»ö« | perl6-projects.org/ | nopaste: sial.org/pbot/perl6 | evalbot: 'perl6: say 3;' | irclog: irc.pugscode.org/ | UTF-8 is your friend!
Set by Tene on 14 May 2009.
00:04 stepnem left 00:05 stepnem joined 00:17 frew joined 00:19 wayland76 left, wayland76 joined
TimToady perl6: [1] ... { [ map &[+], 0, (@$_ Xxx 2), 0 ] } 00:20
p6eval rakudo 9d2934: OUTPUT«Statement not terminated properly at line 1, near "... { [ ma"␤in Main (src/gen_setting.pm:0)␤»
..elf 26879: OUTPUT«Parse error in: /tmp/LIAjpUN4uk␤panic at line 1 column 0 (pos 0): Can't understand next input--giving up␤WHERE: [1] ... { [ map &[+], 0, (@$_ ␤WHERE:/\<-- HERE␤ STD_red/prelude.rb:99:in `panic'␤ STD_red/std.rb:76:in `scan_unitstopper'␤ STD_red/std.rb:224:in `comp_unit'␤
..STD_red…
..pugs: OUTPUT«*** ␤ Unexpected "{"␤ expecting operator␤ at /tmp/vxT4WWX3Fc line 1, column 9␤»
TimToady (pascal's triangle) 00:21
00:24 bacek_ joined 00:27 jonathanturner left 00:49 meppl left 00:50 stepnem left 00:52 stepnem joined, jferrero left 00:56 eternaleye joined 01:03 azawawi joined
azawawi hi 01:03
01:07 nsh left 01:11 ElectricHeavyLan left
azawawi TimToady: how do i access CURPKG and CURPAD after using STD->parse? 01:24
01:39 Whiteknight left
TimToady there is no valid CURPKG or CURPAD outside of the parse. that's what makes it "CUR". you need to store them somewhere in a reduction action 01:45
that's more or less what the YOU_ARE_HERE is for in remembering the CURPAD for a setting 01:46
azawawi hmmm.. so how do i access the symbol table that STD builds? 01:56
01:59 sri_kraih joined
TimToady CURPAD has ::OUTER links all the way up to CORE, and CORE contains ::GLOBAL 02:00
CURPKG has ::PARENT links up to GLOBAL, as well 02:02
internally we also maintain $::CORE and $::GLOBAL 02:04
to get at things from the top
but in theory everything can be reconstructed from $CURPAD
azawawi i see, thx 02:08
02:09 angelixd left
azawawi i think i should turn S:H:P6 to use viv instead... 02:09
TimToady well, the primary purpose of viv is still to replace gimme5, so it will continue to mutate, probably
dinner & 02:10
azawawi enjoy :)
02:10 azawawi left 02:12 hercynium left 02:15 [particle] joined 02:26 dduncan joined 02:27 dduncan left 02:33 sri_kraih_ joined 02:40 sri_kraih left 03:18 skids left 03:20 donaldh left, donaldh joined 03:27 amoc joined
TimToady @tell JDlugosz the lines() function already defaults to $*ARGFILES; that's its only advantage over the method form 03:30
lambdabot Consider it noted.
TimToady rakudo: say lines 03:31
p6eval rakudo 9d2934: OUTPUT«No applicable candidates found to dispatch to for 'lines'␤in Main (/tmp/UymjXLe2qR:1)␤»
03:36 Eevee left 03:41 orafu left 03:42 orafu joined 03:46 Kisu left 04:01 alester joined 04:09 Eevee joined 04:11 REPLeffect joined 04:41 veritos joined
japhb mberends: ping 04:56
04:57 veritos left 05:27 frew left, alester left 05:44 Maghnus_ joined 05:49 cognominal joined
mberends japhb: pong # (backlogging) 05:50
moritz_ wow, longest spectest run ever: 27151 wallclock secs
(there was a suspend-to-disk in between ;-) 05:51
japhb heh
mberends: I'm working my way through HTTP::Daemon, and I was wondering about a couple things in the source that seemed odd. 05:52
05:53 harig joined
japhb For example, why does HTTP::Daemon::ClientConn have several methods that declare their invocant? 05:53
mberends looks
japhb as in method foo ($self:) {... $self.do_something ...}
Also, why is send_status_line a multi method? 05:55
(And it's got a bug, BTW: Instead of \n, it should send crlf of course, as you mention in a comment just below. ;-) ) 05:56
mberends japhb: some of this code is sloppy cruft left over from the NPW Hackathon. It needs cage cleaning :/ 05:57
japhb Care for help with said cage cleaning? 05:58
mberends nice idea. Do you already have an account on Gitorious?
japhb I signed up earlier today, but the email still has not arrived. :-( 05:59
OOC, why gitorious instead of github? 06:00
06:00 Maghnus left
mberends ok, when your account is confirmed, you can become contributor to gitorious.org/http-daemon where I plan to maintain a core version for inclusion into other projects. 06:01
japhb thanks!
japhb wonders why an automated email should take more than a few seconds .... 06:02
mberends Gitorious because I don't believe in putting all eggs into one basket.
japhb Fair enough. Though the nice thing about a DVCS is that the single basket problem gets a lot less troublesome. 06:03
mberends the proto installer also benefits from test cases on different servers. 06:04
japhb Ah, speaking of which, I was going to offer a proto patch that would try cloning personal forks even before writable masters, but I realized I don't understand how the gitorious personal fork naming convention works. 06:06
mberends the aim of this Daemon BTW is to be as faithful as possible to the Perl 5 original. Others such as masak, Tene and ruoso have re-gutted it with custom dispatchers.
japhb: that would be a nice proto enhancement, let's look at that too. 06:07
japhb mberends: OK, I'll keep the Perl 5 source around to compare against while I go. 06:08
mberends japhb: the compatibility is at the API level, keeping to the perldoc. 06:11
I'm planning to do the same with more of LWP. Gisle Aas wrote me that he doesn't mind. 06:12
japhb mberends: Nod. I meant that I planned to favor the actual behavior where it disagreed with the docs.
Why would he mind? Ownership of namespace? 06:13
mberends my broad goal is to help Perl 5 users move to Perl 6 easily.
japhb A worthy goal. :-) 06:14
mberends I did not want to tread on toes if others are already planning their Perl 6 moves.
japhb Good manners, that. 06:15
06:15 masak joined
mberends For example, the DBI side for Perl 6 has active developers. 06:15
masak greetings, perl6ers. 06:16
mberends masak: good day, barefoot butterfly-catcher!
japhb mberends: Oh? Where? I was wondering if I needed to throw tuits at that
masak: greetings!
masak :)
mberends sorry, afk & #@home 06:17
06:19 azawawi joined
azawawi good morning p6 :) 06:19
masak azawawi: mornin' 06:20
std: class Dog {}; my @mammals; grep Dog, @mammals;
p6eval std 26879: OUTPUT«ok 00:02 37m␤»
masak rakudo: class Mammal {}; class Cat is Mammal {}; class Dog is Mammal {}; my @mammals = map { .new }, Cat, Dog, Cat, Cat, Dog; say (grep Dog, @mammals).perl 06:21
p6eval rakudo 9d2934: OUTPUT«No applicable candidates found to dispatch to for 'grep'␤in Main (/tmp/7ezBsN9qE9:1)␤»
masak that's what I thought. 06:22
now, let's see what the spec says about this.
moritz_ it probably says Matcher
06:22 iblechbot joined
moritz_ well, maybe it's an easy fix... 06:23
masak it says Matcher.
furthermore, Matcher is 'defined' in S29, but then used all over S32.
moritz_ runs spectest
masak some enterprising mind might want to re-mention it in S32.
rakudo: class Mammal {}; class Dog is Mammal {}; say Dog.can('ACCEPTS') 06:24
p6eval rakudo 9d2934: OUTPUT«1␤»
moritz_ I'm trying to simply remove the type constraint, and instead of $test($_) I do $_ ~~ $test
now running spectest if it's that simple. 06:25
masak rakudo: subtype Matcher of Object where { .can('ACCEPTS' }; class Mammal {}; class Dog is Mammal; say Dog ~~ Matcher
p6eval rakudo 9d2934: OUTPUT«Statement not terminated properly at line 1, near "of Object "␤in Main (src/gen_setting.pm:0)␤»
masak rakudo: subtype Matcher of Object where { .can('ACCEPTS' }; class Mammal {}; class Dog is Mammal {}; say Dog ~~ Matcher
p6eval rakudo 9d2934: OUTPUT«Statement not terminated properly at line 1, near "of Object "␤in Main (src/gen_setting.pm:0)␤»
moritz_ masak: sub*set*
masak rakudo: subset Matcher of Object where { .can('ACCEPTS' }; class Mammal {}; class Dog is Mammal {}; say Dog ~~ Matcher
p6eval rakudo 9d2934: OUTPUT«Statement not terminated properly at line 1, near "('ACCEPTS'"␤in Main (src/gen_setting.pm:0)␤»
masak moritz_: I often do that mistake lately.
