»ö« | perl6-projects.org/ | nopaste: paste.lisp.org/new/perl6 | evalbot: perl6: say 3;' | irclog: irc.pugscode.org/ | UTF-8 is our friend!
Set by moritz_ on 27 June 2009.
00:00 lumi_ joined 00:06 justatheory joined 00:07 justatheory left 00:19 TimToady joined 00:23 rhr_ joined 00:25 justatheory joined 00:27 justatheory left 00:34 justatheory joined 00:36 rhr left, asciiville left 00:59 shinobi-cl left, alester_ joined 01:11 orafu left, orafu joined 01:13 shinobi-cl joined 01:23 Whiteknight left 01:45 xomas_ left 01:46 xinming_ joined 01:47 shinobi-cl left 01:52 xomas_ joined 01:53 missingthepoint joined 01:55 alester_ left 02:00 xinming left 02:03 wayland76 joined 02:04 dKingston left 02:08 lucs left 02:12 flexibeast joined 02:13 explorer left 02:22 alester_ joined 02:33 molaf joined 02:36 cognominal left 02:45 meppl left 02:50 jauaor joined 03:04 sri_kraih joined 03:05 hoelzro joined 03:17 hoelzro left 03:20 donaldh left, donaldh joined 03:24 cotto left 03:26 hoelzro joined 03:29 dukeleto joined 03:38 hoelzro left 03:42 hoelzro joined, hoelzro left 03:43 molaf left 03:45 hoelzro joined 03:59 synth joined 04:04 hoelzro left 04:08 dukeleto left 04:18 hoelzro joined 04:19 hoelzro left 04:20 synthEEEE joined 04:21 synth^3 joined 04:24 ihrd joined 04:25 nihiliad left 04:30 dukeleto joined
dukeleto hola 04:30
04:30 dukeleto left
wayland76 phenny: ? 04:39
04:41 synth left, synth joined
wayland76 halo ? 04:43
:)
phenny: tell pmichaud I've found some problems with the "install" section of the ins2 makefile. I hope to have a patch by the end of (my) day 04:44
phenny wayland76: I'll pass that on when pmichaud is around.
04:44 synthEEEE left 04:49 synth^3 left 05:03 kane___ joined 05:13 szabgab joined 05:17 kane_ left 05:23 arejay left, orafu left, Patterner left, rjh left, c1sung left, antiphase left, shachaf left, eternaleye left, justatheory left, Aisling left, arthur-_ left 05:28 Aisling joined, arthur-_ joined, orafu joined, Patterner joined, arejay joined, eternaleye joined, shachaf joined, rjh joined, c1sung joined, antiphase joined 05:39 cotto joined
eternaleye (way, way, _way_ backlogging) re the discussion over dies_ok and specific error types, why not use junctions? none(X::Internal) would be the 'anything but Null PMC' equivalent, with any() also DWIM-ing, possibly with a multi that accepts a single type for the 'simple case' 05:47
It would even be compatible with using smartmatch internally to do the comparison 05:49
05:51 synthEEEE joined 05:58 payload left 06:03 synth left 06:10 FurnaceBoy left 06:15 synth joined, rfordinal left, rfordinal joined 06:21 synth left
moritz_ eternaleye: the real problem is that we need to define the things like X::Internal first, and what they are etc 06:21
06:23 szabgab left 06:24 kane___ left, synthEEEE left 06:32 alester_ left, japhb_ joined 06:33 Iqbal joined, Iqbal left 06:34 rfordinal left 06:37 japhb left 06:48 agentzh left 06:49 agentzh joined
wayland76 phenny: tell pmichaud Ok, patch for ins2 branch added to rt.perl.org/rt3/Ticket/Display.html?id=63360 06:50
phenny wayland76: I'll pass that on when pmichaud is around.
06:55 iblechbot joined 07:01 kane_ joined
missingthepoint newbie question: 1. how do i tell rakudo to load libicu? 07:01
2. why do i need to? shouldn't it load by default if it's available? 07:02
moritz_ you have to tell parrot's Configure.pl to use it
but it should be used by default.
if available
07:02 snearch joined
missingthepoint hmm 07:02
wayland76 missingthepoint: What distro/OS? 07:06
missingthepoint is there a rakudo make target that cleans out the built parrot?
('realclean' doesn't) 07:07
wayland76: the os of maximal perversity
wayland76 What, you mean you're trying to run Windows on top of AmigaOS? 07:08
Or are we just talking about plain, ordinary perversity, such as using Windows?
missingthepoint plain, ordinary perversity, which is plenty for me today
moritz_ missingthepoint: delete the parrot/ dir, and you're done 07:09
missingthepoint moritz_: should that be included in any of the existing make targets, or is it done v. infrequently? 07:10
i guess it doesn't matter if rakudo moves to building against system parrot anyways
moritz_ right; I rarely do it
wayland76 missingthepoint: just what I was going to say
moritz_ (and on my system parrot is just a symlink to a git-svn checkout) 07:11
wayland76 I'm hoping my patch to the ins2 branch will clear up most of the remaining problems
missingthepoint can you do symlinks anyhow on windows xp?
missingthepoint pines for a real OS
wayland76 missingthepoint: I thought shortcuts were equivalent, but not sure 07:12
missingthepoint wayland76: nah, they're, well... perverse 07:13
Unix fs symlinks are kinda the simplest possible thing that could work... but they also turn out to be the best IMHO 07:14
moritz_ missingthepoint: I'm not on PerverseOS 07:16
missingthepoint moritz_: i envy you. unfortunately work and uni stuff mean Windows Xtremely Pervers is my easiest option
07:18 Su-Shee joined
Su-Shee good morning. :) 07:18
moritz_ good morning
missingthepoint: I feel with you.
missingthepoint good morning, Su-Shee :)
07:20 donaldh left, dakkar joined
wayland76 missingthepoint: I've been Linux-only since I stopped using Windows 98, at least at home 07:20
I used Win2K at work for a while, but that's it 07:21
07:21 donaldh joined
wayland76 So XP and Vista are things I've used only on others' computers 07:21
missingthepoint wayland76: i'd be *nix-only in a jiffy, but emulated Photoshop <censored> 07:23
wayland76 What about Mac OS? Or is that too $$$?
missingthepoint honestly, haven't investigated much. that could even be ideal - OS X and up are beautiful within and without 07:24
07:25 rfordinal joined
wayland76 I know a guy whose job is Windows Server, and at home, he's using Mac and Linux 07:26
He was fine with Windows, until Vista came along
missingthepoint heh
Su-Shee It's coming in handy not gaming at all.. I'm Linux/BSD for 14 years now. ;)
missingthepoint high-fives Su-Shee 07:27
wayland76 I play games -- web-based games :)
missingthepoint wishes he were stronger 07:28
wayland76 missingthepoint: exercise works better for that than wishes :)
Su-Shee missingthepoint: I'm very passionate about open source and open access, so I use it.
it's somewhat like the first waves of treehuggers or green parties: until "being green" was convenient, you had to carry your empty bottle a kilometre to the next container... 07:29
eternaleye missingthepoint: Well, Linux ought to have PCI passthrough support for KVM virtual machines before too long, which means accelerated graphics in VMs. And someone already got OSX working in a VM
And Windows in a VM already works fine 07:30
missingthepoint Su-Shee: in my case, "carrying the bottle a kilometre" means not drinking (unless Gimp can handle PSD files)
eternaleye missingthepoint: It can handle some of them 07:31
Su-Shee PSD?
eternaleye partial support
PhotoShop Drawing
missingthepoint photoshop documents
07:31 ihrd left
wayland76 And with all the money you save by not buying Windows, you can get better hardware :) 07:31
missingthepoint (drawing) :)
eternaleye missingthepoint: It was a guess anyway
Su-Shee ah
07:32 szabgab joined
missingthepoint eternaleye: maybe i can switch soon. i would like that very much... :) 07:32
07:33 zamolxes_ left
Su-Shee gimp somewhat sucks in terms of usability, but it's rather powerful. 07:34
missingthepoint Su-Shee: i think you just described a lot of open source stuff ;)
(though that has improved a LOT since i started using Linux)
Su-Shee I decided to focus on the "powerful" thing.. :) 07:35
missingthepoint good choice. i reckon it's easier to customize something powerful yourself... 07:36
than to wrestle with a deficient foundation.
missingthepoint peers at windows accusingly 07:37
Su-Shee well I was somewhat lucky in a strange way, because I had to learn Linux the hard way in 1995, so I was never pampered by any attempt to make Linux read for the desktop or easier to use anyway.
ready.
missingthepoint Su-Shee: somewhat similar, though a little later (1999) 07:38
Su-Shee I annoyed 10 people for 3 month with extremely stupid questions and that was it. :)
missingthepoint heh... i didn't annoy anyone, but learned very slowly as a result 07:39
i wish i'd had people to annoy, or at least someone to say "read man pages, learn shell scripting, then learn C"
Su-Shee well I wasn't really faster that way. ;) 07:40
well I ignored shell and C and started with Perl.
because it was the programming language having a somewhat userfriendly book. ;)
(and not some academic computer science thing..)
missingthepoint TimToady++ :)
moritz_ Su-Shee: "learning" or "programming" perl? 07:41
wayland76 The power vs. user-friendly thing depends on how much you use it
Su-Shee moritz_: neither. the pink perl 4 book
missingthepoint thinks linguistics + programming languages is a happy marriage
wayland76 I only use gimp once or twice a year, and never approach its full power
moritz_ Su-Shee: never heard of that; that's way before my programming time :-) 07:42
Su-Shee missingthepoint: I can honestly say, that I probably wouldn't have gotten into programming if it wasn't for Perl and the kind of documentation it offered.
wayland76 I actually started as a Solaris user on my Uni's machines, also in 1995
szabgab can someone point me to where is anything regarding sockets ? specifically lwp like things ?
wayland76 But it wasn't until I got the hang of the admin side of things that I could ditch WIndows
Su-Shee moritz_: I bought it more or less when 5 came out, so I got the blue "programming perl" next.
wayland76 ...and until there was a decent package manager :)
moritz_ szabgab: t/spec/S32-io/IO-Socket-INET.t tests rakudo's sockets 07:43
szabgab ty
do you know what happened to the LWP::Simple of Cosimo ?
wayland76 I got into Perl just as Perl 5 was getting widespeard, so I was mostly able to avoid P4 07:44
Su-Shee szabgab: there was a p6 lwp?
szabgab Su-Shee, use.perl.org/~cosimo/journal/38826
eternaleye (way, way, _way_ backlogging still) Khisanth: re parsing HTML being ugly: why? Just use single quotes: `rule tag {'<' $<name>=(\w)[<.ws><attribute>]* [['>' [ <content> | <tag> ] '</' $<name> '>'] | '/>'] }` looks pretty readable to me
szabgab he implemented it while I was teaching Perl 6 to TimToady ;-)
Su-Shee szabgab: totally missed it. 07:45
szabgab: and I commented 2.5 hours yesterday. ;)
szabgab yes, thanks
moritz_ I don't think the source has mad it to any repostiory yet :(
szabgab cosimo, ping ping , what happened to LWP ?
07:46 frettled left
Su-Shee moritz_: I really asking myself how much p6 we're missing. 07:47
missingthepoint Su-Shee: I think I would have gotten into programming regardless, I waded through a C++ for dummies book to teach myself C 07:48
Su-Shee: but Perl 5 was a special treat ;)
moritz_ Su-Shee: you mean there's a Perl 6 DarkPAN? :-)
Su-Shee I'm sure of it. :) 07:50
missingthepoint: *shiver* :)
missingthepoint Su-Shee: you don't know the half of it.
07:51 cognominal joined
Su-Shee missingthepoint: I really never got friends with C. I can handle it and I like to do more for Arduino and X stuff, but that's it. 07:51
moritz_: did you release all experiements in p6 you made? 07:52
Khisanth eternaleye: heh well generally the main issue with html isn't parsing perfectly well-formed documents ...
07:52 cognominal left
moritz_ Su-Shee: surely not, but I deleted nearlly all of them sooner or later 07:53
or blogged them :-)
or sent to RT :-)
missingthepoint Su-Shee: C was my first love. I don't use it much, but it was my first language...
eternaleye Khisanth: Well, in that case it's more of an 'ugly in, ugly out' situation ( or, 'ugly in, ugly parse' ;D ) 07:54
Su-Shee moritz_: ah yes. the deleted Perl 6 part is also very large on my hd. :)
missingthepoint: mine the .. 4th or so..
missingthepoint thinks he'll keep his Perl6 learning attempts for later lulz
Su-Shee if rakudo really is released in spring '10, I think I have to actually put something out there. 07:55
moritz_ Su-Shee: I hope you know that it's not a "Perl 6 is done" release, but rather a "we've got so far yet, check it out" release 07:57
Juerd_ expects Perl 6 to be done by the first christmas after spring then :) 07:58
Su-Shee moritz_: the more additional written code is available, the easier people will adopt it, I expect.
