Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2013-05-16

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:07 btyler joined #perl6
00:18 leprevost joined #perl6
00:36 berekuk joined #perl6
00:44 adu joined #perl6
00:53 LoRe_ joined #perl6
00:55 btyler joined #perl6
00:58 preflex joined #perl6
01:05 xilo joined #perl6
01:11 bbkr joined #perl6
01:13 anuby joined #perl6
01:19 FROGGS joined #perl6
01:21 kurahaupo joined #perl6
01:25 xilo joined #perl6
01:34 adu joined #perl6
01:48 prevost joined #perl6
02:06 DreamingInCode joined #perl6
02:09 bonsaikitten joined #perl6
02:13 lustlife joined #perl6
02:47 diakopter google+, y u completely different from 3 hours ago
02:49 Pleiades` imitating facebook?
03:05 lizmat_ joined #perl6
03:05 census joined #perl6
03:08 smash_ joined #perl6
03:08 dagurval joined #perl6
03:09 ivan``_ joined #perl6
03:09 odoacre_ joined #perl6
03:10 szbalint_ joined #perl6
03:10 atrodo_ joined #perl6
03:10 nebuchad` joined #perl6
03:11 rhr_ joined #perl6
03:11 nyuszika7h joined #perl6
03:11 mls joined #perl6
03:12 [Coke]_ joined #perl6
03:12 jlaire_ joined #perl6
03:14 pmichaud joined #perl6
03:14 Rix joined #perl6
03:27 labster rn: enum Toggle <MAGIC MORE_MAGIC>;  multi infix:<=> (Toggle $a, Toggle $b) { say "$a | $b"; nextsame; };  my Toggle $a = MAGIC; $a = MAGIC;
03:27 camelia niecza v24-51-g009f999: OUTPUT«Toggle() | MAGIC␤Unhandled exception: Illegal control operator: nextsame/nextwith␤  at /home/p6eval/niecza/lib/CORE.setting line 1533 (nextsame @ 4) ␤  at /tmp/fqiMJQEnPN line 1 (infix:<=> @ 6) ␤  at /tmp/fqiMJQEnPN line 1 (mainline @ 5) ␤  at /home/p6eval…
03:27 camelia ..rakudo 570d1d:  ( no output )
03:28 labster .oO ( Maybe this work work if I set it to MORE_MAGIC? )
03:31 bpalmer joined #perl6
03:55 rindolf joined #perl6
03:58 adu joined #perl6
04:05 djanatyn joined #perl6
04:20 adu joined #perl6
04:24 lue labster: here's the sub line for = in src/core/operators.pm:  sub infix:<=>(Mu \a, Mu \b) is rw {
04:25 Psyche^ joined #perl6
04:26 lue (It's not a multi in the core, so I doubt that would work)
04:45 kurahaupo joined #perl6
04:49 anuby_ joined #perl6
04:49 fhelmberger joined #perl6
04:51 konundra joined #perl6
05:00 thou joined #perl6
05:03 dmol joined #perl6
05:12 erkan joined #perl6
05:12 erkan joined #perl6
05:23 rindolf joined #perl6
05:30 rindolf joined #perl6
05:38 rindolf joined #perl6
05:48 rindolf joined #perl6
05:56 fhelmberger joined #perl6
05:58 SamuraiJack joined #perl6
05:58 rindolf joined #perl6
06:03 rindolf joined #perl6
06:47 lizmat_ good *, #perl6!
06:48 sorear lizmat_!
06:53 kaleem joined #perl6
07:05 * jnthn yawns
07:05 moritz \o *
07:06 jnthn Teaching today, so only occasionally about.
07:07 domidumont joined #perl6
07:11 bpalmer left #perl6
07:13 brrt joined #perl6
07:40 FROGGS joined #perl6
07:42 domidumont joined #perl6
07:48 sqirrel joined #perl6
07:58 berekuk joined #perl6
07:58 kivutar joined #perl6
08:02 labster good morning
08:03 sorear o/ labster
08:04 brrt hi #perl6
08:07 kaleem joined #perl6
08:24 djanatyn joined #perl6
08:26 Tene joined #perl6
08:26 Tene joined #perl6
08:41 dakkar joined #perl6
09:04 donaldh joined #perl6
09:05 daxim joined #perl6
09:14 domidumont joined #perl6
09:23 labster \o sorear
09:23 cognominal__ joined #perl6
09:23 labster delayed reaction there
09:24 xinming joined #perl6
09:44 donaldh Argh, Java classpath separator on Windows is ; but classpaths require quoting for this to work on Cygwin
09:44 donaldh Do we want to support nqp / jvm on Cygwin?
09:45 labster I'm pretty sure that someone here is using cygwin.
09:46 donaldh Yeah, I was just trying it in the absence of a Windows compiler toolchain.
09:46 donaldh But realize it really behaves like a third platform.
09:46 labster Also, I feel your pain on multiple platform weirdness.
09:47 tadzik donaldh: are you working on rakudo/jvm runner?
09:47 donaldh tadzik: yeah
09:47 tadzik donaldh: perhaps something could be stolen from nqp? It has a runner too, maybe the problem is solved there
09:48 donaldh I've queued a pull request already. It's not tested on Windows because I don't have a toolchain
09:48 donaldh And yes, it's largely stolen from nqp.