japhb Missing right paren
masak rakudo: subset Matcher of Object where { .can('ACCEPTS') }; class Mammal {}; class Dog is Mammal {}; say Dog ~~ Matcher 06:26
p6eval rakudo 9d2934: OUTPUT«Re-declaration of type Matcher at line 1, near "; class Ma"␤in Main (src/gen_setting.pm:2668)␤»
masak :/
06:26 justatheory left
masak hasn't woekn up correctly yet, evidently 06:26
rakudo: subset Matcher of Object where { .can('ACCEPTS') }; class Mammal {}; class Dog is Mammal {}; say Dog ~~ Matcher
p6eval rakudo 9d2934: OUTPUT«Re-declaration of type Matcher at line 1, near "; class Ma"␤in Main (src/gen_setting.pm:2668)␤»
moritz_ rakudo: say Matcher
pugs_svn r26880 | azawawi++ | [S:H:P6] bundled latest std in 0.56 06:27
p6eval rakudo 9d2934: OUTPUT«()␤»
moritz_ masak: it exists already
masak oh.
moritz_ that's why you et a Re-declaration error
*get
masak rakudo: say Dog ~~ Matcher
p6eval rakudo 9d2934: OUTPUT«Could not find non-existent sub Dog␤»
masak erm.
rakudo: class Mammal {}; class Dog is Mammal {}; say Dog ~~ Matcher
p6eval rakudo 9d2934: OUTPUT«1␤» 06:28
masak it can ACCEPTS.
moritz_ $ perl6 -e '([], 3, 4, []).grep(Int).perl.say' 06:30
[3, 4]
that works locally here
masak moritz_++ 06:31
japhb Where is the Tags.pm that creates HTML tags? proto doesn't seem to know about it ...
masak japhb: it's in web right now.
japhb: not because it uses anything else in that project, but because it was created in the context of that project. :)
our policy on whether to put 'third-party' things in the web repository or outside of it is a bit hazy still... 06:32
japhb nod. 06:33
But where is it? I can't seem to find it in the web repo that proto checks out ... is proto out of date WRT to the correct repo location/owner?
moritz_ lib/Tags.pm 06:34
github.com/masak/web/blob/d5d439e9f...ib/Tags.pm
masak japhb: no, proto should be up-to-date on that account. 06:35
japhb hmmmm: ~/git/web$ ls lib
Test.pm Web Web.pm
masak japhb: update?
japhb Just did the initial checkout earlier today. But sure, why not? ;-)
masak japhb: either 'git pull origin master' in the web dir, or './proto update web' in the proto dir.
japhb: I'm pretty sure we have more than that in lib/ :) 06:36
japhb Hum.
It appears to have somehow gotten locked into a fork I did a LONG time ago: 06:38
$ git pull origin master
From [email@hidden.address]
japhb kicks it in the head 06:39
Oh yeah, that looks MUCH better
PEBCAK
masak japhb: might it be that there was already a web dir where proto tried to install the official one?
japhb: if proto didn't signal an error at such a situation, that's an error. :) 06:40
japhb masak: Yes, it's possible. My ~/git is getting rather bloated.
masak when I go './proto install web', it says 'Won't install web: already installed.' 06:41
japhb masak: I had started the whole party with './proto install all' which seems to need a bit of hand-holding. It is quite possible I held them incorrectly. 06:42
masak oh, that's not something I've tested properly.
but ISTR that it skips things already installed.
so you'd have wound up with your web dir untouched.
azawawi moritz_: any idea if this is a rakudo bug, sial.org/pbot/36698 ? 06:44
moritz_: it only print 1 on my win32 machine
masak azawawi: here, too. 06:45
azawawi: works if I do s/pod1/pod/, though.
azawawi masak: yeah but it doesnt print out any error or anything... 06:46
masak: silently ignores the rest...
masak azawawi: you're right, that is odd.
azawawi: submit a rakudobug.
japhb Ack. Pbbbt. It looks like web and http-daemon have near-identical Daemon.pm files. Which one is canonical? 06:48
masak http-daemon, at least after mberends++ finishes his refactor. 06:49
the one in web is going away.
(because Web.pm is supposedly engine-agnostic) 06:50
japhb Refactor of what? Web.pm's use of an engine, or the copy of Daemon.pm in http-daemon?
(I don't want to start hacking on http-daemon if the code is about to get mutilated out from under me.)
azawawi masak: how do i submit a rakudo bug, rakudobug at perl.org? 06:51
masak japhb: in any case, you should confer with mberends, if you haven't done so already. 06:52
azawawi: aye.
japhb masak: I was a bit ago (before you came in), but he had to run apparently
azawawi masak: and where can i view current rakudo bugs?
masak azawawi: rt.perl.org/rt3/
azawawi hates RT :) 06:53
masak japhb: all I know is that the HTTP::Daemon in web works. I don't know whether the one in http-daemon is dated or improved.
japhb: maybe the changelog will tell you that.
japhb I diffed. The only difference is four lines of POD. 06:54
masak japhb: :)
06:55 DemoFreak joined
masak japhb: I'd recomment hacking on http-daemon, since that's the one not going away. 06:55
japhb However, I guess no time like the present to begin. :-)
nodnod
masak mberends: let's say A depends on B. if I go './proto install A B', with none of them installed, the following will happen: proto will fetch A, fetch B, and build them. then it will unnecessarily update B, because it's still in the list of things to be installed-or-updated. this is (I guess) acceptable, but a bit... unnecessary. 07:05
07:05 payload left 07:08 charsbar left
moritz_ forgot to fix the sub form of grep, so spectesting again... 07:08
masak mberends: hm, seems like that's solvable. fetch-and-build-projects-and-their-deps simply doesn't take this case into account, but it has all the information needed to do so. I might fix during the day. 07:09
07:11 charsbar joined
moritz_ www.perlmonks.org/?node_id=764687 # Contexts in Perl 6 07:14
in case anybody missed it ;-)
masak looks like a good discussion. 07:17
07:20 donaldh left 07:21 donaldh joined 07:22 payload joined 07:25 cognominal left
azawawi submitted rakudobug #65782 07:32
07:33 azawawi left
moritz_ I don't know if that's actually a bug. 07:34
pugs_svn r26881 | moritz++ | [t/spec] grep with non-Code matcher 07:36
masak moritz_: me neither, but I think it's something that at least needs clarification. 07:37
moritz_++ # r26881 07:38
dalek kudo: d0fbfbe | moritz++ | src/setting/Any-list.pm:
accept more general matchers in grep
masak yay! 07:41
moritz_ doesn't quite understand masak's excitement about such a simple patch ;-)
masak moritz_: it's quite significant, I think. 07:42
moritz_ well, having built-ins written in Perl 6 makes that really trivial ;-) 07:43
masak hm, the usage example of map in S32/Containers.pod:
.. @addresses = map { %addresses_by_name{$_} }, @names;
(the two dots were to un-confuse lambdabot) 07:44
moritz_ can of course be done with a slice
masak yes, that was my point.
moritz_ @test
masak making it a slightly un-good example for didactic reasons.
moritz_ a space is enough already ;-)
masak moritz_: ah, yes, will do that next time.
moritz_ well, feel free to think of a better example ;-) 07:45
masak I will.
actually, I'm thinking of just cutting it for now.
