Camelia, the Perl 6 bug

IRC log for #perl6, 2009-11-11

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs

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

All times shown according to UTC.

Time Nick Message
00:02 synth joined #perl6
00:04 synth joined #perl6
00:04 dalek rakudo: f16c9e2 | pmichaud++ | docs/spectest-progress.csv:
00:04 dalek rakudo: spectest-progress.csv update: 451 files, 32696 (85.2% of 38389) pass, 15 fail
00:04 dalek rakudo: Failure summary:
00:04 dalek rakudo: S02-lexical-conventions/unicode.rakudo aborted 5 test(s)
00:04 dalek rakudo: S12-introspection/methods.t aborted 10 test(s)
00:04 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/f​16c9e2763e6326e2c6f2ea875717e1e11d6b75f
00:05 synth joined #perl6
00:06 meppl joined #perl6
00:08 jnthn S12-introspection/methods.t # unexpected... :-/
00:10 pnate joined #perl6
00:10 [synth] joined #perl6
00:15 [synth] joined #perl6
00:33 orafu joined #perl6
00:34 * jnthn pokes zaslon
00:39 pmichaud s1n/others:  note that I'll be late to p6m tonight
00:39 pmichaud afk
00:41 japhb jnthn, is it expected that blizkost no longer builds against current Parrot (at least, Plumage can't build it)?
00:44 jnthn japhb: Yes, I think the PCC changes probably hit it quite hard.
00:44 japhb jnthn, ah, OK.  I wanted to make sure I hadn't broken something in Plumage it needed.
00:44 jnthn No, almost certainly not.
00:49 jnthn OK, sleep for me...night o/
00:55 Woodi jnthn: ...
00:56 Woodi 1Gb memory looks not so much for Rakudo build now...
00:56 Woodi 820 Mb used for settings...
00:56 Wolfman2000 ...and among the new pieces of stress to my convoluted puzzle...I apparently now have two full time job opportunities while I'm trying to get through college...and neither job seems to allow the flexibility of staying with college.
00:57 Wolfman2000 I'm starting to wonder whether you guys are good luck for me or not.
00:58 Woodi Wolfman2000: easy, collage is important and Perl6 ALPHA :)
00:58 Wolfman2000 Woodi: I don't get how ALPHA fits into "important"
00:59 Woodi whatewer you do Perl6 will wait :)
00:59 Woodi probably :)
00:59 Woodi "first things first"..
01:03 Wolfman2000 Woodi: Hmm...true. Still, I want to at least get a dedicated Perl 6 pastebin up.
01:03 Wolfman2000 I'm actually almost ready to test that part...slowly.
01:04 Woodi you write it entirelly in Perl 6 ?
01:05 Wolfman2000 Woodi: nope. This is in Perl 5. masak is working on a Perl 6 version. Eventually we'll combine forces or something.
01:06 Wolfman2000 I still don't know enough about Web.pm to even attempt.
01:07 Woodi Rakudo master at least builds :)
01:09 Wolfman2000 Woodi: Would you at least know how much the master branch has been updated within the past...week?
01:11 Woodi Wolfman2000: probably no updates in last few weeks :) but required Parrot version changed today
01:12 Woodi and debian upgraded libicu sho i needed to rebuild
01:16 Woodi k, gn #perl6
01:24 justatheory joined #perl6
01:25 synth joined #perl6
01:39 [synth] joined #perl6
01:44 colomon Wolfman2000: do you want to know how much the master branch has updated recently?
01:45 Wolfman2000 colomon: I'm curious.
01:45 colomon One sec...
01:46 lisppaste3 colomon pasted "rakudo git log for november" at http://paste.lisp.org/display/90149
01:47 colomon Most of the changes are quite minor, but there's been about one a day in November.
01:49 emma 08:56 < Wolfman2000> masak: ...that means we're going to be competing then
01:49 emma 08:57 <@masak> Wolfman2000: competition in here is usually quite mutually beneficial.
01:50 Wolfman2000 colomon: understood
01:50 emma ^ Is this the start of a code-off?
01:50 Wolfman2000 emma: no clue.
01:50 emma "we'll settle this with regular expressions"
01:51 diakopter my PAST interpreter in JavaScript proceedeth.  it passes the first 10 nqp-rx test files
01:51 diakopter whee.
01:55 diakopter and 16-ternary.t and 36-callable.t
02:01 seanstickle joined #perl6
02:04 eternaleye joined #perl6
02:05 Exodist joined #perl6
02:06 eternaleye joined #perl6
02:09 oZ] joined #perl6
02:13 nihiliad joined #perl6
02:25 PZt joined #perl6
02:48 oZ] joined #perl6
03:04 * diakopter reads golang.org
03:13 astrojp joined #perl6
03:22 snarkyboojum joined #perl6
03:22 frew joined #perl6
03:24 astrojp left #perl6
03:26 pattm joined #perl6
03:26 pattm ned:~$ Database last updated on 10-Nov-2009                                       21:33:02 EST. Domain servers in listed order: www.greatestate.com www.westonmass.net www.myefact.com www.efactusa.com 205.178.190.14 bash: Database: command no                                      t found
03:27 pattm whoops
03:27 pattm sorry, accidental cutpaste :(
03:27 pattm stupid putty buttons
03:35 crazed_ joined #perl6
03:43 elmex joined #perl6
03:45 seanstickle joined #perl6
03:51 jaldhar joined #perl6
03:52 ihrd joined #perl6
03:52 ihrd left #perl6
03:55 kst joined #perl6
04:00 envi^office joined #perl6
04:02 hercynium joined #perl6
04:08 meppel joined #perl6
04:09 kst joined #perl6
04:13 Wolfman2000 Back from 2nd dinner. 1st dinner sucked. eternaleye: first attempt at pasting went...unsuccessful. Trying to understand what went wrong.
04:14 eternaleye Wolfman2000: Okay. Have you got the API the commandline client will use figured out? If so, maybe I can start with getting an actual CLI client coded up.
04:14 Wolfman2000 eternaleye: Didn't get that far yet. Sorry.
04:14 Wolfman2000 I need to make sure the thing pastes and can be viewed first.
04:14 eternaleye Wolfman2000: Not a problem; I'm in no rush :D
04:14 Wolfman2000 What I can tell you: whatever your client does, it has to be able to send POST data.
04:16 Wolfman2000 Also, while I'm here: is there a permanent way to force MySQL into a sort of strict mode?
04:22 eternaleye Wolfman2000: Global sql_mode is probably what you want: http://dev.mysql.com/doc/refma​n/5.0/en/server-sql-mode.html - session is for that connection; global is for _all_ connections
04:22 eternaleye (I googled 'mysql persistent strict')
04:22 eternaleye (I can haz the googlemancy powerz)
04:23 eternaleye Also, POST is easy enough with either curl or wget
04:24 eternaleye I'm thinking to use curl or wget, wrapped in perl for the first client, then transition to pure-perl so that it works on Windows
04:24 Wolfman2000 eternaleye: what I'm also debating...do we allow anonymous pastings with the CLI?
04:25 eternaleye Wolfman2000: Yes.
04:25 Wolfman2000 right: check credentials before parsing the files.
04:25 eternaleye Wolfman2000: Although I'd recommend setting up Apache to deny unreasonable numbers of connections from individual IPs
04:26 Wolfman2000 eternaleye: Don't know how to do that yet, not worried at this point
04:26 eternaleye where 'unreasonable' is > 20..30
04:26 kst joined #perl6
04:27 eternaleye Wolfman2000: http://dominia.org/djao/limitipconn.html
04:27 eternaleye There are other modules that do the samwe thing.
04:27 oZ] left #perl6
04:27 eternaleye It's a popular wheel to reinvent ;D
04:28 Wolfman2000 eternaleye: sorry, on slight brain burnout. will read that later.
04:28 eternaleye Wolfman2000: Mainly pasted it so it's there for you to come back to in the backlog
04:28 Wolfman2000 eternaleye: that means I'd have to remember to check the backlog
04:29 eternaleye Wolfman2000: irclog.perlgeek.de has a nice search function
04:30 Wolfman2000 let's see if I can at least set the global modes without causing problems.
04:33 Wolfman2000 ...figures
04:51 gfx joined #perl6
04:53 jweisgram_ joined #perl6
04:56 xinming joined #perl6
05:01 pmichaud hugme: tweet rakudoperl The Rakudo-ng branch -- enode 7|mst 8|chanserv 9|frew]
05:01 * hugme hugs pmichaud; tweet delivered
05:01 pmichaud grrrrr
05:02 pmichaud hugme: tweet rakudoperl The Rakudo-ng branch -- http://use.perl.org/~pmichaud/journal/39874
05:02 * hugme hugs pmichaud; tweet delivered
05:02 pmichaud better.
05:03 pmichaud time for sleep
05:03 pmichaud bbl
05:29 cottoo joined #perl6
05:30 metal joined #perl6
05:32 metal !help
05:36 Wolfman2000 metal: what's the problem?
05:37 am0c joined #perl6
05:39 orafu joined #perl6
05:43 metal no
05:45 frew joined #perl6
05:54 Wolfman2000 Juerd: If you are there, I will require DateTime::Format::MySQL installed please.
05:54 Wolfman2000 ...wait, use the bots
05:54 Wolfman2000 phenny: tell Juerd to please install DateTime::Format::MySQL on feather.
05:54 phenny Wolfman2000: I'll pass that on when Juerd is around.
05:57 frew joined #perl6
06:05 nbrown joined #perl6
06:06 justatheory joined #perl6
06:16 quietfanatic_ joined #perl6
06:17 quietfanatic_ Oh, whoops, still logged on in the kitchen.
06:17 quietfanatic_ I gues I'll use that to backlog huh
06:18 diakopter I thought it was in the dining room, with the lead pipe.
06:19 diakopter by Miss Scarlet.
06:30 Wolfman2000 phenny: tell Juerd I may want postgresql at this rate. Mysql + Catalyst + DBIC = Trainwreck at this stage.
06:30 phenny Wolfman2000: I'll pass that on when Juerd is around.
06:30 Wolfman2000 ...pastebins are NOT supposed to be this compliated, are they?
06:40 PZt joined #perl6
07:10 JimmyZ joined #perl6
07:12 payload joined #perl6
07:12 Su-Shee joined #perl6
07:14 zaphar_ps joined #perl6
07:15 pmichaud joined #perl6
07:15 dukelet0 joined #perl6
07:15 mdxi joined #perl6
07:23 mathw Morning
07:54 zaslon loljnthnhazblogged! jnthn++ 'Starting to get over the hump': http://use.perl.org/~JonathanWor​thington/journal/39873?from=rss
07:57 zaslon lolpmichaudhazblogged! pmichaud++ 'The Rakudo-ng branch': http://use.perl.org/~pmich​aud/journal/39874?from=rss
07:59 iblechbot joined #perl6
08:06 lisppaste3 joined #perl6
08:06 _eMaX_ joined #perl6
08:06 IllvilJa joined #perl6
08:06 Bucciarati joined #perl6
08:06 silug joined #perl6
08:10 barney joined #perl6
08:11 moritz_ good morning
08:15 moritz_ Wolfman2000, Juerd: I installed DateTime::Format::MySQL on feather
08:18 payload joined #perl6
08:22 Woodi morning
08:24 Woodi moritz_: should Plumage be default installer for all Parrot based languages or just for things wroted in Parrot language ?
08:25 mathw morning
08:27 moritz_ Woodi: I don't know
08:27 moritz_ #parrot is probably the better place to ask
08:28 moritz_ good morning mathw
08:28 mathw hi moritz_
08:32 mariuz joined #perl6
08:39 rgrau` joined #perl6
08:41 NorwayGeek joined #perl6
08:42 NorwayGeek joined #perl6
08:45 NorwayGeek joined #perl6
08:47 NorwayGeek joined #perl6
09:04 PZt joined #perl6
09:09 pdcawley joined #perl6
09:10 pugs_svn r29050 | lwall++ | [STD] refactor most P6 rules out to STD::P6 subgrammar
09:10 yahooooo joined #perl6
09:12 reid05 joined #perl6
09:30 bryan[c1] joined #perl6
09:31 bryan[c1] can someone recommend a good article or something to enhance my knowledge of regular expressions?
09:33 Su-Shee bryan[c1]: "mastering regular expressions" or something perl 6 specific?
09:37 jaldhar joined #perl6
09:41 mberends joined #perl6
09:46 Woodi mberends: ping
09:48 jnthn morning, folks
09:48 Woodi o/
09:48 moritz_ o morning
09:49 frettled Good morning!
09:50 jnthn Woodi: Yes, that memory usage seems rather...high. :-/
09:50 mberends Woodi: pong
09:51 Woodi mberends: just trying to install http-daemon with not--gen-parrot rakudo...
09:51 Woodi and had problems with Configure.pm
09:52 mberends Woodi: right, there was something in the backlog about the directory layout being neither parrot-in-rakudo or rakudo-in-parrot
09:52 Woodi looks things change...
09:52 Woodi i would like what is "normal"...
09:53 moritz_ "normal" is installed.
09:53 Woodi so /opt/perl/bin/perl6 is right one if parrot there too ?
09:53 moritz_ IMHO yes
09:54 mberends at short notice there is no remedy except to restructure your Rakudo installation... next week I shall have some time to fix the thing properly.
09:54 Woodi rakudo_dir will be in that case /opt/perl/bin... but rakudo Test.pm is moved deeply into parrot-dev install
09:55 mberends afk& # teaching Perl 5 on Padre in London :)
09:55 Juerd Good morning
09:55 phenny Juerd: 05:54Z <Wolfman2000> tell Juerd to please install DateTime::Format::MySQL on feather.
09:55 phenny Juerd: 06:30Z <Wolfman2000> tell Juerd I may want postgresql at this rate. Mysql + Catalyst + DBIC = Trainwreck at this stage.
09:55 Woodi gl :)
09:56 * moritz_ installed that module already
10:01 * moritz_ would also like to know what kind of trainwreck that is
10:02 TiMBuS joined #perl6
10:03 jnthn pmichaud++ # good post
10:13 kst joined #perl6
10:18 Woodi rakudo: my $tpm = %*VM<config><libdir> ~ %*VM<config><versiondir> ~ "Test.pm"; say "$tpm"; if $tpm ~~ :f { say "ok" };
10:19 p6eval rakudo f16c9e: /home/p6eval//p2/lib/parrot/1.7.0-develTest.pm␤
10:19 Woodi rakudo: my $tpm = %*VM<config><libdir> ~ %*VM<config><versiondir> ~ "/Test.pm"; say "$tpm"; if $tpm ~~ :f { say "ok" };
10:19 p6eval rakudo f16c9e: /home/p6eval//p2/lib/parrot/1.7.0-devel/Test.pm␤
10:20 Woodi any idea how to automatically get rakudo test.pm install path ?
10:22 pointme joined #perl6
10:22 zaslon joined #perl6
10:22 mubot joined #perl6
10:22 moritz_ my best bet would be 'use Test;' and then inspect %*INC
10:23 Woodi err :)
10:24 moritz_ err, what's wrong?
10:25 Woodi if it can be use(d) it why Makefile tries to install it ?  :)
10:25 jnthn .oO( didn't err used to be a Perl 6 operator? )
10:25 moritz_ Woodi: which Makefile?
10:26 Woodi json one :)
10:26 Woodi http-daemon particulary
10:27 moritz_ json is quite a different beast than http-daemon
10:27 Woodi prove need test.pir maybe ?
