tech.slashdot.org/story/09/07/03/1447237/XHTML-2-Cancelled
09:52 jaffa8 masak: I mean the importance of what I think of moritz_
09:52 ejs joined #perl6
09:52 braceta left #perl6
09:52 wayland76 Ah, well, I looked into the whole thing earlier this year, so I guess I missed that one :)
09:52 wayland76 Thanks :)
09:53 masak jaffa8: I'm just trying to guide you into seeing more aspects of moritz_. your current view seems a bit... one-sided.
09:54 wayland76 Does that mean that we never get XForms?
09:54 masak wayland76: we already have XForms, no?
09:56 wayland76 Do we?
09:56 wayland76 What standard is it part of?
09:56 masak I always thought it was a standard on its own.
09:56 Matt-W it is
09:56 masak but frankly, I'm not well versed in XForms.
09:56 Matt-W XForms can fit into any XML application
09:56 Matt-W thanks to the wonder of namespaces
09:57 Matt-W but it was kind of intended for XHTML2
09:57 masak :/
09:57 wayland76 Yes, but what about using them on the web?
09:57 Matt-W which was, in itself, pretty damn cool
09:57 Matt-W oh well you need a browser which supports them...
09:57 Matt-W there was a firefox extension which added xforms support at one point
09:57 Matt-W but that was a long time ago
09:57 wayland76 Yeah, I don't care too much about XHTML2 itself, more about XForms and things like that.
09:57 Matt-W sadly I think they'll be gone in favour of ordinary crappy forms and lots of javascript :(
09:57 wayland76 There's still a Firefox/Seamonkey extension that does XForms
09:57 masak theoretically, some browser could just go ahead and add full XHTML2 support, right?
09:57 Matt-W oh yes
09:57 Matt-W but they won't
09:58 masak :(
09:58 wayland76 Well, apparently there's a namespace conflict between XHTML2 and HTML5
09:58 Matt-W well that was stupid wasn't it
09:58 masak namespace conflict?!
09:59 Matt-W shouldn't XHTML2's be www.w3.org/$year/xhtml2
09:59 wayland76 Oh, seems like I ingested some FUD
09:59 Matt-W there's a lot of FUD about XHTML2
09:59 wayland76 one guy said namespace conflict, another said not
10:00 Matt-W some people saw a lack of backward compatibility and went mad
10:00 Matt-W it's a bit like Perl 6 in that respect
10:00 Matt-W only Perl 6 will not be scrapped
10:00 masak *phew*
10:01 wayland76 Oh well, when Perl 6 is done, XHTML2 will be a simple matter of programming :)
10:01 * masak wonders if there'll be a "save HTML2" movement
10:02 Matt-W everyone's looking at HTML5
10:02 Matt-W bloody web applications
10:02 Matt-W mumble mumble
10:02 Matt-W I do NOT want to write my desktop software in HTML!
10:02 Matt-W it's not meant for it!
10:02 masak it's inevitable.
10:02 Matt-W no it's not
10:03 Matt-W especially if people stop saying it's inevitable
10:03 masak :)
10:03 Matt-W I don't mind so much the idea of using browser tech
10:03 Matt-W but HTML is ridiculously unsuitable
10:03 Matt-W it's not even very good at normal documents
10:04 wayland76 Includentally, XHTML2 works partly now: http://w3future.com/weblog/gems/xhtml2.xml
10:04 Matt-W I know people do impressive things with it, but it's usually because they're very clever or very persistent rather than because the tools are good
10:04 jaffa8 masak: ok
10:05 masak jaffa8: sorry if I seemed overboard in any way. :) it was not my intent.
10:05 wayland76 ...it's just the way he is :)
10:05 masak 哈哈
10:05 wayland76 That's why he fits in here :)
10:05 masak Perl 6: all overboard!
10:06 carlin I don't mind HTML, but JS/AJAX is painful
10:06 masak carlin: it's mainly the DOM that's painful.
10:06 masak JS in itself is a quite endurable language... with a bad rep.
10:06 wayland76 Well, XHTML2 would seriously reduce the amount of JS due to having cooler forms features :)
10:07 wayland76 The problem with JS is that it's not Perl 6, but apart from that, it's not too bad :)
10:08 masak wayland76: I could dig out the guy who said it, but I forgot his name right now: "given that JavaScript was promoted to the ubiquitous programming language almost completely by accident... we could have ended up much worse."
10:09 jaffa8 masak, what is the meaning of overboard in this context?
10:09 masak jaffa8: exaggerated.
10:09 jaffa8 And what have you exaggerated?
10:10 wayland76 Life, the universe, and everything :)
10:10 masak I have exaggerated my breakfast. and yet, I'm hungry.
10:11 jaffa8 ok, that I am one-sided?
10:11 jaffa8 Or the importance of all of this?
10:11 jaffa8 or what?
10:11 masak jaffa8: the importance of all of this.
10:12 patrickas joined #perl6
10:13 masak jaffa8: it's nice talking to you and all... but I get the feeling sometimes that you either do not have a 100% grip on the language, or you like living under bridges and taking toll fees from people who pass.
10:13 wayland76 Btw, I often wear a hat like that teddy bear :)
10:13 donaldh joined #perl6
10:13 masak wayland76: I think that might suit you very well. :)
10:14 masak wayland76: in the future, I will picture you with that kind of hat when we chat. :)
10:14 wayland76 jaffa8: Watch this
10:14 wayland76 @karma-all
10:14 lambdabot "ethanbot2"          1454
10:14 lambdabot "moritz"              978
10:14 lambdabot "pmichaud"            643
10:14 lambdabot "jnthn"               584
10:14 lambdabot "lwall"               577
10:14 lambdabot [1922 @more lines]
10:15 wayland76 I don't know how ethanbot2-- got up there, but you'll notice that moritz_ has the highest karma around here
10:16 wayland76 masak: Are you on Facebook?
10:16 masak wayland76: heavens, no.
10:16 masak I don't agree to their idea that the Web should have walls.
10:16 wayland76 No?  Ok.  I don't follow it much, but it's useful for finding people that I kind of know, but not well enough that I see them often
10:17 wayland76 Not even walls with graffiti on them?
10:17 masak maybe that kind of wall.
10:17 masak but not walls which you cannot see over.
10:17 masak Larry Wall is allowed, too.
10:19 wayland76 So you don't like that lots of Facebook is password protected?
10:19 wayland76 Did you know that you can make your own part not password protected?
10:19 masak wayland76: I don't not like it, I just stay away from it. it's not my Web.
10:20 wayland76 Ok.
10:20 masak phenny: tell jnthn that I hear some people wanting user-defined FETCH/STORE. is that on some schedule? I don't see it in docs/ROADMAP...
10:20 phenny masak: I'll pass that on when jnthn is around.
10:20 zamolxes joined #perl6
10:21 masak wayland76: same reasoning as writing open-source software, actually. it lasts longer and feels more meaningful.
10:21 Matt-W I basically only use Facebook to communicate with a certain group of people
10:22 Matt-W I no longer view it as a means to create content
10:22 wayland76 No, I don't regard it as a way of creating content
10:22 Matt-W I did for a brief, foolish few days
10:23 wayland76 But, for example, I met a guy recently, and all I know is his name and his brother.  I haven't seen or heard from the brother in months, and have no contact details for either, but because the brother is my facebook friend, I could write and ask for the other brother's e-mail address
10:23 Su-Shee I clicked me with the same people I already clicked in orkut, xing and so on and now everybody has clicked each other again so we can go back to business. ;)
10:24 wayland76 I'm basically only on Facebook.  I have a MySpace, but nothing ever happens there, except the occasional message from a band I like
10:24 meppl joined #perl6
10:25 Matt-W I despise MySpace
10:25 Matt-W Not on a conceptual level
10:25 Matt-W Because that's not really any different to any of the others
10:25 Matt-W but because their design and markup is so incredibly horrible
10:26 wayland76 Yeah, I've noticed that.  I had to put in a good solid effort to make it look alright :)
10:26 Matt-W I didn't even try
10:26 Matt-W I've only got an account to help me poke around in other people's :)
10:26 Matt-W my profile contains nothing at all
10:27 Matt-W Facebook I approve of more because they start out with a pretty decent design and don't let people break it
10:27 Matt-W shame about the apps platform
10:27 Su-Shee ah well.. I have irc. :)
10:27 wayland76 What's wrong with the apps platform?
10:27 Matt-W silly games
10:28 Matt-W I get an endless stream of invites from games and quizzes my contacts are playing
10:28 Su-Shee which mostly already were silly web games ;)
10:28 Matt-W yes but I could ignore them more effectively then
10:28 wayland76 I'm messaging a few people my myspace URL because I don't want it in the logs, etc :)
10:30 wayland76 masak: Another good reason to get Facebook ... there are pictures of my hat :)
10:30 Su-Shee iiiih wayland76 - it's _translated_ :))
10:30 masak wayland76: as irresistable as that sounds...
10:32 wayland76 Su-Shee: Is that a problem?
10:33 Su-Shee wayland76: oh yes :) because it's an automated translatation. sounds plain "off" :)
10:33 wayland76 Ah, well, just look at my picture then :)
10:33 alanhaggai joined #perl6
10:33 Su-Shee looks very country. :)
10:34 wayland76 :)
10:34 wayland76 With a cravat?
10:34 wayland76 Well, maybe :)
10:35 wayland76 But if you put that hat from the teddy bear on me.... :)
10:40 Su-Shee my office view says "swimming pool", "cafe", "river" and all this in bright sunlight... how very motivating...;)
10:42 rfordinal joined #perl6
10:44 riffraff joined #perl6
10:52 jnthn dobry den, #perl6
10:52 phenny jnthn: 10:20Z <masak> tell jnthn that I hear some people wanting user-defined FETCH/STORE. is that on some schedule? I don't see it in docs/ROADMAP...