07:45 payload left
masak the spec should be succinct, IMHO. and the example doesn't say anything the signature didn't already convey. 07:46
moritz_ +1
masak makes it so
pugs_svn r26882 | masak++ | [S32/Containers.pod] removed examples from 'map' 07:50
r26882 | masak++ |
r26882 | masak++ | The examples didn't particularly add anything that the synopsis didn't
r26882 | masak++ | already make clear. Also, the first example could have been written better
r26882 | masak++ | with a hash slice, making it didactically non-ideal.
r26883 | masak++ | fixed a typo and a missing comma
masak I used to love SVN because it was not CVS. now, only git-svn can make SVN bearable for me. funny how that works. 07:51
moritz_ @karma masak
lambdabot masak has a karma of 184
Matt-W It's because you realised that SVN is just CVS done less badly 07:52
moritz_ yes, I'm sold for DCVS' too
Matt-W The fundamental flaws are still there
masak Matt-W: they are indeed.
moritz_ there's a nice rant from linus torvalds on youtube (or google tech talk)
masak Matt-W: but I can feel smug for getting some advantages out of git while working against SVN.
moritz_ he's got a point, but he really behaved like an asshole 07:53
masak moritz_: I was torn over that video too.
moritz_: first time I watched it, I thought 'asshole'.
moritz_ well, I still think that
masak moritz_: second time, I realized that he was basically painting his enemy really clearly.
moritz_: still an asshole, I guess, but I see now that it was very deliberate. 07:54
moritz_ just because somebody hasn't been exposed to a different technology doesn't mean he's stupid or brain-dead or whatever
masak moritz_: true.
moritz_: being nuanced is hard. he's being very clear, and trading humility and politeness for clarity. 07:55
I wouldn't. but then again, I'm not him.
he's known for doing the same on mailing lists etc. 07:56
moritz_ ah well. I still use his software ;-)
though I wouldn't mind having a few different kernels to plug in into my debian, just to have alternatives 07:57
there's a debian-freebsd port, and an unoffical opensolaris port (called nexenta)
I hope those two work out
masak is a Hurd lover at heart
07:58 jferrero joined
moritz_ do you actually run it? 07:58
rakudo: [1, 2, 3.4, [], {}].grep(Num).perl.say 07:59
p6eval rakudo d0fbfb: OUTPUT«[1, 2, 3.4]␤»
08:00 jferrero left
japhb mberends: gitorious account finally acquired. Turns out their account verification email came from a misconfigured server, and the anti-spam rules at my hosting provider rejected it. 08:04
masak moritz_: no, but I've always wanted to. I hear there's a debian-hurd in the works too. 08:05
japhb masak: has been for AGES.
masak japhb: sounds like something the github team might want to hear about. 08:06
japhb (in the works I mean)
masak japhb: if you say so. :) I only heard about it recently.
japhb masak: Already spoke to gitorious support.
masak japhb++
their page says it's 'not officially released yet'. www.debian.org/ports/hurd/ 08:07
japhb masak: yep. Been that way since IIRC even before Hurd could boot. 08:09
Debian was one of the first distros to try it.
masak Debian++
I think it's my favourite distro by far.
the more I learn about it, the more I like it.
japhb Ditto that. 08:13
masak: './proto update http-daemon' does not seem to make the .pir files, even though it claims to have built it. 08:16
08:16 ejs joined
japhb do a 'make clean' in http-daemon, you can see what I mean. 08:17
08:17 ejs1 joined
masak japhb: yes, there's a flaw in proto that causes that. it doesn't build any project correctly right now. 08:17
japhb masak: any idea what it is? Or should I just go wildly spelunking? :-)
masak japhb: on my box, it has to do with 'env'. 08:18
japhb raises an eyebrow
masak or rather, the paths sent to env causes env to die. 08:19
so all my make.log files contain the same two lines: 'env: projects: No such file or directory', repeated twice.
08:22 ejs left
japhb masak: bug found, cleaning up my diff 08:23
masak wow. japhb++
I bet it was somewhere close to line 332, 'XXX: Need to have error handling here'
japhb A couple lines above it. :-) 08:24
masak :)
japhb Rakudo is now the second thing that makes me want a faster CPU in my laptop. The first being flash. 08:25
pasteling "japhb" at 76.191.190.8 pasted "Fix for proto project building problems, plus trivial inline refactoring to allow (commented out) verbosity" (31 lines, 1.3K) at sial.org/pbot/36699 08:27
masak japhb: thanks. and, um, "d'oh!". 08:30
japhb *chuckle*
masak japhb: I'll accept the part of the patch that fixes the problem. I try to avoid both commented-out code and verbosity in proto.
08:31 ejs2 joined
japhb masak: No problem. But you might consider dropping the commented warn, but keeping the $cmd refactoring ... since I find I often come back to wanting to debug commands I'm about to execute. 08:32
masak japhb: understood.
will do.
japhb: there, pushed. thanks. 08:38
(I added you to the AUTHORS file as well.)
08:41 ejs1 left 08:43 Maghnus_ left
japhb masak: thanks! 08:44
masak no, thank _you_. 08:45
japhb ... Mr. Acavano!
08:46 mikehh_ joined
masak Wise Guys, 1986. 08:46
I had to Google for the quote. :)
japhb heh 08:47
08:47 ejs1 joined
masak seems like a nice movie. 08:47
08:49 cognominal joined
japhb I recall it being fun ... but then again that was almost a quarter century ago. :-) 08:52
08:52 revdiablo left 08:53 clintongormley joined 08:54 ejs2 left
moritz_ only knows "Wise Guys" as a German a-capella band 08:59
08:59 mikehh left
Matt-W moritz_: are they any good? 09:00
moritz_ Matt-W: very good IMHO
Matt-W: very funny, if you understand the lyrics
Matt-W moritz_: do they have a website?
moritz_ www.wiseguys.de/ 09:01
dalek kudo: 222993f | moritz++ | docs/ChangeLog:
[docs] update ChangeLog
 09:03
Matt-W moritz_: Wow. They are fantastic. 09:04
09:04 ejs joined
Matt-W I'm only understanding about a quarter of the lyrics of course 09:04
moritz_ understandable
I've been on two of their concerts 09:05
the first one was fine (they weren't well known at that time), but the second one was really great
Matt-W I think one requires a very high level of language to understand songs in a second language 09:06
09:06 clintongormley left
moritz_ aye. It really helps to read the lyrics once, IMHO 09:06
Matt-W Probably
I can definitely appreciate their technical skill though
I have to tell my singing teacher about this 09:07
She'd love it
being a former Swingle Singer
09:08 payload joined
Matt-W moritz_: even better, I can get their stuff from Amazon UK :) Thanks for mentioning them 09:09
jnthn H H
Matt-W HAI jnthn! 09:10
u in ar rakudo makin us feeturs?
moritz_ Matt-W: you're welcome... have fun listening! ;-) 09:11
masak jnthn: hi! 09:12
Matt-W moritz_: hmm CD for £16 or MP3 256kbps for £8.79... I have to think about that for, oh, two or three seconds
moritz_ Matt-W: IMHO the two newest albums ("Frei" and "Radio") are the best
09:13 azawawi joined
azawawi moritz_: ping 09:13
moritz_ azawawi: pong
azawawi mortiz_: regarding, perlgeek.de/blog-en/perl-5-to-6/14-main-sub.html ; i cant get STD to compile it... although rakudo does 09:14
09:14 M_o_C joined
azawawi moritz_: im learning perl6 from your blogs :) 09:14
masak std: sub MAIN($path, :$force, :$recursive, :$home = glob("~/")) { } 09:15
p6eval std 26883: OUTPUT«##### PARSE FAILED #####␤Can't put optional positional parameter after variadic parameters at /tmp/1iLajbu25I line 1:␤------> $force, :$recursive, :$home = glob("~/")) { }␤ expecting any of:␤ standard stopper␤ terminator␤ whitespace␤Undeclared routine:␤
.. glob used a…
moritz_ IMHO that's a STD parsing bug.