07:58 masak joined
Su-Shee moritz_: and be it only to illustrate how to handle perl 6. 07:59
missingthepoint are there plans to port bits of CPAN to P6? beyond Web.pm?
moritz_ missingthepoint: Tim Bunce is planning to port DBI to parrot
missingthepoint (and the nebulous LWP::Simple)
awesome! 08:00
moritz_ and I volunteered to add a Perl 6 layer
Matt-W Morning
moritz_ oh, and there's an SVG module somewhere, a templating engine that masak works on etc.
Su-Shee missingthepoint: let me put it that way: no matter how cool a programming language in itself might be - what use is it, if there is nothing like .. let's say SSL, database access, GUI bindings...
missingthepoint double awesome.
masak morn'
moritz_ speaking of the devil... :-)
Matt-W missingthepoint: We intend that there will be libraries. We're hoping we won't have to write them all though :)
missingthepoint Su-Shee: yes. bindings matter. hence the massive CPAN win. (that's what attracted me to Perl initially, actually) 08:01
Matt-W masak: I liked your last Web.pm blog post
masak Matt-W: thank you. it was a bitch to write. :)
moritz_ missingthepoint: github.com/masak/proto/blob/96b06b0...jects.list
08:01 frettled joined, kst left
Su-Shee missingthepoint: I think the reddit switch from lisp to python because of the environment missing is a lesson. 08:02
08:02 kst joined
missingthepoint moritz_: thankyou, that is great! 08:04
masak I haven't backlogged, but the discussion you're having seems _very_ interesting.
08:04 explorer joined
masak providing a kick start for web-related libraries is one very important goal for Web.pm. 08:04
missingthepoint masak: what's missing in p6 web-related libraries right now? 08:05
masak missingthepoint: many things. I'd like Expat bindings, for example.
missingthepoint that's eminently sensible.
masak I'd like an XPath search library, and a CSS search library.
an HTML tagsoup parser. 08:06
most of these things are importable from the outside.
Su-Shee rss, atom, web services APIs
missingthepoint masak: you mean like HTML::Parser does currently, or something else?
and what do you mean "importable..."?
masak missingthepoint: haven't looked at HTML::Parser, but I guess yes.
missingthepoint: I mean I don't care if they're pure-Perl6 at this point.
missingthepoint ah, ok 08:07
cosimo szabgab: afaik, LWP::Simple it's working. It was last time I tried.
szabgab but where is it ?
cosimo szabgab: you can find it in the S32-settings lib, IINM
missingthepoint html parser docs <-- "We have tried to make it able to deal with the HTML that is actually 'out there'"
cosimo szabgab: hold on, i can send you a link
masak this weekend I had a guy called arthur-_ come and volunteer for writing Astaire, our Sinatra clone. that makes me very happy.
missingthepoint: yes, like that. 08:08
08:09 huf left
szabgab cosimo, it would be better as I cannot find it 08:10
Matt-W masak: Not that it means anything, but I considered volunteering, I just don't have time... need to do more work on Form with the tuits I have
masak Matt-W: not that it means anything either, but I'd love to have your tuits on the project. but I'd like more to see Form commits. :)
08:12 fridim_ joined
cosimo szabgab: ok, will send a link Soon(tm) 08:12
Su-Shee szabgab: here's another great example of perl totally being ignored... arstechnica.com/open-source/guides/...html-5.ars 08:13
szabgab Su-Shee, that is in Python 08:14
moritz_ it ignores all other languages too 08:15
Su-Shee szabgab: exactly my point.
szabgab: arstechnica generally makes its examples and hacks NOT in perl even though it would be perfectly possible. 08:16
szabgab Su-Shee, so now you have the platform to build it :-) over at #padre ..
Matt-W masak: you'd just like more of everything wouldn't you?
masak Matt-W: aye! keep it coming.
08:17 cognominal joined
szabgab Su-Shee, I don't know how arstechnica decides in which language to show soething, isn't that authors writing that they post? 08:18
so if somone wants more perl content there (5 or 6) she should write articles get it posted there 08:19
or at a minimum find out what is needed to post an article and blog about it
Su-Shee szabgab: ryan paul is either an editor anyway or the "I write the tutorials and examples" gui. the point I'm trying to make is that this is a very well received site and they do their examples not in Perl. if you read this, you wouldn't even guess that Perl has great GUI bindings.
szabgab: it's a perception thing par excellence. 08:20
szabgab thank you :-)
Su-Shee oeh.. guy I mean. :)) too much GUI here. :) 08:22
szabgab I have a working LWP::Simple now on my HD 08:30
but I'd better reuse and extend the one cosimo wrote 08:31
moritz_ inverse NIH syndrome? :-)
Su-Shee moritz_: this perlblog.org thing... can I have one?
moritz_ Su-Shee: sure, as soon as I have set up something working 08:32
Su-Shee moritz_: I see.. I wanted to avoid that I have to get something working on my webhost.. ;)
szabgab moritz_, HIN
Hate Inventing New 08:33
moritz_ Su-Shee: there seems to be a "move away from use.perl.org" wave that I want to catch. But first I have to get something up and running :/ 08:34
Su-Shee moritz_: don't let me pressure you, I'm just to lazy to install movable type on my webhoster.. 08:35
08:35 mj41_ left 08:36 huf joined
szabgab moritz_, written in perl 6 ? 08:42
moritz_ szabgab: I don't think I'll get something decent running in Perl 6 soon 08:43
considering that we don't have database access yet... 08:44
Matt-W And it's really slow... 08:49
moritz_ aye :/ 08:50
Su-Shee *click* *move* *movemore*
ah. great. I so much love all this blog-in-webhoster-package-mysql-customer-frontend stuff... 08:53
08:57 zamolxes joined 08:58 mj41 joined 08:59 agentzh left 09:00 agentzh joined 09:01 rfordinal left
eternaleye I just had an idea for a fun way to explain .wrap etc: make a variant of the 'stone soup' story, where you start with -> {;} and make something useful by adding stuff over time 09:07
masak eternaleye: sounds like a good theme for a blog post. :) 09:12
eternaleye: telling the story and interleaving Perl 6 code.
eternaleye ENOBLOG
moritz_ that can be changed. 09:13
masak indeed it can.
eternaleye If I had a blog, I might be expected to update it. And it would disappoint me (and maybe others) when I inevitably forgot.
masak eternaleye: you could guest-post at somebody else's blog. 09:14
eternaleye Mmm. Maybe
szabgab as ironman seems to be down let me spam you with my blog entry here:-) szabgab.com/blog/2009/07/1248597502.html 09:15
eternaleye moritz_: BTW, next time davef is in here, talk to him about database stuff. He's working on porting JDBC via autoconversion and elbow grease.
Su-Shee eternaleye: well there is a simple solution to that: don't pressure yourself and blog only if you have something to say you think is blogging worthwhile. who cares how often you update? 09:16
eternaleye: it's not a contest, it's content. ;)
eternaleye It's no DBI (though much like the proposed p6 DBDI), but it might be useful
Su-Shee: Well, yes, but there's also the fact that I may have taken the Perl Virtues a little too strongly to heart. Especially the one that starts with 'L'... 09:17
masak eternaleye: you're too lazy even to say the whole word? :P 09:18
eternaleye ;D
moritz_ L-ness driven to extremes :-) 09:19
eternaleye Laziness and OCD interact in an interesting manner. It's not that I'm excessively lazy, it's that I know I go OCD on my writing and it turns into actual work
masak hah, I just created a local Rakudo branch that parses Python-like for/else constructs! :)
eternaleye: I know the feeling. but I still like the results, so it's worth it most of the time. 09:20
if I get the for/else to DTRT too, I think I'll blog about this. :) 09:21
eternaleye masak: Nice
moritz_ masak: parse, or actually execute?
eternaleye s/$/!/
masak parse, thus far.
it's a trivial patch, actually: gist.github.com/156123 09:22
moritz_ masak: I expected nothing more complicated, actually :) 09:23
masak moritz_: me neither. still, it's nice to see that there were no hidden complexities. :)
there might be in the semantics, though.
moritz_ indeed 09:24
when is the else block run?
masak moritz_: when no if block inside the for loop was triggered. 09:25
so there's a hidden state variable there.
moritz_ oh.
masak one that belongs (dynamically) to that for loop.
I actually don't particularly like the feature in Python either. :P 09:26
but I thought I'd give a shot at implementing it, just to get a feel for it.
moritz_ dynamically?
jnthn had thought it was just, if there were no iterations of the loop. 09:27
moritz_ so if it calls a builtin written in Perl 6, and that uses an 'if'...
jnthn for @results { .render } else { say "omg no results!" }
moritz_ jnthn: that would make some sort of sense, zyes 09:28
jnthn wonders exactly what the spec says 09:29
Su-Shee do you reach that? -> sushee.schreibsturm.org/
eternaleye 'Su-Shee's Blog' 09:30
09:30 Jedai joined
eternaleye Red at the top 09:30
Su-Shee fine. :)
moritz_ Su-Shee: I hope you're preserving your old postings somehow? 09:32
Su-Shee moritz_: yes.
moritz_: the fourth archive, then. ;)
09:33 kimtaro joined
moritz_ Su-Shee: maybe we need a blog archiving framework :-) 09:33
masak this text makes it seem like the 'else' block is always executed... docs.python.org/reference/compound_...-statement
Su-Shee moritz_: definetely. I went from old wordpress to new wordpress to textpattern to..
moritz_ Su-Shee: something ueber simple like a JSON dump with a defined structure 09:34
masak ah! that's it! the only way to avoid the 'else' is to do a 'break'!
09:34 ihrd joined
Su-Shee moritz_: it's extremely annoying ever changing a blog engine. usally, things like tags and categories don't really match. 09:35
09:37 meppl joined, M_o_C joined
masak oh, and speaking of Rakudo 1.0 -- I'm not sure Perl 5/CPAN compatibility is on anyone's roadmap, but that would be one feature which would make me much less hesitant to tag Rakudo as 1.0. 09:49
jnthn has it somewhere in the back of his mind. 09:52
Su-Shee hm, movable type is really not bad.. 09:56
10:00 M_o_C left
missingthepoint can anyone explain how the STD.pm <-> evalbot is impl here? (moritz_?) 10:00
curious, because STD.pm uses Cursor, which is written in P5 10:01
10:02 kimtaro left
missingthepoint it's compiled to Perl 5, no? 10:02
elf?
masak missingthepoint: I think TimToady wrote his own translator, gimme5. 10:03
missingthepoint masak: thanks 10:04
10:11 rgrau joined
masak jnthn: how do I make my 'else' block execute at the end of any non-breaked for loop? 10:13
jnthn masak: Heh, I was wondering how you planned to do taht. :-) 10:17
masak :P 10:18
jnthn I guess you need something that handles the break.