10:27 moritz_ anyway, the likely answer is "because it's out of date"
10:29 Woodi i'm aware of it... just wondering is agrement new path/names conventions...
10:32 Woodi for %*INC.kv -> $k, $v { say "$k: $v"; }
10:32 Woodi rakudo: for %*INC.kv -> $k, $v { say "$k: $v"; }
10:32 p6eval rakudo f16c9e: Temporal: 1␤Hash: 1␤Rat: 1␤Junction: 1␤Block: 1␤Pair: 1␤Bool: 1␤NYI: 1␤Num: 1␤Signature: 1␤traits: 1␤IO: 1␤Complex: 1␤List: 1␤Array: 1␤Parameter: 1␤Whatever: 1␤Any: 1␤Int: 1␤Mapping: 1␤Range: 1␤Match: 1␤IO/Socket/INET: 1␤IO/Socket: 1␤Attribute: 1␤Object: 1␤Str: 1␤Code: 1␤Buf: 1␤Safe:
10:32 p6eval ../hom…
10:32 Woodi $keys are @ ?
10:36 Woodi is p6 Data::Dumper available ?
10:39 Woodi moritz_: what you say works perfectly - use and then inspect :)
10:39 rgrau` rakudo: my $b = 4,5,6; my $a=1,2,3,$b; say $a.perl
10:39 p6eval rakudo f16c9e: [1, 2, 3, [4, 5, 6]]␤
10:41 Woodi rgrau`: .WHAT works too
10:42 Woodi coffe &
10:55 rgrau` rakudo: my $b = 4,5,6; my $a=1,2,3,$b; say $a.WHAT , "(.WHAT asks for the Class AFAIK)"
10:55 p6eval rakudo f16c9e: Array()(.WHAT asks for the Class AFAIK)␤
10:57 Woodi right
10:58 pnu joined #perl6
10:58 Woodi .perl can be used as serialisation or like json for js ?
10:59 envi^home joined #perl6
11:00 jnthn Woodi: Yes, that's the right kinda idea.
11:01 jnthn .perl is like Data::Dumper built in.
11:01 elmex joined #perl6
11:01 Woodi so i can say .perl | perl -e STDIO ?
11:02 mathw jnthn: I was going to ask you today about helping with the ng branch. Unfortunately my internet connection's insanely unstable at the moment.
11:02 payload joined #perl6
11:02 mathw jnthn: so if I ask things and then go quiet, that's why
11:03 pnu joined #perl6
11:03 kst joined #perl6
11:04 jnthn mathw: OK. :-)
11:04 mathw But there is a very simple question which probably has a very nasty answer: is there anything I can pick up?
11:05 jnthn Flu?
11:05 jnthn ;-)
11:05 mathw umm
11:05 jnthn mathw: Oh, you didn't want a nasty answer... :-)
11:05 mathw something that will still allow me to sing in the gig I'm in tonight might be preferable
11:06 jnthn mathw: We're actually at the point where we're starting to bring back the setting. :-)
11:06 mathw ahah
11:06 jnthn mathw: That is, we actually can write operator decls in there now.
11:06 jnthn And classes and other bits.
11:06 mathw woo
11:06 mathw I can do that
11:06 jnthn *nod*
11:06 mathw as it's in Perl 6 :)
11:06 jnthn However, we've still got some missing bits that will cause ouch.
11:07 mathw no doubt
11:07 jnthn I think "for" loops are still kinda missing.
11:07 jnthn I fear given/when is too.
11:07 mathw umm
11:07 mathw okay
11:07 mathw those are quite important
11:07 jnthn ;-)
11:07 jnthn given/when should be easy enough to put back.
11:07 jnthn Mostly because I doubt in terms of the PAST they make they would need to change at all from master.
11:08 jnthn The for loops are a bit more interesting, because laziness plays in here.
11:08 jnthn And the relation between for and map.
11:08 mathw hmm
11:08 mathw I think I shan't look at for loops
11:08 jnthn pmichaud++ has those bits in his head much more than I, so I'm inclined to leave him to put that back.
11:08 jnthn If you want to put something core back though, given/when should be fairly approachable.
11:08 mathw mmm
11:10 jnthn mathw: You can almost copy-paste the grammar decls from STD for given/when/default to get the parsing.
11:10 mathw I shall take a look
11:10 jnthn Note statement_control:given needs to become statement_control:sym<given>.
11:10 jnthn (We don't have that bit of sugar.)
11:12 * mathw adds rules
11:13 Woodi running http-daemon gives: the int() sub and .int method have been replaced by the .Int method..
11:13 Woodi it's from rakudo internal ?
11:14 mathw jnthn: so you think the action methods for given/when/default are going to be much the same?
11:16 pnate2 joined #perl6
11:16 jnthn mathw: I'd think so, yeah.
11:19 Woodi Int works; what can i do with: invoke() not implemented in class 'Integer' ?
11:19 moritz_ where does this error come from?
11:19 Woodi http-daemon: make run
11:20 Woodi i replaced int(..) to Int and it helps. but cannot find any invoke
11:20 moritz_ you need (...).Int instead
11:21 Woodi kind conversion ?
11:21 mathw hmm
11:22 mathw rakudo's just segfaulting
11:22 mathw stashed my changes, it's still segfaulting
11:22 kst joined #perl6
11:22 mathw booooo
11:22 moritz_ rm -rf parrot_install; make realclean; build again
11:22 mathw that's what I just did
11:22 mathw and it works
11:22 mathw yay
11:22 mathw now to test
11:23 mathw woohoo
11:23 mathw it's broken
11:23 mathw that's good, I'd have been most astonished if it wasn't
11:24 jnthn mathw: Note that you can't just copy-paste the actions without changing the method names.
11:24 NorwayGeek joined #perl6
11:24 jnthn If you didn't already do that.
11:24 mathw yes, I figured that bit out
11:24 mathw I've got
11:25 mathw Method 'blocktype' not found for invocant of class 'PAST;Op'
11:25 mathw current instr.: 'perl6;Perl6;Actions;statement_control:sym<when>' pc 88129 (src/gen/perl6-actions.pir:825)
11:25 Woodi moritz_: s/Int\((.+)\)/($1).Int/ resolved :)
11:25 moritz_ mathw: is that the ng branch?
11:25 mathw moritz_: yes
11:26 jnthn mathw: Can you nopaste your diff so I can glance at it?
11:27 mathw sure
11:27 mathw I'm finding the error stack highly confusing
11:28 mathw http://gist.github.com/231876
11:28 mathw no doubt something needs migrating to new-way that I don't know about
11:30 moritz_ mathw: uhm, where is the EXPR rule whose capture you're accessing in there?
11:31 mathw moritz_: I'm not sure, actually. It's mentioned several times in the grammar, but doesn't seem to be defined in there
11:31 jnthn mathw: OK, does given work?
11:31 mathw jnthn: I'll check
11:31 jnthn Let's do one at a time. :-)
11:32 mathw heh no
11:32 mathw non-existent sub declare_implicit_function_vars
11:32 mathw is there an alternative to that?
11:33 jnthn mathw: As a first effort, try just ripping that line out.
11:33 jnthn I think we're handling it automatically now as part of .finishpad or so.
11:33 mathw oooh
11:33 mathw that would be fancy
11:34 mathw okay, something of an improvement
11:34 mathw it runs
11:35 jnthn \o/
11:35 mathw I'm not sure it's right though
11:35 jnthn But does $_ get set? :-)
11:35 mathw given 1 { say $_; }
11:35 mathw produces
11:35 mathw Object()
11:35 jnthn Ah.
11:35 mathw this seems wrong
11:35 jnthn I think that's a no.
11:37 jnthn mathw: I think maybe look at the output of --target=pir
11:38 jnthn mathw: And see what the signature code for the inner block looks like
11:38 jnthn There'll be a set_signature_elem line, which if you paste here I can explain.
11:38 mathw set_signature_elem does not appear in the PIR
11:40 jnthn Oh?
11:40 jnthn That's...odd.
11:41 pnate joined #perl6
11:42 Woodi PERL6LIB keeps *.pm ? where *.pir should go ?
11:44 kst joined #perl6
11:44 moritz_ I think rakudo searches for .pm and .pir files in the same directories
11:48 mathw jnthn: want to see the PIR?
11:48 mathw (had a net glitch again, am on hold with tech support)
11:48 mathw jnthn: http://gist.github.com/231884
11:49 jnthn mathw: thanks, looking
11:50 mathw as far as I can see, it looks up $_ from the outer scope, but there's nothing to do anything given-y with it
11:53 mathw but then, I'm not sure there's anything to cause that in the action method
11:53 * mathw doesn't understand the swapping <EXPR> and <xblock> thing
11:54 * moritz_ doesn't understand wher <EXPR> comes from
11:57 moritz_ ah, from xblock
11:57 IllvilJa joined #perl6
11:58 moritz_ but why is it $<EXPR> and not $<xblock><EXPR>?
12:00 Woodi must be automagic :)
12:00 mathw moritz_: no I think you're correct
12:01 mathw however, that's only mentioned in when, so it's not the problem with given :)
12:01 mathw bah, tech support are useless as always
12:01 mathw are you using a router
12:01 mathw yes
12:02 mathw plug the modem directly into your computer, it's probably the router's fault
12:02 * mathw scowls
12:07 kst joined #perl6
12:21 jnthn mathw: Sorry, got dragged in to sort out some flash/perl interaction bug.
12:22 mathw jnthn: no problem
12:23 mathw As far as I'v egot is convincing myself that there's no mechanism in the PIR which can possibly allow $_ to be set properly inside the given, by which I assume I'm missing something in the action method which would do that
12:24 jnthn mathw: looking at the PIR, I agree.
12:24 jnthn The block is not getting a signature / optional $_.
12:25 moritz_ rakudo: given 5 -> $x { say $x }
12:25 p6eval rakudo f16c9e: 5␤
12:25 moritz_ rakudo: given 5 -> $x { say $x; say $_ }
12:25 p6eval rakudo f16c9e: 5␤Use of uninitialized value␤␤
12:26 mathw std: giv{
12:26 p6eval std 29050: [31m===[0mSORRY![31m===[0m␤Unable to parse subscript; couldn't find final '}' at /tmp/m8DUR2ZoAd line 1 (EOF):␤------> [32mgiv{[33m⏏[31m<EOL>[0m␤Undeclared routine:␤        giv used at line 1␤FAILED 00:01 101m␤
12:26 mathw argh
12:26 mathw damned cat
12:26 mathw std: given 5 -> $x { say $x; }
12:26 jnthn :-)
12:26 p6eval std 29050: ok 00:01 104m␤
12:27 mathw interestingly
12:27 mathw my local rakudo-ng-with-given can't parse that
12:27 mathw but I used the rules from STD
12:30 jnthn That's...odd.
12:30 mathw yes
12:30 mathw especially since I can't see why STD can handle it either
12:31 payload joined #perl6
12:31 kst joined #perl6
12:31 moritz_ does rakudo's xblock rule do the same as STD's?
12:32 jnthn Yeah, that's what I'm pondering. It may do the wrong thing.
12:32 jnthn Also, fixing that may fix the other issues.
12:32 mathw rakudo says
12:32 mathw token xblock($*IMPLICIT = 0) { <EXPR> <.ws> <pblock($*IMPLICIT)>
12:33 mathw }
12:33 mathw STD's is messier
12:33 moritz_ that looks rightish
12:33 mathw token xblock {
12:33 mathw :my $*GOAL ::= '{';
12:33 mathw :my $*BORG = {};
12:33 mathw <EXPR>
12:33 mathw { $*BORG.<culprit> //= $<EXPR>.cursor(self.pos) }
12:33 mathw <.ws>
12:33 mathw <pblock>
12:33 mathw }
12:34 mathw the difference is the <pblock> rule
12:34 mathw STD's looks for signature
12:35 moritz_ right, I just found that too
12:35 mathw rakudo's calls <.newpad> and then goes to <blockoid>
12:35 moritz_ rakudo-ng's pblock is really just block
12:35 mathw yes
12:35 mathw that did strike me as odd
12:36 jnthn Oh.
12:36 jnthn That...might just be the source of the problem.
12:36 mathw :)
12:37 moritz_ somehow I thought of pblock as "pointy block", not "parameterized block"
12:38 mathw either way it's still not the same as a normal block
12:40 jnthn Aye
12:40 * mathw experiments
12:41 jnthn OK, I think I'm done sorting out the "fun"
12:41 jnthn *sigh* I can't help but feel that a browser that allows you hit "back" and return to a page that did a POST to without prompting you to reload, is buggy.
12:41 jnthn Or at least breaking protocol.
12:42 moritz_ what kind of browsers do that?
12:42 jnthn Firefox was.
12:42 jnthn Well, does
12:42 jnthn IE doesn't.
12:42 moritz_ well, it does ask me
12:43 huf it usually asks you if you want to do it
12:43 huf might be a bug, firefox has a few of those i hear ;)
12:43 jnthn It was showing a cached copy of the page.
12:43 jnthn Why on earth it thought it was OK to cache a page resulting from a POST, I've no idea.
12:43 jnthn Ah well, issue worked around.
12:44 mathw ah yes it does sometimes from the cache when it realy shouldn't
12:44 jnthn Yeah.
12:45 jnthn It feels wrong-ish to me.
12:45 * mathw fails to comprehend how to fix pblock
12:45 jnthn mathw: Looking.
12:48 jnthn mathw: eh, I think our pointy blocks just ain't very pointy at all. :-)
12:53 whooosh joined #perl6
12:54 kst joined #perl6
12:54 jnthn mathw: I sorta need to feed myself a bit before I collapse. :-)
12:54 jnthn mathw: I'll have a crack at fixing pblock shorly.
12:54 jnthn *shortly
12:55 masak joined #perl6
12:56 masak good too-late-to-even-pretend-it's-morning, #perl6.
12:57 moritz_ rakudo: say "good {time}, masak"
12:57 p6eval rakudo f16c9e: good 1257943750.76453, masak␤
12:58 masak :)
12:59 Juerd rakudo: say date
12:59 Juerd ;)
12:59 lisppaste3 joined #perl6
13:00 * moritz_ wonders why rakudo on p6eval's server is so abysimal slow
13:00 p6eval rakudo f16c9e:  ( no output )
13:06 pmurias joined #perl6
13:08 Woodi should Configure.pm used for build be in project lib/ ?
13:11 masak aye.
13:11 masak as soon as mberends shows up next time, I'm going to ask him what he thinks about Configure.pm under the new installed-modules regime.
13:14 pugs_svn r29051 | pmurias++ | [mildew-js] load Failure.pm, t/fail.t passes
13:14 Woodi masak: building modules should be possible from cmd too.
13:14 pmichaud hello
13:14 masak Woodi: yes.
13:15 Woodi and just cleaned http-daemon. now looking into web and have de ja vu or something...
13:15 masak Woodi: I've always been a proponent of proto making things possible, but not making projects dependent on it.
13:15 masak pmichaud: \o
13:15 SmokeMachine joined #perl6
13:16 Woodi hi :)
13:16 pmichaud people are having issues with xblock/pblock...?
13:16 mathw pmichaud: yes
13:16 mathw pblock appears to be the same as block
13:16 jnthn pmichaud: Our pblock isn't...pointy :-)
13:16 moritz_ pblock in ng doesn't match a pointy block
13:17 pmichaud right, I didn't add that in yet.