TimToady: see above. It says "optional positional parameter', but :$home for sure is named 09:16
masak yes.
azawawi thx :) 09:17
jnthn Matt-W: Well, not so much features, but I got a good fix in last night (Null PMC Access and others now get backtrace with HLL source line/file.) 09:18
09:18 ejs1 left
Matt-W jnthn: My hair thanks you for something which is bound to reduce the number of times I try to pull it out in frustration 09:19
mberends japhb: welcome, you are a committer on gitorious.org/http-daemon/mainline
.oO( so much to backlog, so little time )
 09:21
masak jnthn: Null PMC access backtraces, cool! 09:27
09:27 mikehh_ is now known as mkehh
azawawi jnthn++ 09:28
jnthn rakudo: sub foo { ::T = 42; }; sub bar { foo() }; bar();
p6eval rakudo 222993: OUTPUT«Null PMC access in getprop()␤in sub foo (/tmp/4ar4uyZcbV:1)␤called from sub bar (/tmp/4ar4uyZcbV:1)␤called from Main (/tmp/4ar4uyZcbV:1)␤»
jnthn Well, line number output is a litle boring on a one-liner ;-)
mberends yes, jnthn++ diagnostics much appreciated 09:29
jnthn pmichaud++ and Tene++ for doing parts of what it's taken to get us here too. :-)
moritz_ I wonder, should p6eval filter out back traces? (or at least partially?) 09:30
or are they usually suffieciently short so that we don't care? 09:31
masak moritz_: I'm in ur STD, checking your bug. the error message is issued on line 3758, and it seems to my untrained eye as if the check done is 'is this an optional parameter?', when it should really be 'is this an optional positional parameter?'
azawawi mortiz_: regarding perlgeek.de/blog-en/perl-5-to-6/16-enums.html, the example does not working also on rakudo/std
09:31 payload left
azawawi std: enum bit Bool <False True>; 09:31
p6eval std 26883: OUTPUT«##### PARSE FAILED #####␤Syntax error (two terms in a row?) at /tmp/yiYHgk7Yra line 1:␤------> enum bit Bool <False True>;␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard stopper␤ statement modifier loop␤ terminator␤ trait␤FAILED
..00:02 37m␤»
mberends moritz_: (filter the backtrace)++ 09:32
azawawi rakudo: enum bit Bool <False True>;
p6eval rakudo 222993: OUTPUT«Statement not terminated properly at line 1, near "Bool <Fals"␤in Main (src/gen_setting.pm:0)␤»
masak moritz_: there's something to be said for not modifying the output that p6eval shows.
moritz_ rakudo: enum Bool <False True>
p6eval rakudo 222993: OUTPUT«Re-declaration of type Bool at line 1, near ""␤in Main (src/gen_setting.pm:689)␤»
moritz_ azawawi: that's because Rakudo doesn't know about bit, and because `Bool' is already declared in the prelude 09:33
masak: I'll look into it later, lunch now
azawawi moritz_: anyway, there is also undeclared $arbitrary_value 09:34
mberends masak: on your barefoot analogies, what maps to 'blister'? a rakudobug?
moritz_ azawawi: not all examples are intended as self-contained runnable code
azawawi moritz_: ah my mistake :)
masak moritz_: funny enough, the line that issues the error was last changed in r22746: '[STD] fix parameter parsing problem noted by moritz++' 09:35
mberends: among other things.
azawawi moritz_: given it was a tutorial, i thought they were :)
09:35 payload joined
masak mberends: it actually carries over quite well, because blisters are the way to thick soles. 09:35
and we're all about being soleful here in the Perl 6 community. 09:36
mberends masak: touch􏿽xE9 09:37
09:37 meppl joined
azawawi masak: what's the status on u4x docs? 09:38
09:39 jjiang joined
masak azawawi: well, two things: I'm going through P2T1A, filling in the things necessary for getting a grasp on what's needed of the hierarchy. I've done one sprint out of two on that. also, literal is about to start his GSoC project, implementing grok, the u4x documentation client. 09:40
azawawi masak: i heard about that from szabgab; any links? 09:41
moritz_: and where can i find the keyword index that szabgab published? 09:42
masak azawawi: svn.pugscode.org/pugs/docs/u4x/
azawawi masak: thx
masak enjoy.
09:42 meppl left 09:44 bacek_ left
azawawi rakudo: if 'ab cd ef' ~~ mm/ (..) ** 2 / { say $1; } 09:47
p6eval rakudo 222993: OUTPUT«Statement not terminated properly at line 1, near "/ (..) ** "␤in Main (src/gen_setting.pm:0)␤»
azawawi std: if 'ab cd ef' ~~ mm/ (..) ** 2 / { say $1; }
p6eval std 26883: OUTPUT«ok 00:03 37m␤»
masak azawawi: 'm//' and 'mm//' NIY in Rakudo as far as I know. 09:48
mberends general question: what work has been done on P5 -> P6 source code translation? Pugs seems to have only a regex converter. Apparently TimToady++ will make the definitive one, but perhaps only after STD and Christmas. Asking as research for my YAPC::EU talk, and toying with the idea of making something small on that topic. 09:49
japhb mberends: thank you for the commitbit; I've pushed my first block of changes
azawawi masak: thx
mberends japhb++ :) thanks podnuh 09:50
japhb ya betcha sweet britches.
masak mberends: MAD.
mberends: uh, props. 09:51
mberends masak: oops
masak mberends: it's a setting when building Perl 5.
japhb OK, it's painfully late here. Must attempt sleep before alarm goes off in <4 hours. Bleah. 09:52
09:59 harig left
azawawi std: subset Squares of Int where { .sqrt.int**2 == $_ };multi sub square_root(Squares $x --> Int) { return int $x.sqrt; } 10:00
p6eval std 26883: ( no output )
azawawi rakudo: subset Squares of Int where { .sqrt.int**2 == $_ };multi sub square_root(Squares $x --> Int) { return int $x.sqrt; }
p6eval rakudo 222993: ( no output )
azawawi std: my $foo; 10:01
p6eval std 26883: OUTPUT«ok 00:02 36m␤»
azawawi std: subset Squares of Int where { .sqrt.int**2 == $_ }; 10:02
p6eval std 26883: OUTPUT«ok 00:02 36m␤»
azawawi std: subset Squares of Int where { .sqrt.int**2 == $_ }; multi sub square_root(Squares $x --> Int) { return int $x.sqrt; }
p6eval std 26883: OUTPUT«##### PARSE FAILED #####␤Syntax error (two terms in a row?) at /tmp/wbL0szDsxg line 1:␤------> e_root(Squares $x --> Int) { return int $x.sqrt; }␤ expecting any of:␤ infix or meta-infix␤ infix stopper␤ standard stopper␤ statement modifier loop␤
..terminator␤FAILED 00…
10:04 payload left 10:05 fridim_ left 10:07 pmurias joined 10:08 pmurias left
azawawi moritz_: what about subset Squares of Int where { .sqrt.int**2 == $_ }; multi sub square_root(Squares $x --> Int) { return int $x.sqrt; } ? 10:12
moritz
mortiz_: i think this is an STD parsing error...
azawawi lunch & 10:13
10:14 payload joined 10:16 bacek joined
cognominal does someone successfully build rakudo on a Mac unibody? 10:23
10:28 mib_76sowd joined 10:29 mib_76sowd left 10:33 dakkar joined 10:35 cotto left 10:37 bacek left 10:39 bacek joined, payload1 joined
mberends cognominal: what's the matter? surely it's just a Mac? (no, I don't use one myself) 10:40
10:40 ashizawa left 10:44 payload left 10:46 payload joined, payload1 left 10:51 icwiener joined 10:54 azawawi left 10:57 jjiang left
cognominal mberends, I sent a bug report with the stack tracd 10:57
mberends cognominal: had your system previously successfully built rakudo? 10:59
cognominal yes 11:00
mberends rt#65784? that Stage 1 compiler had aborted on me too under different circumstances 11:03
this is different, it has not managed to make perl6_s1.pbc yet :( 11:04
11:04 payload left 11:20 donaldh left 11:21 donaldh joined 11:26 ElectricHeavyLan joined
masak I'm still kinda curious about this "Stage 1" business. is there a Stage 2? 11:39
moritz_ @tell azawawi I never advertised my blog as a tutorial, and it's not really intended as one (see <use.perl.org/firehose.pl?op=view&id=7894> for the announcement); that said improvements are always welcome 11:40
lambdabot Consider it noted.
moritz_ masak: Stage 2 = pir + setting
masak: Stage 1 = pir only
masak ooh.
moritz_ wrt builtins, that is
masak yes.
moritz_: if your blog isn't a tutorial, I don't know what is. :) 11:41
moritz_ masak: it's an attempt to show off cool Perl 6 features, teaching those along the way
masak ...as opposed to...? :)
moritz_ masak: it's not an attempt to teach programming in Perl 6 11:42
masak fair enough.
moritz_ it's mostly a matter of choosing the topics
however I guess it could be turned into a tutorial 11:43
jnthn masak: We have a s1 compiler because we need a Perl 6 compiler to be able to compile Perl 6, and the setting is written in Perl 6. :-) 11:44
masak moritz_: I find the idea of "Learning Y assuming X" to be quite nice. it means one can start from something other than the very basics. 11:46
jnthn: I see.