10:52 masak lolitsjnthn
10:52 jnthn masak: It may not be too hard to do, just need to work out exactly how.
10:53 masak \o/
11:04 rjh joined #perl6
11:14 eMaX joined #perl6
11:20 donaldh joined #perl6
11:29 brrrrr joined #perl6
11:33 jaffa8 wayland76, what has popularity have to do with anything?
11:36 jnthn Gah, my brane won't work today. :-/
11:41 szabgab joined #perl6
11:49 wayland76 jaffa8: Well, he gets the karma when he's been helpful in some way
11:49 wayland76 Although the number of helps is what counts -- a big help counts for the same as a small one :)
11:50 masak ...which promotes long-term helpfulness, not just one-off stunts.
11:50 * wayland76 offers jnthn an oil filter for his brain
11:51 wayland76 Well, my point is, moritz_ does lots of little things around, and that's what got him the karma.  Our higher-level mages don't have the same level, but tend to produce their help in larger blocks.
11:56 masak (and karma gets reset occasionally, making it even more approximate)
11:58 takadonet joined #perl6
11:58 takadonet morning all
11:58 yves joined #perl6
11:59 masak takadonet: o/
11:59 takadonet masak: How are you doing today?
12:00 masak takadonet: I am doing adequately, thank you. and you?
12:00 takadonet masak: Doing good
12:07 jaffa8 Is there a webpage where I can try long programs on std
12:07 jaffa8 ?
12:08 polettix rakudo: my @a; my $x = 1; @a.push($x); $x++; @a.say;
12:09 polettix uhm, no
12:09 carlin polettix: the bot is offline
12:09 masak polettix: evalbot's server is down. :/
12:09 polettix ach
12:09 polettix I get 2 printed in rakudo, I'd expect 1
12:09 jnthn known
12:09 polettix ok
12:09 jnthn (yes, bug)
12:09 masak polettix: but thanks. :)
12:10 jnthn I did try fixing it once and oddly it broke Other Things. :-/
12:10 polettix fine, I'll add an "as copy" then :)
12:11 polettix s/as/is/
12:13 masak polettix++ # works around
12:14 * jnthn should maybe try and fix it tomorrow.
12:14 polettix complaint-driven development? :)
12:15 moritz_ jnthn: are you doing a rakudo day tomorrow?
12:15 moritz_ polettix: we partly do that, yes.
12:15 jnthn moritz_: Aye, plan to.
12:16 polettix well, I'm not complaining anyway, playing with perl6 is awesome
12:16 carlin it's the best kind of development, given that everything no one uses works fine ;)
12:16 * jnthn prefers to call it "feedback driven development" :-)
12:16 polettix [performance apart of course :)]
12:17 abra joined #perl6
12:18 masak carlin: :)
12:18 masak polettix: Rakudo is awesome, the performance is a...w...e...s...o...m...e...
12:20 takadonet man.... it really suck trying to keep 3 computers with latest version of rakudo. Always confuses me why some tests pass on certain computers and not others. Till I realizes forgot to update rakudo.
12:20 jnthn Good job pmichaud and I are both planning on improvements that will aid performance in various areas then. :-)
12:20 takadonet excellent
12:21 masak takadonet: make it a one-keypress action to update.
12:21 jnthn Plus the Parrot folks are building profiling tools to help us know why we're slow.
12:21 Matt-W ooh good for them
12:21 jnthn But yes, I fully agree - performance is sucky.
12:21 takadonet masak: I do but still forget to update :)
12:21 masak takadonet: update every morning with your morning yoghurt. I do. :)
12:22 takadonet I should
12:22 * moritz_ doesn't eat yoghurt regularly, but still updates his rakudo regularly :-)
12:22 * masak smiles and plays "Living on the edge"
12:23 * Matt-W bleeds a lot
12:24 carlin it's a sign of a balanced lifestyle; maintaining a good yoghurt:rakudo-updates ratio
12:24 wayland76 I eat yoghurt more regularly :)
12:27 masak we have a fair number of branches on http://github.com/rakudo/rakudo/ -- are there any outdated ones that can be removed?
12:28 moritz_ the autounfudge-with-limits and mail-patch branches are still valuable
12:28 jnthn Probably
12:29 moritz_ I doubt that ins and ins2 are of any use though (maybe pmichaud can comment)
12:29 jnthn In half an hour, that github page will have loaded, and I'll tell you.
12:29 masak :)
12:30 moritz_ jnthn: git-branch -r also shows all remote branches
12:30 * jnthn got the page with the crying thingy that looks like a cat and an octopus had a kid together
12:31 wayland76 Wait, doesn't the latest rakudo include the changes from ins2?
12:31 jnthn wayland76: yes
12:31 wayland76 ins is dead
12:31 moritz_ wayland76: which is why we don't need that branch anymore
12:31 wayland76 Ah, I see the question now :)
12:41 payload joined #perl6
12:57 jaffa8 Whare are talking about the speed of the compiler or the speed of generated code?
12:57 jnthn jaffa8: Both.
12:57 moritz_ mostly of the generated code, I think
12:57 moritz_ but many performance improvements will affect both
12:59 jaffa8 I wonder
12:59 carlin joined #perl6
12:59 jaffa8 Is there any optimasion in parrot?
12:59 jaffa8 I mean automatic one.
13:01 jaffa8 What is the objective performance of Perl 6?
13:01 jaffa8 Has anyone measured that?
13:02 moritz_ the JIT runcore can be seen as optimization
13:02 moritz_ Perl 6 is a language specification, and as such has no performance
13:02 moritz_ only implementations have performance
13:02 moritz_ and rakudo is currently ~500 times slower than perl5
13:02 jaffa8 I mean implementation
13:03 _jaldhar joined #perl6
13:03 jaffa8 In what?
13:03 jaffa8 What is slower?
13:03 jaffa8 The code execution?
13:04 moritz_ running a simple program in both
13:04 beggars joined #perl6
13:04 moritz_ including compilation and run time
13:04 moritz_ anyway, feel free to do your own benchmark
13:04 jaffa8 what about runtime in itself without compilation?
13:04 moritz_ (and be sure to compile parrot with --optimize for fairness)
13:04 moritz_ I don't know, never tried.
13:05 jaffa8 I suspect compilation accounts for most of it.
13:05 moritz_ don't suspect. Try it!
13:06 jaffa8 It toook 5:40 seconds
13:06 jaffa8 to compile 3000 lines program
13:06 jaffa8 and give an error message
13:07 jaffa8 I have a look
13:07 jnthn Is that 5 mins 40 seconds or 5.4 seconds?
13:07 jaffa8 5 minutes
13:07 jnthn wow
13:07 masak that's... slow.
13:08 jnthn Unusually slow.
13:08 masak aye.
13:08 masak compilation here takes seconds, not minutes.
13:08 jaffa8 if I just use configure.pl
13:09 jaffa8 What kind of code would that produce?
13:09 jaffa8 Debugged or optimised?
13:09 moritz_ debugging
13:10 moritz_ jaffa8: try perl Configure --gen-parrot --gen-parrot-option=--optimize
13:10 moritz_ and then rebuild
13:10 moritz_ and try again
13:10 moritz_ should be about 30% faster, or so
13:10 jaffa8 30%?
13:10 jaffa8 that is not mch
13:11 jaffa8 masak: How big is your code?
13:11 molaf joined #perl6
13:11 masak jaffa8: good question. hold on.
13:14 jaffa8 it is 5:30 seconds.
13:14 jaffa8 now
13:14 jaffa8 What have you expected?
13:14 jaffa8 How should It be?
13:14 jaffa8 How fast should It be?
13:15 jaffa8 Should it compile it in 10 seconds?
13:15 jaffa8 Is it backtracking parser?
13:16 jaffa8 Is rakudo a backtracking parser?
13:16 takadonet Looks like someone is on a coffee high...
13:17 masak :)
13:20 jnthn jaffa8: No, it (hardly ever) backtracks.
13:21 jnthn jaffa8: I'd expect an order of magnitude performance gain on the parsing once we have proto-regexes in place.
13:21 jnthn And are using those.
13:21 jnthn Since those drastically cut down the search space.
13:21 jnthn At the moment we try a lot of things when parsing that we coulda ruled out in advance.
13:22 jaffa8 no cofee, just me.
13:22 takadonet jaffa8: Good to hear
13:25 Front_slash joined #perl6
13:26 orafu joined #perl6
13:27 masak jaffa8: the November codebase is ~2000 lines of code (incl empty lines and comments) spread over ~20 .pm files. it takes 50 seconds for Rakudo to parse it all on my computer,
13:29 masak so, about 40 lines per second.
13:31 wayland76 'night all
13:31 masak wayland76: o/
13:32 Su-Shee does p6 have a goto? :)
13:32 jaffa8 Whare you?
13:32 jaffa8 Whare are you?
13:32 moritz_ Su-Shee: afaict yes
13:32 wayland76 Australia
13:32 masak buubot: spack goto
13:32 buubot masak: Couldn't match input.
13:32 masak :/
13:33 masak Su-Shee: anyway, it's in S04.
13:33 mtkatwork joined #perl6
13:35 abra joined #perl6
13:35 Su-Shee hrmpf. ;)
13:35 Su-Shee now I can't reasonably say "it has to go, won't exist in p6 anymore anyway!" ;)
13:37 masak Su-Shee: goto is a reasonable, very simple construct. we have abstracted away most of its uses in loops and if statements, but it can still be useful in its bare form.
13:38 moritz_ masak: simple? I doubt that
13:38 moritz_ masak: you know there are scopes to consider...
13:38 masak moritz_: that is true.
13:38 jnthn tbh, I haven't got the slightest clue how on earth we're going to make goto work.