13:17 jnthn pmichaud: So we've been discovering. :-)
13:17 pmichaud but see the definition of pblock in nqp-rx for an example of what it should look like
13:17 masak people interested in new languages might want to check out the new Go language: http://developers.slashdot.org/sto​ry/09/11/11/0210212/Go-Googles-New​-Open-Source-Programming-Language
13:17 mathw pmichaud: I'll take a look
13:18 jnthn masak: Spent half an hour this morning glancing over that. :-)
13:18 moritz_ Go is an interesting mixture of really good and really bad ideas, IMHO
13:18 Woodi masak: perl-sqlit3 works ?
13:18 masak Woodi: to my knowledge, yes.
13:18 Su-Shee moritz_: "but it's from google!1!"
13:18 moritz_ the bad idea is that they don't have exceptions
13:18 masak moritz_: oh, that's too bad. :/
13:18 moritz_ and checking the return value of each function is so web 1.0
13:19 Woodi masak: it's just wrapper from parrot ?
13:19 mathw pmichaud: that looks far more comprehensible than the pblock in STD
13:19 moritz_ (actually it's pre-web 1.0)
13:19 masak Woodi: aye. but don't say "just". :)
13:19 pmichaud mathw: yes, I would suggest going with it to begin with
13:20 pmichaud STD's pblock is a bit more complex because it's doing some advanced error handling.  I totally agree with the error handling, but at this point it's more important to get correct functionality in pblock, I think; and I'm also hoping to find a slightly cleaner way to handle the errors.
13:20 Woodi masak: ok :) maybe there is wrapper for ncurses ?
13:20 moritz_ http://golang.org/doc/go_lang_faq.html#exceptions
13:20 pmichaud "Everything works in Go.  No exception!"
13:20 masak Woodi: I've seen ncurses in Parrot, so at least it's possible.
13:20 masak pmichaud: :)
13:20 moritz_ basically they say "it didn't fit in our design, so we left it out"
13:21 pmichaud designs are much easier when you don't have to handle the exceptional cases.  :)
13:21 mathw I've read a bit about Go this morning, as my dodgy internet has let me do so
13:21 mathw I'm not convinced
13:21 jnthn I'm not fond of the declaration syntax for functions.
13:21 moritz_ their concurrency ideas sound interesting
13:22 jnthn De-values the name of the thing a bit.
13:22 moritz_ I'm curious if it works out well in practice
13:22 jnthn moritz_: Yes, that was what I found most interesting from it.
13:23 mathw pmichaud: I'm curious - pblock's action method in both NQP-rx and Rakudo-ng just does make $<blockoid>.ast. Shouldn't it be doing something with any pointiness it might have found?
13:23 pmichaud mathw: method signature will have already taken care of it
13:23 mathw ooh
13:23 mathw cunning
13:23 mathw okay
13:23 mathw of course, my given statement is still missing something
13:23 pmichaud at least, that's the case in nqp-rx
13:24 * mathw wonders if it's really actually using the new grammar
13:24 pmichaud I'm not sure what will be needed to fully handle the signature in Rakudo
13:25 pmichaud might need a $*IN_DECL setting in pblock
13:25 pmichaud oh, looks like not.
13:28 pmichaud pblock's action method has to be a bit special, though, as it's the thing that has to handle $_ as an implicit parameter.
13:28 jnthn pmichaud: Ah, that belongs in pblock's action? OK...I was wondering where it was meant to go.
13:28 mathw still something wrong with the parsing of it here :(
13:28 mathw every time it sees -> it gets confuesd
13:29 pmichaud mathw: we have to tell infix:sym<->  and  prefix:sym<->  not to match there
13:29 mathw oh
13:29 pmichaud when we have full ltm with fate we won't have to do that, but at the moment we do
13:29 mathw okay
13:29 pmichaud (see nqp for examples of that also)
13:31 pmichaud actually, that probably shouldn't be an issue here -- EXPR should be able to deal with it already, but possibly doesn't
13:31 mathw NQP doesn't have anything for infix:<-> on that
13:31 pmichaud well, try it just for prefix:sym<->
13:31 pmichaud really it shouldn't happen at all, though
13:31 mathw ideally no
13:31 pmichaud but I think EXPR may have a bug in it
13:32 moritz_ std: sub x -> $x {  }
13:32 mathw where is EXPR defined?
13:32 p6eval std 29051: [31m===[0mSORRY![31m===[0m␤Malformed block at /tmp/yM6jiysi55 line 1:␤------> [32msub x [33m⏏[31m-> $x {  }[0m␤    expecting any of:␤        block␤  routine_def␤        terminator␤      trait␤FAILED 00:01 101m␤
13:32 pmichaud (I meant nqp-rx's EXPR, not STD's)
13:32 NorwayGeek joined #perl6
13:32 mathw hmm that didn't seem to help
13:32 pmichaud mathw: okay, I'll take a closer look in just a minute
13:32 pmichaud EXPR is in src/cheats/hll-grammar.pir
13:33 pmichaud (in nqp-rx)
13:33 mathw I feel dreadful, I try to help out and I end up just finding more things for you and jnthn to fix
13:33 pmichaud oh, don't feel dreadful about that
13:33 pmichaud I at least know that we still have a lot to fix :)
13:33 pmichaud and I'm very glad you're trying to help out and learning in the process
13:33 masak mathw++ # dreadfully help1
13:33 masak s/1/ful!/
13:33 mathw What will be really learning is when I see what the fixes for this lot are
13:34 moritz_ std: sub x (-> $x ) { }
13:34 pmichaud it also (pardon the pun)  points out what people find most interesting/useful/important at any (argh!) given time :)
13:34 p6eval std 29051: [31m===[0mSORRY![31m===[0m␤Malformed parameter at /tmp/K7BWGQHEnW line 1:␤------> [32msub x ([33m⏏[31m-> $x ) { }[0m␤    expecting any of:␤  name␤   parameter␤  signature␤      whitespace␤FAILED 00:01 106m␤
13:34 mathw pmichaud: jnthn mentioned that he thought adding given/when/default might be something suitable for me to look at, it wasn't so much that I found it out myself
13:35 pmichaud he probably didn't realize that pointies weren't done yet
13:35 mathw no I don't think he did
13:35 jnthn pmichaud: I thought they were, yes.
13:35 pmichaud or implicit $_, whichis needed for the non-pointy block :)
13:35 jnthn pmichaud: I remember some discussion we had about implicit $_ in ng a while back.
13:35 mathw it'll parse given 1 { say $_; }
13:35 mathw but it just says 'Object()' at the moment
13:35 * moritz_ can't read, and confused with -> with -->
13:36 jnthn pmichaud: And I'd kinda assumed that you'd done it at that point. :-)
13:36 pmichaud jnthn: no, I still hadn't figured out the parameter handling in ng yet
13:36 jnthn Ah.
13:36 mathw so I need to wait for someone clever to do pblock
13:36 mathw and then I can continue
13:37 pmichaud I'll do pblock immediately after I write my informal hague grant report to the grants managers
13:37 pmichaud (I'm trying to close out my grant this month)
13:37 jnthn pmichaud++ # ok, I'll leave that one for you. :-)
13:37 mathw :)
13:37 mathw pmichaud++
13:38 pmichaud jnthn: I may need a bit of help on the $_ param
13:38 pmichaud but let me get the rest in place first
13:38 jnthn pmichaud: Oh, that bit is probably not quite so bad.
13:40 kst joined #perl6
13:41 pmichaud jnthn: probably not -- I had a little difficulty following when I looked at it last week (but I was also extremely tired at the time)
13:49 pnate2 joined #perl6
13:50 * moritz_ didn't get any new emails for the last few hours, not even spam or mailing list mails, and wonders if his mail server is broken
13:51 * pmichaud sends moritz' mail address to a spam bot to help him test his mail server.
13:52 moritz_ pmichaud: no problem if you receive the delivery failures :-)
13:58 kst joined #perl6
14:04 * mberends shows up, ears burning
14:04 * jnthn takes a fire extinguisher to mberends ears
14:04 * pmichaud throws water onto mberends ears.
14:05 * mberends wipes the excess moisture off with two towels
14:05 * moritz_ joins the towel party
14:05 pmichaud Wasn't that a Tolkein novel... "The Two Towels"?
14:06 * pmichaud thinks he misplaced a consonant.
14:06 mberends pmichaud: a toweletta
14:06 masak mberends: hiya, Hot Ears.
14:06 mberends masak: you'e the one who set fire to my ears!
14:07 masak mberends: I did? sorry about that... :/
14:07 mberends forgiveness > permission
14:07 masak was it something I blogged? :)
14:07 jnthn .oO( must never lend masak my blowtorch again )
14:07 mberends s/blogged/backlogged/
14:08 mberends about Configure.pm, you threatened to bash my ears about that
14:08 PerlJam good morning
14:08 masak mberends: I'm pretty sure I didn't say it like that. :)
14:08 masak mberends: just wanted to discuss it, peacefully.
14:08 mberends must be my paranoia playing up again ;)
14:09 * mberends gets peaceful
14:09 masak just because your ears aren't burning doesn't mean they're not out to bash 'em...
14:09 mberends yeah
14:10 mberends masak: will you have time to plan installed-modules and Configure.pm this evening?
14:10 masak yes.
14:11 payload joined #perl6
14:11 moritz_ speaking of this evening... do we want a #perl6book meeting tonight?
14:11 * masak does
14:11 moritz_ would you mind doing it a bit later than usual, 19:00 UTC?
14:12 moritz_ (this question also goes to pmichaud, jnthn, PerlJam)
14:12 masak a bit later is fine by me.
14:12 mberends masak: good, any time after 18:00 UTC, gotta &work() now &
14:12 PerlJam 1900UTC is fine by me.
14:12 masak mberends: have the appropriate amound of fun! see you tonight!
14:12 * jnthn books that time into his schedule
14:12 PerlJam jnthn: do you have punitis?
14:12 * moritz_ sighs
14:13 moritz_ our punbot :-)
14:13 whooosh joined #perl6
14:14 jnthn I'd like to voice my concern about this mode change...
14:14 jnthn ;-)
14:15 obra_ 'morning
14:15 moritz_ hi obra_, what's up?
14:15 jnthn hi obra_
14:15 masak obra_: \o
14:15 PerlJam jnthn: I'm not sure if you're corny or campy today, but it's some c-word.
14:15 obra_ masak: me, barely
14:15 PerlJam crazy perhaps?
14:16 masak obra_: that was for moritz_, I presume. :)
14:16 moritz_ m<tab> :-)
14:16 masak obra_: welcome to the club of people who mis-tab on me and moritz_ :)
14:16 jnthn PerlJam: Putting var traits in tends to leave one that way. :-)
14:16 PerlJam :-)
14:16 moritz_ but mathw and masak are even worse, the mis-tab with two leading characters
14:16 masak oh, true.
14:17 obra_ masak: sorry about that :)
14:17 moritz_ one of you should .flip :-)
14:17 masak that's why I tend to throw a cat at mathw when I intend to say something. :)
14:17 jaffa4 joined #perl6
14:17 masak obra_: it's ok. :)
14:18 pmichaud later is much better than me
14:18 pmichaud much better for me
14:18 pmichaud (sorry, too many conversations again)
14:18 moritz_ 1900 UTC it is then
14:19 moritz_ .oO( pmichaud is a conversative :-)
14:19 kidd` joined #perl6
14:19 moritz_ wow, that word actually exists
14:19 reid05 joined #perl6
14:19 moritz_ I just wanted to make a pun on conservative
14:19 masak moritz_: isn't 19:00 UTC the usual time?
14:20 * masak confused
14:20 moritz_ didn't we have 18:30 UTC usually?
14:20 pmichaud 18:30 is parrotsketch
14:20 moritz_ or was that #ps?
14:20 masak yes.
14:20 pmichaud 19:00 is book
14:20 masak still fine by me, though.
14:20 masak just not later than usual. :P
14:21 pmichaud personally, I think there's just far too much Perl 6 activity taking place for a dead language.
14:21 moritz_ any starting time between 19H and 20H UTC is fine by me, then
14:23 masak pmichaud: well, a minimum of activity is needed, so we can change the name of the language everywhere.
14:23 pmichaud I've already done my bit:  "Rakudo"  :-)
14:24 masak also, don't worry, because it's all *academic* activity, which doesn't count.
14:24 PerlJam masak++
14:24 pmichaud Besides, I think we should give Perl 5 a name.  "Classic Perl", perhaps.  :-)
14:24 mj41 joined #perl6
14:24 pmichaud "Perl:  The Original Series"
14:25 masak Vanilla Perl. oh wait.
14:25 pmichaud We have a winner!
14:25 PerlJam as long as we don't accidentally create the "New Perl" meme
14:25 jaffa4 Perl Reloaded.
14:25 masak "Perl 6: a New Deal"
14:25 pmichaud Perl Overloaded.
14:25 jaffa4 Revenge of The Perl.
14:25 pmichaud "I for one, welcome our new Perl overloads."
14:26 masak I, for one, welcome these new Perl 6 Overloads.
14:26 moritz_ Vanilla Perl and Stracciatella Perl!
14:26 masak dang, pmichaud is faster :)
14:26 jaffa4 Perl Squared.
14:26 PerlJam Perl to the Perl
14:27 PerlJam Perl is dead.  Long live Perl!  (oh wait, we've already been there)
14:27 moritz_ I thought it wasn't dead after all :-)
14:27 jaffa4 Perl on a new horse.
14:27 pmichaud The Legend of the Black Perl.  (Oh, wait)
14:27 masak Perl: Imerial Edition.
14:28 jaffa4 Perl Strikes back.
14:28 masak Perl: Second Coming.
14:28 jaffa4 Perl Renasaince.
14:30 kst joined #perl6
14:30 * Woodi fells something will happen soon...
14:30 PerlJam something is always happening.
14:31 Woodi brainstormes usually have some/side effects :)
14:32 * jnthn shows up with a patch for var traits, after an adventure involving making POST::Compiler recurse infinitely. Oops.
14:34 * PerlJam looks at ng's Grammar.pm and Actions.pm for a few minutes
14:35 pdcawley_ joined #perl6
14:41 colomon Is there a guide for mere mortals to build ng somewhere?
14:41 pmichaud At the moment, no.
14:41 pmichaud It will become much simpler once nqp-rx is in Parrot.
14:42 pmichaud Perhaps I should do that before working on pblock
14:42 pmichaud so that more people can be building ng
14:42 moritz_ colomon: build parrot, install it. Build nqp-rx, install it (into the same location)
14:42 moritz_ colomon: then configure ng, and build
14:42 masak not necessarily the same location. just configure nqp-rx to find parrot.
14:43 masak (with --parrot-config to Makefile.PL)
14:43 PerlJam colomon: perl Configure.pl --parrot-config=/path/to/parrot_config  # very useful
14:43 moritz_ well, rakudo and nqp-rx should use the same parrot ;-)
14:43 PerlJam colomon: you use that same line for both nqp-rx and ng :)
14:44 jnthn pmichaud: ping
14:44 colomon danke, everyone.  will try it now.
14:44 masak right, Configure.pl of course, not Makefile.PL
14:44 * masak is behind the times :)
14:45 pmichaud jnthn: pong
14:45 jnthn pmichaud: I've got code-gen for variable traits working
14:45 pmichaud \o/
14:45 jnthn my $x of Int # now dispatches to the right place
14:45 jnthn Anyway, my Int $x should really do the same thing.
14:46 jnthn My first thought was "well fine, just build us an AST that we can tweak"
14:46 jnthn But now I'm looking at it and thinking "gah, we parse a multi_declarator, and thus a declarator, after this"
14:46 PerlJam colomon: make sure your parrot revision is the same as in rakudo/build/PARROT_REVISION if you're not using the --gen-parrot option to Configure.pl
14:46 jnthn And wondering if we really want to be capturing <typename>+ into some contextuval
14:46 jnthn *contextual
14:47 colomon PerlJam: I'm too lazy to do anything other than --gen-parrot.  :)
14:47 jnthn So the nested decls can use them.