And maybe sets a "don't run this" variable
That probably needs to be lexical.
masak I actually don't have much of a clue about PAST. I blindly copied stuff and was happy when Rakudo built and even ran my script.
jnthn: ok, that makes sense. I'll try that. 10:19
jnthn You can call .unique on any PAST node to get a unique identifier and even supply a prefix.
masak if I succeed, this'll be a blog post.
jnthn: that's useful. I guess I can combine that with --target=PAST to see what I get. 10:20
jnthn masak++ # getting into rakudo guts hacking
masak just some trivial branch hacking to amuse myself. :)
10:21 flexibeast left 10:24 arejay left, zamolxes left, japhb_ left, rjh left, Patterner left, antiphase left, shachaf left, eternaleye left, c1sung left, orafu left, Jedai left, cotto left, Aisling left, arthur-_ left 10:25 Jedai joined, cotto joined, arthur-_ joined, Aisling joined 10:26 zamolxes joined, japhb_ joined, orafu joined, Patterner joined, arejay joined, eternaleye joined, shachaf joined, rjh joined, c1sung joined, antiphase joined, szabgab left 10:27 xinming_ is now known as xinming 10:34 ihrd left 10:47 nsh left, sbp left, nsh_ joined, sbp joined, nsh joined
eternaleye (way, way, _way_ backlogging still) pmichaud: re ins2 branch, most of what I've seen in 'packager-friendly' projects is that 'what to install' is unimportant outside of the project itself - the packager just cares that DESTDIR=foo on the 'install' make target works. Gentoo and Exherbo both use this as the basis of packaging. 11:03
masak oh, right. backlogging. 11:04
11:06 ruoso_ left
moritz_ @seen pmichaud 11:07
lambdabot pmichaud is in #perl6. I last heard pmichaud speak 19h 14m 12s ago.
moritz_ @seen diakopter
lambdabot I saw diakopter leaving #perl6 13h 25m 1s ago, and .
moritz_ I was just wondering where the ', and .' went, but here it is again
masak lambdabot is a mysterious creature. 11:09
moritz_ I wonder if it's actually written in perl :-)
masak :)
moritz_ just like... that xkcd.. :-) 11:10
masak moritz_: is there any chance you could pick another yellow for the ⏏ arrow in the logs?
the current nuance doesn't agree very well with my eyes. 11:11
moritz_ masak: patches welcome :-)
masak to where?
moritz_ misc/irclog/lib/IrcLog/WWW.pm line 21 11:12
masak gotcha. 11:13
moritz_ it's used as a CSS color description
masak ah, so those are CSS color values?
moritz_ aye 11:14
masak moritz_++
pugs_svn r27749 | moritz++ | [irclog] gah, left in way too many debugging warn()s 11:15
moritz_ you'll need to svn up / git svn rebase first :/
masak hokay.
moritz_ sometimes I wish that 'git svn rebase' would automatically stash my local modifications, rebase, and then stash apply them 11:16
afk
masak moritz_: write an alias. 11:17
pugs_svn r27750 | masak++ | [misc/irclog/lib/IrcLog/] changed arrow color to an eye-compatible one
11:20 donaldh left 11:21 donaldh joined, Jedai left 11:22 snearch left
jnthn afk for a couple of hours 11:25
moritz_ masak: I updated the server to latest svn, so newly generated pages will display the new colour 11:51
11:51 rgrau left
moritz_ std: +++ 11:51
p6eval std 27750: OUTPUT«===SORRY!===␤Preceding operator expects term, but found infix + instead at /tmp/kXjJMzUp2v line 1 (EOF):␤------> +++⏏<EOL>␤ expecting any of:␤ noun␤ prefix or meta-prefix␤ standard stopper␤ term␤ terminator␤ whitespace␤FAILED
..00:02 39m␤»
masak moritz_++
moritz_ masak++ # JFDI
11:58 M_o_C joined 11:59 Psyche^ joined 12:02 Patterner left, Psyche^ is now known as Patterner 12:04 Psyche^ joined 12:05 Patterner left, Psyche^ is now known as Patterner 12:07 ingmar is now known as Ingmar
moritz_ radar.oreilly.com/2009/07/oscon-sta...-crow.html 12:23
"women make up only 1.5% of contributors to open source projects overall. They're 5% of the perl community and 10% of Drupal."
masak the Drupal community seems like a fun one. 12:24
kudos to them.
moritz_ aye 12:25
masak this Drupal community member in particular impresses me to no end: www.youtube.com/watch?v=8mwKq7_JlS8 12:26
like, I mean... wow!
moritz_ www.roblambert.com/wp-content/uploa...dee-os.png
masak I hope I'm not too age-ist by thinking that that video is very impressive. I simply wish that I were as knowledgeable at 12. 12:27
I hope this kid represents the future in some way. 12:28
moritz_ and I hope it does not, in some other way :-) 12:29
Su-Shee masak: the higher range of female participiants within web projects isn't what it seems to be. 12:31
masak Su-Shee: in what sense?
Su-Shee masak: web projects simply have a much larger count of designers - which are often women.
masak moritz_: to be more specific, what I like is that a 12-year-old obviously found his place as a productive member in a development community. that makes me very happy.
Su-Shee masak: and let's face it - there's participiant and .. developer. ;)
12:32 pmurias joined
masak Su-Shee: still, I think it's a positive sign. 12:32
I still don't see myself as a Rakudo committer. I'm more of a Rakudo complainer. :)
moritz_ masak: that aspect I agree with. OTOH when I want to get a job as a programmer, I'm not keen on competing with somebody who was so proficient at that age already 12:33
Su-Shee masak: well, statistically it's not really, because those numbers are pretty constant for 10 years or so. certain areas around "core technology" attracts some women. but they never break the barrier.
moritz_: doesn't matter. who cares what you coded in the age of 12 which was (I'm assuming) some 15 years ago...
masak moritz_: that's just the backside of the immensely positive effect of incorporating 12-year-olds in the dev community. 12:34
12:34 zamolxes left
masak Su-Shee: but you're a dev, no? 12:34
moritz_ Su-Shee: if he continues to develop like this, I'll be no match for him in any way :/ 12:35
Su-Shee but that's a common phenomen in many scene that people start out at the age of 12 - it really doesn't matter in the long run...
moritz_ masak: you're right. I wouldn't complain if he hacked on Rakudo or Parrot :-)
Su-Shee moritz_: you egal "length of experience" with "depth of experience"
masak moritz_: I'm afraid we'll have to find our own 12-year-old. Drupal got dibs on him. :) 12:36
Su-Shee masak: yes. but look here: I've made the website of perl6 projects - I didn't wrote a 100 tests or the next 3 important modules or something..
moritz_ Su-Shee: well, if I had started playing table tennis with 12 years I'd be loads better by now
masak Su-Shee: point taken.
Su-Shee: so, when will we see a module written by you? :) 12:37
Su-Shee moritz_: I started politics and writing with 12. It might have opened one or two doors, but that's it. when we all become 30, noone cares for the kind of past you've got but only for what you can do _now_
masak: when I started the new job and relaxed a little mentally. I'm not very creative while having certain kinds of stress. 12:38
12:38 fridim_ left
masak understood. 12:38
12:38 SmokeMachine joined
Su-Shee moritz_: let's set aside sports, I think this is something different due to physical questions. 12:38
masak: but yes, I have to jump through certain hoops of socialization to play with the boys. ;) 12:39
12:39 zarko joined, synth joined
masak Su-Shee: I guess you do. my mother tells of similar things from her boys-only work environment 20 years ago. 12:40
12:40 zamolxes joined
masak she used to work at an oil refinery. 12:41
she told me that's where she learned to tell and enjoy/endure really crude jokes.
moritz_ masak: did she pass on those on to you? :-) 12:42
masak :)
I don't know!
do I have a foul mouth, in your opinion?
moritz_ s/foul/fast/ :-) 12:43
Su-Shee the most important thing is learning not to take critizism personally. men plainly critize totally different than women.
moritz_ indeed
Su-Shee besides that I really don't get the reluctance to do technology.
masak aye.
I wish more people would think like that. 12:44
Su-Shee one day I realized that I won't get "cool technology" served on a silver platter. either I take/make it or it just passes over me. 12:45
and the first female med students or physics student didn't go there because of the nice and women-friendly athmosphere of 1900. ;)
masak guess not. 12:46
Su-Shee besides that: the open source community is rather fluffy and cuddly compared to other scenes.. 12:47
moritz_ unless you talk to the GNU libc maintainers, of course :-)
masak they're not cuddly? who'd have thought! 12:48
moritz_ Su-Shee: actually 1900 was quite a good number (guessed? remembered?) - that's when Emmy Noether started studying in my home town
synth Su-Shee: learning not to take criticism personally, in what context? sorry I didnt catch the whole convo but I am intrigued about what you're saying 12:53
12:53 snearch joined
synth is a man and plainly criticizes everything :) wifey takes it personally though.. 12:54
pmurias masak: but they are extra fluffy to compensate for the lack of cuddliness
masak pmurias: :)
Su-Shee synth: for example if you critize my code as bad, unsecure, inefficient, plainly not elegant or don't like my indention style or find that my choice of editor/distribution is plain bad. ;) 12:55
moritz_: not guessed, 2 years feminist theory and women's studies at universities. ;)
synth Su-Shee: criticisms in that context are never personal is what you're saying? 12:56
they're strictly about the matter at hand?
or shouldn't be?
Su-Shee synth: I have several choices in that case: shut up, go and cry and never come back. shut up, think you're an idiot and don't care. take what's correct and leave the rest as "he's just an annoying alpha male" or ignore you completely.
moritz_ Su-Shee: wohoo
synth Su-Shee: gotcha 12:57
Su-Shee synth: even if it is: _who cares_? everyone has to learn that not everybody likes you.
12:57 molaf joined
synth Su-Shee: *nod* if people cant realize that they are in for a world of hurt 12:57
synth is overtly critical about everything in his life, to a fault 12:58
Su-Shee synth: yeah, but women simply handle this entirely different. the say it differently - if they say anything at all, they suggest improvements rather then to critize..
synth Su-Shee: yeah thats my wife for sure
I tell her, criticize, I can take it.. I need someone to tell me my shit stinks some times 12:59
Su-Shee (we're talking prototypes here - of course there's very social and polite coders and real bitches out there..)
synth: yeah, but now imagine you're the only one being used to a certain social rule and you're thrown within 100 people doing it your way.
synth Su-Shee: heh yeah 13:00
pmurias it shouldn't be to hard to figure out that the social rule doesn't apply
synth you'll learn quick, Id imagine 13:01
ick keyboard is sticky.... laptop old... I need an upgrade
Su-Shee pmurias: yeah but _experiencing_ it is something else entirely.
13:02 nsh_ left
Su-Shee anyway. the open source community is what it is and there will only be more women if women really want it. and even if they hate the community, you can still write code alone or with friends and plainly release it and having never to put up with a 600 alpha male chat channel. ;) 13:03
13:05 explorer left
masak does 'alpha male' have to amount to 'pushy, bullying and inconsiderate'? 13:05
Su-Shee masak: in principle it hasn't, no. but I don't wanna really know what happened up until the pumpking recently resigned... 13:07
masak Su-Shee: you make it sound like some king-of-the-hill wrestling happened. :) what really happened was that disagreements reached a tipping point, and the pumpking didn't find that the equation of being pumpking landed on the plus side of things. 13:09
but if you mean that it was alphamaleish behaviour to react like that, then maybe yes.
13:11 skids_ joined
masak this post is the most pointed example of the less mature sides of the debate: use.perl.org/~pozer/journal/39242 13:11
though I like reading the support for chromatic in the comments. :) 13:12
13:12 fridim_ joined
pmurias masak: did you read through some of the p5p threads? 13:12
masak pmurias: no, I didn't. maybe that would make things clearer.
are they online? 13:13
pmurias searches...
masak actually, I saw one or two, but they weren't so bad.
certainly not the-Hulk-like fights. 13:14
pmurias masak: rafael posted links to some of the threads where they refuted some of the chromatics proposals 13:15
masak that might have been the ones I read.
13:15 [particle] left 13:16 mj41 left
Tene moritz_: parrot has support for mysql. it would be pretty easy to get that wrapped up for use by HLLs. 13:17
masak it's even been done once. 13:18
but the code was bit-rotten last time I looked.
Tene Huh.
masak examples/nci/mysqltest.p6 13:19
feel free to get it running again. :)
pmurias masak: but i feel they shouldn't have started "debating" with chromatic on the blogs instead of pointing out the relevant threads and make him looks silly 13:20
masak pmurias: you mean that there are a set of relevant threads that would make chromatic look silly? out of curiosity, do you have any links to those? 13:21
13:21 jonjj joined 13:24 jonjj left
pmurias the class feature he proposed was discredited 13:32
13:32 hoelzro joined, [particle] joined
masak doesn't feel informed enough to opine in the debate on p5p, chromatic and rafael 13:35
13:37 zarko1 joined 13:39 colomon left 13:46 zarko left
Su-Shee If someone does code a perl 6 blog: please make a better userinterface than movable types. 13:49
masak I'm hoping to be able to shove Web.pm under the feet of people who eventually write a Perl 6 blog. 13:50
(by making it laughable to attempt it without Web.pm)
13:50 kidd_` joined
Su-Shee masak: I also still have it in mind. But I managed recently to mvc myself into a corner when I started to consider making a userinterface in javascript - which also should be mvc.. 13:51
masak Su-Shee: well, you know what they say. start simple.
complexity is a feature that can be added stepwise, in working increments. :) 13:52
Su-Shee masak: if you get your architecture right. ;)
masak well, re-architecturing is what tests are for. :)
Su-Shee well, we will have a perl6 blog engine some day. I'm sure of it. 13:53
masak oh, it won't be long.
half a year, tops.
there's already Yarn in its various incarnations.
mberends++ has promised to dogfood Yarn. he hasn't yet, but I have a feeling he will. 13:55
Su-Shee looks forward to again converting blog content from A to B :)
14:03 lucs joined 14:05 colomon joined
missingthepoint night all :) 14:05
14:05 iblechbot left, PacoLinux joined 14:06 missingthepoint left, colomon left 14:11 KyleHa joined 14:21 zarko joined 14:22 FurnaceBoy joined 14:23 szabgab joined 14:24 cls_bsd joined, diakopter joined
KyleHa @seen KyleHa 14:25
lambdabot You are in #perl6. I last heard you speak just now.