13:38 moritz_ if it were all that simple, rakudo would support it by now
13:39 jaffa8 yes, goto looks simple
13:39 Matt-W Make it only work within the same scope
13:39 jaffa8 there is goto in PIR
13:39 Matt-W But don't tell anybody
13:39 masak moritz_: one could support a limited form of it... in-scope goto or some such.
13:39 jnthn jaffa8: lol
13:39 * moritz_ wonders if TimToady insists on having a general goto
13:39 Su-Shee let it silently die. :)
13:39 moritz_ or mark it as post-6.0
13:39 jnthn jaffa8: The goto in PIR is also block-limited.
13:40 Matt-W general goto is just too painful
13:40 jnthn sub foo { label: if 42 { goto label; } } # own't work, goto is in a different block
13:40 jnthn *won't
13:40 Matt-W also almost completely useless
13:41 moritz_ so can be omitted altogether :-)
13:41 jaffa8 what about the assembler?
13:41 jaffa8 pasm?
13:41 moritz_ PIR doesn't have scopes as Perl 6
13:41 moritz_ neither does PASM
13:41 carlin Whenever conversation starts about goto I'm reminded of this: http://xkcd.com/292/
13:42 Matt-W that might have to go up on my desk at work
13:44 jaffa8 jnthn,,   moritz_ says there are not scopes in PIR
13:45 Matt-W no, he said there aren't scopes like Perl 6's
13:45 Matt-W doesn't mean it doesn't have its own
13:45 jnthn jaffa8: Perl 6 block == Parrot Sub.
13:49 jaffa8 I guess there is if block as well
13:49 jaffa8 and loop block
13:50 jnthn A form of goto that only lets you jump within the current scope or to a lexical outer scope would probably be do-able without excessive pain, and cover a lot of the use cases.
13:51 Matt-W it's the only time I ever use it
13:51 jaffa8 why? SHould goto be able to jump into another procedure?
13:51 Matt-W noooooooooooooo
13:51 carlin But not implementing it would be completely painless and lower raptor deaths
13:52 Matt-W jnthn: put it down for after Rakudo Star
13:52 jaffa8 programming ne pain. IMO.
13:52 jnthn Matt-W: I suspect it already was. :-)
13:52 Matt-W jnthn: no doubt
13:52 Matt-W far more important things to worry about
13:53 jnthn Matt-W: Unless some enterprising soul comes along and implements it, it's unlikely to happen before then.
13:53 jaffa8 if ($programming eq "pain") { print "this is something wrong"}
13:53 jnthn I'm uninterested in implementing it.
13:53 Matt-W jnthn: I'm not going to take that bait
13:53 Su-Shee I'll refactor my gotos here into perl nirvana anyway. I've never once used goto in Perl.
13:54 jaffa8 Su-Shee, it is only about you, it is about humanity
13:54 jaffa8 s/it is/it is not/
13:55 Su-Shee yeah, but I just asked for it. ;)
13:55 carlin humanity would be better off without goto :)
13:55 Matt-W humanity does not require goto in Perl 6
13:55 jaffa8 Are you sure?
13:55 Matt-W yes
13:56 Matt-W however, there may be some benefit from a sane version
13:56 jaffa8 I think it is like hyperjump, Stargate, instantenous travelling.
13:56 Su-Shee jaffa8: yeah and years later everybody asks themselves how they ended up here. ;)
13:56 carlin Yes, pity it's difficult to do that travelling while reading the code
13:57 jaffa8 With a good ide, you can travel.
13:57 jaffa8 within a sec.
13:57 Matt-W I'm not sure there's any kind of IDE which can help you make sense of goto-spaghetti
13:58 jaffa8 there may be visual representation of the code
13:58 jaffa8 Also, who said it should be overused?
14:00 Su-Shee my goto here is a replacement for checking for some settings first. instead, it jumps into goto "when the setting situation comes along" instead of checking it first and depend everything else later on the setting.
14:02 Matt-W I am admittedly usually the last person to argue in favour of avoiding a potentially troublesome language feature
14:02 Matt-W But unrestricted goto is just too horrible to contemplate
14:04 moritz_ "As in Perl 5, it is possible to goto into a lexical scope, but only for lexical scopes that require no special initialization of parameters"
14:04 moritz_ that's not unrestricted at all
14:04 KyleHa joined #perl6
14:04 moritz_ it means you can't goto into a method, for example
14:05 Matt-W that's acceptable :)
14:05 * moritz_ recommends reading S04
14:06 moritz_ "You may not go into a given or a for, though, because that would bypass a formal parameter binding"
14:06 moritz_ uhm, but doesn't 'if' also do formal parameter binding?
14:06 moritz_ if $x -> $y { say $y }
14:07 nihiliad joined #perl6
14:08 Matt-W It does if you do that form of if
14:08 Matt-W but surely the ordinary form doesn't
14:08 Matt-W if $x { say "cow"; }
14:10 eMaX joined #perl6
14:14 polettix left #perl6
14:14 M_o_C joined #perl6
14:25 jnthn The branch re-dispatch-1 can go.
14:26 Psyche^ joined #perl6
14:28 pmurias joined #perl6
14:30 masak jnthn: done.
14:31 jnthn masak++
14:31 masak the syntax is `git push origin :re-dispatch-1` for some inscrutable reason.
14:32 masak "strangely inconsistent" :P
14:33 moritz_ do you have to delete them first locally?
14:33 masak moritz_: not that I know.
14:34 masak but I didn't have a local copy of that branch, so maybe.
14:35 masak ah. reading the manpage for 'git push' makes it a little less arbitrary having the colon there.
14:36 ruoso joined #perl6
14:39 masak what happens if you pass a positional, and then the same argument as a named parameter?
14:40 mikehh joined #perl6
14:40 moritz_ presumable the same as if you passed the same twice as named
14:43 masak last-one-wins?
14:43 moritz_ dunno
14:43 masak ISTR that was the behaviour for two nameds.
14:43 moritz_ rakudo: sub a (:$b) { say $b }; a(:b(4), :b(1))
14:43 moritz_ DAMMIT
14:43 masak :)
14:44 EDevil joined #perl6
14:45 payload joined #perl6
14:45 moritz_ mlenz@wvbh074:~$ perl6 -e 'sub a (:$b) { say $b }; a(:b(4), :b(1))'
14:45 moritz_ duplicate named argument - 'b' not expected
14:45 moritz_ jnthn++ # don't know if that's how it's specced - but the error message is awesome
14:46 jnthn moritz_: Heh, that's a Parrot error message actually. ;-)
14:46 moritz_ oh
14:46 jnthn moritz_: BTW, is that the correct semantics?
14:46 moritz_ parrot++ then
14:46 moritz_ jnthn: no idea.
14:47 pmurias ruoso: hi
14:48 moritz_ S06 doesn't seem to mention it
14:48 masak "Perl 6 allows multiple same-named arguments, and records the relative order of arguments with the same name. When there are more than one argument, the C<@> sigil in the parameter list causes the arguments to be concatenated [...] Other sigils bind only to the I<last> argument with that name"
14:48 moritz_ oh
14:49 masak I had to look a while for it, too.
14:49 jnthn ouch
14:49 jnthn That's going to be fun to implement. :-)
14:49 masak :)
14:49 jnthn Good job we're moving away from Parrot's binder. :-)
14:49 moritz_ well, maybe there's room for negotiation here
14:50 moritz_ I don't think it's particularly useful
14:50 Matt-W I can see uses for it
14:50 moritz_ it can generate hard to debug errors
14:50 Matt-W but yes, it could
14:50 moritz_ foo(:a, :b, [ten more lines] :!a)
14:51 moritz_ "I called it a with :a, why is :$a false?"
14:51 Matt-W because you aren't reading your code properly!
14:51 Matt-W at least we *have* named arguments
14:52 Matt-W maybe multiple named arguments should be something the routine has to say it accepts
14:52 Matt-W not entirely sure about the utility of accepting multiple ones when it just uses the last silently
14:52 jnthn BTW, for those interested in module install etc toolchain stuff for Perl 6, Alias just pointed out on #parrot the #toolchain channel, which may be a source of input.
14:52 masak Matt-W: flattened hashes.
14:53 pmurias where can i find the remains of the perl 5 to perl 6 translator?
14:53 moritz_ masak: that's the one use case I could think of
14:53 jnthn (It was mentioned in the context of the Parrot module stuff that japhb was playing with, but may be applicable here.)
14:55 jaffa8 What is the story of perl 5 to perl 6 translator?
14:57 moritz_ jaffa8: Perl 5 has a backend (MAD) which emits something XML-like for perl 5 code
14:57 pmurias it used to be a gsoc project
14:58 moritz_ jaffa8: and there was a gsoc project to translate that into Perl 6 code
14:58 pmurias it lived in the pugs repo from what i can remember
14:58 moritz_ iirc the source code is stil in the pugs repo
14:58 ruoso hi pmurias
14:58 Matt-W I didn't know about that
14:59 jaffa8 Is Mad complete?
15:00 pmurias ruoso: do you know where should we replace the smopp5 runloop with the polymorphic one?
15:00 pmurias jaffa8: it used to preserve all info
15:01 pmurias re translator misc/pX/Common/P5_to_P6_Translation
15:03 pmichaud good morning, #perl6
15:03 masak morning, pm.
15:03 moritz_ oh hai pmichaud
15:04 Matt-W oh hai!
15:06 masak if I could make a modest proposal for a spec simplification, it would be this: for nameds not coming in through hashes, disallow duplicate nameds, period. allow a flattened hash to either override (if it comes after) or be overridden by (if it comes before) the explicit named params.
15:06 moritz_ is that really simpler?
15:07 masak I think so.