jnthn masak: In theory, the only difference between them is that one lacks the built-ins that are defined in the setting. 11:47
moritz_ masak: when I started writing some Perl 6 tutorials in German, I actually wrote three; one for beginners, one for perl 5 programmers and one for programmers in general :-)
jnthn masak: In practice, I fear that we (accidentally) have another difference somewhere too. But I couldn't work out what yet. :-(
11:48 azawawi joined
masak jnthn: huzzah for striving to combine theory and practice :) 11:49
11:51 iblechbot left 11:53 azawawi left 12:03 Trey joined 12:17 REPLeffect left 12:20 mizioumt joined
mberends masak++ proto/logotype/proto.svg # looks like a bowling ball :) 12:23
masak mberends: during my barefoot walk the other day, I saw some graffiti that looked like that. I decided it would be a fine logotype for proto. :) 12:24
12:25 lichtkind joined
moritz_ rakudo: 3.4.int.say 12:25
p6eval rakudo 222993: OUTPUT«3␤»
masak rakudo++
12:25 ruoso joined 12:26 alester joined 12:33 iblechbot joined
masak rakudo: class A { method postcircumfix:<( )>() { 42 } }; my $a = A.new; say $a() 12:34
p6eval rakudo 222993: OUTPUT«invoke() not implemented in class 'A'␤in Main (/tmp/Hmi8wCLtQY:1)␤»
12:39 fridim_ joined 12:43 alester left 12:45 abra joined 12:53 bacek left
Matt-W masak: awww 12:59
masak Matt-W: so close! :/
moritz_ ticket it!
erm, I meant "masak it!"
Matt-W I guess it's not noticing the custom operator before it decides to do a sub call 13:00
masak whips out a TODO rakudobug
mberends the [] one works 13:01
masak mberends: somehow, it's not the same. :)
Matt-W no it's not at all the same 13:03
13:03 mizioumt left
masak sulks 13:03
13:04 mkehh left 13:05 mkehh joined 13:08 mizioumt joined 13:16 bacek joined
jnthn masak: That inovke one is hard to fix. :-( 13:19
Blocks on Parrot calling convention refactor...
masak jnthn: in that case, creating a TODO was the right thing to do. :)
13:22 M_o_C left, |MoC| joined
Matt-W Did I read somewhere that Allison's looking at calling conventions? 13:23
jnthn Matt-W: Yes, she is. 13:26
Matt-W: She's started on the refactor already.
Matt-W Excellent 13:27
13:28 mizioumt1 joined 13:30 clintongormley joined 13:38 gbacon left, wayland76 left 13:40 exodist joined 13:41 gbacon joined 13:47 skids joined 13:48 mizioumt left 13:53 smtms left 13:59 cj left 14:01 REPLeffect joined 14:08 justatheory joined 14:09 bacek_ joined 14:10 DemoFreak left 14:11 xinming_ is now known as xinming 14:18 azawawi joined
azawawi std: sub sub($foo) { say "From div!\n"; }; sub(1); 14:19
lambdabot azawawi: You have 1 new message. '/msg lambdabot @messages' to read it.
azawawi @messages
lambdabot moritz_ said 2h 39m 41s ago: I never advertised my blog as a tutorial, and it's not really intended as one (see <use.perl.org/firehose.pl?op=view&id=7894> for the announcement); that said
improvements are always welcome
p6eval std 26883: OUTPUT«ok 00:04 37m␤»
azawawi std: sub sub($foo) { say "From div!\n"; }; sub(1); 14:20
rakudo: sub sub($foo) { say "From div!\n"; }; sub(1);
p6eval std 26883: OUTPUT«ok 00:04 37m␤»
rakudo 222993: OUTPUT«Malformed routine definition at line 1, near "(1);"␤in Main (src/gen_setting.pm:0)␤»
14:20 ruoso left
masak 'sub(1)'? 14:20
azawawi yeah is that valid in perl6?
rakudo: sub if($foo) { say "From if!\n"; }; if(1); 14:22
p6eval rakudo 222993: OUTPUT«From if!␤␤»
masak azawawi: it should be. 14:23
azawawi scratches his head :)
masak azawawi: care to submit a rakudobug?
azawawi yeah why not :) 14:25
masak that's the spirit! 14:26
azawawi proper syntax highlighting coming soon
that handles such cases...
14:27 ejs left 14:29 mkehh left
azawawi submits a rakudobug 14:30
14:31 mkehh joined, bacek left
masak azawawi++ 14:31
azawawi moritz_: your "Perl6 programming" tutorial blog really helped me a lot in my "perl6 programming" learning quest. thx. moritz_++ :)
masak: lol 14:32
masak: so what's new with november? 14:33
masak azawawi: well, we're working to put it on top of Web.pm.
most of the effort is going to Web.pm right now. 14:34
mberends: gist.github.com/114132
azawawi masak: nice 14:36
masak: so proto can be used to install perl6 programs on all platforms that rakudo works on, right?
masak azawawi: well... it's basically not tested at all on Windows. 14:37
azawawi hmmm.. i can do that
masak azawawi: I would wager it contains a fair amount of Unixisms.
azawawi: yes, please. :)
azawawi: be prepared to encounter trouble, though.
azawawi: a good set of use cases might be the ones in the README file. 14:38
azawawi masak: bring it on :)
masak azawawi: good luck. :)
azawawi masak: and where does proto live these days? 14:39
masak github.com/masak/proto
if you have proto installed, you can also do './proto install pro'... oh, wait. 14:40
azawawi i dont have it :)
masak no. :) and that wouldn't work if you have it installed either. :)
masak tries
indeed, proto complains that proto is already installed. :) 14:41
14:41 riffraff joined
azawawi masak: ok i got it 14:50
masak great. 14:51
azawawi masak: it installs perl6 rakudo each time when you want to install to a module?
masak azawawi: no, it installs it if you don't have PARROT_DIR and/or RAKUDO_DIR set.
azawawi: but only after giving you a chance to configure it first. 14:52
azawawi ah ok
ok i need git then 14:53
masak yes. 14:54
azawawi masak: sial.org/pbot/36706
masak: i will work on it from home then 14:55
14:55 H1N1[A][Again] left
azawawi is going home & 14:55
masak: thx
masak azawawi: sounds nice.
azawawi: again, good luck!
14:55 azawawi left 14:56 H1N1[A][Again] joined 14:59 eternaleye left
masak what's the difference between a Hash and a P6hash mentioned in S12? 15:01
oh, wait. the latter is a value to the :repr parameter. nevermind. 15:02
pugs_svn r26884 | lwall++ | [STD] fix masak++'s named-parameter-with-default-assumes-positional bug 15:04
15:05 tulcod joined
masak a two-character fix. :) 15:05
pmichaud good morning, #perl6 15:06
masak ahoj, pmichaud.
15:06 justatheory left
jnthn Привет, pmichaud 15:07
TimToady @tell azawawi there is no builtin "int" function, and you can't use the type name "int" as a listop, though int() would work 15:14
lambdabot Consider it noted.
15:20 donaldh left 15:21 donaldh joined 15:22 smtms joined 15:30 cls_bsd left
Tene Morning, all. 15:30
masak Tene: o/
Tene hi masak
masak Tene: I hope to have time to write some initial Genshi tests in the next 24 hours. 15:32
um, s/Genshi/Hitomi/
tulcod "real_exception (severity:2 error:40): Malformed string" when trying a simple print statement in a pasm with simple unicode characters 15:36
as well as "likely reason: argument count mismatch in main (more than 1 param)" 15:37
15:38 Psyche^ joined
jnthn tulcod: maybe that you need unicode:"the string" or something. 15:39
tulcod: PIR is more normal to write than pasm these days.
tulcod jnthn: woot, it works using "unicode:"
jnthn: well, i'm kinda slowly working my way through the parrot documentation 15:40
started with the first chapter, which starts with pasm
jnthn tulcod: :-)
tulcod and nqp, which i didn't get :(
jnthn tulcod: Ah, OK, gotcha. :-)
nqp is like, a subset of Perl 6.
tulcod yeah, i know
but it didn't get installed by paludis
15:43 Chillance joined
tulcod what version of parrot do i need for nqp? 15:43
mberends masak: good proto work, sir. it would be really nice to have windows users as well. 15:44
masak mberends: it would indeed. 15:45
jnthn tulcod: Well, NQP has been in Parrot for quite a while, but it may just be that it's not in the main package...
tulcod 0.6.1?
jnthn Huh?!
tulcod is that old?
jnthn We're on like 1.1 now :-)
Yes, very.
tulcod okay, i guess i should kick the package maintainers' butts :P
masak mberends: what do you think, should we create circumfix:<` `> in proto after the next monthly release of Rakudo? 15:46
mberends: I think it's an abstraction which might come in handy in many places. right now we're a bit hampered by the administration with getting output from processes. 15:47
mberends yes! a fake backtick!
masak mberends: it can even be made forwards-compatible if we make sure to always write `$cmd` after putting the right command string in $cmd. 15:48
15:48 iblechbot left
pmichaud I think that output from processes may have some support now in Rakudo. 15:48
s/Rakudo/Parrot
mberends sweet, as in sugar 15:49
masak pmichaud: wow!
pmichaud: is there something I can do? can we get backticks before Thursday? :)
pmichaud We just need to test it and add the relevant bits to Rakudo.