KyleHa Nyuck nyuck nyuck.
14:25 grawity joined, grawity left
moritz_ wonders why nobody else replied to Ben Morrow's mail on p6l - seems like the perfect topic for bikeshedding 14:25
14:26 gbacon left
KyleHa I, for one, did not feel qualified to answer. 14:27
I'm interested in the answers, however, since they'd make good test fodder.
It's also possible I didn't actually read Ben's email. 14:28
masak moritz_: I thought of replying, but found I didn't have anything to add to your email. :)
moritz_ I thought more of the usual suspects like Tsa, Mark Reed, Jon Lang...
masak: ah, one form of warnocking :-) Glad you said it now on IRC 14:29
szabgab once I tried to recommend the use of X<> tags in the docs but it seems nearly noone adds them, are they discouraged or shall I keep adding them ?
masak moritz_: is it really warnocking after you answered? :) 14:30
szabgab: I must admit to not having seen the light yet as to the X<> and C<> tags.
moritz_ masak: you can warnock replies also :-) 14:31
szabgab oh I think C tags just make the text emphasized - and thus they are good target for indexing but they are accidental 14:32
masak moritz_: I think I'll need to re-read your reply. apart from the final 'no idea', I don't recall it as being full of open-ended things.
szabgab X tags were specifically planned to be used to add entires to generate an index
masak szabgab: ah, ok. 14:33
szabgab for example I was just looking for the defintion of <ws>
searching for <ws> did not help
same with the word 'for' 14:34
moritz_ szabgab++ # adding an index 14:35
szabgab so I keep adding them and I hope literal will add them to grok soon 14:36
14:37 donaldh left, frettled left, charsbar left, ilbot2 left, rbuels left, Tene left, patmat left, zarko1 left 14:39 M_o_C left, xinming left, xinming joined
hoelzro I suppose I should only publish the client-side UI for my "Try Rakudo!" project, and leave the server side stuff private, huh? 14:41
14:41 eiro left, eiro joined, Exodist joined
moritz_ hoelzro: why do you think so? 14:42
14:42 M_o_C joined
hoelzro moritz_: that seems to be the general impression I'm getting about creating secure web-REPLs 14:42
moritz_ hoelzro: we're always happy to see source code that works with Rakudo somehow, and maybe you'll even get some contributions
hoelzro I'd be more than willing to share it
I've no intention to horde it =) 14:43
moritz_ hoelzro: "security by obscurity"?
14:43 donaldh joined, frettled joined, charsbar joined, rbuels joined, ilbot2 joined, Tene joined, patmat joined
hoelzro moritz_: partially 14:43
14:43 charsbar left
hoelzro I'd like more hardcore Rakudo/Parrot devs to review it before making it more public, at the very least 14:43
moritz_ hoelzro: if you don't want any of your boxes compromized, I can offer you shell access on the server where p6eval runs
hoelzro: you can run it there if you want
hoelzro moritz_: thanks, but pmichaud already made me that offer =)
also, I've only done UI work so far
you can check it out here if you'd like: hoelzro.net/try-rakudo 14:44
it's still very much in infancy
but I feel that I have some good ideas about securing the client side
s/client/server/
but those ideas will require modifications to both Rakudo and Parrot 14:45
14:45 Exodist left, charsbar joined
moritz_ hoelzro: I'm sure that if it doesn't effect normal working they will be accepted 14:46
hoelzro: I'm all for a secured rakudo/parrot, because that would make me hosting p6eval much safer too :-)
hoelzro moritz_: well, for the moment, I was just going to remove certain opcodes/classes
but I've been thinking about PDD 18 (security) a lot
moritz_ sounds more like local patches then
14:47 explorer joined
hoelzro so if I can contribute something that helps security for Parrot in general, I'll do that 14:47
it'll make maintaining my try-rakudo branch easier =)
same goes for Rakudo
removing classes from Rakudo may be trickier, though 14:48
I should say, allowing Rakudo to start up, and then telling it to cut off all access and recollection of classes A...B 14:49
that'll be no picnic.
14:49 charsbar_ left 14:50 explorer left, gbacon joined
moritz_ I think you can get away with disallowing inline PIR 14:51
hoelzro moritz_: what about things like file I/O? 14:52
moritz_ the other dangerous opcodes (qx, run(), open() etc) are easily removed by monkey-patching
pugs_svn r27751 | szabgab++ | add X<> tags to Predefined Subrules
hoelzro (acutally, file I/O on a virtual file system would be cool)
moritz_ hoelzro: look at lib/Safe.pm in the Rakudo repo
hoelzro moritz_: ok
ah 14:54
moritz_ rakudo: open('foo')
p6eval rakudo 4c31fb: OUTPUT«operation not permitted in safe mode␤in Main (lib/Safe.pm:25)␤»
jnthn back 14:56
masak rakudo: sub foo(&bar = { $^id }) { say bar("OH HAI") }; foo 14:58
p6eval rakudo 4c31fb: OUTPUT«OH HAI␤»
masak Rakudo++
Su-Shee grrr. movable type isn't really the best blog I've ever installed.
moritz_: I would think very hard if you want this for your site. 14:59
moritz_ rakudo: sub foo(&foo = sub infix<MOEP>($a, $b) { say "$a MOEP $B }){ 3 MOEP 5 }; foo()
pmichaud good morning, #perl6
phenny pmichaud: 04:44Z <wayland76> tell pmichaud I've found some problems with the "install" section of the ins2 makefile. I hope to have a patch by the end of (my) day
pmichaud: 06:50Z <wayland76> tell pmichaud Ok, patch for ins2 branch added to rt.perl.org/rt3/Ticket/Display.html?id=63360
p6eval rakudo 4c31fb: OUTPUT«Malformed routine definition at line 2, near "infix<MOEP"␤in Main (src/gen_setting.pm:3343)␤»
hoelzro pmichaud: good morning 15:00
moritz_ Su-Shee: ok, thanks for the info
masak rakudo: sub foo($n, &bar = &foo) { return unless $n; say $n; bar($n-1) }; foo(5)
hoelzro pmichaud: would you like to see the "Try Rakudo!" UI as it stands right now?
p6eval rakudo 4c31fb: OUTPUT«5␤4␤3␤2␤1␤»
masak morning, pm.
Su-Shee moritz_: you'll have to post-install plugins for certain common features and this is somewhat.. zaeh. ;)
moritz_ rakudo: sub foo(&foo = sub bar($a, $b) { } ) { bar() }
pmichaud hoelzro: yes!
p6eval rakudo 4c31fb: ( no output ) 15:01
pugs_svn r27752 | kyle++ | [t/spec] Test for RT #63778
hoelzro pmichaud: hoelzro.net/try-rakudo
no server-side yet, and only the command line part is "done"
but I feel it's off to a good start
masak rakudo: sub foo(&foo = {;}) { say "OH HAI"; foo }; foo 15:02
p6eval rakudo 4c31fb: OUTPUT«OH HAI␤»
jnthn ...= sub infix<MOEP>($a... # note missing :
pmichaud hoelzro: the interface looks excellent. Should it be running the things I type in to the command line yet?
masak rakudo: sub foo(&foo = &foo) { say "OH HAI"; foo }; foo
hah!
p6eval rakudo 4c31fb: OUTPUT«Null PMC access in isa()␤in sub foo (/tmp/d6DbO2Evkd:2)␤called from Main (/tmp/d6DbO2Evkd:2)␤»
hoelzro pmichaud: it should just alert you
masak submits rakudobug 15:03
15:03 explorer joined
pmichaud okay, that's what it is doing, yes. 15:03
hoelzro good =)
any thoughts on the interface (layout, coloration, etc) would be welcome
I'm going to put the UI stuff in a Git repo after work today
masak rakudo: sub foo(&foo = &foo) {}; foo 15:04
p6eval rakudo 4c31fb: OUTPUT«Null PMC access in isa()␤in sub foo (/tmp/YNtZwS8Tl8:2)␤called from Main (/tmp/YNtZwS8Tl8:2)␤»
moritz_ rakudo: sub foo(&foo = &foo) {};
p6eval rakudo 4c31fb: ( no output )
moritz_ ah, happens while calling
masak aye.
KyleHa Is that supposed to live?
masak I think so.
pmichaud hoelzro: if you want a git repo in the rakudo or perl6 accounts on github, let me know.
(or it can go into perl6-examples) 15:05
masak KyleHa: the first &foo is the param, the second is the sub.
moritz_ so the first &foo would lexically override foo()?
hoelzro pmichaud: ok, I'll ask after I get home from work
KyleHa It seems the second &foo is taking a reference to foo before it's defined.
moritz_ sub foo(&foo = { say 3 }) { foo() };
rakudo: sub foo(&foo = { say 3 }) { foo() }; 15:06
p6eval rakudo 4c31fb: ( no output )
moritz_ that should print out 3\n, no?
masak? :-)
jnthn Probably.
pmichaud no, you never called foo()
moritz_ ouch.
KyleHa rakudo: sub foo(&foo = { say 3 }) { foo() }; foo;
masak :)
p6eval rakudo 4c31fb: OUTPUT«3␤»
jnthn oh
:-)
moritz_ rakudo++
pmichaud rakudo ftw
moritz_ indeed
jnthn Rakudo is becoming smarter than us. ;-)
masak yes, rakudo++ caught most of the screwballs I threw at it. I had to try hard to find a bug. :) 15:07
hoelzro pmichaud: I also had some insight into the server-side security, with some help from bpetering; check yesterday's logs for details
pmichaud hoelzro: will do. It may be a little while before I get to backlogging (lots of non-rakudo things happening around this house today)
hoelzro sounds good 15:08
masak had to think long to get the bug title right: "[BUG] Null PMC Access when calling a sub &foo with a parameter &foo defaulting to &foo in Rakudo" 15:10
:)
pmichaud does it work for other subs?
masak yes.
well, not in the sense that &foo is special.
hoelzro quick question about patch submission: git-format-patch patches are just supposed to be sent to the Rakudo bugs e-mail with [PATCH] in the subject, right?
masak but all those three have to be the same.
pmichaud rakudo: sub bar() { say 'bar'; }; sub foo(&x = &bar) { x() }; foo()
moritz_ rakudo: sub foo(&foo = &foo) { foo() }; foo({say "bar" })
p6eval rakudo 4c31fb: OUTPUT«bar␤»
hoelzro I feel like my patches may have gotten lost in the ether
p6eval rakudo 4c31fb: OUTPUT«bar␤»
pmichaud hoelzro: yes 15:11
moritz_ hoelzro: it's just that pmichaud++ had been at oscon, jnthn was busy and nobody else feels qualified to review them
15:11 hercynium joined
hoelzro ok, that's fine, just wanted to get some feedback =) 15:11
moritz_ our bus number is approximately 2.7 or so
hoelzro pmichaud: that reminds me, I also had some insight into retaining lexicals in the REPL last night 15:12
hoelzro thinks maybe he should organize these thoughts into an e-mail
pmichaud hoelzro: what's the insight? 15:13
hoelzro pmichaud: well, I was thinking of the interpreter's eval method should return some sort of lexical context object, and you can provide that object to later calls to eval
pmichaud Parrot doesn't have lexical context objects :-( 15:14
(not yet)
hoelzro pmichaud: just an idea, maybe something we could build on
masak hoelzro: oh, that insight. :)
we have it from time to time.
hoelzro ah.
I thought I was being so smart and creative...
masak yes, that's how I'd like it to be done, too.
pmichaud This falls under the general heading of "turn contexts into PMCs", which has been on the todo list since last November but nobody seems to want to attack it.
pugs_svn r27753 | szabgab++ | add also X<< >> tags to Predefined Subrules 15:15
masak hoelzro: don't get discouraged, I didn't mean to take the enterprising spirit out of you. :)
hoelzro =)
pmichaud: is that todo in RT?
pmichaud hoelzro: you are being smart and creative, if you came up with the same solution that the other smart and creative folks came up with :-)
Parrot todos are in trac.
hoelzro pmichaud: good point =)
15:16 nihiliad joined
KyleHa I'd be tempted to have a REPL just record everything and re-eval everything after each line is added except I'd replace each old 'say' with a noop. Other output-producing commands would be discouraged. 8-) 15:17
pmichaud the todo is trac.parrot.org/parrot/ticket/596
KyleHa: I've thought of that also, but decided it's not quite practical. We'd also have to deal with "rand" and "pick" and other things like that. :-)
KyleHa 'srand $$' at the top. 15:18
masak KyleHa: that works until a command takes a minute to run...