15:07 jnthn pmichaud: morning
15:07 masak it would catch coding errors in the form of accidentally duplicated nameds.
15:08 masak it would also cause saner semantics for nameds whose values are lists.
15:10 Eevee joined #perl6
15:11 pmurias would it be worthwhile to revive the p5 to p6 translator?
15:11 masak pmichaud: I've been admiring compilers/pge/PGE/Exp.pir today. :)
15:11 masak pmurias: I think so.
15:11 moritz_ pmurias: IMHO it would make to wait until the possible targets support more of Perl 6
15:12 moritz_ *make sense*
15:12 pmichaud masak: where "admiring" == "in complete awe of how someone can make such a big mess of things?"   ;-)
15:12 carlin are +<= / +>= mention anywhere in the specs? Couldn't find them in S03 (and it would be a big help if they were in Differences.pod ;) )
15:12 masak pmichaud: it's not a big mess.
15:12 pmichaud :-)
15:12 masak pmichaud: for what it does, it's quite well organized.
15:12 pmurias moritz: rakudo supports most of the common stuff doesn't it?
15:12 pmichaud masak: thanks
15:12 moritz_ pmurias: and it would make sense to write it in Perl 6. Haskell hackers are hard to find within the perl community
15:12 masak pmichaud: of course, PIR-producing PIR is scary in itself... :P
15:12 moritz_ pmurias: for some value of "common", yes :-)
15:13 pmurias moritz_: the orginal code is very poorly written so it's not really worth keeping it
15:13 masak or you could write it in Perl 5, and conv... oh wait.
15:14 jnthn The translator being able to translate itself would be kind the point you know that it's really working. :-)
15:14 moritz_ I see two options for a 5 -> 6 converter
15:15 moritz_ either it works like a normal compiler, and produces code that doesn't have much in common with the source
15:15 masak jnthn: true. you could run both it and it-but-converted on itself, and see if they both produce the same Perl 6 program. :)
15:15 moritz_ or it tries to do some fuzzy translation, which will not always be what the users wants, but much closer to the source, and more readable
15:16 pmurias i think the second choice is preferable
15:16 masak but teh hard.
15:16 * moritz_ too
15:16 moritz_ but it wouldn't be able to bootstrap :-)
15:17 pmurias why not? if you avoid things that don't translate it should bootstrap itself easily
15:18 moritz_ anyway, if it were my project I'd write it in Perl 6 in the first place
15:18 pmurias most of the stuff in p5 could be translated fairly easily
15:19 moritz_ however there's not yet a Perl 5 grammar available in Perl 6, so it might be worth using PPI
15:19 Su-Shee <-- home. later.
15:19 pmurias why not use MAD?
15:19 Su-Shee left #perl6
15:20 masak does MAD work for 5.10?
15:20 moritz_ it should
15:20 moritz_ it's not very convenient though
15:20 moritz_ you have to specially compile your Perl 5 to use it
15:20 donaldh joined #perl6
15:21 moritz_ but of course it's the option that offers the most accurate parse trees
15:22 finanalyst left #perl6
15:24 mtkatwork left #perl6
15:25 jaffa8 Can you explain this line in Perl:*FROM = *$from{FILEHANDLE};?
15:26 nihiliad joined #perl6
15:27 masak jaffa8: this is not #perl, it's #perl6.
15:34 masak we're often off-topic here, but we prefer not to become a general Perl help channel. there are already good such channels.
15:35 jaffa8 no problem, there are other way of seeking help.
15:36 cdarroch joined #perl6
15:36 jaffa8 ways.
15:37 arnsholt joined #perl6
15:38 masak also, in some sense, we prefer missing the future rather than missing the past. :)
15:38 masak er, in the 'longing for' sense.
15:38 frew joined #perl6
15:44 arnsholt masak: It's a lot easier to feel a longing for what you've had and know, than something you haven't and don't
15:44 arnsholt That's my theory, anyways
15:44 masak arnsholt: I have Perl 6.
15:44 arnsholt Hooray! =)
15:44 masak \o/
15:45 arnsholt Is there a scrollback bot, BTW? I'm probably missing out on a certain amount of context to the two lines by you I read
15:46 masak arnsholt: http://irclog.perlgeek.de/perl6/2009-08-24
15:47 arnsholt Gracias
15:48 jaffa8 arnsholt,unless you have seen others having it
15:48 jaffa8 So you may think it will be good for you too.
15:49 arnsholt Good point
15:53 rfordinal joined #perl6
15:57 ruoso pmurias, first we need to define the strategy we're going to use...
15:58 ruoso my initial plan was to add another variable to the current interpreter and have a PolymorphicEvalOP that would execute $continuation.next
16:02 ruoso pmurias, that would also mean that the "goto" implementation would get the current p5 state and save as $continuation.back
16:17 rfordinal left #perl6
16:28 Su-Shee joined #perl6
16:28 charsbar joined #perl6
16:30 rindolf joined #perl6
16:30 rindolf Hi all.
16:30 rindolf Does Rakudo support PERL6LIB?
16:30 rindolf Or RAKUDOLIB or something?
16:30 rindolf Someone is asking us about it on #padre  on irc.perl.org
16:31 moritz_ rindolf: PERL6LIB, yes
16:31 rindolf moritz_: OK.
16:32 takadonet that's me rindolf :)
16:32 moritz_ there's also BEGIN { @*INC.push: 'path' }
16:32 rindolf takadonet: ah.
16:32 takadonet STD.pm in Padre is not finding any of my modules that have created....
16:32 Su-Shee full-service rindolf.. :)
16:33 moritz_ takadonet: well, STD.pm != rakudo
16:33 takadonet I know
16:34 takadonet Was hope to get rid of those silly message saying it cannot find my modules in padre
16:52 jferrero joined #perl6
16:55 Limbic_Region joined #perl6
16:57 moritz_ I don't know what I should make of http://rt.perl.org/rt3/Ticket/Display.html?id=68760 - is it spam?
16:57 ruoso left #perl6
16:59 lambdabot joined #perl6
17:02 adam-pwgsc joined #perl6
17:12 jnthn moritz_: Heh, not sure.
17:12 justatheory joined #perl6
17:12 jnthn moritz_: I'm not sure how you'd accidentally send that... :-|
17:12 jnthn People do manage to achieve all sorts though. ;-)
17:13 TimToady clicked on a mailto: link somewhere?
17:16 * moritz_ currently spectests the patch from http://rt.perl.org/rt3/Ticket/Display.html?id=68296
17:45 frederico joined #perl6
17:55 Daenyth_ joined #perl6
17:55 Daenyth_ o/
17:56 moritz_ welcome Daenyth
17:56 Daenyth How's it going? :)
17:56 TimToady would you like a commit bit? :)
17:56 Daenyth for rakudo on github? I wouldn't object.. :)
17:57 TimToady I only have some pugs repo commit bits, alas
17:57 moritz_ we hand out commit bits for the pugs repo (test suite, STD.pm, synopsis) freely
17:57 TimToady but with those you can hack tests or spec
17:57 Daenyth I think I have one for pugs actually
17:57 Daenyth afk, some errands to run
17:58 moritz_ Daenyth: right (if you hav $nick@gmail.com)
17:58 moritz_ *have*
17:58 TimToady *HAZ*
17:59 moritz_ ah right. KANNOD GET USED TO KNEW SPELING
17:59 jaffa8 Hi
18:12 Su-Shee we can stop all efforts. perl 6 is ready: Su-Shee: Chuck Norris also wrote a complete Perl 6 implementation :)
18:13 moritz_ Su-Shee: but did he release it? ;-)
18:14 Su-Shee :)
18:14 hercynium joined #perl6
18:15 japhb joined #perl6
18:16 jaffa8 Why not Swarzenegger?
18:17 moritz_ because Chuck Norris doesn't write Schwarzenegger jokes
18:17 Su-Shee jaffa8: too weak. only chuck can do it. :)
18:17 jaffa8 Look at how far Swarzengegger have got!
18:18 jnthn If Chuck Norris ran California, it would have recovered from the recession before the recession even started!
18:18 jnthn ;-)
18:18 jaffa8 What makes you think that?
18:18 Su-Shee jnthn: if chuck ran California, it would never had seen a recession.. :)
18:18 jaffa8 I am not sure about that.
18:18 leedo joined #perl6
18:21 frederico joined #perl6
18:21 moritz_ what should a Perl 6 japh for Rakudo Star print?
18:21 moritz_ "Just another Perl compiler,"?
18:21 jaffa8 Sushi?
18:22 TimToady Just another Perl
18:22 moritz_ TimToady++ # nice idea
18:24 dalek rakudo: 7512462 | (Gianni Ceccarelli)++ | src/ (2 files):
18:24 dalek rakudo: move pick to setting
18:24 dalek rakudo: Signed-off-by: Moritz Lenz <moritz@faui2k3.org>
18:24 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/7512462c718a2f4eb978d7110f4832a38946d25e
18:26 pioto joined #perl6
18:27 justatheory joined #perl6
18:28 frederico joined #perl6
18:57 frettled He probably got it from Chuck Norris.  ;)
19:00 rindolf She was quoting me.
19:00 rindolf http://perl.net.au/wiki/Perl_Humour#Chuck_Norris_and_Perl_6
19:01 masak joined #perl6
19:02 jnthn lolitsmasak
19:02 jnthn :-D
19:02 masak :)
19:02 * masak had a nice swim
19:02 jnthn Cool
19:03 jnthn Bet it wasn't in the sea. :-)
19:06 masak no, it was in a cool pool.
19:06 lisppaste3 moritz_ pasted "Stary obfu" at http://paste.lisp.org/display/85972
19:07 masak moritz_++
19:07 masak scary indeed.