14:47 pmichaud does STD.pm do that?
14:47 jnthn pmichaud: No, it doesn't.
14:47 masak jnthn: contextuval = contextual + nothingmuch :)
14:47 pmichaud (in general I'm not happy with the multi_declarator/declarator sequence of rules in STD.pm)
14:47 PerlJam masak: heh, I was thinking the same thing :)
14:47 jnthn pmichaud: Wait, does STD do which bit?
14:47 kst joined #perl6
14:47 pmichaud does STD provide a contextual for traits
14:48 pmichaud er, <typename>+
14:48 jnthn pmichaud: No
14:48 jnthn pmichaud: I can do it either way I guess.
14:48 jnthn I'm just looking at the code I'm about to write and thinking "oh gee, this is going to look horrible by the time we're done"
14:48 PerlJam colomon: I don't know that nqp-rx and ng keep build/PARROT_REVISION in sync.  You'll want to use the parrot revision that Rakudo expects though, I think
14:49 pmichaud rakudo's PARROT_REVISION is generally more up to date
14:49 jnthn Because the type name there could be any of...
14:49 jnthn * Variable declarator
14:49 jnthn * Bunch of variables declared through a variable declarator
14:49 masak wohoo! suddenly 'make test' on the ng branch looks much better!
14:49 Woodi colomon: you can use ready Makefile for build parrot and rakudo, http://perl.lunski.pl/perl6/spi/
14:49 jnthn * Routine or some kind, but we don't actually pass that back up the tree anyway, we emit code to install it somewhere
14:49 jnthn *of
14:50 jnthn pmichaud: It just gets massively neater if the individual actions have the typename available.
14:50 jnthn Well, the match node and thus ast of...
14:50 Woodi do rakudo soon need to build nqp-ng ?
14:50 pmichaud jnthn: yeah, I know what you mean.  That whole section of declarators feels like it wants a refactor somehow.
14:50 pmichaud jnthn: anyway, go with whatever seems cleanest or best to you
14:51 jnthn I mena, for now in scoped I can look for PAST::Var, but I know that in a day or two, I'm just going to he tearing my hair out over this. :-)
14:51 jnthn std: my Int $x of Str;
14:52 p6eval std 29051: ok 00:01 103m␤
14:52 jnthn heh
14:52 jnthn Well, guess semantic check.
14:52 pmichaud maybe that's really the same as   "my $x of Int of Str"
14:52 jnthn pmichaud: Yeah, it would be.
14:52 jnthn oh
14:52 jnthn wait
14:52 * jnthn ponders how that associates
14:53 PerlJam std: my $x of Int of Str
14:53 p6eval std 29051: ok 00:01 101m␤
14:53 jnthn Is that of Str going as a trait mod on $x?
14:53 jnthn e.g. if I wrote my $x of List of Book
14:53 PerlJam std: my $x of Int of Str of Int of Str
14:53 p6eval std 29051: ok 00:01 101m␤
14:54 jnthn Then that's parametric like List[Book], iirc.
14:54 pmichaud jnthn: correct
14:54 jnthn So no, it's not quite the same. :-)
14:54 pmichaud so  I'd take     my Int $x of Str  to be parametric like   Int[Str]
14:54 pmichaud which doesn't make sense for Int
14:54 pmichaud but
14:54 jnthn Oh?
14:54 jnthn Hmm. And ouch.
14:54 pmichaud my List $x of Str  would be   List[Str]
14:55 pmichaud that would be my expectation at least, based on present knowledge.
14:55 * jnthn groans
14:55 jnthn Yeah, I guess that can be made to work. :-)
14:55 pmichaud I'm not sure that's actually correct.
14:55 jnthn Me either.
14:55 jnthn It sure isn't what current Rakudo does.
14:56 jnthn I suspect current Rakudo silently discards one of the types.
14:56 pmichaud in that sense, the <typename>+ should really end up being the equivalent-ish of an "of Type" trait
14:56 jnthn Which is of course wrong too.
14:56 jnthn Well, we compile it (even in master) to a trait dispatch, yes.
14:56 jnthn The question is just about its relationship with any other "of Type" trait.
14:56 pmichaud whatever currently handles the trait dispatch in (variable_declarator?)  should probably have some way of seeing "oh, there was also a typename trait, add that as well)
14:57 jnthn Well, I was just going to push a call to trait_mod:<of> onto the traits list for the variable.
14:57 pmichaud wfm
14:57 jnthn Yes, but it won't quite work, because we need to "merge" the two.
14:57 pmichaud thus my comment
14:58 pmichaud whatever does the "normal"   "of XYZ" handling should probably be made smart enough to also see the traits supplied by <typename>+
14:58 jnthn *nod*
14:59 jnthn Sure, I can do that. It's just a bit magical. :-)
14:59 pmichaud i.e.,  basically treat     my Int $x   as syntactic sugar for   my $x of Int
14:59 jnthn Sure, I have been - I just hadn't been handling the merging up until now.
14:59 pmichaud okay
14:59 PerlJam where does the implementation type go?
15:00 pmichaud that's an "is" trait
15:00 pmichaud my $x is Scalar of Int of ....
15:00 pmichaud I'm not sure we handle that yet.  Basically the implementation type is implied by the sigil if not explicitly given.
15:01 jnthn No, we've never handled that yet.
15:01 pmichaud shouldn't be too difficult when we get to it :)
15:02 moritz_ 1) spec it properly 2) test it 3) implement it 4) ... 5) profit!
15:03 moritz_ of course they are partially parallelizable :-)
15:03 masak I still don't grok what a Scalar is.
15:03 KyleHa joined #perl6
15:03 pmichaud It's a container
15:03 pmichaud just like Array and Hash are containers
15:03 masak kind of a one-element array?
15:03 pmichaud yeah
15:03 masak ok.
15:03 pmichaud it's basically the least restrictive form of container
15:03 kst joined #perl6
15:06 Wolfman2000 masak: The way my luck has been with Catalyst, DBIC, and Mysql last night...I think a part of me wants Web.pm to be closer to done so that I can actually have a pastebin on Perl 6 written in Perl 6
15:06 PerlJam where is it specced that "of Foo of Bar" is the same as "Foo[Bar]" ?
15:06 masak Wolfman2000: that's what I'm aiming for.
15:06 masak Wolfman2000: maybe I'll spend some time on it today.
15:07 jnthn PerlJam: Probably S14.
15:07 Wolfman2000 masak: I would...rather not explain the hurdles I failed to get over. But I need to anyway for context.
15:07 masak yes, please.
15:07 Wolfman2000 masak: However you plan on handling database joins...make them consistent, and make them WORK.
15:07 PerlJam jnthn: ah,  thanks.
15:08 jnthn =head2 Relationship Between of And Types
15:08 masak Wolfman2000: what prompted the need for joins in a pastebin?
15:08 PerlJam yeah, reading now  :)
15:08 Wolfman2000 masak: the pastebin I'm designing allows for registering, tags, annotations/grouping.
15:08 Wolfman2000 Each key part requires a separate table.
15:08 masak Wolfman2000: oh, cool!
15:09 masak I didn't plan to be that ambitious on the first go.
15:09 Wolfman2000 masak: Only reason I'm trying to be this ambitious...well, http://www.pumpproedits.com <-- I made that. Granted, that's in Python/Pylons, but we're talking similar complexity.
15:10 Wolfman2000 Actually, should be less complex, but...well, that website was in Pylons and Postgresql, while I'm trying Catalyst and I'm stuck with Mysql.
15:10 colomon Two of the "make test" tests failed on my ng build, is that expected at the moment?
15:10 jnthn colomon: yes
15:11 PerlJam only 2?  ;)
15:11 moritz_ test, not spectest
15:11 PerlJam still.
15:11 colomon PerlJam: make spectest is failing gloriously at the moment.  :)
15:12 moritz_ it should pass about 90 tests (not files, single tests)
15:12 jnthn Glorious fail!
15:12 masak Wolfman2000: I think I'm drifting over to the TDD/release-early camp more and more. that's why I'm surprised your adding everything at the outset. YMMV. http://use.perl.org/~masak/journal/39483
15:13 KyleHa Does it run the spectests?
15:13 moritz_ yes
15:13 KyleHa Far out.  That's progress!
15:13 Wolfman2000 masak: ...The Cult of Dome? Sorry, I can't say I'm understanding it.
15:13 pmichaud fsvo "run"  :)
15:14 masak Wolfman2000: not 'dome', 'done'. :)
15:14 moritz_ for "run" as in "fail" :-)
15:14 jnthn We can run the spectests in the sense that I can run a marathon.
15:14 jnthn It'll be quite some work before a successful running. :-)
15:15 Wolfman2000 masak: ah. I'll give it another read later: right now taking care of another assignment, non perl related I'm afraid.
15:17 KyleHa Well, I didn't think it PASSED the spectests, but I'm impressed that it's up to running them.
15:18 lisppaste3 moritz_ pasted "passing spectests in ng branch" at http://paste.lisp.org/display/90175
15:18 KyleHa Seven whole files!
15:18 jnthn Wow.
15:19 moritz_ this reminds me of the good old days when I used to run 'make spectest' (which is now spectest_all or so) to see if any new tests would pass
15:19 moritz_ and submit patches to spectest.data
15:19 moritz_ (what is now spectest was called spectest_regression or so)
15:19 KyleHa Sounds like fun.
15:20 Psyche^ joined #perl6
15:20 moritz_ now tools/update_passing_tests_data.pl does the same, with much less noise and more information
15:21 kst joined #perl6
15:29 zamolxes joined #perl6
15:29 fax joined #perl6
15:30 moritz_ masak: that could the kind of style question you're interested in: http://www.perlmonks.org/?node_id=806515
15:31 jaldhar joined #perl6
15:31 masak moritz_: it is indeed. thanks.
15:32 japhb Woodi: Plumage will be the default installer for anything on parrot (language, library, whatever).  Of course languages are welcome to port their own "native" installers, but they need not do so.
15:35 jaldhar joined #perl6
15:36 KyleHa I'm running that tools/update_passing_test_data.pl now.  I'm surprised by how much it says works (in 'master').  Has this not been done in a while, or does it not work the way it used to, or am I not reading the results right?
15:38 jnthn pmichaud: OK, the only thing I think is left in the typed variables (not attributes) thing now is filling out the trait_mod.
15:38 moritz_ KyleHa: it seems to have problems with files marked with #icu
15:38 pmichaud jnthn++
15:38 KyleHa Intensive Care Unit?
15:38 jnthn pmichaud: Any requests for "what next"?
15:38 moritz_ KyleHa: and reports them even though they are in t/spectest.data already
15:38 moritz_ KyleHa: unicode library
15:38 pmichaud I think I'll do pblock, then work on nqp-in-parrot
15:39 KyleHa Oh, I see.
15:39 pmichaud jnthn: no requests that I can think of at the moment
15:39 jnthn 'k
15:39 pmichaud I haven't even tried running "make spectest" yet to see the fireworks :)
15:39 kst joined #perl6
15:39 pmurias joined #perl6
15:39 jnthn pmichaud: Expect glorious fail.
15:40 moritz_ KyleHa: also some test files run all tests, and pass some of them, but there's no point putting them in t/spectest.data because they simply don't implement that feature yet
15:40 moritz_ for example there's an LTM test
15:40 masak moritz_: I put in a reply: http://www.perlmonks.org/?node_id=806515
15:40 moritz_ which rakudo runs, and it passes some sanity tests in there
15:41 moritz_ KyleHa: but if you find some low-hanging fruits for fudging and inclusion, feel free :-)
15:42 nihiliad joined #perl6
15:43 pmurias masak: does a lot of people actually want to keep the parenthesis in if's?
15:43 masak pmurias: maybe not a lot. but a vocal minority.
15:43 pmurias i find it similiar to wanting to unpack @_ manually
15:44 KyleHa If minorities weren't vocal, you'd never hear from them.
15:44 masak pmurias: believe it or not, there are proponents for the mutating sigils in Perl 5 as well. :)
15:45 pmurias mutating sigils actually have some sense
15:45 masak they do.
15:45 pmurias extra parens in ifs are just extra cruft
15:45 moritz_ they are needed for Perl 5's context model
15:45 masak moritz_: that sounds like material for a blog post.
15:46 masak moritz_: along with how Perl 6 does without them.
15:46 pmurias you could add extra parens everywhere (but you might just switch to lisp)
15:46 moritz_ (the mutating sigils, not the parens)
15:46 moritz_ masak: nice idea, yes
15:47 justatheory joined #perl6
15:47 masak moritz_: it's all yours. :)
15:52 KyleHa Is t/spec/S32-trig/trig.t obsoleted by all the other trig tests?
15:53 moritz_ we should ask colomon
15:56 pmurias perl6: say (do {if 0 {}}).perl
15:56 p6eval elf 29051, pugs: undef␤
15:56 p6eval ..rakudo f16c9e: 0␤
15:56 pmurias is rakudo right here?
15:56 moritz_ no
15:56 moritz_ IMHO it should really be Nil
15:57 pmurias Nil?
15:57 dalek rakudo: a02a26e | (Kyle Hasselbacher)++ | tools/update_passing_test_data.pl:
15:57 dalek rakudo: tools/update_passing_test_data.pl: ignore spectest.data comments better
15:57 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/a​02a26ea384145c812b3d8d6b5274bec2c7f66dd
15:57 moritz_ the thing that's an empty list in list context, and undef in scalar context
15:58 kst joined #perl6
15:58 payload joined #perl6
16:02 KyleHa What's bitwise negation in Perl 6 (that is, prefix ~ in Perl 5)?
16:03 moritz_ rakudo: say +! 4
16:03 p6eval rakudo f16c9e: 0␤
16:03 kst joined #perl6
16:03 moritz_ rakudo: say +^ 4
16:03 p6eval rakudo f16c9e: -5␤
16:03 moritz_ that's it
16:03 moritz_ rakudo: say ~^ 4
16:03 p6eval rakudo f16c9e: 0 1 2 3␤
16:03 Woodi masak: ping
16:03 moritz_ erm
16:03 masak Woodi: pong
16:04 KyleHa Hmmm.  In Perl 5, I get 'say ~0' as a huge number, but in Perl 6, 'say +^ 0' is -1.
16:04 Woodi masak: Squerel isn't part of Web ?
16:04 jaldhar joined #perl6
16:04 PerlJam rakudo: say ~^4
16:04 p6eval rakudo f16c9e: 0 1 2 3␤
16:04 PerlJam heh, that's a rakudo bug
16:05 PerlJam (maybe)
16:05 PerlJam rakudo: say ~^~4
16:05 p6eval rakudo f16c9e: 0 1 2 3␤
16:05 PerlJam no, definitely
16:06 masak Woodi: it is, currently.
16:06 masak Woodi: though the relation 'part of Web.pm' is a bit problematic, and has never been wholly nailed down.
16:06 masak Woodi: I consider Web.pm as a sort of incubator for projects that will be needed for web-related modules.
16:07 Woodi masak: k
16:08 kst joined #perl6
16:12 pmichaud pointy blocks should now "work" in ng, for various values of "work"
16:12 pmichaud we still don't handled implicit $_ yet
16:12 pmichaud *handle
16:13 kst joined #perl6
16:13 pmichaud afk, lunch
16:16 alester joined #perl6
16:17 alester Where's my backup admins at? https://twitter.com/clonezone/status/5622151808
16:17 kst joined #perl6
16:18 * PerlJam wasn't aware that there were any backup admins for rakudo.org
16:19 KyleHa We have backups?