KyleHa If brute force doesn't work, you're not using enough of it. 15:19
masak I mean, suddenly you have a minute-long delay for all the subsequent commands.
KyleHa (I didn't say I thought this was a GOOD idea.)
masak: Yeah, that would stink too.
pmichaud I think I have a basic plan for getting the REPL to work in Parrot; it just needs tuits to either (1) execute the plan or (2) write down the plan for someone else to execute.
masak KyleHa: fwiw, I've toyed with that idea too. :) I _really_ want to see a good Rakudo REPL.
15:20 donaldh left
masak pmichaud: is (2) a lot of work? otherwise it seems like a good plan to do that first, and then possibly (1). 15:20
moritz_ we don't have enough pmichauds around here :-)
hoelzro pmichaud: thanks
pmichaud masak: so far it feels like (2) is approx the same work as (1)
moritz_ then (1) is preferred, of course :-) 15:21
15:21 donaldh joined
pmichaud otoh, (2) has the benefit of drawing someone else into development :-) :-) 15:21
15:21 snearch2 joined
pmichaud either way, the tuits haven't been present to date 15:21
15:22 snearch2 left
pmichaud however, fixing the repl would be a really good Rakudo Day task 15:22
hmmmmmmm
pmichaud reviews his schedule for this week
szabgab pmichaud, is any Rakudo hacking planned for Lisbon ?
hoelzro pmichaud: what needs fixing with the repl?
pmichaud szabgab: jnthn++ and I are planning to do hacking on Sunday and Thursday 15:23
masak hoelzro: we don't have arrow keys, history, tab completion...
hoelzro ah
pmichaud masak: I have history and arrow keys
that part already works
szabgab I'll be there starting from Friday afternoon
masak oh right. scratch that.
pmichaud szabgab: yes, I was thinking I might sit in on the Perl 6 training course on Saturday 15:24
szabgab not that I can help you hack, but I can carry the beer :-)
pmichaud that would be awesome
pmichaud well, my primary goal for Lisbon is not code as much as putting together the detailed plan for Rakudo's "major release"
szabgab I will want to show you how we can now highlight Python code in Padre
masak maybe we should call it 'awesome.0' instead of '1.0' :P 15:25
pmichaud actually, that should read "publishing the detailed plan"
szabgab using PCT
pmichaud szabgab: yes, I read the announcement about it, and I think that's so cool. I look forward to seeing it :-)
masak szabgab: cool!
pmichaud masak: that's not a bad name (awesome.0)
overall I think I would prefer something that isn't "1.0", yes 15:26
masak pmichaud: I wasn't 100% serious, though. :)
KyleHa class, module, package...is there another chunk-o-code type construct I'm forgetting?
pmichaud grammar
role 15:27
masak it might actually send the wrong signals to have such a cocky version 'number'...
KyleHa Role, yeah.
masak KyleHa: grammar.
oh, pmichaud got that one first...
KyleHa Thanks, masak++ and pmichaud++
masak (karma for being redundant)++ 15:28
hoelzro hmm...in Parrot, closures are first-class objects, right?
jnthn aye
well, subs are
pmichaud rakudo: my $closure = { 'first class' }; if $closure() { say 'yes' } 15:29
p6eval rakudo 4c31fb: OUTPUT«yes␤»
hoelzro all, a closure is just code + a lexical context
jnthn There's no Closure type any more, it's just a clone.
szabgab pmichaud, actually I'd like to get the example grammer you are going to show when you talk about PTC so I can make sure I can create the highlighter for it, spontaneously, on the spot
hoelzro jnthn: a clone of a Sub?
jnthn Of a sub attatched to a context.
Yup
pmichaud szabgab: you mean the parrot compiler toolkit ? 15:30
szabgab yes
hoelzro jnthn: thanks
szabgab ah wrong letter order
pmichaud szabgab: afaik, I'm not scheduled to give the PCT talk at lisbon. Just "Hacking Rakudo Perl"
szabgab oh :-( 15:31
pmichaud (apparently there were only enough slots for me to be able to do one talk this year; I submitted three :-)
15:31 rfordinal joined
szabgab you can do a lightening talk with it :-) 15:31
pmichaud I've already submitted my lightning talk
(and that's not it) 15:32
szabgab yeah I think they had 50% more submissions than time
I also submitted talks and got only one slot, but have not written that yet
but I'll mostly do live demo anyway 15:33
hoelzro so I'm looking through the code for Parrot subs, and I was thinking: maybe we could decouple the context part of a Sub from Sub, encapsulate that into a LexicalContext class, and have Subs refer to LexicalContext objects?
15:33 snearch left, masak left
jnthn hoelzro: I think there's a longer-term plan that a Parrot context, which is currently just a C structure, would become a PMC. 15:34
hoelzro jnthn: that's kind of what I was going for =)
jnthn Thing is, it adds more GC pressure. 15:35
hoelzro curses
pmichaud hoelzro: see trac.parrot.org/parrot/ticket/596 (I think that's much the same plan :-)
hoelzro reading it over
pugs_svn r27754 | kyle++ | [t/spec] Test for RT #63826 15:36
hoelzro I'm feeling a bit like Leibniz or Wallace right now... 15:39
s/or/and/ 15:40
lisppaste3 szabgab pasted "STD parsing error" at paste.lisp.org/display/84273 15:44
szabgab that thing works in rakudo but STD complains 15:45
Unrecognized regex metacharacter at (eval) line 11:
at least whet we have in Padre complains that way
KyleHa Line 11 is the last line? 15:46
szabgab one before
the closing of the comment token 15:47
KyleHa Is it the double dollar that does it?
TimToady it's the \# 15:48
15:48 charsbar left
TimToady I think it's taking it as a comment, so you lose the ] 15:49
moritz_ std: token foo { \# }
p6eval std 27753: OUTPUT«===SORRY!===␤Regex missing terminator at /tmp/ShtpoNlTso line 1 (EOF):␤------> token foo { \# }⏏<EOL>␤ expecting any of:␤ regex atom␤ rxinfix␤ termish␤ unspace␤ whitespace␤ ws␤FAILED 00:02 36m␤»
moritz_ std: token foo { '#' bar }
pmichaud maybe try '#'
p6eval std 27753: OUTPUT«ok 00:02 36m␤»
pmichaud moritz++
KyleHa rakudo: token foo { \# } 15:50
p6eval rakudo 4c31fb: ( no output )
15:50 charsbar joined
moritz_ rakudo++ # we found an instance where it parses better than STD.pm :-) 15:50
15:51 rfordinal left
Lorn in rakudo i have something like a list of object methods? like say @array.list_method ? 15:51
szabgab same problem with '#'
this breaks the syntax highlighting :-(
switching to PGE highlighting :-)
moritz_ rakudo: my @a; say @a.^methods().map({.name}).sort.join(", ") 15:52
p6eval rakudo 4c31fb: OUTPUT«, Array, Scalar, delete, exists, item, pop, push, shift, splice, unshift, values␤»
Lorn moritz_: great! thanks
szabgab though I am not sure why I call it PGE
Lorn moritz_: why they not show .WHAT method? 15:53
moritz_ Lorn: it seems inherited methods aren't shown that way
rakudo: my @a; say @a.^methods(:all).map({.name}).sort.join(", ")
p6eval rakudo 4c31fb: OUTPUT«, Array, Scalar, delete, exists, item, pop, push, shift, splice, unshift, values␤»
moritz_ rakudo: my @a; say @a.^methods(:tree).map({.name}).sort.join(", ") 15:54
pmichaud (also .WHAT isn't truly a method in Perl 6, although Rakudo has it that way...)
p6eval rakudo 4c31fb: OUTPUT«, Array, Scalar, delete, exists, item, pop, push, shift, splice, unshift, values␤»
moritz_ right, it's a macro
15:56 zamolxes left
Lorn all uppercase method is a MACRO? 15:57
TimToady szabgab: you say same problem with '#', but moritz++ showed it working above
szabgab I think it was a caching issue in Padre
now it seems to work
15:57 abra joined
pmichaud Lorn: "all uppercase" usually indicates something "special" (but not necessarily a macro 15:58
15:58 justatheory joined
Lorn pmichaud: ok, thanks 15:58
16:00 PZt left, Trey left, ascent_ left, integral left 16:02 dakkar left, Trey joined, ascent_ joined
szabgab thanks for the previous fixe, I have a question related to that regex , what is the recommended way to create a grammer of a line-based language? where some of the tokens represent full lines 16:04
I used this ^^ stuff $$ \n?
but I guess there is something better 16:05
16:06 SmokeMachine left 16:07 colomon joined
KyleHa This ticket perplexes me: rt.perl.org/rt3/Public/Bug/Display.html?id=62900 16:09
Shouldn't it die there? The error message may be less that awesome, but it seems to me the death is inevitable. 16:10
16:11 colomon_ joined 16:12 colomon left, colomon_ left, Juerd_ left, integral joined
KyleHa rakudo: role A::B {}; say 42 ~~ A; 16:12
p6eval rakudo 4c31fb: OUTPUT«invoke() not implemented in class 'NameSpace'␤in Main (/tmp/Mxj4uHyzvt:2)␤»
KyleHa rakudo: say 42 ~~ A; 16:14
p6eval rakudo 4c31fb: OUTPUT«Could not find non-existent sub A␤»
KyleHa Hmm. Maybe it is a bug.
pmichaud Rakudo doesn't recognize "A" as a namespace there 16:18
(when compiling)
so it treats it like a subroutine call
TimToady STD just guess for the message, based on uppercase
std: say 42 ~~ A
16:18 colomon joined
p6eval std 27754: OUTPUT«Undeclared name:␤ A used at 1 ␤ok 00:02 36m␤» 16:18
TimToady *guesses 16:19
std: say 42 ~~ a
p6eval std 27754: OUTPUT«Undeclared routine:␤ a used at 1 ␤ok 00:02 36m␤»
pmichaud std: role A::B {}; say 42 ~~ A
p6eval std 27754: OUTPUT«Undeclared name:␤ A used at 1 ␤ok 00:02 36m␤»
TimToady hmm
yes, only the stash is created, not the type object 16:20
16:22 moritz_ sets mode: +o TimToady
TimToady levels up 16:22
16:25 lisppaste3 left
TimToady biab & 16:27
16:30 szabgab left 16:33 Psyche^ joined, molaf left, cdarroch joined 16:34 alester left, alester joined 16:39 lisppaste3 joined 16:44 Patterner left, Psyche^ is now known as Patterner 16:46 zarko left 16:54 zarko joined 16:56 rfordinal joined 17:00 [particle]1 joined 17:05 sparc left, desertm4x joined 17:10 payload joined, SmokeMachine joined 17:18 [particle] left 17:23 KyleHa left 17:26 zarko left 17:27 zarko joined, [particle]1 is now known as [particle] 17:28 zarko left 17:35 KyleHa joined 17:38 nbrown_ joined, zarko joined 17:39 japhb_ left, zarko left 17:42 nbrown_ left 17:43 nbrown left 17:45 hercynium left, abra left 17:46 nbrown joined 17:58 b2gills joined 18:00 japhb joined 18:01 payload left 18:04 iblechbot joined
KyleHa rakudo: sub a { 1 }; eval 'sub a { 2 }'; say a; 18:09
p6eval rakudo 4c31fb: OUTPUT«1␤»
18:10 kane_ left
KyleHa perl6: sub a { 1 }; eval 'sub a { 2 }'; say OUR::a(); 18:10
p6eval pugs: OUTPUT«*** No such subroutine: "&OUR::a"␤ at /tmp/KZPLIeiFs4 line 1, column 38-46␤»
..rakudo 4c31fb: OUTPUT«Null PMC access in invoke()␤in Main (/tmp/IUgBKBGFHs:2)␤»
..elf 27754: OUTPUT«Subroutine a redefined at (eval 129) line 3.␤Undefined subroutine &GLOBAL::circumfix_O_32_41 called at (eval 126) line 7.␤ at ./elf_h line 5881␤»
moritz_ wow, three very different error messages ;-) 18:11
18:11 nbrown left
KyleHa std: sub a { 1 }; eval 'sub a { 2 }'; say OUR::a(); 18:11
p6eval std 27754: OUTPUT«Undeclared name:␤ OUR::a used at 1 ␤ok 00:03 36m␤»
TimToady hmm
must be a bug in OUR:: 18:12
KyleHa I figured it was because OUR::a was defined in the eval.
STD doesn't go into eval, does it?
moritz_ no. 18:13
TimToady but it ought to see the outer sub a in the current package
wolverian STD isn't just syntax, it does name lookups too?
TimToady yes
18:13 hercynium joined
moritz_ wolverian: it has to know if a name is a type to do the parsing right 18:14
wolverian interesting. though perhaps obvious
right.