19:07 moritz_ somehow the last few characters are a bit too obvious
19:07 moritz_ I had some fun fiddling these stars together ;-)
19:08 mikehh_ joined #perl6
19:08 masak moritz_: that one alone is worth a blog post, I'd say.
19:09 moritz_ I'm now fiddling a bit with the quotes
19:09 moritz_ turns out I can use <...> instead
19:09 jnthn omfg
19:09 masak I think there are two Rs in 'starry', btw.
19:09 jnthn I think so.
19:09 masak unless you meant stare-y.
19:09 moritz_ I probably meant both ;-)
19:09 jnthn I parsed it as the Slovak word for "old" first :-)
19:10 moritz_ and that (3) galls me - how can I turn a * into a number?
19:10 masak jnthn: hah, me too!
19:10 masak jnthn: but Russian.
19:11 masak moritz_: pass in a list (*,*,*) and numify it.
19:12 moritz_ masak: thanks
19:13 moritz_ except that I'll spell it +<* * *> ;-9
19:13 Daenyth back
19:13 moritz_ no need for the + actually
19:14 * jnthn is trying to write in Russian at the moment to his language partner.
19:15 rindolf left #perl6
19:18 TimToady rakudo: [*]()+[*]()+[*]()
19:19 Daenyth so what's this I hear about a possible «make install» for rakudo?
19:19 PerlJam Daenyth: not only a possible, but an actual.
19:19 Daenyth :D
19:19 Daenyth so that means I can package it finally!
19:20 moritz_ Daenyth: there's a README in rakudo that's fairly up to date
19:20 Daenyth moritz_: I'll start off with a -git package
19:21 donaldh joined #perl6
19:23 Daenyth hmm
19:23 Daenyth is ICU a compile time optional depend or runtime?
19:23 moritz_ compile time
19:23 Daenyth mmk
19:23 moritz_ and *strongly+ recommended
19:23 Daenyth I'll probably package that if it isn't already then
19:24 moritz_ (actually parrot-configure-time)
19:24 Daenyth ok
19:24 TimToady1 joined #perl6
19:24 Daenyth well, for now rakudo/parrot will probably be bundled in one package, unless there's a better way you'd advise me to do it... can rakudo compile off a system install of parrot or does it need the svn checkout?
19:25 moritz_ it needs an installed parrot
19:25 moritz_ (+ make dev-install)
19:25 moritz_ erm, install-dev
19:25 Daenyth ok.. so a parrot package by itself would be a good thing
19:26 jnthn Yes.
19:26 Daenyth oh wait.. someone beat me to it
19:26 Daenyth rakudo is already in AUR
19:26 Daenyth :<
19:26 guest_8976 joined #perl6
19:26 moritz_ is it parrot-1.5?
19:26 Daenyth http://aur.archlinux.org/packages/rakudo/rakudo/PKGBUILD
19:26 guest_8976 Computer Repair Nampa - http://nnucomputerwhiz.com Any Dog Breed - http://anydogbreed.com
19:26 jnthn Just to point out ICU is a compile-time dependency for Parrot, rather than Rakudo.
19:26 Daenyth looks like the guy didn't really know what he was doing making it..
19:26 reqamst Daenyth: oh, nice
19:27 Daenyth that PKGBUILD is screwed up on so many levels
19:27 PerlJam Daenyth: so fix it!  :)
19:27 Daenyth PerlJam: yeah, I'll email the maintainer and ask if he wants to let me adopt it
19:27 moritz_ is it custom to include such things upstream?
19:28 Daenyth http://aur.archlinux.org/packages/rakudo-git/rakudo-git/PKGBUILD ok, this package looks much more sane
19:29 Daenyth except for the extra clone..
19:30 * jnthn nom
19:30 Daenyth mm
19:30 Daenyth I'll email both maintainers and see if they want to disown
19:30 iblechbot joined #perl6
19:35 takadonet welcome back TimToday
19:36 carlin joined #perl6
19:38 Daenyth So I was under the impression that pugs was dead?
19:38 TimToady it's "stable"
19:38 PerlJam as a perl 6 compiler, it's dead.
19:38 moritz_ hibernating, I'd say
19:38 Daenyth I see
19:38 moritz_ given the last blog entry on http://pugs.blogs.com/audrey/
19:38 masak not changing, I'd say.
19:39 TimToady the basic problem was that its bus number never exceeded 1
19:40 Daenyth bus number?
19:40 moritz_ we have that problem in more areas
19:40 moritz_ Daenyth: number of people how, when hit by a bus, will make development halt
19:40 Daenyth ahh
19:41 pmichaud low bus number == your project could end very quickly
19:41 moritz_ http://perlgeek.de/blog-en/perl-6/starry-obfu.html there you go
19:41 masak moritz_++ delivers!
19:42 * Daenyth glazes
19:42 pmichaud Perl 6 has arrived.
19:42 masak moritz_: which feature? I ask immediately.
19:43 moritz_ masak: hm`
19:43 adam-pwgsc moritz_: holy ****
19:43 moritz_ "but I'm
19:43 moritz_ not aware of any feature"
19:43 Daenyth is <> a quoting thing nowadays?
19:43 masak moritz_: oh, I read too quickly... :/
19:43 moritz_ Daenyth: yes, it's like qw( ) only shorter ;-)
19:43 masak Daenyth: yes, it's what used to be qw()
19:43 Daenyth gotcha
19:43 Daenyth ok, now this starts to make more sense
19:43 PerlJam moritz_: it's needs another *** *-* ** *+*/*+*/* at the end of the string though  :)
19:43 Daenyth ...barely
19:44 moritz_ PerlJam: newlines are optional ;-)
19:45 masak moritz_: it prints 'ust another Perl' and then nothing more... :)
19:45 masak er, s/ust/Just/
19:45 moritz_ PerlJam: and *** *-* ** *+*/*+*/* can be written as ****-* ** *+*/*+*/* (one space removed)
19:45 moritz_ masak: that's because Rakudo is just another Perl.
19:45 adam-pwgsc haha that's brilliant. I love it. I love perl so much.
19:45 Daenyth I have no clue how it works
19:45 masak moritz_: ah. :)
19:45 Daenyth but great! :D
19:46 PerlJam moritz_: yeah, I wasn't golfing so much.
19:46 moritz_ Daenyth: let me explain... when you use '*' as term it constructs a closure
19:46 * Daenyth glazes
19:46 moritz_ Daenyth: so 1 + * generates a block -> $x { 1 + $x }
19:46 moritz_ or * * * generates -> $x { $x * $x }
19:46 Daenyth I mean, I can see the splitting on *** bit
19:47 moritz_ inside the eval the (<* * *>) invokes that closure with a three-item list
19:47 PerlJam moritz_: turn the explanation into another blog post  :)
19:47 Daenyth +1
19:47 moritz_ ok ;-)
19:47 Daenyth can you link me to the relevent spec? Maybe I'd get it more if I read that
19:48 * PerlJam wonders how hard it would be to make a program that discovered the starry form of all numbers from 1 to 100 using <* * *>
19:48 moritz_ S02:812
19:49 Daenyth okies
19:49 moritz_ the IRC logs turn that into a direct link to the correct line
19:49 Daenyth irc logs?
19:49 moritz_ see /topic
19:49 Daenyth aha
19:49 Daenyth gotcha
19:51 moritz_ PerlJam: actually I think that much space could be safed by factoring out a common term like **** ** * * *
19:52 Daenyth * is going to take me a while to wrap my head around ._.
19:52 KyleHa joined #perl6
19:53 PerlJam Daenyth: I have to say, that even knowing what * does, it surprised me the first time I saw something like (* ** *)(2)
19:55 moritz_ masak: I suppose you did fill out your gsoc evaluation for literal++?
19:56 masak moritz_: I did.
19:56 moritz_ great.
19:56 masak and I only needed to be harassed by googlebots once, too!
19:56 moritz_ ;-)
19:57 jnthn Daenyth: You wouldn't by any chance be the person who gave a Rakudo + FreeBSD talk at UPW back in...march/february-ish this year, would you?
19:57 Daenyth jnthn: I doubt it
19:57 Daenyth unless I've suddenly taken to traveling for lectures in my sleep >.>
19:57 jnthn Oh wait, is pkgbuild the BSD stuff?
19:57 Daenyth arch
19:57 * jnthn is clueless
19:57 jnthn Ah
19:57 Daenyth freebsd does ports
19:57 jnthn Then in that case, it makes it even less likely.
19:57 jnthn Sorry.
19:57 Daenyth which is a similar concept
19:57 jnthn Yeah, you're right.
19:57 Daenyth arch PKGBUILDs are inspired partially by ports
19:57 * jnthn is not a *nix guy.
20:00 adam-pwgsc Is it hard doing perl dev on a windows box?
20:00 masak no, not nowadays.
20:00 jnthn No.
20:00 TimToady apart from perl dev being always hard everywhere... :)
20:00 masak that's just programming being hard.
20:01 masak as far as I'm concerned, Perl makes it easier. :)
20:01 PerlJam adam-pwgsc: are you familiar with padre?
20:01 Daenyth so wait
20:01 Daenyth a closure is like code interpolated in a string?
20:01 masak Daenyth: yes, in strings it is.
20:01 PerlJam Daenyth: no, a closure is like a subroutine that knows some things about the environment in which it was created .
20:02 Daenyth ahh
20:02 Daenyth now things begin to make more sense
20:02 masak rakudo: my $a = { say "I am a closure" }; $a()
20:02 Daenyth >.>
20:03 Daenyth is evalbot not here?
20:03 masak seems not.
20:03 TimToady it seems to be "stable"
20:03 masak I think it's the evalbot server that's stable.