16:19 Woodi masak: may I move Test.pm and Configure.pm from web/lib into somewhere else ? web/tools or web/plib ?
16:19 moritz_ I think I've mentioned that a few times here on the channel
16:19 masak Woodi: oh dear, is there a Test.pm still in there? :/
16:20 masak Woodi: what is it you're doing, by the way? packaging something?
16:20 Woodi making moduless are possible to build...
16:21 Woodi and make way to put make install into modules...
16:21 kst joined #perl6
16:22 moritz_ that's what the installed-module branch in proto aims at
16:22 Woodi it is root of all evil...
16:23 mariuz joined #perl6
16:23 Woodi it makes installation code not required...
16:23 Woodi so no way to install
16:24 Woodi assume you have /usr/lib/perl6 for modules...
16:24 moritz_ the branch I'm talking about doesn't
16:24 Woodi none of modules is possible to install... ENOTSEEYET
16:25 Woodi moritz_: branch cannot do magic if module do not designate what .pm files belong to module...
16:26 moritz_ Woodi: right, but that can be declarative
16:26 Woodi not at the moment...
16:27 Woodi i think module/lib should be only for important modules, not build tools...
16:28 Woodi maybe plib for private module libraries ?
16:29 moritz_ some Perl 5 modules use 'inc' for build tools, iirc
16:30 Woodi works for me
16:30 [particle] inc/Module/Install.pm
16:30 moritz_ that's what I meant, [particle]++
16:30 Woodi and blib is ?
16:31 PerlJam Woodi: staging area for installation "build lib" or somesuch
16:31 Woodi so install copies from blib ?
16:31 [particle] it's the install tree
16:32 [particle] files are copied to blib (the install tree), and testing is performed from there
16:32 masak I don't know why, but every time I think of questions related to 'what proto should do', a great weakness comes upon me.
16:32 [particle] that way, you're testing things as they will be installed, not as they are in source
16:32 Woodi good for packaging too (tgz, etc)
16:33 Woodi masak: proto_project_dir can be good for versioning
16:34 masak Woodi: versioning? please, please do not make proto handle versions. make something that works instead.
16:34 Woodi masak: except it should be downloading directory imo...
16:34 masak cf. pigs, lipstick, etc.
16:34 cotto_work joined #perl6
16:34 [particle] mmm, pork.
16:34 Woodi masak: why ? :)
16:34 masak Woodi: because we deserve something that works.
16:35 moritz_ masak: I've written the blog post now... will be published in a week :-)
16:35 masak and proto isn't it.
16:35 masak moritz_: :)
16:35 moritz_ so that I can stay in the ironman competition a bit longer :-)
16:35 Woodi masak: 1. name is superb; 2. proto works - ppls use it atm
16:36 moritz_ fsvo "works"?
16:36 Woodi masak: and i just removed perl6 installing things from proto and started (after month) working to me too
16:36 masak Woodi: I know. see http://use.perl.org/~masak/journal/38876
16:37 Woodi masak: but then i had stuck with modules "design"...
16:37 PerlJam masak: proto may be a victim of its own success relative to Woodi   :)
16:37 masak PerlJam: frothing at the mouth and mumblings of 'Austria' are early symptoms.
16:38 pugs_svn r29052 | kyle++ | [t/spec] fudge S04-phasers/in-eval
16:38 Woodi the most successful Perl 6 installer of the decade is true as 100% :)
16:38 masak I rest my case.
16:38 PerlJam he
16:39 PerlJam heh even
16:41 Wolfman2000 moritz_: thanks for installing DateTime::Format::MySQL. I'll mess with it once I actually get the initial query I'm trying to do to work.
16:42 Woodi masak: so i will move config.pm and test.pm from lib to inc and add inc to perl6lib, k ?
16:42 masak Woodi: I have no idea whether that is 'k' or not. :/
16:42 Woodi o(k) ?
16:42 masak I don't feel I have an overview of the problem landscape.
16:42 alester No, we don't have backup admins, but last time there was something b0rken on rakudo.org, there was much hue and cry over the need for them.
16:43 NorwayGeek_ joined #perl6
16:43 Woodi backups ? was looking on github in perl projects and rakudo was: moust forked thing :)
16:44 moritz_ alester: it's an old problem, and the cries have long since faded
16:44 moritz_ (the RSS feed thing)
16:44 alester I figured.
16:44 PerlJam alester: Well ... make me an admin and i'll fix broken things if I can when I can.
16:45 pugs_svn r29053 | pmurias++ | [mildew-js] empty blocks have a value (TODO make it Nil)
16:45 pugs_svn r29054 | pmurias++ | [mildew-js] added t/dollar_bang.t to TESTS-js
16:45 masak someone complained about rakudo.org's feed today on Twitter: http://twitter.com/clonezone/status/5622151808
16:45 PerlJam deja vu
16:47 Woodi re-mildew Perl6 compiler need a lot of Perl5 Moose to install...
16:47 pmurias Woodi: you mean re-mildew needs Moose, yes
16:48 Woodi a lot, a lot of keypressing in -e CPAN :)
16:48 jaldhar joined #perl6
16:48 pmurias Woodi: i think it's possible to make CPAN install stuff without asking
16:49 Woodi hmm, to late...
16:49 Woodi what module Mouse do ? :)
16:49 pmurias Woodi: by the way, installing {mildew-js,re-mildew} is a bit rough so if have problems ask
16:49 pmurias Woodi: Moose without meta stuff
16:49 Woodi oki
16:50 PerlJam Woodi: Mouse is a tiny Moose
16:50 Woodi but re-mildew emits plain old C ?
16:50 KyleHa Is there a way for me to determine from Rakudo whether my CPU is 32 bit or 64 bit?
16:50 moritz_ KyleHa: probably over %*VM<config>
16:50 KyleHa More to the point, I want to know when an int will overflow.
16:51 jaffa4 what is *?
16:51 PerlJam jaffa4: contextual
16:51 moritz_ an asterisk
16:51 pmurias re-mildew it emits C but it needs SMOP for the runtime to work
16:51 PerlJam jaffa4: assuming you're referring to %*VM<config>
16:51 Woodi to link ?
16:51 jaffa4 contextual? yes,.
16:51 pmurias Woodi: yes
16:51 jaffa4 How does that differ from local or global declaration?
16:52 moritz_ it's lexotic :-)
16:52 Woodi pmurias: let modules install and make all works then i see what happens :)
16:52 PerlJam moritz_: heh, you were reading my mind (I was trying hard not to say that :)
16:53 oZ] joined #perl6
16:53 KyleHa Would it be bad for spectests to use some Perl 5?
16:54 moritz_ it should be avoided whenever possible
16:54 pmurias Woodi: re-mildew doesn't support much features yet, i'm working on mildew-js now (which is a javascript backend for re-mildew) and features re-mildew after it passes the re-mildew tests
16:54 KyleHa Yeah, that was my thought too.
16:54 pmurias KyleHa: how do you want o use Perl 5?
16:54 KyleHa pmurias: I want to get its value of '~0'.  8-)
16:55 pmurias the only place where using Perl 5 is acceptable is in perl5 interop tests
16:55 KyleHa I see S03-operators/overflow.t works for me, but that may be because it's aimed at 32 bit procs, and I have a 64 bit proc.
16:55 Woodi pmurias: but it is one language more in chain p6->js->c...
16:56 masak &
16:57 pmurias Woodi: yes, but js can run in browsers
16:58 Woodi pmurias: how many MB m-js is now ? :)
16:58 oZ] joined #perl6
16:59 pmurias Woodi: you mean how big is the source code?
16:59 jaffa4 How does that differ from local or global declaration?
16:59 Woodi pmurias: no, compiler compiled to js
17:00 pmurias it's not yet bootstraped
17:00 Woodi i do not expect something comparable to css, but.... realy no idea what to think...
17:01 pmurias css = ?
17:01 Woodi pmurias: but what should be expected ? mean: for what can be used, if big then not for plain www users...
17:01 moritz_ jaffa4: read r29049 of the pugs repo, it explains that in great detail
17:02 Woodi p6 compiler should weight more then cascade style sheets...
17:02 Woodi or for server side usage ?
17:03 pmurias you would compile p6 to js on the server and if you don't use eval the user doesn't have to download the compiler
17:04 pmurias just compiled code + runtime
17:04 Woodi super. but it implies i have no idea how big picture looks like :)
17:05 TimToady pmurias: regarding if 0 {} returning Nil, see S04:197 and S04:266
17:05 pmurias you compile p6 code with mildew-js to js and the the user downloads the js
17:06 TimToady this was an explicit change from P5 so that we could write list comprehensions without special syntax
17:06 pmichaud back from lunch
17:06 dalek rakudo: 50e495b | (Kyle Hasselbacher)++ | t/spectest.data:
17:06 dalek rakudo: [spectest.data] add a couple runnable files
17:06 dalek rakudo: review: http://github.com/rakudo/rakudo/commit/5​0e495b8dd96c1df579ff64a4d549ffb9cb62952
17:06 Woodi pmurias: ah, that explains everything :)
17:07 jnthn pmichaud: w/b. I'm finally gearing up to put parametric roles back.
17:07 pmurias TimToady: thanks
17:07 jaffa4 I see
17:07 pmurias the actuall problematic case was if 1 {}
17:08 pmurias TimToady: does if 1 {} return 1 or Nil?
17:09 TimToady KyleHa: the difference on bitwise complement between p5 and p6 results from the fact that p6's Int is reallly BigInt, so if we do unsigned complement we end up with Inf
17:09 KyleHa Aha.
17:09 TimToady so signed complement seems more useful
17:09 TimToady pmurias: it returns Nil too
17:09 cdarroch joined #perl6
17:13 pugs_svn r29055 | pmurias++ | [S04] clarify the return value of {}
17:13 pmurias TimToady: forgive me if i'm wrong ;)
17:14 Wolfman2000 phenny: tell Juerd Sorry to make another request, but mysql is causing too many problems for me. Any chance of installing postgresql, or will I have to adapt to using SQLite instead?
17:14 phenny Wolfman2000: I'll pass that on when Juerd is around.
17:14 TimToady pmurias: well, I'll just have forgive you for the editing collision :)
17:15 colomon KyleHa: trig.t is officially obsolete.  I think I told rakudo not to run it?
17:16 KyleHa Yes, it doesn't run, but it's still in the pugs repo.  Can it go away completely, or is there something in there we need?
17:17 colomon As far as I know it can go away completely, I was just paranoid about making it so.
17:18 colomon I suppose it will still be hiding in svn history anyway...
17:18 pugs_svn r29056 | lwall++ | [S04] clarify Nil semantics of empty blocks and missing branches, pmurias++
17:18 KyleHa Yes.
17:18 KyleHa Thanks, colomon++
17:18 colomon Want me to get rid of it?
17:19 KyleHa If you don't, I will.
17:19 colomon I'll take it.
17:20 pugs_svn r29057 | colomon++ | [t/spec] Delete dead trig.t tests.
17:20 _jaldhar joined #perl6
17:22 colomon KyleHa++ # for appropriate prodding
17:24 TimToady moritz_: context vars aren't precisely lexotic, but they are dynlexic.
17:24 TimToady lexotic is specifically about preferring the lexical search over the dynamic search
17:25 TimToady and context vars are purely dynamic search, albeit into lexpads
17:25 * colomon is wondering what percentage of the sentences of #perl6 would have been meaningless gibberish just the week before...
17:25 moritz_ colomon: aren't the still? ;-)
17:26 moritz_ TimToady: got it, thanks
17:26 moritz_ s/the/they/
17:26 explorer joined #perl6
17:27 cognominal TimToady is clearly a lexomane
17:27 moritz_ not a dynlexomane? ;-)
17:27 pugs_svn r29058 | kyle++ | [t/spec] refudge S10-packages/basic.  One more pass since parrot rev bump
17:29 cognominal ...and a syno maniac too, probably what perl is sinful chinese to many
17:33 TimToady 哈哈
17:34 TimToady I guess a laugh/yawn is something that "fits in your mouth"
17:37 TimToady and makes a "ha" sound
17:41 cognominal anyway, with OO in Perl 6, TimToady is a role model
17:42 TimToady you're pretty classy yourself
17:42 jnthn You both have your good attributes. :-)
17:42 * pmichaud wonders if he is a Failure, albeit a very interesting one.  :-)
17:42 [particle] if you both introspect, i think you'll find you're not objective enough
17:44 * TimToady is picturing Introspector Clouseau
17:44 TimToady "Cato, what have you done to my brain?"
17:45 TimToady mb
17:49 [particle] v++
17:51 KyleHa smolder.plusthree.com not feeling well today?
17:51 * jnthn finally finds and catches the irritating buzzy insect that's been distracting him on and off all afternoon
17:51 TimToady debugging, eh?
17:52 jnthn *groan*
17:52 PerlJam jnthn: come on!  That was *much* better than some of your earlier puns.  :)
17:53 jnthn It was an admiring groan. :-)
17:56 TimToady .oO(it wasn't that good...)
17:57 payload joined #perl6
17:58 TimToady y'know, the bug metaphor isn't very good.  with a meatspace bug, it's more efficient to just let it bite you and fly away again.
17:58 TimToady unlike with a computer bug, that keeps biting over and over
17:59 TimToady well, assuming you don't get malaria, which isn't very efficient either, except in the way that it is...
18:00 moritz_ oh, I've been bitten by some meatspace bugs multiple times
18:00 TimToady well, fleas are special
18:06 Woodi TimToady: any ideas about filesystem layout for versioned modules ? or it should be db or git or ... ?
18:09 stephenlb joined #perl6
18:20 pugs_svn r29059 | kyle++ | [t/spec] move a test that passed only because previous test autovivified
18:23 * pmichaud decides he really ought to write an article about how modules aren't likely to be tied to a filesystem.  Except he doesn't want to touch that tarbaby.
18:24 * jnthn -> store, bbs
18:28 pmichaud ugh, requiring default values to be in their own blocks is making $_ a pain
18:31 Wolfman2000 brb/afk
18:35 pmichaud jnthn: ping  (when you get back)
18:43 dalek book: 30057bc | masak++ | src/basics.pod:
18:43 dalek book: [basics] more minor fixes
18:43 dalek book: review: http://github.com/perl6/book/commit/30​057bc9d95b780681353cfc99cc433bfdff91b2
18:43 dalek book: a6cee3d | masak++ | :
18:43 dalek book: Merge branch 'master' of github.com:perl6/book
18:43 dalek book: Conflicts:
18:43 dalek book: src/basics.pod
18:43 dalek book: review: http://github.com/perl6/book/commit/a6​cee3de231b9325f654768182e6af3dcf0b138b
18:45 mberends joined #perl6
18:52 jnthn pmichaud: pong, I'm back
18:52 pmichaud I'm a little concerned about handling $_ as a default parameter
18:53 jnthn (just unpacking shopping, but go ahead)
18:53 pmichaud a bare pblock is generally equivalent to      <-> $_ is ref = OUTER::<$_> { ... }
18:54 pmichaud (perhaps not 'is ref', not relevant to this discussion)
18:54 pmichaud but since default parameters are now always required to be Parrot subs, we're likely to end up with a *lot* of Parrot subs that do nothing but look up two outer scopes for $_
18:55 jnthn pmichaud: Right. It did run through my mind earlier that this may be an issue for us.