TimToady and it also does settings "correctly", where UNIT::OUTER === SETTING
wolverian I'm implementing a typed scheme right now, parsing isn't high on my reading list...
TimToady std: my SETTING::Int $x 18:15
p6eval std 27754: OUTPUT«ok 00:02 36m␤»
TimToady std: my CORE::Int $x
p6eval std 27754: OUTPUT«ok 00:02 36m␤»
TimToady std: my UNIT::OUTER::Int $x
moritz_ TimToady: sorry, I don't know if I asked you before, and if so what the answer was... do you object to removing want() from the spec?
p6eval std 27754: OUTPUT«ok 00:02 36m␤»
TimToady hmm, maybe I'm fooling myself
KyleHa rakudo: enum A <a b>; say a;
p6eval rakudo 4c31fb: OUTPUT«0␤» 18:16
TimToady std: my Foo::Bar $x;
p6eval std 27754: OUTPUT«===SORRY!===␤In "my" declaration, typename Foo::Bar must be predeclared (or marked as declarative with :: prefix) at /tmp/Hzy4GWBxKG line 1:␤------> my Foo::Bar⏏ $x;␤FAILED 00:02 36m␤»
KyleHa rakudo: enum A <a b>; eval 'sub a { 3 }'; say a;
p6eval rakudo 4c31fb: OUTPUT«a␤»
TimToady I think want should probably go away
jnthn I want it to. 18:17
TimToady unless it gets implemented as a lazy continuation that waits till the result is bound :)
jnthn Maybe in 6.0.1 ;-)
moritz_ notices the smiley at the end, and is happy
wolverian by the way, why is enum's list argument a quoted list? it seems weirdly runtime-ish to me.
moritz_ std: my %h; enum %h;
p6eval std 27754: OUTPUT«ok 00:03 39m␤» 18:18
moritz_ std: my %h; enum A %h;
p6eval std 27754: OUTPUT«Undeclared name:␤ A used at 1 ␤ok 00:02 39m␤»
TimToady many declarations run code at compile time
jnthn wolverian: Certainly in Rakudo, we figure out what the type names are at compile time for named enums, and we get upset if we can't.
TimToady (in theory)
moritz_ wolverian: I think you can give it anything that evaluates to a list or a list of pairs
TimToady though STD cheats on enums
wolverian okay, so it's compiletime runtime, that's fine. thanks. )
TimToady like the args to 'use' 18:19
jnthn rakudo: my %h; enum Foo %h;
p6eval rakudo 4c31fb: OUTPUT«Could not find non-existent sub h␤»
jnthn ...as I said, upset. ;-)
TimToady not to mention confused
but it looks like STD is backtracking on enum when it should have commited 18:20
jnthn TimToady: Aye. Just not smart enough to knwo it's confused yet. :-) 18:21
moritz_ so you have to commit a commit fix ;-)
jnthn rakudo: my %h; sub h { say "srsly wtf" }; enum Foo %h;
p6eval rakudo 4c31fb: OUTPUT«srsly wtf␤Null PMC access in get_number()␤in Main (/tmp/RYyqcc1SKd:2)␤»
jnthn wow 18:22
moritz_ where's masak when you need him? ;-)
ok, I'll submit
jnthn thanks.
TimToady std: sub a { 1 }; eval 'sub a { 2 }'; say &OUR::a();
p6eval std 27754: OUTPUT«ok 00:03 54m␤»
TimToady that one works
jnthn rakudo: sub a { 1 }; eval 'sub a { 2 }'; say &OUR::a(); 18:23
p6eval rakudo 4c31fb: OUTPUT«invoke() not implemented in class 'Undef'␤in Main (/tmp/WNfVeWHBjC:2)␤»
jnthn oh, we don't know about OUR yet.
TimToady but maybe it's an accident
std: sub a { 1 }; eval 'sub a { 2 }'; say &OUR::b();
p6eval std 27754: OUTPUT«ok 00:03 54m␤»
TimToady yeah, an accident, probably
here's another bad backtrack: 18:25
std: rx/ x \# /
p6eval std 27754: OUTPUT«Undeclared routines:␤ rx used at 1 ␤ x used at 1 ␤ok 00:03 39m␤»
jnthn Yeah
moritz_ ou.ch
jnthn Grammar bug, or grammar engine bug?
TimToady well, backtrack bug is independent of \# bug
jnthn True 18:26
TimToady backtrack bugs are usually trivial to fix
jnthn I meant the backtrack.
TimToady usually means it's missing a || <.panic: "phooey">
jnthn Aye, Rakudo has taken various of those fixes from STD.pm. Often they bring nicer errors too.
18:33 kane_ joined 18:38 fridim_ left
jnthn -> dinner 18:39
pmichaud back 18:40
18:52 zamolxes joined, masak joined
pugs_svn r27755 | moritz++ | get rid of want(), it's not implementable 18:54
r27755 | moritz++ |
r27755 | moritz++ | See <www.nntp.perl.org/group/perl.perl6....2.html>
r27755 | moritz++ | for a discussion where nobody stood up in defense of want().
18:55 molaf joined
masak EDIDNOTWANT 18:56
pugs_svn r27756 | moritz++ | [t/spec] remove tests for want(), which was removed a few minutes ago
pmurias hoelzro: if you put it try rakudo on moritz box you won't feel guilty for security holes 18:57
(as evalbot's security is purely trust based) 18:58
moritz_ s/purely/mostly/ ;-)
(I know, I know... just leave me my illusions... ;-)
pmichaud I recommend using moritz' box also 18:59
hoelzro pmurias: I don't care if whose box we put it on, really; I just want to make it really secure for fun =)
(also as a matter of practicality)
19:02 charsbar left 19:03 jrtayloriv joined, jrtayloriv left
pugs_svn r27757 | moritz++ | [t/spec/TODO] new task: remove want() from assignment tests 19:04
moritz_ rakudo: sub a() { 3, 4 }; my $x; my @a = ($x = a(), 4); say @a.perl
p6eval rakudo 4c31fb: OUTPUT«[[3, 4, 4]]␤»
19:04 charsbar joined
moritz_ is that correct? or should it be [[3, 4], 4]? 19:05
masak thinks it correct 19:07
oh wait.
rakudo: sub a() { 3, 4 }; my $x; my @a = ($x = a(), 4); say $x.perl
p6eval rakudo 4c31fb: OUTPUT«[3, 4, 4]␤» 19:08
19:08 mberends joined
masak moritz_: :) 19:08
moritz_ masak: that doesn't really answer my question.
masak moritz_: well, things flatten, don't they? 19:09
moritz_ masak: there's also the question of relative precedence of , and = 19:10
masak moritz_: I believe S03 is very clear on that.
b2gills rakudo: sub a() { 3, 4 }; my $x; my @a = (($x = a()), 4);
p6eval rakudo 4c31fb: ( no output )
b2gills rakudo: sub a() { 3, 4 }; my $x; my @a = ($x = a(), 4); say $x.perl 19:11
p6eval rakudo 4c31fb: OUTPUT«[3, 4, 4]␤»
19:11 rjh left
pmurias moritz_: mostly? don't you fully trust us? ;) 19:11
b2gills rakudo: sub a() { 3, 4 }; my $x; my @a = (($x = a()), 4); say $x.perl
p6eval rakudo 4c31fb: OUTPUT«[3, 4]␤»
moritz_ masak: if naiively read the precedence table in S03 then it tells me that = is tighter, and therefore $x should only get the elements from a()
pmurias hoelzro: releasing the code will make it more secure by having the exploits found sooner 19:12
masak moritz_: no, there's a whole section on list assignment.
hoelzro pmurias: then that's what I'll do
masak moritz_: let me find it for you.
line 1785. 19:13
moritz_ S03:1785 # too lazy to open an editor
masak and it carries on explaining until S03:1861. 19:14
pmurias hoelzro: if you want true obscurity you can deploy something else than you release (not recommending it) ;)
jnthn moritz_: + liste => { ... }, # typo 19:16
moritz_ jnthn: oops 19:17
jnthn would patch myself, but cooking
moritz_ I'll do it
19:18 rjh joined
pugs_svn r27758 | moritz++ | [S06] typo found by jnthn++ 19:18
pmichaud if the left hand side of eq is a scalar, it's an item assignment, and = binds tighter than comma 19:19
at present Rakudo doesn't do that -- it treats all ='s as list assignment
19:19 snearch joined
pmichaud sorry, it parses all ='s as list assignment 19:19
so Rakudo will incorrectly see $x = a(), 4 as being $x = (a(), 4) instead of ($x = a()), 4
19:20 donaldh left
moritz_ pmichaud: ok, that answers my question 19:20
19:20 pyrimidine joined
moritz_ and that's indepent of a possible @a = to the left of the whole expression, right? 19:20
pmichaud correct.
19:20 donaldh joined
moritz_ thank you. 19:20
19:21 sjohnson joined 19:22 ruoso joined
masak so... it's a bug? :> 19:22
moritz_ aye. 19:23
masak submits
pmurias ruoso: hi 19:24
19:25 masak left
ruoso very happy to see "want" go away 19:25
hoelzro was want the Perl 6 version of wantarray? 19:26
moritz_ yes
but it was designed to do even more than that
diakopter moritz_: idea: irc.pugscode.org syntax highlight the rakudo: lines
ruoso pmurias, hi
moritz_ diakopter: you're evil, I love you
diakopter :)
even better idea 19:27
moritz_ all perl6: lines
diakopter syntax highlight them with the grammar/parser as it existed in source repo at the time the line was sent
19:28 kane_ left
moritz_ bah, additional state 19:28
diakopter (kidding)
19:28 kane_ joined
moritz_ it was already an ugly hack to support nickname senstive r\d+ links 19:29
19:29 Guest77012 left, bionoid joined
moritz_ so that rakudo_svn and pugs_svn r\d+ would link to the correct repo each 19:30
diakopter sillier idea: for historical-to-now comparisons, display how the current rakudo outputs each line as well as how it did at the time.
19:30 bionoid is now known as Guest44770
diakopter silliness-- 19:32
19:35 edpratomo joined
moritz_ assign.t is a monster. 19:35
diakopter to clarify, by "outputs", I meant the result of the p6eval
19:35 simcop2387 left
moritz_ that#s what I understood too 19:36
TimToady rakudo: my $silliness = "-100"; $silliness--; say $silliness
p6eval rakudo 4c31fb: OUTPUT«-099␤»
19:37 jauaor left
pugs_svn r27759 | moritz++ | [t/spec] remove some instances of want() from assignment tests 19:38
diakopter rakudo: my $silliness = "-100-"; $silliness--; say $silliness
p6eval rakudo 4c31fb: OUTPUT«-099-␤»
19:39 diakopter is now known as human_fuzzer, simcop2387 joined, human_fuzzer is now known as diakopter
diakopter rakudo: my $silliness = "100-099"; $silliness--; say $silliness 19:40
p6eval rakudo 4c31fb: OUTPUT«100-098␤»
ruoso TimToady, sorry to keep bugging you about the same issues... but... any idea on the S11 issue? 19:42
moritz_ rakudo: say 5.elems
p6eval rakudo 4c31fb: OUTPUT«1␤»
moritz_ rakudo: my $x; @$x = 2, 3, 4; say $x.perl 19:44
p6eval rakudo 4c31fb: OUTPUT«Symbol '@$x' not predeclared in <anonymous> (/tmp/NMTyGe1XuW:2)␤in Main (src/gen_setting.pm:3343)␤»
pmurias perl6: class Foo {multi foo(Int $foo) {say "hi"}};class Bar is Foo {multi foo($foo) {say "hello"}};Bar.foo(1);Bar.foo("hi")
p6eval elf 27759: OUTPUT«No viable candidate for call to multimethod foo($,#) at (eval 132) line 20␤ at ./elf_h line 5881␤»
..rakudo 4c31fb: OUTPUT«Method 'foo' not found for invocant of class ''␤»
..pugs: OUTPUT«*** No such method in class Bar: "&foo"␤ at /tmp/v2OzMvvCPK line 1, column 93-103␤»
moritz_ rakudo: my $x; @($x) = 2, 3, 4; say $x.perl
p6eval rakudo 4c31fb: OUTPUT«2␤»
pmurias perl6: class Foo {multi foo(Int $foo) {say "hi"}};class Bar is Foo {multi foo($foo) {say "hello"}};Bar.new.foo(1);Bar.new.foo("hi")
p6eval rakudo 4c31fb: OUTPUT«Method 'foo' not found for invocant of class 'Bar'␤»
..elf 27759: OUTPUT«No viable candidate for call to multimethod foo(Bar,#) at (eval 134) line 20␤ at ./elf_h line 5881␤»
..pugs: OUTPUT«*** No such method in class Bar: "&foo"␤ at /tmp/8YR3Qastbs line 1, column 93-107␤»
pmurias perl6: class Foo {multi method foo(Int $foo) {say "hi"}};class Bar is Foo {multi method foo($foo) {say "hello"}};Bar.foo(1);Bar.foo("hi") 19:45
p6eval pugs, rakudo 4c31fb: OUTPUT«hello␤hello␤»
..elf 27759: OUTPUT«hi␤hello␤»
edpratomo rakudo: ([0, 0], [1, 1]).grep({say .perl; 1});
p6eval rakudo 4c31fb: OUTPUT«0␤1␤» 19:46
edpratomo pugs: ([0, 0], [1, 1]).grep({say .perl; 1});
p6eval pugs: OUTPUT«\(0, 0)␤\(1, 1)␤»
pmurias ruoso: is rakudo correct in my snippet? 19:47
ruoso pmurias, yes... TimToady recently localized the multis to the current class
pmurias so ClassHOW needs fixing here 19:48
ruoso pmurias, yes... it does 19:49
edpratomo hello 19:50
pmurias hi
edpratomo seems that rakudo is incorrect in ([0, 0], [1, 1]).grep({say .perl; 1});
?