20:03 cj joined #perl6
20:03 TimToady std: 42 43
20:04 Daenyth stably offline? :)
20:04 Eevee joined #perl6
20:04 moritz_ joined #perl6
20:04 * masak plays evalbot
20:04 guest_007 joined #perl6
20:04 carlin_ joined #perl6
20:04 gbacon joined #perl6
20:04 * moritz_ is back from net outage
20:04 masak TimToady: SORRY! Couldn't parse... two terms in a row?
20:04 b_jonas joined #perl6
20:04 TimToady moritz_: we determined that evalbot is down while you were netsplit
20:05 moritz_ TimToady: actually its server is unreachable
20:05 TimToady probably just an asteroid, then
20:06 TimToady or would that be an asterite after it hits?
20:07 TimToady .oO(tonight, chance of light aster showers)
20:07 Daenyth is Q != q (string quoting)?
20:07 TimToady Q is raw quoting, no escapes
20:07 TimToady other forms derive from that
20:07 TimToady q is really Q:q for instance
20:08 Daenyth ok
20:08 Daenyth just browsing the perlgeek blog
20:08 Daenyth the post I'm reading wasn't clear
20:08 TimToady it's hard to explain everything first :)
20:08 * Daenyth nods
20:09 b_jonas I wonder if there are any p5 obfus that accidentally work in perl6 unchanged.
20:09 mikehh joined #perl6
20:10 TimToady probably, I'd guess
20:10 elmex joined #perl6
20:10 TimToady though the comma on map will kill a lot of them
20:11 b_jonas http://www.perlmonks.com/?node=japh+algebra is killed by the =~ for regex match
20:11 TimToady that too
20:11 TimToady anything with bitshifts will die
20:11 b_jonas maybe one of my short quines
20:14 b_jonas does perl6 have here documents? I can't find them in S02
20:14 pmichaud yes, see   q:to
20:14 pmichaud or looks for "heredoc"
20:14 TimToady just another language derived from Q, really
20:14 b_jonas oh, but not with the p5 syntax, so that short quine doesn't work
20:14 TimToady yes, << is the Texas form of « these days
20:15 b_jonas yep
20:15 b_jonas second in http://www.perlmonks.com/?node_id=661934 I mean
20:15 b_jonas third might work though
20:15 b_jonas $a=q(;print"\$a=q($a)$a");print"\$a=q($a)$a"
20:16 moritz_ q() is a sub call in Perl 6
20:16 b_jonas is it, now?
20:16 b_jonas how do you use parens as a delimiter for regexen?
20:17 b_jonas wait, comment syntax in S02 has changed!
20:17 b_jonas great, I didn't like the old one
20:17 b_jonas though it's still not what I hoped for
20:17 masak postcircumfix parens always mean 'sub call' in Perl 6.
20:17 TimToady part of extensibility is making keywords not special, as much as possible.  so foo() is always a function call regardless
20:17 moritz_ rakudo: sub q($x) { say $x }; q("foo")
20:17 moritz_ ouch.
20:17 moritz_ no p6eval
20:18 moritz_ anyway, that prints "foo\n" for me
20:18 b_jonas but q is special anyway before most characters
20:18 masak b_jonas: yes, but parens are more special.
20:20 b_jonas oh well, parens are convenient in p5 for regex quoting only because of p5 regex syntax anyway
20:20 b_jonas how about the seventh then?
20:20 b_jonas print"$_\Q$_\E\"\n"for"print\"\$_\\Q\$_\\E\\\"\\n\"for\""
20:20 b_jonas or the eighth, which is really similar
20:21 TimToady no \Q
20:21 b_jonas makes sense
20:21 b_jonas no point for it really
20:21 [1]Limbic_Region joined #perl6
20:23 Confield joined #perl6
20:26 b_jonas can you use alphabetics as quote delimiters in perl 6?
20:27 TimToady no
20:27 moritz_ no?
20:28 b_jonas am wondering on http://www.perlmonks.com/?node_id=453519
20:28 b_jonas with strictures off
20:28 moritz_ TimToady: STD.pm parses m foof just fine
20:29 moritz_ TimToady: and so does rakudo
20:29 b_jonas no wait, there's no 'no strict subs' in p6 so it can't work
20:29 japhb pmichaud, did you get a chance to look at the NQP section of http://gitorious.org/parrot-plumage/parrot-plumage/blobs/master/TODO ?
20:31 hercynium joined #perl6
20:32 TimToady moritz_: well, maybe we relaxed that at some point
20:33 pmichaud japhb: not yet I hadn't, no.  I'll look at it after I finish the email I'm currently writing.
20:33 japhb pmichaud, OK, thanks!
20:37 TimToady moritz_: I was thinking of S05:1775
20:38 b_jonas http://www.perlmonks.com/?node_id=438050 doesn't work because it uses $_[$n] to access the nth positional parameter
20:38 TimToady which STD also violates
20:39 moritz_ TimToady: if it's something more general than regexes it should be in S02, really
20:39 TimToady agreed
20:39 b_jonas uh huh
20:39 b_jonas "You may not use whitespace or alphanumerics for delimiters.  "
20:39 b_jonas thanks
20:40 TimToady I knew it was in there somewhere though :)
20:41 b_jonas none of my obfus work then I think
20:41 b_jonas unsurprising since perl 5.10 broke like three of them as well
20:42 b_jonas am I too version-dependent in my japhs?
20:52 Limbic_Region joined #perl6
21:07 Whiteknight joined #perl6
21:24 japhb joined #perl6
21:30 pmichaud okay, done with email (finally)
21:31 pmichaud japhb: Q:PIR{...}  should be working.  If it's not working, feel free to submit a bug report.
21:31 japhb pmichaud, OK, will do.  Which bug address?
21:31 japhb (Or trac?)
21:31 pmichaud trac
21:31 japhb k
21:32 pmichaud I don't understand  "make Glue.pir a 'standard option' for NQP"
21:32 japhb pmichaud, Oh, sorry, that referred to a discussion with jnthn
21:32 japhb We were basically saying, we agree that NQP should stay clean, with minimal builtins.
21:33 japhb However, there is a use case for people to get at least the glue builtins, like run() ... and we thought shipping a "standard" PIR file with these builtins would be useful.  That way, anyone that wanted them could just load_bytecode(), and have them ready.
21:34 pmichaud that sounds to me like a Parrot library, not an NQP one.
21:34 japhb pmichaud, I referred to it as an NQP library because my implementation was pretty much a port of PIR from Rakudo to NQP.
21:34 pmichaud sure.
21:34 japhb But it matters not to me, that was merely a suggestion.
21:35 pmichaud NQP should be able to load libraries, yes.  :-)
21:35 pmichaud it's likely that NQP will get its own "use" syntax handler as well
21:35 pmichaud I had been waiting to see how Perl 6 and Rakudo fleshed those out before committing anything to NQP
21:36 japhb Right, and it does, through the existing load_bytecode() (plus Tene's upcoming use support).  I was just suggesting adding a library of common builtins to the supported set ... whether with NQP or with Parrot ... *shrug*
21:36 japhb nodnod
21:36 pmichaud last item:  I agree that NQP should make it possible to declare class attributes
21:36 pmichaud it probably won't be as extensive as Perl 6's declarations -- for example, we might only support declaration of private attributes
21:37 japhb Makes sense to me, fits the minimalist design.
21:37 jnthn I could use thsoe too.
21:38 pmichaud (library of common builtins)  I'm all in favor of libraries of common builtins; I just don't think that NQP wants to be defining them.  :-)
21:38 pmichaud NQP might very well be used to create them, but it shouldn't define them :-)
21:39 mzedeler joined #perl6
21:39 mzedeler Hi
21:40 mzedeler I am preparing a posting to the Perl 6 mailing list and just want to check that this subject hasn't been discussed before.
21:40 pmichaud what's the subject?
21:40 jnthn "IO"
21:40 jnthn ;-)
21:40 mzedeler Ranges and over-countable domains
21:40 japhb pmichaud, I just realized the other reason I was thinking of them as "NQP builtins" -- they include poking objects into the global namespace, with names that are from Perl 6.  See http://gitorious.org/parrot-plumage/parrot-plumage/blobs/master/Glue.pir .
21:40 pmichaud Ranges have undoubtedly been discussed before, although not always in complete detail.
21:41 mzedeler 1.0001 .. 1.01 doesn't really make much sense without specifying some step sizve.
21:41 pmichaud japhb: afaik, there's not really a "global namespace"
21:41 PerlJam mzedeler: sure it does; the step size is 1
21:41 mzedeler And with 1 .. 2i ?
21:41 japhb pmichaud, I'm sorry, I'm very tired and being imprecise.  .namespace []
21:41 pmichaud mzedeler: I can certainly talk about  1.0001 .. 1.01   as a continuous range
21:42 mikehh rakudo (7512462) builds on parrot r40764, make test / make spectest (up to 28056) PASS - Ubuntu 9.04 amd64 (g++)
21:42 mzedeler pmichaud: yes - as interval definition, but expanding to a list...?
21:42 pmichaud japhb: okay... I don't quite see how that affects NQP, though.  If a library wants to put things in the global namespace, that's fine.
21:43 jrtayloriv joined #perl6
21:43 moritz_ as a list it expands into (1.0001, )
21:43 pmichaud mzedeler: since the default interval is 1, that would be a list of one element.  That's fine.
21:43 japhb pmichaud, I think I have failure of brain.  Nevermind.
21:43 pmichaud Perl generally doesn't warn or prevent degenerate cases.
21:43 pmichaud japhb: no problem :-)
21:43 pmichaud japhb: my whole past week or so has been "failure of brain" mode.  :-)
21:43 japhb heh
21:44 mzedeler pmichaud: but you cant really compare 1 and 2i. Its like having a range from 'banana' to 3.14.
21:44 PerlJam Though mzedeler does have a point that 1.0001..1.01 as a list is just 1.0001, but 1.0006 ~~ 1.0001..1.01 is true.  That might seem strange to some people
21:44 mzedeler Pun intended.