18:55 jnthn If nothing against an extra flag we set on the signature, that the binder recognizes and handles specially.
18:55 jnthn "default from outer" or something.
18:56 pmichaud since this is extremely common, I think I'd prefer that.
18:56 jnthn Yeah.
18:56 jnthn Same.
18:56 pmichaud let me test and check in what I have now
18:56 pmichaud the code that generates the default $_ is at line 128
18:56 pmichaud you can switch it to whatever you decide for the binder
18:57 pmichaud test succeeded.
18:57 jnthn pmichaud: You just create a Parameter object, followed by a Signature?
18:57 pmichaud signature, followed by parameter, but yeah.
18:57 jnthn oh, yes
18:57 jnthn OK, cool
18:57 pmichaud it's pretty obvious how it works when you see it
18:57 jnthn I can probably have this together Real Quick.
18:58 pmichaud I suspect you can :-)
18:58 pmichaud just a "bind from outer" flag that binds the variable to its equivalent outer would be excellent
18:58 jnthn Right.
18:58 pmichaud er, default from outer
18:58 pmichaud whatever
18:58 jnthn I'll just put it as a generic flag.
18:58 jnthn If there's anything else that needs it, it'll be there. :-)
18:59 jnthn (And if not, it's all guts.)
18:59 seanstickle joined #perl6
18:59 pmichaud pushed.
18:59 jnthn (that is, not exposed leakily)
18:59 jnthn oh, maybe apart from introspection...
18:59 seanstickle left #perl6
18:59 jnthn can solve that if we need to.
19:00 dalek book: 25e3c18 | masak++ | src/classes-and-objects.pod:
19:00 dalek book: [OO] added three questions
19:00 dalek book: These questions could eventually be incorporated as exercises (if we decide
19:00 dalek book: to have exercises), or they could perhaps be incorporated in the text itself.
19:00 dalek book: I plan to write full answers to these three questions as well.
19:01 dalek book: review: http://github.com/perl6/book/commit/25​e3c186a7b883107cb732268e86d658740c8035
19:02 pmichaud hugme tweet rakudoperl rakudo-ng now has pointy blocks again
19:02 * hugme hugs pmichaud; tweet delivered
19:03 jnthn oh no!!!
19:03 jnthn when I went to the store, I forgot to buy beer. :-/
19:03 [particle] hugme: hug jonathan
19:03 * hugme hugs jonathan
19:03 jnthn :-)
19:03 jnthn Eh, it's open for another 2 hours...
19:03 * pmichaud watches as rakudo development grinds to a complete stop.
19:04 pmichaud Q:  "Why did Rakudo fail?"    A: "Insufficient beer."
19:04 [particle] EBEERRUN
19:05 * KyleHa has a new respect for the pugs/VICTUALS file.
19:05 pmichaud I'm not sure which concerns me more -- the fact that jonathan is running short of beer, or the fact that he went to the store AND FORGOT TO BUY BEER.
19:06 pmichaud :)
19:06 KyleHa I have a dream that someday, I too might consume something worthy of inclusion in VICTUALS.
19:13 Wolfman2000 joined #perl6
19:13 payload joined #perl6
19:18 dalek book: 5a61748 | masak++ | src/classes-and-objects.pod:
19:18 dalek book: [OO] added answer to the first question
19:18 dalek book: review: http://github.com/perl6/book/commit/5a​61748a3030fd8748973c8f5a60230aa0e7209c
19:22 Wolfman2000 ...okay, I think I'm doing a little better. These past few days have been slightly stressful.
19:22 SmokeMachine joined #perl6
19:22 masak joined #perl6
19:25 Wolfman2000 masak: I am starting to not like Catalyst. or MySQL.
19:25 masak Wolfman2000: sorry to hear that.
19:25 Wolfman2000 mysql--; postgresql++; sqlite++;
19:26 Wolfman2000 @karma mysql
19:26 Wolfman2000 ...the bot's not here? BAH
19:27 moritz_ zaslon: karma mysql
19:27 zaslon Sorry, I don't understand that command
19:27 KyleHa Pg > MySQL
19:27 moritz_ mubot: karma mysql
19:27 mubot mysql has a karma of -1
19:27 moritz_ mubot: karma postgresql
19:27 mubot postgresql has a karma of 1
19:27 Wolfman2000 ...hold on a second...one of my scripts is acting up.
19:27 cdarroch joined #perl6
19:28 Wolfman2000 moritz_++: thanks.
19:34 Wolfman2000 masak: What are your thoughts of sqlite anyway?
19:34 NorwayGeek_ joined #perl6
19:35 masak Wolfman2000: that it's what currently works with Web.pm :)
19:36 Wolfman2000 If nothing else, SQLite does seem like a good fallback. But...is it a good primary?
19:37 PerlJam fallback?
19:37 Wolfman2000 ...great: I've angered a fan.
19:37 PerlJam not at all
19:37 Su-Shee I like sqlite very much. small, fast, convenient. great project.
19:38 moritz_ Wolfman2000: sqlite is very good, unless you have lots of concurrent accesses
19:38 PerlJam sqlite is great if you need a self-contained RDBMS with a small set of tables with a small amount of data in them.  :)
19:38 PerlJam (where you get to pick what "small" means here)
19:38 Su-Shee PerlJam: though small is relative relative ;)
19:38 Wolfman2000 PerlJam: ...you just answered my question.
19:39 Su-Shee no, he didn't really.
19:39 Wolfman2000 I'm debating chucking the mysql database that I have planned for the Perl 6 Pastebin I'm making and using sqlite instead. It will probably work better with DBIC and Catalyst.
19:39 Su-Shee for the average joe web developer and website and minor shops, sqlite would be a great choice.
19:39 Wolfman2000 masak: If I run into further trouble, I'll probably find a way to email you the schema I made.
19:40 Wolfman2000 Su-Shee: You can't read my mind.
19:40 masak Wolfman2000: oki.
19:41 Su-Shee Wolfman2000: ah well whatever. use what you want. none of my business anyway.
19:41 cognominal joined #perl6
19:42 Wolfman2000 Su-Shee: I'm thinking of sqlite. I'm not expecting postgresql to be installed on feather...that, and I probably will need to learn how to work with sqlite at some point anyway.
19:47 spinclad joined #perl6
19:47 cls_bsd joined #perl6
19:47 constant joined #perl6
19:47 sri__ joined #perl6
19:47 phenny joined #perl6
19:47 jiing_ joined #perl6
19:47 hatseflats joined #perl6
19:47 pugs_svn joined #perl6
19:47 krunen joined #perl6
19:47 dukeleto joined #perl6
19:47 Woodi joined #perl6
19:47 parduncia joined #perl6
19:47 frew__ joined #perl6
19:48 japhb joined #perl6
19:48 _jaldhar joined #perl6
19:48 oZ] joined #perl6
19:48 reid05 joined #perl6
19:48 pointme joined #perl6
19:48 barney joined #perl6
19:48 zaphar_ps joined #perl6
19:48 xinming joined #perl6
19:48 jan_ joined #perl6
19:48 szabgab joined #perl6
19:48 colomon joined #perl6
19:48 [particle] joined #perl6
19:48 PacoLinux joined #perl6
19:48 Infinoid joined #perl6
19:48 carlin joined #perl6
19:48 araujo joined #perl6
19:48 arthur-_ joined #perl6
19:48 yath joined #perl6
19:48 hicx174 joined #perl6
19:48 Juerd joined #perl6
19:48 Maddingue joined #perl6
19:48 omega_ joined #perl6
19:48 ssm joined #perl6
19:48 daemon joined #perl6
19:48 athomason joined #perl6
19:48 cosimo joined #perl6
19:48 lambdabot joined #perl6
19:48 jnthn joined #perl6
19:50 TimToady wow
19:52 Wolfman2000 TimToady: surprised at the split?
19:52 TimToady just surprised it recovered so well  :)
19:53 dalek book: 5992bee | masak++ | src/classes-and-objects.pod:
19:53 dalek book: [OO] added answer to the second question
19:53 dalek book: review: http://github.com/perl6/book/commit/59​92beed104b1bec29f97191ef087c9376ebb6ab
19:54 rapacity_ joined #perl6
19:54 buu_ joined #perl6
19:54 Tene_ joined #perl6
19:54 broquaint joined #perl6
19:54 Juerd Yeah, it recovered very well. Uh-huh.
19:54 phenny Juerd: 17:14Z <Wolfman2000> tell Juerd Sorry to make another request, but mysql is causing too many problems for me. Any chance of installing postgresql, or will I have to adapt to using SQLite instead?
19:54 Wolfman2000 Juerd: ignore that request
19:55 Juerd Okay
19:55 Juerd I'm glad you say that because I have no experience with Pg from a sysadmin perspective.
19:55 Wolfman2000 I'm going to try to use sqlite for my Perl 6 Pastebin instead
19:55 Juerd Why would pastes even need a database?
19:55 Wolfman2000 Catalyst + DBIC + MySQL = *KABOOM*
19:55 Wolfman2000 Juerd: registration, tags, expiring pastes, etc
19:55 Juerd It's write-once, then readonly, and typically single id'ed/indexed.
19:56 Juerd Registration?
19:56 Juerd As in user accounts?
19:56 Wolfman2000 Yes. It's optional though.
19:56 Wolfman2000 This way, no one else can use wolfman2000 for me
19:57 Juerd Hm. Sounds a bit over-designed to be honest :)
19:58 Wolfman2000 Juerd: www.pumpproedits.com <-- I'm capable of it. That was designed using Python, Pylons, and PostgreSQL. Use of peeing three times not intended. Use of the lame joke fully intended.
20:00 TimToady phone
20:03 Su-Shee I like a little syntax highlighting and the possibility to edit the paste, that's it.
20:03 Wolfman2000 Su-Shee: syntax highlighting is planned. editing the paste? Well, it would work best if I implemented user accounts for that. Otherwise, I plan on having a limited form of annotations/paste grouping.
20:05 Su-Shee why would I need an account to edit _pastebin-ed_ code?!
20:05 Wolfman2000 Su-Shee: Because I don't like the idea of anonymous people editing other people's pastes.
20:05 PerlJam Wolfman2000: who said anything about editting?
20:06 Su-Shee I pastebin a problem and juerd - instead of endlessly elaborating - adds a few lines of code - problem solved.
20:06 Wolfman2000 Su-Shee: wait for Google Wave.
20:06 PerlJam oh, Su-Shee did
20:06 PerlJam Hmm
20:06 Su-Shee I already have google wave.
20:06 Wolfman2000 Su-Shee: Well I don't.
20:06 Su-Shee and several pastebins support editing.
20:07 masak mberends: do you remember your novemer-wiki.org password?
20:07 Wolfman2000 Su-Shee: name one that allows you to edit anyone else's paste
20:07 oZ] There's a lot of pooping on ideas here.
20:08 mberends masak: yes, shall I broadcast it to the googlesphere?
20:08 PerlJam Wolfman2000: so ... why have accounts at all anyway?  Each edit creates a new post.
20:08 masak mberends: no; could you try logging in?
20:08 oZ] Wolfman2000: You could allow those maintaining the same session to be able to edit the pastes they made without authenticating.
20:09 oZ] Su-Shee: Why wouldn't you just create a new paste with the updates?
20:09 diakopter oZ]: yeah; several problems apparently there: 1) #perl6 attracts the constipated; 2) #perl6 has a laxative effect; and 3) everyone's ideas in #perl look like toilets.  :/
20:09 diakopter figuratively, of course, for all 3.
20:10 Wolfman2000 ...how simple do you guys want a syntax highlighting Perl 6 pastebin?
20:10 Wolfman2000 This is meant for all of you, after all.
20:11 * diakopter thought there was one that used the spectests to seed it...
20:11 oZ] diakopter: lol
20:11 Su-Shee ah well..
20:11 Su-Shee left #perl6
20:11 mberends masak: http://www.november-wiki.org login failed :(
20:12 masak mberends: yes, that's what I'm experiencing too.
20:12 diakopter oZ]: oops.  I meant #perl6 in 3), of course... :/
20:12 diakopter freudian slip.
20:12 masak mberends: lichtkind++ discovered that one. and I can't find anything wrong! it's like, the ultimate debugging problem.
20:12 masak mberends: do you have a feather account yet?
20:12 mberends masak: yes, but fairly inactive
20:12 KyleHa A pastebin should be very simple, like paste.lisp.org.  The hardest thing on it should be the capcha.
20:13 Wolfman2000 KyleHa: should I include a captcha?
20:13 oZ] But a pastebin used by many users should be able to scale up to protect.
20:13 moritz_ not until you need it
20:13 oZ] IRC should be simple, but we have Nickservs.
20:13 mberends nom &
20:13 * diakopter likes perl6.pastebin.com
20:13 KyleHa If I have to log into it, I won't use it (I can already do that on github).
20:13 oZ] moritz: which is why accounts shouldn't be required.
20:13 Wolfman2000 KyleHa: You don't have to log into it
20:14 Wolfman2000 Accounts are optional.
20:14 oZ] but available.
20:14 Wolfman2000 oZ]: That's my goal.
20:14 Wolfman2000 Accounts are available, but not required.
20:14 Wolfman2000 You can post as anonymous, and you can also claim any name that is not password protected.
20:14 KyleHa Wolfman2000: You don't need captcha right off the bat, but it might be good to plan for it.
20:14 oZ] Integrating reCaptcha is pretty easypeasy.
20:15 Wolfman2000 KyleHa, oZ]: I'll worry about captcha later.
20:15 Wolfman2000 Right now, I just want...well, a working database. mysql-- sucks.
20:15 oZ] Agreed.
20:16 Wolfman2000 I'll throw my support to postgresql++ and sqlite++ at this time.
20:16 Wolfman2000 db2-- seems to cost money, and mssql-- is Microsoft.
20:16 oZ] Depending on use, I'll get behind that.
20:16 oZ] I'll even Oracle++ compared to mysql--.
20:17 Wolfman2000 ...then it's official. Time to recreate my tables...using sqlite
20:17 Wolfman2000 And hope I don't blow it.
20:17 PerlJam oZ]: oracle is fat enough already  ;)
20:17 jnthn fwiw, mssql++ ain't so bad, in my experience.
20:17 jnthn I'd choose it over mysql.
20:17 jnthn erm, mysql-- I meant
20:17 jnthn ;-)
20:18 oZ] mssql isn't terrible, but the fact that I have to run it on a Windows server combined with the ridiculous DBD play I have to do leaves it behind. :)
20:18 Wolfman2000 ...if I was allowed to set the topic, I would make it say "UTF-8 is our friend! mysql-- is our enemy!"
20:19 masak hugme: hug Wolfman2000
20:19 * hugme hugs Wolfman2000
20:19 Wolfman2000 hugme++; masak++;
20:19 Wolfman2000 ...hugme is a female bot, right?
20:19 jnthn oZ]: Aye. I've never chosen to use it in a project, I've just inherited a couple that did, and that already had other choices tying them to Windows anyway.
20:19 masak Wolfman2000: it has bot gender.
20:19 Wolfman2000 ...
20:19 jnthn neuter? :-)
20:20 Wolfman2000 ...at least it's not male.
20:20 masak no, that's a fourth one. :)
20:20 moritz_ it has the appropriate gender, whatever that might be for you
20:20 masak make, female, neuter and bot.
20:20 masak s/k/l/
20:20 masak sheesh, don't they teach you anything in grammar class these days? :P
20:21 Wolfman2000 masak: reminder needed. Does sqlite enforce NOT NULL?