19:51 pyrimidine left
moritz_ edpratomo: List.perl is not fully correct, yes 19:51
pmichaud we're still waiting for a clear spec on parens, captures, and commas
edpratomo i mean, $_
moritz_ rakudo: my $x = \(3, 4); say $x.perl; say $x.WHAT 19:52
p6eval rakudo 4c31fb: OUTPUT«[3, 4]␤List()␤»
moritz_ edpratomo: and what do you think is wrong there?
edpratomo looks correct 19:53
the $_ should contain [0,0] and [0,1], ith ink
[0,0] and [1,1]
pmichaud rakudo: say ([0,0], [1,1]).perl; 19:54
p6eval rakudo 4c31fb: OUTPUT«[[0, 0], [1, 1]]␤»
pmichaud I suspect it's an issue with .grep
edpratomo yes
pmichaud rakudo: say ([0,0], [1,1]).map( { say .perl } )
moritz_ edpratomo: you're right, yes
p6eval rakudo 4c31fb: OUTPUT«[0, 0]␤[1, 1]␤11␤»
19:54 synth left
pmichaud rakudo: ([0,0], [1,1]).grep( { say .perl; } ) 19:55
p6eval rakudo 4c31fb: OUTPUT«0␤1␤»
pmichaud rakudo: ([0,0], [1,1]).grep( { say .WHAT; } )
p6eval rakudo 4c31fb: OUTPUT«Int()␤Int()␤»
19:55 hercynium left
moritz_ edpratomo: care to submit a bug report to [email@hidden.address] 19:55
pmurias ruoso: are there cases where can returns more than one method yet we know which one to choose?
edpratomo hmm never did that before
19:55 szabgab joined
edpratomo moritz_: never did that before 19:56
moritz_ edpratomo: it's enough if you copy&paste the lines from IRC, and a short line what you expected instead
edpratomo moritz_: ok
ruoso pmurias, yes... with subsets...
pmichaud rakudo: my $test = { say .perl; }; say [0,0] ~~ $test
ruoso pmurias, if more than one subset fits, the first wins
p6eval rakudo 4c31fb: OUTPUT«0␤1␤»
pmichaud it's a smartmatch bug.
moritz_ it autothreads? 19:57
pmichaud here's a better one
rakudo: my $test = { say .perl; }; [0,0] ~~ $test
pmurias ruoso: i don't understand, example please?
p6eval rakudo 4c31fb: OUTPUT«0␤»
pmichaud something is happen to the [0,0] in the smartmatch to cause it to become its first element
rakudo: my $test = { say .perl; }; [2,4] ~~ $test
pugs_svn r27760 | kyle++ | [t/spec] Test for RT #63882
p6eval rakudo 4c31fb: OUTPUT«2␤»
KyleHa Some more experienced eyeballs on that one would make me happy.
moritz_ rakudo: [2,4] ~~ .elems.say 19:58
19:58 synth joined
p6eval rakudo 4c31fb: OUTPUT«24␤» 19:58
moritz_ KyleHa: will take a look, sure
KyleHa Thanks!
pugs_svn r27761 | moritz++ | [t/spec] rewrite more assignment tests to not use want() 19:59
pmichaud rakudo: [0,0] ~~ { say .perl }
p6eval rakudo 4c31fb: OUTPUT«0␤»
ruoso pmurias, if a parameter matches the type in two candidates which use subsets, it won't be an "ambiguous exception", but the first will win
19:59 zamolxes left
moritz_ KyleHa: looks fine 20:00
KyleHa Thanks, moritz++
20:11 snearch left
pmurias ruoso: i meant the ClassHOW.^can 20:14
ruoso pmurias, what's the difference? 20:16
jnthn AFAIK, .^can returns an iterator 20:18
That (perhaps lazily) works out what methods are available to be called.
20:23 sri_kraih_ joined 20:25 icwiener joined 20:34 hudnix_ joined
pmurias what i don't understand is when does .^can returns multiple methods 20:34
jnthn pmurias: What do you mean by "multiple method"? 20:35
pmurias multiple elemens
* elements
jnthn pmurias: AFAIK, it returns an iterator that lets you get all possible methods that could be called.
pmurias there are no iterators in Perl 6
20:36 edpratomo left
jnthn In Rakudo, it'll do it (plan to implement this Real Soon Now...) using the same mechanism that is used to do ordinary dispatch and deference. 20:36
OK, something that does the Iterator role then. :-)
pmichaud I found the issue with smart matching 20:37
fixing now.
jnthn pmichaud: Nice.
pmichaud: About to start trying to untangle my bunch of patches.
pmichaud yay
PerlJam pmurias: It sure is weird to have a whole synopsis on iterators if they aren't in the language :) 20:38
pmurias sorry, i think i'm mistaken 20:39
20:40 sri_kraih left
jnthn pmurias: I didn't quite figure out what I'm going to return just yet tbh. 20:40
Might have to make a first approximation and see where masak breaks it. :-)
pmurias are the iterators supposed to be visible in the language or are they just internal plumbing 20:41
jnthn As far as I can follow it, the Iterator role exposes an interface that things that want to act iterator-y should provide.
pmurias that's my understanding too 20:42
jnthn But what goes under the hood I guess varies depending on what you're itereating.
e.g. it'll look quite different for a filehandle than what .^can returns, I guess.
pmurias but what i'm unsure of is are we supposed to use the iterator as lazy lists or using the Iterator role 20:43
jnthn In the specific case of .^can? 20:44
pmurias as lazy lists are just iterators which remember past values
jnthn: in the general case of Perl 6 programming
ruoso pmurias, they can be visible as well... there are a lot of iterators that are invisible to the user, but they are only invisible while you don't look for them
pmurias, at the moment you start looking for them, they'll become visible
jnthn ruoso: There's probably a decent bit of truth in that.
Back to .^can though, I guess you probably should be able to say 20:45
for $foo.^can('bar') -> $meth { ... }
ruoso while if you do my @a = $foo.^can, it will eagerly return the list and store on the new array 20:46
jnthn The only thing I'm not sure of is if you're also meant to be able to do $foo.^can('bar').()
Right, but that's a property of array assignment, and nothing special to .^can
ruoso jnthn, the crude return of $foo.^can is a capture, so it can intercept .() atoo
jnthn atoo? 20:47
oh, typo of too
ruoso s/a//
jnthn tries to see acronyms in everything :-/
Yeah, to invoke.
That's kinda what falls out of Rakudo's current approach. Kinda.
pmurias ruoso: what does $foo.^can return? 20:48
ruoso it returns something that DWIM to call the desired method
but can be used to iterate in all the possible candidates as well
20:49 hercynium joined
ruoso $foo.^can is a proto-invocation 20:49
jnthn is glad ruoso saw it that way too.
Given that's what I implemented. :-)
(not for .^can yet, but it's the underlying plumbing)
If it's really that simple, .^can might be about to be very easy to implement in Rakudo... 20:50
ruoso there's one extra bit to that...
pmurias ruoso: possible candidates being all the inherited methods?
ruoso which is that the invocation $foo.^can is a multi, not private to that object
jnthn That and also also multi candidates.
ruoso: erm? 20:51
It's a call on the metaclass, no?
ruoso jnthn, sorry... that didn't look right
the dotty is lexically bound
it's really like if "infix:<.>" was an operator 20:52
getting the list of valid candidates is up to the metaclass
deciding how to invoke the list is up to the current lexical scope 20:53
jnthn Right, but it's the metaclass that returns the list (or iterator that would produce the list).
ruoso at least that's how I understood some TimToady wordings in irc
jnthn, exactly
jnthn OK, I think we're agreeing. :-)
ruoso it's about nextsame and stuff 20:54
jnthn I'm not quite sure infix:<.> is really a multi-dispatch though.
ruoso doing it in the caller side makes it easier
jnthn, it's *kinda*
jnthn, TimToady said something about it a while ago
jnthn Oh sure, the metaclass provides the candidate list, and nextsame just consumes it. In Rakudo, nextsame ain't really all that special.
The nasty is all in the dispatcher. 20:55
ruoso yeah... it becomes just a special contextual variable
$*CURRENT_INVOCATION_CANDIDATE_LIST or something longer 20:56
and harder to type
jnthn Aye. Right now, it's kept hidden away in Rakudo.
It's stored as a lexical for sure.
But under an unsigl'd name.
ruoso I'm not yet sure how that works with the idea that not all dispatching happens through .^can 20:57
jnthn In Rakudo, it pretty much does.
ruoso but maybe that's just optimization
20:58 pmurias left
ruoso and .^can should return a proto-call even if that can't be decomposed into a list 20:58
jnthn It's the most natural way to build Perl 6 dispatch semantics on the primitives Parrot provides.
...proto-call? 20:59
ruoso jnthn, yeah... considering $foo.^can('method').() 21:00
jnthn I don't understand the term "proto-call". 21:01
ruoso it's a "soon-to-be-a-call"
jnthn Unless you just mean "something callable"?
Ah, OK.
Well, if you bother to call it rather than just checking you can. But yes. :-)
ruoso jnthn, but in the low-level, it's probably going to be called unless known otherwise 21:02
since dotty semantics are complicated 21:03
i.e.: nextsame and stuff
jnthn Oh yes, I'm not saying we optimize away it's callability or anything.
There'd be no saving from a Rakudo angle anyway. 21:04
21:04 molaf left
ruoso the point is that this is a pessimization for SMOP 21:04
jnthn To make it also callable?
ruoso since SMOP can handle direct calls
jnthn, having to do it through .^can in order to support nextsame
ruoso decommute & 21:05
ruoso later &
jnthn Ah, I see
cu
21:05 ruoso left 21:06 Su-Shee left 21:07 desertm4x left 21:14 Whiteknight joined, icwiener left
jnthn does a make spectest and wonders what the fallout will be 21:22
21:33 [particle]1 joined 21:37 lumi__ joined, lumi_ left
jnthn awww...fails :-( 21:40
21:42 maja_ joined 21:47 cdarroch left 21:50 [particle] left 21:51 KyleHa left 21:52 skids_ left 21:54 maja__ left 21:59 hoelzro left
jnthn rakudo: my $x = "\x[391]\x[391]\x[391]"; say (--$x ~~ Failure) 22:00
p6eval rakudo 4c31fb: OUTPUT«1␤» 22:01
jnthn rakudo: my $x = "\x[391]\x[391]\x[391]"; say --$x)
rakudo: my $x = "\x[391]\x[391]\x[391]"; say --$x
p6eval rakudo 4c31fb: OUTPUT«Syntax error at line 2, near ")"␤in Main (src/gen_setting.pm:3343)␤»
rakudo 4c31fb: OUTPUT«Decrement out of range␤␤»
jnthn Aww, how on earth have I managed to turn that into a Null PMC Access locally...