I might even be able to do that today.
masak awesome!
mberends (fake^H^H^H^Hreal backticks)++ 15:50
masak lately, even the improvements in Rakudo have been getting better. :)
I bet the proto logotype knew this all along, and that's why it's so smug... 15:51
Tene pmichaud: is it possible to get rakudo to assume that the argument to -e is UTF-8?
pmichaud Tene: well, it's really Parrot that provides the arguments.
Tene ah
pmichaud (i.e., the args array in 'main') 15:52
I'm not sure what encoding those assume by default.
Let's check. 15:53
TimToady note that real backticks in STD should be quote: rather than circumfix:
since the insides are more like a string than an expression
Tene I was thinking of trying again to add support for term:<∅> and such to rakudo. I failed the last time I tried.
TimToady not that STD does backticks, but that's how it *would* do them
15:53 Patterner left, Psyche^ is now known as Patterner
jnthn TimToady: Does Perl 6 do backticks? 15:54
pmichaud Tene: It's very unlikely to work until we revamp the parser.
TimToady jnthn: no
Tene ah
jnthn Ah. :-)
Tene AFK driving
TimToady but it does do qqx//
backticks are reserved for user syntax
15:55 dakkar left
pmichaud Tene: Parrot assumes that command-line arguments are fixed_8 encoding. 15:57
If we can find a way to convert the fixed_8 to utf8, then that might work.
16:00 nixphreak joined
nixphreak is there a parrot channel ? 16:00
or is it called something else
PerlJam nixphreak: #parrot on irc.perl.org
nixphreak looking for more information using different languages on parrot
ok thanks 16:01
wow perl has its own ircd
16:02 hercynium joined
tulcod yeah, and ircd is so hard to setup! :O 16:02
nixphreak that wasn't the point 16:03
tulcod oh :P
pmichaud Tene: more to the point, parrot thinks the command-line arguments are ascii
PerlJam tulcod: no need to be snarky whatever the point was.
tulcod j/k guys
pmichaud I'm not at all sure how to convert them to be unicode/utf8.
16:04 nixphreak left
pmichaud gist.github.com/114189 16:05
masak tulcod: so, have you tried Rakudo yet?
tulcod trying to get things to work... currently waiting for parrot 1.1.0 to compile
masak ah, been there. :) 16:06
tulcod: well, good luck!
masak heads home to make nom
tulcod yeah, thanks :{
16:06 masak left 16:07 azawawi joined 16:14 bacek_ left 16:16 iblechbot joined 16:25 donaldh left
Infinoid pmichaud: On unix at least, we might be able to change that based on $ENV{TERM} 16:26
It's a fair bet that if the terminal is set up to handle utf8 output, it also provides utf8 input
16:28 presh left
Infinoid That will help for interactive command line use, at least. I don't have a good answer to the question of which encoding a calling script was written in 16:28
16:33 azawawi left 16:36 cotto joined 16:39 riffraff left 16:40 riffraff joined, pmurias joined, pmurias left
jnthn pmichaud: Pondering doing Rakudo Day tomorrow, to help get a few more bits in shape for the release, if that works for you. 16:42
pmichaud that works for me.
Thursday and Friday won't be so good.
jnthn Friday would be the other choice. But, sounds good.
pmichaud Much of Friday I'll be on airplanes. 16:43
jnthn Ah, flying off to remote job?
pmichaud no, visiting in-laws for niece's graduation
jnthn Oh, nice. :_)
*:-)
jnthn can't quite believe it's coming up on 3 years since his own graduation 16:44
Feels strange to think it's been a year here in Slovakia too... 16:45
Tene can believe it's coming up on four years since dropping out of uni
jnthn pmichaud: Even if I get the branch for switching us over to Perl6MultiSub completely ready before the release, I'm thinking it may be sensible to merge it in afterwards. 16:47
Unless you see a strong reason to try and make merging that a target for the release. 16:48
pmichaud I don't.
(I.e., I agree it's okay to wait for the release.)
We should also remember that we'll have a PARROT_REVISION freeze between parrot's release today and Rakudo's release. 16:49
jnthn pmichaud: Yes. I got a fix into Parrot for hll_map of MultiSub yesterday.
pmichaud I'm thinking it's not likely I'll get the combine-loadinit-into-one-sub before Parrot's release, unless the release is much later today.
jnthn Though it may not be the only one that's needed. But OK, I won't push on it quite so much.
Well, if it goes in a bit after we can bump Rakudo up to it after the Rakudo release. It'll be a nice performance win, but it's not critical as such. 16:50
pmichaud Agreed.
It should go in shortly after. 16:51
jnthn Works for me.
pmichaud (as in, this week)
I'm also expecting to do the rakudo release very late tomorrow
jnthn I think it should be quite a good release feature wise.
pmichaud it'll still be "Thursday", but it'll be 00h00 Thursday instead of much later :-)
jnthn ;-) 16:52
Thursday is a kinda vague definition anyway.
There's plenty of your Wednesday that you could do the release that would be my Thursday. ;-)
pmichaud well, I consider valid Thursday to be anytime from 00:00 UTC Thu to 05:00 UTC Fri :-)
(where 05:00 UTC Fri corresponds to 00:00 CDT Fri :-) 16:53
jnthn :-)
jnthn afk for a bit, will update changelog a bit and probably other Rakudo bits later 16:54
pmichaud although I _will_ hold the release if I don't have the ROADMAP ready by then :-|
(which means I really want to get the ROADMAP done.)
jnthn Yes, that would be really good to have.
pugs_svn r26885 | lwall++ | [STD] allow only =begin END to omit its =end
r26885 | lwall++ | [STD] better fix for masak++'s bug
jnthn I wonder how much of the ROADMAP did we already did since we discussed what should be on it in Oslo. ;-) 16:55
pmichaud heh
some, but not all.
jnthn If we'd managed all of it, that would be pretty scary. 16:56
pmichaud we did get some big items done this month, though.
And the next month is looking quite good as well.
moritz_ aye, the ChangeLong is fairly impressive
pmichaud time for lunch here. 16:57
moritz_ nom is a good idea 16:58
jnthn aye, that's why I was about to go afk too :-)
16:59 cdarroch joined
dalek kudo: ac60b66 | pmichaud++ | docs/spectest-progress.csv:
"2009-05-19 00:00",9d2934e,11297,0,389,2199,13885,16603,391
 17:00
pmichaud oops, wrong line in log. Oh well. 17:01
Tene heh
jnthn rakudo: say 11297 - 10467 17:03
p6eval rakudo ac60b6: OUTPUT«830␤»
moritz_ that's still above average (discounting the regex peak from previous release)
17:06 jisom joined, jisom left 17:15 iblechbot left 17:23 szabgab joined
Tene rakudo: my $a = PAST::Node.new(); say $a.WHAT(); 17:23
p6eval rakudo ac60b6: OUTPUT«PAST::Node()␤»
17:28 pmurias joined 17:29 sri_kraih_ left 17:33 azawawi joined
azawawi TimToady: thx for the previous two STD fixes... TimToady++ 17:34
lambdabot azawawi: You have 1 new message. '/msg lambdabot @messages' to read it.
azawawi @messages
lambdabot TimToady said 2h 20m 19s ago: there is no builtin "int" function, and you can't use the type name "int" as a listop, though int() would work
17:35 azawawi left 17:44 justatheory joined 17:53 pmurias left 17:57 justatheory left 18:00 hcchien left, Khisanth left, ascent_ left, ingy left, ascent__ joined, hcchien joined, ingy joined, Khisanth joined 18:04 meppl joined 18:09 Su-Shee joined, sri_kraih joined, brunov joined
Su-Shee hi. 18:09
moritz_ hi
Su-Shee I've got a build error since "bratislava" concerning perl6_ops_cg... 18:10
which I don't understand: pastie.org/483039 18:11
(build with perl Configure.pl --gen-parrot with perl 5.10)
18:12 abra left
moritz_ never seen that one before 18:12
Su-Shee: does that still happen for a current version of Rakudo?