21:44 pmichaud mzedeler: sure, that part of the specification is somewhat imprecise, I agree.
21:44 mzedeler PerlJam: Yes. Same point for character ranges.
21:45 moritz_ 1..2i will probably be the empty list, or fail()
21:45 pmichaud I'm also not entirely comfortable with the range specification for strings; I posted a message about it but I think it was warnocked.
21:45 mzedeler aaa is between aa and ba, but not when you use a Range.
21:45 PerlJam pmichaud: what are you unhappy about for string ranges?
21:45 mzedeler Same here.
21:45 moritz_ I think it's generally desirable to fail() if start- and endpoint are of different type, or not comparable
21:45 b_jonas perl5 does have a range from 'banana' to '3.14' and it's empty, but 'a' to '3.14' has size 26**4 I think
21:46 pmichaud PerlJam: parts of string ranges are very unintuitive and have bizarre effects
21:46 b_jonas hmm, wait
21:46 nbrown_ joined #perl6
21:46 b_jonas strange, "a"..3.14 and "a".."3.14" give different results in perl5
21:46 mzedeler b_jonas: I am talking about a range from the Str 'banana' to the Num 3.14.
21:46 b_jonas sure, it doesn't make sense
21:47 jrtayloriv I'm getting the following when I run 'make spectest': http://pastebin.com/d5530b6c2 ... is this anything I should be concerned about / post a bug report on?
21:47 moritz_ jrtayloriv: is that the 2009-08 release of rakudo?
21:47 mzedeler Sorry for stirring up things. I'll post my rants at the mailing list right now. Its already done. It sounds like it is appropriate.
21:47 pmichaud jrtayloriv: the autoincrement-range.t tests aren't expected...  the other two are.
21:47 PerlJam jrtayloriv: known bugginess
21:47 moritz_ heh, three answers, all of them different
21:48 jrtayloriv moritz_: :)
21:48 jrtayloriv moritz_: I just did git pull today (an hour or so ago)
21:48 moritz_ the problem in autoincrement-range.t seems that libicu is not available
21:48 japhb Brilliant.  Last night I couldn't get Q:PIR{} to work.  Today I can't make it fail.
21:48 pmichaud if so, then autoincrement-range.t needs to be marked # icu in spectest.data
21:48 japhb I think I'm going to bed and pretend this afternoon doesn't exist.
21:49 mzedeler Bye for now.
21:50 moritz_ jrtayloriv: what does './parrot_install/bin/parrot_config has_icu' from within the rakudo dir say?
21:50 jrtayloriv 0
21:51 moritz_ jrtayloriv: thanks
21:51 ruoso joined #perl6
21:52 moritz_ jrtayloriv: ok, I committed a "fix" (which is not to run that test if the ICU library is not available)
21:52 jrtayloriv moritz_: thank you
21:52 PerlJam pmichaud: oh, while I'm thinking of it again ...  www.rakudo.org has many blank pages for me.  Do you know why or how to fix ot?
21:52 PerlJam s/ot/it/
21:53 moritz_ it seems to be related to the presence of cookies
21:53 dalek rakudo: 9d9d416 | moritz++ | t/spectest.data:
21:53 dalek rakudo: [t/spectest.data] autoincrement-range.t needs the # icu flag, jrtayloriv++
21:53 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/9d9d41675dab91fbeb6dcf29998210e00baecf97
21:53 moritz_ when I delete all cookies from rakudo.org the next page load works fine
21:53 * PerlJam t ries it
21:54 PerlJam nope.
21:54 PerlJam http://www.rakudo.org/documentation is still blank
21:55 PerlJam I also get a blank page when I try to login.
21:55 PerlJam But http://www.rakudo.org/how-to-get-rakudo and http://www.rakudo.org/how-to-help show up fine.
21:55 PerlJam (maybe someone enabled a rule in drupal to only show pages that start with "how"  :-)
21:56 moritz_ ;-)
21:56 moritz_ let's try "why" too
21:57 jrtayloriv I am having the same problem -- http://www.rakudo.org/community is blank for me
21:57 moritz_ bah.
21:59 PerlJam Is it possible to get a login on the box that hosts rakudo.org?
22:00 jaffa8 I think
22:01 jaffa8 rakudo should have been written in perl
22:01 moritz_ it is, mostly
22:01 jaffa8 what else?
22:01 PerlJam more and more of it is every day in fact
22:01 moritz_ PIR, C
22:01 jaffa8 I mean perl 5
22:01 PerlJam jaffa8: see STD
22:02 jaffa8 what for?
22:02 moritz_ well, Perl 6 is much more suited for writing a compiler than Perl 5
22:02 moritz_ we've had various attempts at writing Perl 6 is Perl 5
22:02 jaffa8 in?
22:02 carlin v6.pm
22:02 moritz_ s/is/in/
22:02 jaffa8 is=in?
22:02 PerlJam yeah, and they've been *so* successful!  :)
22:02 moritz_ kp6, v6.pm
22:02 moritz_ for example
22:02 jaffa8 who?
22:03 moritz_ they all suffered from the same problem
22:03 jaffa8 Which is?
22:03 moritz_ jaffa8: I'm about to explain, but explaining takes longer than asking questions
22:03 moritz_ so... either they implemented only a small part of Perl 6 which coudl be directly translated to Perl 5
22:04 moritz_ or they became horribly slow.
22:04 moritz_ because they had to add layer upon layer on perl 5
22:04 jaffa8 What parser have they used?
22:04 Tene pmichaud: ping
22:04 moritz_ and then perl 5 was just not fast enough
22:04 moritz_ custom parsers
22:04 moritz_ but I'm mostly talking about execution speed
22:04 Tene pmichaud: japhb asked me to add support for 'use' and 'eval' to NQP.  Do you have any inclination to comment on what you'd like to see there before I write and commit it?
22:05 jaffa8 Of what?
22:05 moritz_ of these implementations.
22:05 PerlJam jaffa8: you're welcome to try to write a perl 6 compiler in perl 5 if you want.
22:06 PerlJam jaffa8: the spec is more mature now than when those others tried.  That's got to help a little  :)
22:06 PerlJam and you've got several other implementations to crib ideas from if you get stuck on how to handle something
22:06 pmichaud Tene: I want them to continue the minimal nature of NQP -- i.e., to provide essentially the semantics that HLLCompiler is providing
22:07 * Tene nods.
22:08 pmichaud I expect to be improving HLLCompiler's eval capability to resolve the REPL issues.
22:08 Tene Do you want me to send you whatever patch I write before committing it, to make sure I don't add too much?
22:08 pmichaud Tene: if you send me a patch beforehand I'll be glad to review.  If you just want to commit, I might revert.
22:08 pmichaud (but I'm fine with that if you are)
22:08 pmichaud I'd say forgiveness better than permission here :-)
22:08 jaffa8 I have used Perl 5 for compilation purposes.
22:08 Tene I rather prefer that you just revert, yes.
22:08 jaffa8 I used Parse:Yapp
22:08 pmichaud Let's do that then.
22:09 moritz_ jaffa8: so you already have experience. That's great, and will make success more probable
22:09 Tene i generally assume that if I commit something stupid, it'll just get reverted and I'll get decremented, as that's how I'd prefer anyone else contributed to something I cared about, except in exceptional cases.
22:09 pmichaud Tene: that's fine with me.
22:09 pmichaud I also wouldn't want it to block on my review.
22:10 pmichaud however, I'm now done with trips for the year, and the kids are back in school, and I don't have anyone coming to visit, so I _hope_ I'll have a ton more time than I've had for the past several months
22:11 * Tene schedules a visit to pmichaud's roof.
22:11 pmichaud yes, if anyone wants to come visit, we can likely get some extensive hacking done :-)
22:11 pmichaud and I do have a spare bedroom
22:11 PerlJam pmichaud: somehow I doubt that.
22:12 pmichaud PerlJam: oh?  why's that?
22:12 Tene I'm wondering if I should try to arrange hacking with allison and particle here in seattle.  We've already arranged for dinner...
22:12 PerlJam pm: for me, only if the wife and kids don't come. :)
22:12 pmichaud oooh, that's right, the members meeting is tomorrow.  I should see if I'm authorized to vote yet.
22:13 jaffa8 pmichaud, where do you live?
22:13 pmichaud jaffa8: Plano, Texas   (just north of Dallas)
22:13 jaffa8 Is it windy?
22:14 pmichaud sometimes, yes, but generally no.
22:14 jaffa8 Is it hot?
22:14 Tene Taking time off work to visit you to help work on the PGE rewrite/refactor/rewhatever is actually an extremely attractive idea to me.
22:14 pmichaud Tene: that could work, definitely.
22:14 pmichaud if you can get as far as a dallas airport, I can get you the rest of the way :)
22:15 jaffa8 I cannot imagine you are serious.
22:15 Tene Why not?
22:15 jaffa8 Group hacking?
22:15 jaffa8 Does that work?
22:15 pmichaud sure.  f2f work is often very productive (more)
22:15 jaffa8 What is f2f?
22:15 pmichaud jnthn++ and I are extremely productive when we're in the same room working on things
22:15 Tene face-to-face
22:15 PerlJam jaffa8: have you ever heard of pair programming?
22:16 jaffa8 yes
22:16 PerlJam jaffa8: have you ever done it?
22:16 Tene I'd have to get time off work... and I don't have any savings to speak of right now.
22:16 jaffa8 yes
22:16 pmichaud Tene: we'll keep it in mind over the next couple of months.  We might be able to work something out.
22:16 jaffa8 What would you do in pair?
22:17 pmichaud many times it's easier to explain or explore design issues in person in front of a whiteboard than to try to do it via email or irc
22:17 jnthn Yes, face to face meetups are very useful.