22:02 mj41 joined
jnthn oh 22:02
I haven't
jnthn was looking at the wrong tests 22:03
(it waws the hebrew ones, and they were todo...) 22:04
rakudo: my $x = "\x[5d0]\x[5d0]\x[5d0]"; say --$x 22:05
p6eval rakudo 4c31fb: OUTPUT«אאא␤»
22:11 missingthepoint joined, M_o_C left
jnthn rakudo: say Callable ~~ Any 22:13
p6eval rakudo 4c31fb: OUTPUT«1␤»
22:13 alester left 22:15 explorer left
missingthepoint rakudo: say Dolphin ~~ Fish 22:15
p6eval rakudo 4c31fb: OUTPUT«Could not find non-existent sub Fish␤»
missingthepoint hmm :(
jnthn rakudo: class Fish { }; class Dolphin is Fish { }; say Dolphin ~~ Fish; # Rakudo just needs a lil education. 22:16
22:16 iblechbot left
p6eval rakudo 4c31fb: OUTPUT«1␤» 22:16
missingthepoint thanks for edumacatin' Rakudo. :) 22:17
jnthn rakudo: say "we don't need to edukashun" 22:19
p6eval rakudo 4c31fb: OUTPUT«we don't need to edukashun␤»
22:19 SmokeMachine left
pmichaud o/ all in all it's just another brick from The Wall o/ 22:20
dalek kudo: 53059d1 | pmichaud++ | (4 files):
Correct Code.ACCEPTS and Regex.ACCEPTS . [1,2,3] ~~ { ... } should now work properly.
22:21
missingthepoint pmichaud: haha :=)
jnthn pmichaud: I'm almost there with being ready to merge...
pmichaud: I discovered a curiosity though. 22:22
rakudo: say Object ~~ Bool::True
missingthepoint giggles at that little exchange...
p6eval rakudo 4c31fb: OUTPUT«1␤»
jnthn pmichaud: Well, this fails when we stop lying.
Since
method ACCEPTS($topic) { 22:23
return self;
}
Guess what? Now that Object !~~ Any (phew!) the sig don't bind.
So I've changed it to
method ACCEPTS(Object $topic) {
return self;
}
moritz_ so it won't autothread
jnthn But that now means it won't auto-thread. :-/
(we were actually wanting Object ~~ Bool::True to work somewhere in the setting.) 22:24
(not even in a test...but in the setting.)
pmichaud it's okay with me if ~~ Bool::True doesn't autothread, because the result is the same.
The bigger question is whether other .ACCEPTS methods need to accept Object parameters 22:25
jnthn Right.
I'm not feeling quite awake enough ATM to ponder if there's wider fallout.
moritz_ rakudo: say ?(none(1, 2) ~~ True)
p6eval rakudo 4c31fb: OUTPUT«0␤»
jnthn I pass all of the junctions tests anyways.
moritz_ that will give 1 if it stops autothreading.
pmichaud where in the setting are we looking for Object smart matching against true
jnthn pmichaud: WALK 22:26
See sig
Matcher :$omit = False, Matcher :$include = True
22:26 kst left
jnthn That is, we default to omitting nothing and including every class's methods. 22:26
pmichaud moritz_: I can argue that rakudo is wrong in the example you just gave
jnthn It doesn't feel like a crazy thing to do to me... 22:27
22:27 kst joined
pmichaud for the moment, I think the correct thing is for Object ~~ Bool::True to fail. 22:27
jnthn The thing is 22:28
It doesn't just return false
pmichaud The smart matching table says "Any".
moritz_ pmichaud: the smartmathing table says 'Any Bool', not 'Object Bool'
jnthn It....explodes.
pmichaud moritz_: exactly.
jnthn: we should fix WALK then to avoid smart matching Object against truth.
what line is doing it now?
jnthn if $include.ACCEPTS($class) && !$omit.ACCEPTS($class) { 22:29
pmichaud (what file?)_
jnthn For WALK?
Object.pm
I can easily work around it. 22:30
Just defauting to Ture (always matches) and False (never matches) felt like the Right Thing.
pmichaud jnthn: it may indeed be the Right Thing. But for now I think we go with a literal interpretation of the smart matching table ("Any", not "Object") and perhaps ask for clarification from p6l 22:31
jnthn OK
Beyond that, tests are looking quite good after I fix up other fallout. 22:32
moritz_ rakudo: say undef ~~ Any
pmichaud because whatever we decide for boolean could be argued for many of the many other tests
p6eval rakudo 4c31fb: OUTPUT«1␤»
jnthn Aye.
In that case, it may be a can of worms better left closed.
pmichaud and it's pretty clear to me that the others need to autothread
in .WALK I think I would just skip over Object 22:33
although I guess that doesn't quite work
jnthn ?
No, it doesn't. 22:34
pmichaud afk, dinner 22:35
jnthn Time to make dalek flood... 22:37
pugs_svn r27762 | jnthn++ | [t/spec] Start cleaning up/correcting some traits tests. Think Rakudo should be able to run this one soon, or at least much of it. 22:38
jnthn rakudo: { .^can('n') }.signature.perl.say 22:39
p6eval rakudo 4c31fb: OUTPUT«:(Object $_?)␤»
22:40 explorer joined, Limbic_Region joined
moritz_ 53059d1c5435fce71029ecfad98b0b6f350d5d1e seems to break t/spec/S06-multi/type-based.t 22:41
pugs_svn r27763 | jnthn++ | [t/spec] ::FOO gives a Null PMC back. .does is now more sensitive to this (probably now that it delegates to the right place). Anyway, was probably slightly bogus pass in the first place. 22:42
r27764 | jnthn++ | [t/spec] .does(Enum::Value) is on shakey ground, since the values are not roles, like the enum itself. Having investigated, we passed this for extremely bogus reasons before now. But I suspect these may not even be valid tests. Skip 'em for now with a note saying as much. 22:44
jnthn moritz_: Is that pmichaud's latest push? 22:45
moritz_ jnthn: yes
jnthn moritz_: lol. 22:46
moritz_: I also broke and then fixed that test. ;-)
In the bunch of stuff I'm about to push.
moritz_ will test again tomorrow
jnthn checks his and pm's code at least build together. 22:47
yes
push! 22:48
o hrovno
yes, I see the breakage too.
pmichaud: See above... t\spec\S06-multi\type-based.t br0ked. :-( 22:49
Limbic_Region jnthn - are you in Japan yet?
jnthn Limbic_Region: Not yet
Limbic_Region: Not for a while yet.
Limbic_Region so Korea comes after that? 22:50
22:50 nihiliad left
jnthn Limbic_Region: Just trying to learn some stuff ahead of going. :-) 22:50
Right.
Having about 2 weeks in each.
And hitting YAPC::Asia while I'm in Japan.
Limbic_Region do you know what parts of Korea you will be in? I might make some recommendations on things to do and places to go
jnthn So far I know I will arrive into Busan. 22:51
Other than that, I didn't plan much yet.
Got the Japan bit pretty planned out, but not sure about Korea.
Suggestions welcome.
Limbic_Region oh, well - that place is know for it's Russian Juicy Girls :-)
jnthn ...do I even want details? ;-)
Limbic_Region waste of money - nuff said 22:52
regarding where to go and what to do - I will ping you on FB - best to leave the channel to somewhat on topic
jnthn is almost afraid to google
Limbic_Region ok - if you must know 22:53
jnthn "Hot girls pictures from Pakistan and celebrities juicy pictures..."...yeah, thanks Google.
Limbic_Region in the cities where ever the US has US military, the bars employee what they call juicy girls
jnthn Ah.
Limbic_Region a juicy girl works for the bar and will sit with you, and get rather friendly as long as you are buying her drinks
the drinks are expensive - probably 10,000 won or more 22:54
jnthn ...which are probably sold to you at inflated prices, in order to pay for said girl?
dalek kudo: 82b59c4 | jnthn++ | :
Merge latest.
kudo: d043978 | jnthn++ | src/pmc/perl6multisub.pmc:
Fix a bug in the narrowness analyzer for multi-dispatch (not being narrower is a necesary but not sufficeint condition for being tied).

spectest-progress.csv update: 423 files, 12043 passing, 0 failing
22:54 dalek left
Limbic_Region but regardless of what you order - the girl is likely only drinking juice 22:54
jnthn lolibrokedalek
Limbic_Region this is so she can continue drinking all night long
jnthn Cunning plan.
22:54 dalek joined
jnthn For her. Not for her drunkee. 22:55
"This time is generally like a date in that the girl sits with the customer, asks questions about his life and pretends to be generally interested." 22:56
I love the insinuation that on a date, a girl only pretends to be interested. ;-)
wikipedia++
OK, thanks for the warning.
erm, any...good...recommendations for Korea? ;-) 22:57
Limbic_Region plenty
regardless of what your tastes are
jnthn Not Russian Juicy Girls, it seems.
Limbic_Region buddhists temples, cuisine, night life, etc
they are a waste of money - if you want that kind of entertainment, I could point you in the right direction 22:58
jnthn No, I don't really feel a need to pay people to pretend to like me. I'm not *that* bad. :-)
Cuisine...now THAT is what I like.
Limbic_Region well, the beer is mediocre at best - better to drink the soju if you like it 22:59
Limbic_Region is eating but promises to ping you on FB about food later 23:00
jnthn Not tried it, but I live in an east european country, so am plenty used to things along those lines ;-) 23:01
Cool
Probably better there than filling #perl6 with it. ;-)
Tene searches for jnthn on FB 23:02
jnthn rakudo: say "I { ("want", "don't want").pick } a Russian Juicy Girl";
p6eval rakudo 53059d: OUTPUT«I don't want a Russian Juicy Girl␤»
jnthn Well we did something right in Rakudo. 23:03
(After tonights flurry of patches, Rakudo now lies to itself a bunch less too. Now tomorrow I can get back to doing interesting things.)
pmichaud jnthn: type-based.t passes on my box
jnthn pmichaud: orly? 23:04
Hmm. :-S
pmichaud perhaps I forgot to commit something?
no, looks like it's all there to me.
gist.github.com/156813 23:05
jnthn pmichaud: Something else must be adrift. 23:06
Oh!#
pmichaud: Think I didn't realclean.
And re-Configure 23:07
pmichaud That might do it.
jnthn And you'd added a file.
(Code.pm)
pmichaud yes, that would certainly do it.
jnthn Yeah, I get fails all over the place.
That'll be why.
offs
Parrot revision r40294 required (currently r40291)
Where does it get the rev number from?
pmichaud build/PARROT_REVISION
jnthn And can we make this a warning rather than a omg-i-won't-write-makefile
No, the current revision.
I find if I've comitted that latest revsion 23:08
And I go to re-configure
Without re-configuring and re-building my Parrot
I get this issue.
It's annoying.
pmichaud I don't follow (more)
the build/PARROT_REVISION I have is 40282
I haven't updated recently
jnthn OK, the latest is the one reported there
However, svn info in my Parrot build dir reports 23:09
Parrot revision r40294 required (currently r40291)
Tene jnthn: you need to reconfigure Parrot too.
jnthn oops
Tene: Yes, I know.
pmichaud it gets 40291 from the parrot_config executable.
jnthn Right, so I see.
I guess I can go re-Configure my Parrot.
I'd just prefer to see this as a warning that I might have a problem than a refusal to write a makefile. 23:10
pmichaud for the average person, it needs to be stronger than a warning.
Otherwise we end up with the case that you just ran into with Code.pm, but on a much larger scale
(i.e., you overlooked the warning that was telling you your Makefile is out of date :-) 23:11
jnthn :-P
Anyway, hopefully the missing Code.pm in my makefile was the issue. 23:12
pmichaud very likely.
jnthn The tests that failed in that test would not be fun ones to fix. ;-)
pmichaud Code.ACCEPTS was moved out of src/classes and into src/setting
jnthn Ah, OK.
I noticed it got shorter in the process. :-)
pmichaud so anything that would attempt to smartmatch against a closure would fail
jnthn Ah
Those tests that failed all would have.
pmichaud it got short yes, but that's because we *were* doing the Regex.ACCEPTS logic in Code.ACCEPTS 23:13
i.e., dating from the time when we couldn't distinguish regexes from other code blocks
jnthn Ah, you broke those two out now?
Right.
Cool
pmichaud and Array ~~ Regex is different from Array ~~ Code
jnthn That probably fixes other lingering odnesses.
pmichaud which is why [1,2,3] ~~ { .say } was producing the wrong result
the Regex.ACCEPTS probably deserves to be written in setting as well 23:14
but I figured I'd just get it working for now.
jnthn working is good.
Are you planning to work on the PGE/NQP refactors this week?
pmichaud either that or on the REPL 23:15
likely both
jnthn prefers PGE/NQP refactors
That'll unblock the lexicals refactor.
Which I really want.
pmichaud well, the REPL may be my Rakudo Day
I need to catch up on those, too.
jnthn *nod*
I will do one this week.
pmichaud I'm going to do 2-per-week until I catch up
conference season has kept me busy :-) 23:16
jnthn Confirm type-based.t works now
23:16 hyfen joined 23:20 donaldh left, donaldh joined 23:23 Whiteknight left 23:24 missingthepoint left
pmichaud good, glad I didn't hose anything. 23:25
23:25 ihrd joined
23:27 hoelzro joined
pmichaud afk for a bit # walk 23:27
jnthn -> sleep 23:31
23:51 synth left