Su-Shee didn't found anything with google either.. yes, it does.
just pulled freshly from git
18:13 alester joined
moritz_ Su-Shee: could you also paste the output from Configure.pl please? 18:13
Su-Shee sec, clean it and re-do. 18:14
18:14 rewtv6 joined 18:15 rewt left, lichtkind_ joined
Su-Shee moritz_: checks out parrot, builds parrot without errors.. up until "Reading configuration information from parrot/parrot_config" and telling me to invoke gmake. 18:17
pugs_svn r26886 | azawawi++ | [S:H:P6] version 0.57 contains two important STD fixes
moritz_ Su-Shee: you could please nopaste it nonetheless?
Su-Shee moritz_: ok.
jnthn Su-Shee: That's a bug in GCC. 18:19
No known workaround. 18:20
Su-Shee "oh".
jnthn Other than use a different version of GCC.
I saw this happen to someone else who lives near me.
Different gcc version resolved it.
It appears it tries to be too clever in doing a structure copy. 18:21
And then exhausts the registers it can use.
18:21 DemoFreak joined
jnthn We have a macro in Parrot that attempts to deal with it, but it appears that some version of gcc it doesn't help. :-( 18:21
(see STRUCT_COPY)
Su-Shee well, perl6 is more important to me than a specific gcc version, so I'll update. 18:22
jnthn I've been there and tried all of pointer assignment, memcpy, memmove...and no luck wiht any of them. :-(
Sucks...
Su-Shee: What gcc version are you on atm btw?
Su-Shee gcc (GCC) 4.1.2
jnthn That rings a bell from last time. 18:23
The may be a workaround, but I've tried all I can think of already. 18:24
And as you said, not much help to find on Google...
Su-Shee well, I suspected something like this, this linux isn't the most recent one..
jnthn Anyway, you're not the first person I've seen hit this one, at least. :-) 18:25
Su-Shee well another very good reason to update.. I want my fresh perl6. :) 18:26
jnthn :-)
18:27 iblechbot joined
pugs_svn r26887 | mberends++ | add Temporal.t (and Temporal.pm within) currently fails with perl6_s1.pbc 18:28
18:32 hercynium left 18:33 lichtkind left
japhb jnthn, pmichaud: Do you have dependency info for the ROADMAP milestones? I just taught myself graphviz for doing a task dependency graph at $day_job, so if you've got the data, I could use the practice, and if it works you'd get a nice (not shiny, just functional) graphic out of it. 18:43
18:44 amoc left 18:52 skids left 19:12 rewtv6 is now known as rewt 19:14 icwiener_ joined 19:16 icwiener-_- joined, icwiener_ left 19:17 riffraff left 19:18 icwiener left 19:20 ejs joined 19:21 donaldh joined 19:23 ruoso joined
ruoso Hello! 19:24
moritz_ hi ruoso
TimToady mberends: that's not legal Perl 6 19:25
ruoso what's up!
TimToady std: sub infix:\x{00ab}<=>\x{00bb}( Temporal::Time $left, Temporal::Time $right ) {...}
p6eval std 26887: OUTPUT«##### PARSE FAILED #####␤Malformed block at /tmp/b1QyZyZEIS line 1:␤------> sub infix:\x{00ab}<=>\x{00bb}( Temporal::Time $lef␤ expecting signature␤FAILED 00:02 35m␤»
TimToady you can't just replace any old character with a \x sequence, and even if you could, \x{...} is P5 notation, not P6 19:27
std: my ($offset = 8765, $t1, $t2 ); # 19:32
p6eval std 26887: OUTPUT«##### PARSE FAILED #####␤Can't put required parameter after optional parameters at /tmp/TQZPnmMT11 line 1:␤------> my ($offset = 8765, $t1, $t2 ); #␤ expecting any of:␤ parameter␤ post_constraint␤ trait␤ whitespace␤FAILED 00:02 37m␤»
TimToady there's an interesting failure 19:33
jnthn That's a fail?
Isn't it a signature after my?
19:34 Trey left
mberends TimToady: [red-faced] er, it was French quotes when I started several editors ago, it's not what I intended either. Assuming French quotes are fine, will re-commit. Thanks for the eagle eyeing. 19:34
TimToady no eagle eyes involved, except STD.pm
jnthn: yes, syntactically it's parsed as a sig, but nothing is ever bound to it, so the zones kinda don't matter 19:35
jnthn rakudo: my ($offset = 8765, $t1, $t2 ); say $offset; 19:36
p6eval rakudo ac60b6: OUTPUT«8765␤»
jnthn neh neh ;-) 19:37
TimToady rakudo: sub foo ($offset = 8765; $t1, $t2) {...}
p6eval rakudo ac60b6: ( no output )
TimToady rakudo: sub foo ($offset = 8765, $t1, $t2) {...}
p6eval rakudo ac60b6: ( no output )
TimToady mums the word 19:38
std: sub foo ($offset = 8765, $t1, $t2) {...}
p6eval std 26887: OUTPUT«##### PARSE FAILED #####␤Can't put required parameter after optional parameters at /tmp/fQGGRG3WNB line 1:␤------> sub foo ($offset = 8765, $t1, $t2) {...}␤ expecting any of:␤ parameter␤ post_constraint␤ trait␤ whitespace␤FAILED 00:02 37m␤»
jnthn Awww.
ruoso @tell pmurias you removed the init_destr_2 test, is there any test that reproduces that behavior in its place?
lambdabot Consider it noted.
moritz_ that error message seems quite sane to me
TimToady yes, but not on the my 19:39
moritz_ right
TimToady though it would be sane if followed by := ...
so I'll probably let it stand 19:40
STD also complains about the second 'use v6' for some reason
pugs_svn r26888 | mberends++ | Temporal.t fixed infix typos picked up by TimToady++ and Std.pm 19:42
19:44 Trey joined
pugs_svn r26889 | lwall++ | [Temporal.t] couple more tweaks to keep STD happy 19:45
r26890 | lwall++ | [autoderef.t] missing =end 19:48
19:53 jferrero joined
dalek kudo: 595d364 | pmichaud++ | src/parser/grammar.pg:
Update =begin/=end handling slightly (RT #65782).
 19:59
20:00 jrockway left, jrockway joined 20:07 meppl left 20:16 mizioumt1 left 20:25 masak joined
masak std: my ($a = 1, $b = 2) = 3, 4; 20:25
p6eval std 26890: OUTPUT«ok 00:02 38m␤»
masak rakudo: my ($a = 1, $b = 2) = 3, 4; say $a, $b;
p6eval rakudo 595d36: OUTPUT«34␤» 20:26
masak so, different rules hold within signatures-in-declarations and signatures-in-parameter-lists?
mberends: oh, oh! return sprintf '%02d:%02d:%02d', .hour, .minute, .second given self; 20:28
oh course!
s/h/f/
std: =begin pod␤=end pod1␤=end pod 20:30
p6eval std 26890: OUTPUT«Undefined␤ok 00:02 35m␤»
masak TimToady: I argued in the ticket that this shouldn't be defined either. forgot to cc p6c.
rt.perl.org/rt3/Ticket/Display.html...txn-565612 20:32
er, s/defined/permitted/ 20:33
20:33 meppl joined
mberends masak: better golf it is. checking with stage 1 compiler we are unable. 20:33
masak mberends: note that I don't entirely focus on shortness; if so, I'd have preffered the $_-as-invocant version. 20:34
rather, I'm trying to make it read well. :P
mberends this one does read well :) 20:35
masak another variant is (.hour, .minute, .second).fmt('%02d', ':'), that's also quite nice.
masak ♥ Perl 6 20:36
mberends where's the join implied in that last one? 20:37
masak mberends: there are four .fmt methods, one is on List.
(the other three are on Any, Hash and Pair)
s/Hash/Map/ 20:38
ruoso decommute &
moritz_ hasn't really looked at .fmt at all
masak it's just sugar.
20:38 ruoso left
masak but oh how sweet! 20:38
moritz_ ;-)
just like @list.grep: Int ;-) 20:39
masak mmm!