20:22 masak Wolfman2000: you seem to assume that I know anything about sqlite... :)
20:22 Wolfman2000 masak: well, you're using it
20:22 masak I just wrote a simple wrapper for it.
20:22 PerlJam Wolfman2000: I use it too, but I still lookup stuff on the web about it.
20:26 Wolfman2000 PerlJam: fair enough
20:27 Juerd Editing a paste? That's weird.
20:28 Wolfman2000 Juerd++: Thank you.
20:28 Juerd It's no longer a paste service then. It's just text file hosting.
20:28 PerlJam A wiki! :)
20:29 masak it could be as simple as 'this paste has been copied and the modified version is at <link>'
20:29 moritz_ that's quite usual
20:29 moritz_ s/usual/common/
20:32 jsut|work joined #perl6
20:32 ShaneC joined #perl6
20:40 donaldh joined #perl6
20:40 diakopter joined #perl6
20:42 PZt joined #perl6
20:49 zamolxes joined #perl6
21:02 quietfanatic Currently
21:02 quietfanatic Currently Gamebase has a rather large memory leak, to the tune of 14m per second
21:03 fax joined #perl6
21:03 masak ouch.
21:03 jnthn epic leak. :-|
21:03 jnthn .oO( at least Rakudo does *something* fast... )
21:04 oZ] lol.
21:04 quietfanatic jnthn: :)
21:04 fax joined #perl6
21:04 quietfanatic It's possible my SDL wrapper is being a little sloppy with UnManagedStructs but not that sloppy.
21:05 jnthn quietfanatic: Well, perl6 -e "while 1 { }" leaks too. :-|
21:05 * jnthn ponders
21:06 quietfanatic Bout the same speed
21:06 jnthn Yeah
21:06 jnthn That's good news and bad news.
21:06 jnthn (e.g. it means there's a small-ish test case that can provoke the leak.)
21:08 jnthn It's doing it in ng too.
21:10 Anon8398 joined #perl6
21:14 jnthn quietfanatic: Whoa. If you want to see something leaking really really fast...
21:14 jnthn .sub 'main'
21:14 jnthn lab:
21:14 jnthn 'foo'()
21:14 jnthn goto lab
21:15 jnthn .end
21:15 jnthn .sub 'foo'
21:15 jnthn .end
21:15 quietfanatic Oh really?
21:15 jnthn That leaks at about tens of megs a second for me.
21:15 quietfanatic I only did a loop in one sub
21:15 jnthn No Perl 6 runtime, just a Parrot loop.
21:15 Juerd Memory is practically free nowadays, so don't bother :D
21:15 jnthn So basically, every single sub invocation in Parrot, it seems, leaks.
21:16 jnthn That. Sucks.
21:16 jan_ joined #perl6
21:16 quietfanatic Yes, it's like 60m per second for me
21:16 jnthn I get about 50m or so here.
21:17 lithos joined #perl6
21:17 jnthn ...not that it's a "who can leak fastest" competition. :-)
21:18 quietfanatic What is a sub call doing allocating memory, anyway?  Unless it's for the stack or something
21:18 lithos hello, perl6 community! o/
21:18 jnthn lithos: hi! :-)
21:18 quietfanatic hello, lithos
21:19 jnthn quietfanatic: It'd allocate a call frame, but those are meant to be garbage collectable.
21:19 masak oh hai, lithos!
21:19 lithos I've been following perl6 development for quite a long time. I'd like to help, too
21:19 moritz_ cool
21:20 masak \o/
21:20 moritz_ what would you like to do?
21:20 moritz_ write tests? documentation? a book? compiler?
21:20 masak applications?
21:20 moritz_ make cool websites? write Perl 6 libraries or applications?
21:20 jaffa4 a book
21:20 masak submit triage RT tickets?
21:20 jaffa4 Will you publish it?
21:20 quietfanatic Play with a game engine?
21:21 jaffa4 what game engine?
21:21 quietfanatic Gamebase
21:21 lithos the thing is, I'd like to do something that does not involve copyright issues
21:21 quietfanatic http://github.com/quietfanatic/gamebase
21:21 moritz_ jaffa4: no, we just write it to delete it afterwards
21:21 masak hugme: hug moritz_
21:21 * hugme hugs moritz_
21:22 masak jaffa4: it's being published around April, when Rakudo Star lands.
21:22 quietfanatic (though this memory leak makes it unable to run more than 15 seconds or so right now). :|
21:22 jnthn quietfanatic: Is this a new problem?
21:22 masak quietfanatic: depends how much RAM you have. :P
21:22 quietfanatic I don't know
21:22 lithos so I thought about helping with testing and debugging
21:22 jnthn quietfanatic: That is, you're not really seen it before?
21:22 jnthn OK.
21:22 masak lithos: sounds excellent!
21:22 quietfanatic jnthn: I don't know, I haven't been checking recently
21:23 jnthn quietfanatic: OK.
21:23 jnthn Guess we could always file Parrot bug.
21:23 quietfanatic jnthn: I do know that it's in the last month or so
21:23 goneri joined #perl6
21:23 masak lithos: have you seen the spectests yet?
21:23 lithos yes
21:23 masak good. do you feel up to writing some?
21:23 jnthn quietfanatic: Yeah. Probably worth reporting, if it hasn't already been.
21:24 jnthn quietfanatic: I can, or you can...I don't mind. :-)
21:24 quietfanatic jnthn: where do I report parrot bugs?
21:24 jnthn quietfanatic: trac.parrot.org I believe.
21:24 jnthn quietfanatic: http://gist.github.com/232313 # for easier copy-paste to teh ticket
21:24 quietfanatic jnthn: okay
21:25 diakopter quietfanatic: 'course, there's no link to add a bug ticket
21:25 lithos yes, I think. depends on the area of testing
21:25 jnthn Oh, I think you ahve to makean account and log in. :-|
21:26 lithos masak: do spectests need any kind of copyright assignment? (i ask because my employment contract is nasty in this regard)
21:26 quietfanatic diakopter: if you log in there is
21:27 masak lithos: good question.
21:27 masak no, I don't think so.
21:27 masak making Rakudo commits does, but not committing to the Pugs repo.
21:28 Juerd lithos: Well, it would hurt the project immensely if your employer would start to claim copyright on your work.
21:28 masak jnthn: is there any way, from within Rakudo, to force a Str to be coded in utf-8?
21:28 [particle] i'm guessing if tpf is officially blessing the spec and the spectests as the definition of Perl 6, they'll want copyright
21:29 quietfanatic jnthn: Parrot version would be trunk if I --gen-parrot'ed it, right?
21:29 moritz_ lithos: no copyright assignment necessary, but the test suite is Artistic License 2
21:29 [particle] and will release with Artistic 2
21:29 masak [particle]: I wonder if that's possible.
21:29 Juerd lithos: See http://www.perlmonks.org/?node=Profession​al%20Employees%20and%20Works%20for%20Hire
21:29 [particle] i wonder, too
21:29 moritz_ lithos: so you need to be allowed to publish things under AL2
21:29 Juerd It depends on your contract whether this does or doesn't affect you.
21:29 jnthn masak: Good question.
21:30 jnthn masak: I thought those things were Buf-ish concenrs?
21:30 jnthn *concerns?
21:30 tylerni7 joined #perl6
21:30 masak jnthn: I think I have a situation where I have non-utf8-Strs coming in and making a Parrot-based library confused.
21:30 * lithos takes a look
21:30 jnthn quietfanatic: We're not quite on trunk, but close.
21:30 jnthn quietfanatic: See build/PARROT_REVISION
21:31 masak jnthn: you mean HEAD, not trunk.
21:31 quietfanatic jnthn: okay
21:31 jnthn masak: I didn't say trunk was a version. :-P
21:31 jnthn masak: but yes, I did. :-)
21:31 jnthn masak: Hmm...that's an...interesting...situation. :-|
21:31 jnthn masak: Which Parrot library?
21:31 jnthn And what makes you think that's the problem?
21:32 masak jnthn: hold on, I'll paste a URL to the logs where I already describe the symptoms.
21:32 jnthn k
21:32 masak jnthn: http://irclog.perlgeek.de/nove​mber-wiki/2009-11-11#i_1717654
21:33 masak so, &digest is a wrapper to a Parrot lib.
21:35 jnthn masak: And you're getting the wrong data from...?
21:35 jnthn Sorry
21:35 jnthn I meant where is it you're reading the stuff that gets bad results from digest from?
21:35 masak jnthn: I only get wrong results when pulling in data from $*IN, via November.pm.
21:36 jnthn If you "say" the data, does it look OK?
21:36 jnthn (that is, the thing you feed to digest)?
21:37 lithos Juerd, that article describes my situation well. does this mean i cannot contribute anything without making the project vulnerable? :(
21:37 masak jnthn: yes.
21:37 masak it looks identical.
21:37 masak same length, too.
21:38 masak haven't tried printing the number of bytes, though. will do that now.
21:38 jnthn masak: (horrible though) does a .split('').join('') give back a string that works out any better? :-/
21:39 xenoterracide joined #perl6
21:39 masak jnthn: so far, I've only tried concatenating it with the empty string. will try your variant too. hold on.
21:39 moritz_ what does .PARROT tell you about the string?
21:39 jnthn Oh, that's a good point too.
21:39 masak that also shows no differences.
21:40 jnthn masak: Was that to me or moritz++?
21:40 masak to moritz_.
21:40 masak .bytes also shows no differences. :/
21:40 phenny masak: 'also shows no differences. :/'
21:40 masak phenny: oh, don't you start, too.
21:40 moritz_ wtf?
21:41 jnthn .wtf
21:41 jnthn .wtf huh :/
21:41 masak Fifth Law of Bots: There are no good bots.
21:41 moritz_ hugme: hug masak
21:41 * hugme hugs masak and blushes
21:41 jnthn .thingy does nothing :/
21:41 masak hugme++
21:41 jnthn .bytes triggers it?
21:41 phenny jnthn: 'triggers it?'
21:41 jnthn .chars too mabye?
21:41 japhb joined #perl6
21:41 jnthn ...erm. :-/
21:42 moritz_ .bytes möp
21:42 phenny moritz_: 'm\xc3\xb6p'
21:43 jnthn Wow.
21:43 masak jnthn: your scary suggestion... worked.
21:43 masak jnthn: now I have something to blog about today. :)
21:43 moritz_ .bytes bold
21:43 phenny moritz_: '\x02bold\x02'
21:43 masak jnthn++
21:44 jnthn masak: OK...that's...odd
21:44 jnthn masak: curious. did you try "" ~ $str and $str ~ "" ?
21:44 jnthn e.g. both of them?
21:44 masak jnthn: no, only the former.
21:44 jnthn ok
21:45 masak I wish I had a small test case for this.
21:45 * jnthn would hope that wouldn't make a difference, but... :-|
21:45 jnthn masak: Yeah. :-S
21:45 masak right now, all I have is November.
21:45 jnthn It seems like something that's going to cause pain elsewhere.
21:45 masak jnthn++ jnthn++ jnthn++ # couple more karma for fixing this
21:46 jnthn masak: Huh, I didn't fix anything. :-P
21:46 masak I mean the temporary fix.
21:46 jnthn Ah
21:46 masak now people can log in again.
21:46 jnthn :-)
21:46 jnthn \o/
21:46 * jnthn gradually pieces parametric roles back together.
21:48 moritz_ let me introduce you to a German-English mixed word: "Würgaround" - it's like a work-around, but "würgen" means to choke or to suffocate
21:48 masak moritz_: very nice.
21:48 masak I will treasure this word.
21:49 jnthn Nice! :-)
21:49 * cotto_work needs to hang out in #perl6 more often
21:50 masak everyone needs to hang out in #perl6 more often :)
21:54 * diakopter wants a 17 Mingabit hard disk
21:58 diakopter or the new 2 Lumabit one
21:58 jnthn nqp: say(2*8192)
21:58 p6eval nqp: 16384␤
21:59 jnthn worrevaminga
21:59 jnthn </geordie_slang>
22:01 lmc joined #perl6
22:09 Whiteknight joined #perl6
22:16 masak jnthn: here's a smaller example of where it goes wrong. http://gist.github.com/232360
22:17 masak jnthn: only requires CGI.pm and Digest.pm from the November repo.
22:17 synth joined #perl6
22:18 jnthn pmichaud: The wonderful "default from outer" flag is implemented, pushed and activated for implicit $_. :-)
22:18 pmichaud jnthn++ \o/
22:18 jnthn pmichaud: It was easy, once I remembered to do it. :-)
22:18 pmichaud I feel much better about that -- thanks.
22:18 jnthn np
22:19 pmichaud having a bunch of "grab $_ from outer" subs kinda bugged me.
22:19 jnthn Yeah.
22:19 jnthn Woulda been slow too.
22:19 pmichaud okay, so we now have even better pointy subs.  If someone wants to implement "given", it should be really simple now.  :-)
22:20 * jnthn looks around for mathw
22:20 pmichaud hint:  get $<xblock>.ast
22:20 jnthn Well, I guess I could always jfdi too. :-)
22:20 pmichaud swap the 0 and 1 elements
22:20 Chillance joined #perl6
22:20 pmichaud change the pasttype to 'call'
22:20 pmichaud profit.
22:21 * jnthn leaves someone else to have a chance
22:21 jnthn I'll do it tomorrow if nobody claims it first. :-)
22:21 jnthn I think my latest refactor of roles sucks less than all my previous ones.
22:21 * moritz_ reminds people get t/00-parrot/10-regex.t passing
22:21 moritz_ or formulated differently, I want those freakin' awesome regexes in my Perl 6 ;-)
22:22 * jnthn looks at Mr Regex Guy :-)
22:22 * pmichaud looks around for Mr Regex Guy and doesn't see him.
22:22 masak here I am! :)
22:22 * jnthn hands masak a mirror
22:22 jnthn gah
22:22 jnthn s/masak/pmichaud/
22:22 jnthn ...that was a great tab complete fail...
22:23 moritz_ wow, we have a new regex overlord!
22:23 masak jnthn: classic Freudian slip. :)
22:23 pmichaud suuuuure, blame the tab key.
22:23 pmichaud I, for one, welcome our new regex overlords.
22:23 * diakopter read that as 'overloads'
22:23 moritz_ pmichaud: I'm sure jnthn wanted to type 'pm<tab>', and accidentally missed the 'p' ;-)
22:23 Juerd 22:48 <@moritz_> let me introduce you to a German-English mixed word: "Würgaround" - it's like a work-around, but "würgen" means to choke or to suffocate
22:23 Juerd Or you could say wurgaround, which would be the Dutch-English mixed word, and easier to type
22:24 moritz_ Juerd: now that's cheating ;-)
22:24 Juerd Full legacy character encoding compatibility!
22:24 jnthn Juerd: Yes, but does it involve choking?
22:24 * masak is with Juerd on this one
22:24 Juerd jnthn: Yes, same meaning
22:24 * jnthn is sold
22:24 masak Dutch++
22:24 jnthn Dutch++ indeed.
22:24 Juerd Slightly different pronunciation. As a matter of fact, it sounds a little more like you're being strangled.
22:24 Juerd gggggggggggggggg :)
22:25 diakopter ++Afrikaans
22:25 diakopter erm
22:25 moritz_ doesn't most dutch sounds a bit strangled? *SCNR*
22:25 jnthn Juerd: Well, good job I only have to type it.