22:17 pmichaud irc is relatively low-bandwidth
22:17 jaffa8 how long does it last?
22:17 PerlJam real-life-bandwidth++
22:17 jaffa8 Such a cooperation
22:18 jaffa8 1 day
22:18 jaffa8 ?
22:18 jaffa8 1 week?
22:18 pmichaud depends
22:18 jnthn Depends. 1 day is generally too short though.
22:18 pmichaud jnthn and I have managed to make productive use of two or three days easily
22:18 jnthn I mean, the first 2 hours I've not had enough coffee to do anything useful anyway.
22:18 PerlJam pm: Wasn't there a Dallas.p6m hackathon recently?
22:18 jnthn ;-)
22:18 pmichaud and I found the parrot developers summit last november (2 days) to be extremely productive
22:18 pmichaud yes, we're doing a Dallas.p6m this saturday
22:19 pmichaud we're doing about 2-3 hours for that
22:19 jnthn Two or three days normally leaves me feeling "we got lots done, but we could have done more tomorrow too" :-)
22:19 jaffa8 you could do pair programming with a suitable software
22:19 PerlJam jnthn: I don't see Dallas on your travel map  :)
22:19 pmichaud anyway, if Tene++ arrived here, I have no doubt we would whip out a bunch of interoperability code and resolve several big design issues
22:20 PerlJam jnthn: we've used gobby to good effect here (even with the limited bandwidth)
22:20 PerlJam er, s/jnthn/jaffa8/
22:20 pmichaud if seattle weren't so far away I'd even have considered attending there :-)
22:20 jnthn PerlJam: Dallas is a little far for me to come.
22:20 * pmichaud gets his parrot board votes in.
22:21 pmichaud PerlJam: I much prefer to meet with jnthn++ in some locale other than Dallas :-)
22:21 PerlJam jnthn: All you have to do is find some client that'll pay your way nearby and use that as an excuse to visit Pm for a few days  :)
22:21 PerlJam pmichaud: even CRP?  ;)
22:21 pmichaud sure, even CRP
22:21 jaffa8 Dallas is 8 hours from here.
22:21 pmichaud jaffa8: eight hours driving or flying?
22:21 jnthn Is CRP missing a vowel?
22:21 jaffa8 flying
22:22 jnthn .oO( oh my, I wonder which city I just offended )
22:22 pmichaud CRP = airport code for "Corpus Christi, Texas"
22:22 PerlJam jaffa8: you and I are roughly within the same time locus of Dallas,  though I'd be driving
22:22 pmichaud PerlJam: I'll likely be in CRP sometime in Sep, btw.  :-)
22:22 sri_kraih joined #perl6
22:23 PerlJam pmichaud: if it's the week of the 14th, I'll be in Vegas
22:23 pmichaud ...vegas?
22:23 jnthn Vegas would be a terrible place for a hackathon.
22:23 jnthn Since what happens in Vegas stays in Vegas, nothing would get committed at all.
22:23 pmichaud heh
22:23 PerlJam pmichaud: Rizzo and I are attending a "project management" course there.
22:23 pmichaud PerlJam: suuuuuuuure you are
22:23 tak11 joined #perl6
22:23 jnthn "project"
22:23 jnthn ;-)
22:23 pmichaud "project management" sounds like a euphemism to me
22:24 PerlJam pm: Rizzo set it up, so you can guess his motivation  :)
22:24 pmichaud pj:  okay, I think I nailed it on my first comment then :)
22:24 pmichaud where are you staying?
22:25 PerlJam in fact, the same course is offered at different times in other cities and he argued convincingly that we should go to the one in Nevada.
22:25 pmichaud hey, I'm "managing a project"!  Maybe I should go also?
22:25 PerlJam I actually am not sure where we're staying.
22:25 pmichaud I wonder if CBI/HRI would pay for me to go... :-P
22:26 PerlJam I think it's the Flamingo
22:26 pmichaud we could do project management course by day, and then "hacking" at night.  After all, there's nothing else to do after 5pm in vegas.... :-P
22:26 PerlJam heh
22:27 pmichaud afk for a bit
22:28 * PerlJam saw "The Hangover" 2 nights ago, so Las Vegas has a different brain-feel than it would have otherwise.
22:33 carlin is anyone able to grant commit access to the pugs svn? I have some small changes I want to make to the docs, I was going to email the patches but they're fairly trivial so it doesn't seem worth bothering others with if I can do it myself :)
22:34 jnthn carlin: yes
22:34 moritz_ carlin: just /msg me or jnthn your email address and desired nickname
22:34 jnthn carlin: pm me your email address and desired...
22:35 jnthn If you want comedy, send it to both of us and we'll race.
22:35 moritz_ lol
22:35 moritz_ how neglecting of us not to offer you a commit bit earlier
22:36 carlin I would have taken you up on the offer to have a race but I couldn't send both at the same time so it would be unfair
22:36 mikehh joined #perl6
22:36 jnthn carlin: added you
22:37 mikehh joined #perl6
22:37 moritz_ carlin: it's a custom to add yourself to the AUTHORS file in your first commit
22:37 jnthn carlin: So you'll have password in your email.
22:38 moritz_ carlin: and you can also invite new commiters if you log in to http://commitbit.pugscode.org/admin/project/Pugs/people with your svn password
22:39 carlin viral commit access, nice :)
22:39 moritz_ aye ;-)
22:39 PerlJam carlin: it's an exclusive club ... just look at the list :)
22:40 moritz_ gvanrossum: <guido@python.org> (author, pending)
22:40 moritz_ ("pending" means he never actually comitted something)
22:40 dew joined #perl6
22:42 s1n moritz_: i never added myself to AUTHORS in pugs or CREDITS in rakudo lol
22:42 moritz_ s1n: you should have
22:42 s1n moritz_: i should indeed, maybe i should do that before the hackathon this weekend
22:44 s1n moritz_: i haven't made any patches for rakudo since the switchover though :(
22:45 moritz_ s1n: you should change that ;-)
22:47 s1n moritz_: i haven't been able to keep up, i work best if i have specific tasks to do, fishing for things to work on when i don't know what needs done or know the code that well is tough (for me anyways)
22:49 s1n moritz_: know of anything that needs work?
22:49 moritz_ s1n: t/spec/TODO in the pugs repo is full of stuff that needs work (not Rakudo stuff, I know)
22:50 moritz_ if you want to work on Rakudo, I have a nice idea to speed up the build process
22:50 moritz_ currently the setting is concatenated into one huge Perl 6 file, and then compiled to PIR
22:51 moritz_ it would be more efficient (and easier to parallelize) if all the files would be compiled separately
22:51 moritz_ and the PIR concatenated then
22:51 s1n can you concat PIR files like that?
22:52 s1n and why do you think that would speed things up?
22:52 moritz_ because larger files take over-proportionally long
22:52 moritz_ because they contain growing data structures like symbol tables etc.
22:53 moritz_ and yes, you can concatenate PIR like that
22:53 moritz_ that's how it's done for all the src/classes/*.pir etc.
22:55 s1n jnthn: here's my shortcut, this seems to be a nice stop-gap measure
22:55 s1n jnthn: http://gist.github.com/174282
22:55 moritz_ gotta sleep now, TTFN folks
22:57 alester joined #perl6
22:57 jnthn TTFN?
23:00 carlin ta ta for now
23:05 jnthn s1n: Ah, it's a step on the way to working out how to make %*ENV work. :-)
23:05 jnthn carlin: ah, thanks :-)
23:05 s1n jnthn: yes, but rakudo has no "temporary hacks" area
23:06 jnthn s1n: I'm not really inclined to add a bunch of functions that people will start using, only to have to deprecate them later.
23:06 s1n jnthn: it does indeed call setenv, i stopped it in gdb :)
23:06 s1n jnthn: yes, i understand, but i thought i'd point it out and mention that it doesn't really belong yet
23:07 s1n jnthn: there are no functions in the spec similar to the POSIX function, so it doesn't really belong anywhere
23:10 jnthn s1n: *nod*
23:10 jnthn I think getting %*ENV fixed up is probably the way to go.
23:11 s1n jnthn: i agree, but that'll have to wait until i have a weekend free :)
23:11 jnthn I may try and get Proxy objects done soon-ish so you can do that.
23:11 jnthn (If you want to do that. :-))
23:11 s1n jnthn: honestly, i'd rather do it once, than have to reimplement
23:12 jnthn s1n: Aye.
23:12 jnthn Then I'd wait for Proxy objects.
23:12 s1n (if the needed bits aren't _huge_ that is)
23:14 jnthn TimToady: Is this bit in S06 still correct?
23:14 jnthn set_name $obj: "Sam";   # try $obj.set_name("Sam") first, then
23:14 jnthn # fall-back to set_name($obj, "Sam")
23:15 jnthn (I know a lot of fall-back stuff went away, I'm not sure if this counts.)
23:17 REPLeffect joined #perl6
23:20 donaldh joined #perl6
23:30 Tene japhb: I'm just about done at work for today.  Depending on what the gf has planned for tonight, I'll either get stuff done soon or late tonight before bed.
23:34 jnthn std: ::Everything but KitchenSink
23:34 jnthn aww, no evalbot :-)
23:34 * jnthn just found where in the spec KitchenSink is defined.
23:41 s1n phenny: tell masak did you forget to commit some stuff in Web.pm? i just pulled the tests and either i am testing wrong or they are failing; for example, it can't find Web::Happle
23:41 phenny s1n: I'll pass that on when masak is around.
23:44 nbrown joined #perl6
23:55 payload joined #perl6
23:58 * jnthn just finished a re-read of S06.
23:58 jnthn My word, there's some stuff in there.

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

Perl 6 | Reference Documentation | Rakudo