Su-Shee jnthn: ha! with a slightly newer gcc, everything builds fine. I've just updated to 4.2.4 which was seemingly new enough.
jnthn Su-Shee: Nice! 20:40
you can haz rakudo
masak that should be the name of someone's talk.
Su-Shee well, everyone with a standard slackware 12.0 will have the problem as me then..
and there my shiny new perl6 executable is and everything works just fine fresh from git. 20:42
masak has this question been thoroughly explored yet: now that we have user-ops, what new cool technologies are possible? mathematics is one simple answer; what else? 20:43
mberends where does one share successful procedures such as this Slackware experience?
PerlJam mberends: use.perl.org is one place
masak PerlJam: or it used to be, until everyone left. :) 20:44
moritz_ well, not everyone left
masak I know, I overgeneralized.
I'm still there, for one. 20:45
moritz_ and I foolishly offered to install a shared blogging system on perlblog.org once I'm back from vacations
masak moritz_: cool!
moritz_ (after Ovid's rant)
Su-Shee I'll post it a linux forum where serveral slackware package maintainer are reading.
masak mberends: how's the Yarn idea going?
mberends masak: yes, now that the blogserver RAD storm has blown over, I shall refactor the HTTP bits and add sessions and authentication. 20:47
masak kinda dug that RAD storm
jnthn moritz_: I saw that. If you do so, I'll probably migrate.
If it seems there's going to be a general movement in that direction, anyways
masak mberends: let's keep in touch over those HTTP refactorings. I'm interested in where HTTP::Daemon is going. 20:48
mberends masak: sure. HTTP::Engine is similar, and may be worth porting as well for comparison. 20:49
masak mberends: aye. 20:50
I think it should be the super-easy default for Web.pm.
Apache is nice and all, but not so easy to provide out-of-the-box. 20:51
Su-Shee I've half finished a little blog in rakudo. 20:53
masak :)
20:54 riffraff joined 20:58 jferrero left 20:59 [particle]1 joined, cognominal left
masak Su-Shee: I'd be interested in reading whatever source you have. so would other people in here, I think. please consider pushing it somewhere and/or blogging about it. 21:01
21:01 payload joined
Su-Shee masak: I thought, I'll anounce it on the Perl6 users ML and wrap it into a little tutorial. 21:02
masak Su-Shee: whatever way you prefer, of course. 21:03
this topic naturally interests me, since I've been writing web apps in Rakudo for almost a year, and since I'm now developing Web.pm to make it easier for people to do so. 21:04
Su-Shee masak: I've read about it and I know that the I think russian guy writes a blog engine..
masak Su-Shee: correct. Ilya writes Maya.
Su-Shee right now, I'm mostly confused about the usage of roles. 21:06
masak Su-Shee: I know the feeling...
21:06 cognominal joined
masak Su-Shee: as far as I understand, though, they're a way to introduce reusable behavior, but they're not (as is inheritance) restricted to a tree topology. 21:08
Su-Shee if I take the keyword "does" seriously, I suddenly have more components which "do" something instead of "being" something. :)
masak Su-Shee: that's not necessarily a bad sign.
jnthn Su-Shee: The use of "does" is a good linguistic hing. :-)
*hit
...
*hint
jnthn hopes he recovers his typing skills before Rakudo Day tomorrow
Su-Shee that's what I thought. :)
masak Su-Shee: just keep things flexible and go with what feels right. :) 21:09
masak considers taking up a job writing an advice column on role play
Su-Shee masak: right now, even to the unexperienced without handbook perl6 feels extremely fast and smooth to code. and I'm not even a computer scientist.
masak Su-Shee: full ack.
I'm not a CSist either. 21:10
Su-Shee and it's _compact_ - everything is suddenly short compared to Perl 5.
jnthn Su-Shee: I've talked about roles various times in presentations; you may or may not find some of my slides helpful.
www.jnthn.net/papers/2008-yapc-eu-perl6oo.pdf is perhaps a good one to look at
masak Su-Shee: at least in Ideal Perl 6, it's short... 21:11
jnthn Which reminds me, I still failed to upload my slides from NPW. Gah!
Su-Shee jnthn: thanks.
21:11 pmurias joined
masak NPW slides upload fail. 21:12
I think I've failed to push mine to November. G'ah!
masak does it now
Su-Shee masak: well I so can't wait .. :) and if I understand the features of parrot correctly, grand times are coming, I'm sure. :)
masak Su-Shee: they have been for the past 9 months or so. and they just keep on coming. :)
we're living in truly wondrous times. 21:13
mberends: so, how about a lightning talk about proto at YAPC::EU? :)
Su-Shee I've read some shorter pieces of wrapping the usal c libs in parrot to use them in rakudo and this looks extremely promising. (I think. :)
masak mberends: if we start planning now, it might reach the levels of crazy that proto deserves. :P 21:14
21:14 riffraff left
masak Su-Shee: I would really like such an ability. 21:15
someone, please publish a five-step tutorial to wrapping a C lib in a Rakudo script!
lurkers: ^
mberends masak: start procrastinating, proto's perfect for a live demo!
Su-Shee masak: there's one piece from chromatic I think..
21:15 iassai joined
masak mberends: we might create mock objects, so we can live-demo without a network conn. 21:16
Su-Shee: URL?
21:16 [particle] left
Su-Shee masak: www.oreillynet.com/onlamp/blog/2004...t_sdl.html 21:17
mberends masak: no way! that could so easily fail due to lack of testing, we would have to get stronger drugs ;) 21:18
masak Su-Shee: hm, doesn't go all the way to Rakudo, though...
mberends: it's only May... :) 21:19
the proto logotype looks like it's on pretty strong drugs...
jnthn masak: masak doing mine now :-)
masak jnthn: your slides? excellent. 21:20
21:20 ejs left 21:21 ZuLuuuuuu joined
masak jnthn: so, rakudo day tomorrow, eh? 21:21
mberends masak: safer to bring another laptop with a private github and a crossover cable
masak 'private github'? :)
mberends it's open source, build your own hub 21:22
masak would that one thrash and scale badly like the public one does?
mberends: github is closed source, I think.
mberends masak: ok, that's the mock to do then
masak there's still plenty of time. 21:23
jnthn masak: Yes, tomorrow
Su-Shee masak: HA! I found it. The most complete example I know of is the old SQLLite experiment in the parrot-source tree under "ext/"
masak mberends: of course, proto will only be the eye candy for our then Actual project: a Real package manager for Perl 6. :) 21:24
Su-Shee: indeed. worth taking a look at. again.
jnthn: \o/
21:25 iassai left
Su-Shee all right. now I can sleep happily. ;) happy hackin' everyone. 21:25
masak heads bedwards too 21:26
'night, all.
21:26 masak left, Su-Shee left 21:28 jferrero joined 21:31 Whiteknight joined, payload left
jnthn www.jnthn.net/articles.shtml # for anyone interested in my NPW slides 21:36
21:41 payload joined 21:44 bacek joined
ZuLuuuuuu jnthn: I'm interested, thanks 21:52
21:52 iblechbot left 22:01 mj41_ joined 22:03 |MoC| left 22:05 M_o_C joined 22:08 kidd left 22:17 mj41 left 22:19 pmurias left 22:25 araujo left 22:31 jferrero left, tulcod left 22:33 jferrero joined, icwiener-_- left 22:34 khatar joined, khatar left, mberends left 22:36 alester left 22:41 ZuLuuuuuu left 22:45 Chillance left 22:55 rjh joined 23:01 clintongormley left 23:10 skids joined
TimToady @tell masak STD isn't trying to be a full POD parse yet, and I don't think writing the specs for it via the bug reporting system is going to be terribly efficient :) 23:15
lambdabot Consider it noted.
23:16 payload left, payload1 joined 23:20 donaldh left, donaldh joined 23:39 lichtkind_ left 23:42 exodist left 23:55 donaldh left 23:56 DemoFreak left
japhb I have an array of pairs that I need to process in reverse order, and I naturally want the key and value in each loop iteration. Is there a more colloquial way to do this than: for @rray.reverse.map: {.kv} -> $k, $v { ... } 23:56
jnthn japhb: @rray.reverse.kv doesn't do it? 23:58
japhb jnthn: no, because then you get array position and pair, rather than each pair's key and value
jnthn Oh, yes. 23:59

Powered by Cro and the Raku® Programming Language.

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

Thank you!