22:25 jnthn Juerd: A couple of half-Dutch girls spent immense amounts of time trying to get me able to make that ggggggggggg noise. And failed. :-)
22:25 justatheory joined #perl6
22:26 Juerd Awh :)
22:26 * moritz_ doesn't want to know all the details
22:26 diakopter jnthn: hmm. did they still make you pay them?
22:26 jnthn ...
22:26 jaffa8 joined #perl6
22:26 jnthn ...it was actually a completely innocent linguistic exercise.
22:26 jnthn :-P
22:27 Juerd diakopter: rotflol
22:27 Juerd I can only wonder how that mind of yours works :)
22:28 * Juerd did awful html/css work.
22:28 Juerd And now I have to test it in MSIE, but I'm scared.
22:29 jnthn You did it awful, so it'll probably work fine. ;-)
22:29 Juerd It's awful because the design requires awful techniques
22:30 Juerd I had to script some HTML generating thing to build one part of the layout
22:30 jnthn <table>? ;-)
22:30 jnthn Eww.
22:30 Juerd Almost
22:30 Juerd http://juerd.nl/floater is an example of what that script does.
22:30 jnthn Yeah, I find that whole area quite horrible.
22:30 jnthn .oO( do I really want to click that link? )
22:31 Juerd Sure. I'm not convinced you want to see the source though
22:31 jnthn It was the use of "floater" :-)
22:31 jnthn oh win, the script is called that too. :-D
22:31 Juerd Does that mean something else, besides something that floats?
22:32 jnthn Erm, well, yes. :-)
22:32 Juerd Uh oh
22:32 moritz_ oh, that's *really* scary
22:33 jnthn Full marks on the script though. :-)
22:33 Juerd jnthn: ...what does it mean? :)
22:33 jnthn Juerd: I'm trying to think of a nice way to say this, but failing.
22:33 Juerd (Uh oh)**2
22:34 jnthn Juerd: It's when somebody has a poop, and then doesn't flush. They're said to have "left a floater". :-)
22:34 Juerd Hahaha
22:34 Juerd Right.
22:34 Juerd That's still something that floats though :)
22:34 jnthn I'm not sure how, erm, Yorkshire-specific this is. :-)
22:35 Juerd Does this happen a lot in Yorkshire?
22:35 Juerd I think I've never seen it happen...
22:35 Juerd Oh, may be because of Dutch toilets - duh.
22:36 masak Juerd: no flush? :)
22:36 masak *scnr*
22:36 zaslon lolmasakhazblogged! masak++ 'November 11 2009 -- nobody said it was going to be easy': http://use.perl.org/~masak/journal/39884?from=rss
22:36 masak well, no-one did.
22:36 Juerd masak: They flush alright
22:36 masak Juerd: industry-strength flush?
22:37 Juerd Poop falls onto a, well, let's call it a shelf.
22:37 Juerd You can actually examine (and smell) it before you flush.
22:37 jnthn Juerd: I wouldn't say a lot, but enough to have engrained that meaning in my head forevermore...
22:37 jnthn ...what have I started...
22:38 masak Juerd: ah, so floaters are hardly possible.
22:38 Juerd masak: Well, you could easily get them but it requires, er, a good aim.
22:38 masak now comes the interesting Sapir-Whorf question: did the construction of the toilets affect language, or the other way around? :)
22:39 Juerd Hah
22:39 jnthn lol :-)
22:39 masak see? you can turn even scatology into linguistics!
22:39 Juerd scatology? Is that a word?
22:39 masak 'fraid so.
22:40 * jnthn is intimated by masak's enormous vocabularly
22:40 jnthn gah
22:40 jnthn intimidated
22:40 Juerd It resembles schijt (which is Dutch for shit)
22:40 jnthn *sigh*
22:40 masak jnthn: 'intimated' is a word, too :)
22:40 jnthn masak: Yes, and it *so* wasn't the one I wanted.
22:41 masak :)
22:41 jnthn Juerd: Yay! My first Dutch swear word.
22:41 jnthn Soon my Dutch vocab will be on a par with my Swedish. ;)
22:41 Juerd jnthn: It contains the Dutch hard g though, so you can't pronounce it :D
22:41 jnthn Oh crap.
22:42 Juerd Some dialects pronounce the ch as k
22:42 Juerd Then schijt becomes like skite.
22:42 japhb NQP users: When I need to bind an empty array to a hash element in NQP, I've come up with two ways: 1. '%foo<bar> = my @empty;' which unfortunately is prone to error if you try it twice in the same scope and forget to use a different name for @empty, or 2) '%foo<bar> = {my @empty}();' which fixes the naming problem, but is even uglier than #1.  Any better ideas?
22:42 jnthn That sounds a tad easier. :-)
22:42 masak jnthn: don't worry, I'll teach you the Dutch 'g' in Latvia.
22:43 masak japhb: a sub?
22:43 jnthn japhb: sub make_me_an_empty() { my @empty; @empty }
22:43 * masak is faster o/
22:43 * jnthn is exampler o/
22:43 japhb masak, jnthn: That's what I *used* to have, and what I was trying to get rid of.  ;-)
22:43 pmichaud japhb: what would you like it to be?
22:44 masak japhb: don't get rid of it! it's great!
22:44 japhb pmichaud, []
22:44 pmichaud japhb: how abut    %foo<bar> := ()  ?
22:44 pmichaud or that
22:44 japhb either!
22:44 pmichaud I can make those happen.
22:44 jnthn my @a = pir::split('') # ;-)
22:44 jnthn er, := of course :-)
22:44 pmichaud nqp:  say(());
22:44 p6eval nqp: Confused at line 1, near "say(());"␤current instr.: 'parrot;HLL;Grammar;panic' pc 448 (src/cheats/hll-grammar.pir:197)␤
22:44 japhb pmichaud, the problem is that %foo<bar> := (1,2); works, as does := (1,) ... but not := ();
22:44 pmichaud okay.  Will fix when I get back tonight.
22:45 jnthn pmichaud: I'd find it quite useful too.
22:45 pmichaud hmmm, I wonder
22:45 pmichaud nqp:  say(<>);
22:45 p6eval nqp: ␤
22:45 pmichaud nqp:  say(<>.elems);
22:45 p6eval nqp: Method 'elems' not found for invocant of class 'String'␤current instr.: '_block11' pc 0 (EVAL_1:6)␤
22:45 jnthn aww
22:45 pmichaud nqp:  say(pir::elements(<>));
22:45 p6eval nqp: 0␤
22:45 pmichaud ftw
22:45 japhb So := (<>) then?
22:46 pmichaud nqp:  my %h;  %h<a> := <>;   say(pir::elements(%h<a>));
22:46 p6eval nqp: 0␤
22:46 pmichaud <> works for now :-)
22:46 pmichaud but yes, () should work also
22:46 jnthn rakudo: say <>.WHAT
22:46 p6eval rakudo 50e495: List()␤
22:46 pmichaud rakudo ftw
22:46 jnthn nqp: say(<>.WHAT)
22:46 p6eval nqp: Method 'WHAT' not found for invocant of class 'String'␤current instr.: '_block11' pc 0 (EVAL_1:6)␤
22:46 pmichaud ...string?
22:47 jnthn The String there confuses me.
22:47 pmichaud hmmm. fail.
22:47 japhb pmichaud, yeah, that's suboptimal
22:47 jnthn pmichaud: It said it above for .elems too
22:47 pmichaud right, it should be ResizablePMCArray
22:47 jnthn pmichaud: Right.
22:47 pmichaud elements on a string returns the number of characters in Parrot
22:47 jnthn Oh.
22:47 jnthn Fail.
22:47 pmichaud Fail.
22:47 * pmichaud hangs his head.
22:47 jnthn hugme: hug pmichaud
22:47 * hugme hugs pmichaud
22:47 pmichaud I'll see about fixing it tonight.
22:48 japhb I'm constantly falling over ("foo",) is a list, but ("foo") is a string, which will happily be treated as a container if you try to iterate it!
22:48 pmichaud I'll get () to work
22:48 diakopter nqp: say(< 11>[2])
22:48 p6eval nqp: 1␤
22:48 pmichaud I take little responsibility for Parrot's builtin PMC semantics.
22:48 japhb pmichaud, Oh, not blaming you.  Just saying it's *really* easy to trip over that one.
22:49 pmichaud I'm still trying to figure out exactly how I want @(...) and %(...) to work in the general case
22:50 pmichaud anyway, the answer should be ().  and maybe {} can create an empty hash.
22:50 diakopter nqp: say(<<<<h>>>>[0])
22:50 p6eval nqp: h␤
22:50 * diakopter fires up the fuzzer
22:50 jnthn pmichaud: We could get rid of loads of Q:PIR in Actions.pm with that.
22:50 pmichaud jnthn: then we shall do it.
22:51 japhb pmichaud, masak: FWIW, instead of make_me_an_empty(), I actually have as_array(), which was just a PIR function to slurp up the args and return the slurped array.
22:51 jnthn \o/
22:51 pmichaud okay, gotta take kid to soccer practice.  bbiah or so
22:51 jnthn kk
22:51 japhb pmichaud, thanks
22:52 pmichaud japhb: fwiw, my evening tonight is likely to be nqp stuff, including placing nqp into parrot repo.
22:52 pmichaud (gone)
22:52 japhb YAY!
22:52 japhb pmichaud++ # For when he gets back.  :-)
22:56 Juerd diakopter: What's a fuzzer?
22:56 Juerd It sounds a bit like phaser.
22:56 Juerd But... fuzzier.
22:57 jnthn Ooh. Anonymous roles should work in ng too. :-)
22:57 Juerd Oh. Turns out this is one of those things that you should search for with normal Google instead of Google Images :)
22:58 * Juerd kills time reading what people say about Dutch toilets. It's hilarious :)
22:59 diakopter Juerd: short for fuzz-tester
23:00 diakopter can use a parser's or compiler's input grammar to generate bad (or good!) inputs...
23:01 moritz_ or uses an existing valid input and mutates it randomly
23:01 jnthn or you can just give it to diakopter
23:01 jnthn The Human Fuzzer
23:01 jnthn :-)
23:02 mubot joined #perl6
23:02 scion joined #perl6
23:02 diakopter hugme: hug it more
23:02 * hugme hugs it
23:03 payload joined #perl6
23:07 quietfanatic huh
23:07 masak jnthn: http://gist.github.com/232360 -- voila
23:08 * masak submits rakudobug
23:09 jnthn masak: ewww
23:09 jnthn masak++ # very nicely golfed
23:09 masak my pleasure.
23:09 quietfanatic Now I'm getting a "invoke() not implemented in class 'ResizablePMCArray'" the second time I use the overloaded trait_mod:<is> but not the first...
23:10 moritz_ rakudo: say 'user_name=carl'.split("=")[1].PARROT
23:10 p6eval rakudo 50e495: String␤
23:10 moritz_ rakudo: say 'carl'.PARROT
23:10 p6eval rakudo 50e495: Str␤
23:10 moritz_ there you go.
23:10 masak moritz_: d'oh!
23:11 masak moritz_: I didn't get that in my method.
23:11 masak could the String have been wrapped in a Str or something?
23:11 * masak adds moritz_'s discovery to the ticket
23:11 moritz_ rakudo: say 'user_name=carl'.split("=")[1].trans(['c'] => ['C'])
23:11 p6eval rakudo 50e495: too few positional arguments: 2 passed, 3 (or more) expected␤in Main (file <unknown>, line <unknown>)␤
23:11 moritz_ rakudo: say 'carl'.trans(['c'] => ['C'])
23:11 p6eval rakudo 50e495: Carl␤
23:12 moritz_ masak: t/spec/integration/real-strings.t is full of similar tests
23:12 moritz_ .trans happens to be a good method to test that because it's present both in parrot and rakudo, but uses a different call syntax
23:12 masak nod.
23:12 moritz_ and we already have a test case of $str.split(Regex)
23:13 masak I should have thought of the .trans test for my strings.
23:14 masak so... all evening, I've been chasing a known bug. :/
23:14 moritz_ this particular instance was not known.
23:14 masak no?
23:14 * masak gets hopeful again
23:17 moritz_ I don't understand that
23:18 masak ah. time to sleep, I believe.
23:18 moritz_ Any.split(Any), which is the case that Str.split(Str) handles, just uses .substr to under extract the pieces
23:18 moritz_ rakudo: say 'carl'.substr(0, 2).PARROT
23:18 p6eval rakudo 50e495: String␤
23:18 moritz_ oh.
23:18 masak :)
23:19 moritz_ so that's to blame.
23:19 masak it's a wonder any programs using strings work at all!
23:19 moritz_ aye
23:19 masak 'night
23:19 moritz_ that's probably we extract so often with regexes
23:19 moritz_ good night
23:19 jnthn night masak
23:20 Wolfman2000 ...if I'm lucky, I'll finally be able to relax a little. SQLite work later tonight for the pastebin.
23:23 moritz_ epic
23:23 diakopter rakudo: say (~('carl'.substr(0, 2))).PARROT
23:23 p6eval rakudo 50e495: Str␤
23:23 diakopter rakudo: say (~('carl'.substr(0, 2)))
23:23 p6eval rakudo 50e495: ca␤
23:23 diakopter hee hee
23:23 moritz_ I just tried to change substr to return a PMC instead of string register
23:24 diakopter a fix
23:24 moritz_ Null PMC access in set_string_native()
23:24 moritz_ whiel compiling
23:24 moritz_ while compiling the setting
23:26 moritz_ I don't even want to know what happens if I move it to the setting
23:28 jnthn moritz_: You may need to instantiate the pMC first
23:28 jnthn moritz_: nopaste the diff?
23:28 moritz_ $P0 = $S1
23:28 moritz_ return($P0)
23:28 moritz_ instead of
23:28 moritz_ return($S1)
23:28 moritz_ would $P0 = box $S1
23:28 moritz_ help?
23:29 jnthn no
23:30 jnthn $P0 = new ['Str']
23:30 jnthn $P0 = $s1
23:30 jnthn *S1
23:30 jnthn .return ($P0)
23:30 jnthn Try that.
23:30 moritz_ will do
23:35 oZ] left #perl6
23:37 xomas joined #perl6
23:40 * mberends reads the latest masakblog, presses F5 on http://www.november-wiki.org/in, and gets "You got in!" WOOT!
23:42 * moritz_ closes the bug that prevented november from working without that ugly workaround
23:42 explorer joined #perl6
23:42 moritz_ RT #70415
23:42 moritz_ my spectest didn't finish yet, but I'm rather confident that it won't turn up any new nastinesses
23:45 jnthn moritz_: Excellent. :-)
23:54 mathw evening
23:55 mberends evening mathw, 5 minutes before UTC-morning :)
23:56 mathw yes
23:56 mathw however I just got in
23:56 mathw so it's still evening for me
23:56 jnthn hi mathw :-)
23:57 mathw I just validated all the money I ever spent on singing lessons
23:57 moritz_ mathw: do tell ;-)
23:57 mathw sang at a charity gig with the band from work
23:58 mathw it went down really well
23:58 mathw loads of people came
23:58 mathw and bought raffle tickets
23:58 mathw think the raffle raised about £750
23:58 mathw there was lots of dancing and singing along
23:58 mathw and applause, which we performers do like to hear :)
23:58 moritz_ cool
23:58 mberends mathw++ # IRC applause
23:58 mathw So I'm buzzing
23:59 mathw my singing teacher even came along
23:59 jnthn mathw++ # cool! :-)
23:59 mathw and seemed to have a lovely time
23:59 mathw she was dancing and singing along and getting rather drunk

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

Perl 6 | Reference Documentation | Rakudo | Niecza | Specs