10:02 xinming joined #perl6
10:38 breinbaas joined #perl6
10:48 sqirrel joined #perl6
10:51 xinming joined #perl6
11:10 census joined #perl6
11:11 rking1 joined #perl6
11:44 FROGGS std: use v5; say $]
11:44 camelia std 9906f18: OUTPUT«ok 00:01 53m␤»
11:53 _jaldhar_ joined #perl6
11:54 gudahtt joined #perl6
12:08 lizmat r: my %h=a=>1; for %h.keys { say "{$_.WHAT}" }    # me thinks this is a rakudo bug
12:08 camelia rakudo 570d1d: OUTPUT«use of uninitialized value of type Str in string context  in block  at /tmp/PdHofbm1Cl:1␤␤␤»
12:08 lizmat r: my %h=a=>1; for %h.keys { say $_.WHAT }    # this works
12:08 camelia rakudo 570d1d: OUTPUT«(Str)␤»
12:09 FROGGS r: my %h=a=>1; for %h.keys { say "{$_}" }    # me thinks this is a rakudo bug
12:09 camelia rakudo 570d1d: OUTPUT«a␤»
12:09 lizmat r: my %h=a=>1; for %h.keys { say "{$_}" }    # so does this
12:09 camelia rakudo 570d1d: OUTPUT«a␤»
12:09 FROGGS r: my %h=a=>1; for %h.keys { say "{$_.Str}" }    # me thinks this is a rakudo bug
12:09 camelia rakudo 570d1d: OUTPUT«a␤»
12:09 moritz lizmat: afaict it works according to the specs (more)
12:10 moritz lizmat: your confusion stems from the fact that .WHAT returns a type object, which say() turns into (Type), but stringification leads to empty string + warning
12:10 moritz (say uses .gist instead of .Str for conversion)
12:11 moritz if you want to get the name of a type as a string, use $obj.^name
12:11 FROGGS r: my %h=a=>1; for %h.keys { say "{$_.WHAT.gist}" }
12:11 camelia rakudo 570d1d: OUTPUT«(Str)␤»
12:11 lizmat ok, I now understand why it does this
12:12 lizmat still it feels incorrect...
12:13 FROGGS r: say ~Str
12:13 camelia rakudo 570d1d: OUTPUT«use of uninitialized value of type Str in string context  in block  at /tmp/qVjEuegmaR:1␤␤␤»
12:13 FROGGS then this must feel the same
12:13 lizmat indeed…
12:14 lizmat guess it's one of those things to get used to, as having to put a comma behind the block in grep
12:14 FROGGS yeah
12:15 lizmat moritz++, FROGGS++ for the clarification
12:20 rindolf joined #perl6
12:21 adu joined #perl6
12:26 lizmat r: my %h=a=>1; for %h.kv -> $k,$v is rw { $v=3 }; say %h  # is this NYI ?
12:26 camelia rakudo 570d1d: OUTPUT«("a" => 1).hash␤»
12:26 lizmat expected a => 3
12:26 FROGGS n: my %h=a=>1; for %h.kv -> $k,$v is rw { $v=3 }; say %h  # is this NYI ?
12:26 camelia niecza v24-51-g009f999: OUTPUT«Potential difficulties:â�¤  $k is declared but not used at /tmp/EgQU4BuOmk line 1:â�¤------> [32mmy %h=a=>1; for %h.kv -> [33mâ��[31m$k,$v is rw { $v=3 }; say %h  # is this [0mâ�¤â�¤{"a" => 3}â�¤Â»
12:26 ivan`` joined #perl6
12:27 FROGGS looks like
12:27 btyler joined #perl6
12:27 lizmat r: my %h=a=>1; for %h.values -> $v is rw { $v=3 }; say %h  # same for .values
12:27 camelia rakudo 570d1d: OUTPUT«("a" => 1).hash␤»
12:28 lizmat n: my %h=a=>1; for %h.keys -> $k is rw { $k="b" }; say %h
12:28 camelia niecza v24-51-g009f999: OUTPUT«Unhandled exception: Writing to readonly scalar␤  at /tmp/gy03ZaGbp_ line 1 (mainline @ 8) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4331 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4332 (module-CORE @ 582) ␤  at /home/p6eval/niecza/li…
12:29 lizmat r: my %h=a=>1; for %h.keys -> $k is rw { $k="b" }; say %h  # not expecting this to work, but also not silently fail
12:29 camelia rakudo 570d1d: OUTPUT«("a" => 1).hash␤»
12:36 konundra joined #perl6
12:39 colomon lizmat: it's not silently failing.
12:40 colomon you're changing the list of keys you created, not the original hash
12:40 lizmat r: my %h=a=>1; for %h.values -> $v is rw { $v=3 }; say FAIL if %h<a> != 3
12:40 camelia rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Undeclared name:â�¤    FAIL used at line 1â�¤â�¤Â»
12:40 lizmat r: my %h=a=>1; for %h.values -> $v is rw { $v=3 }; say "FAIL" if %h<a> != 3
12:40 camelia rakudo 570d1d: OUTPUT«FAIL␤»
12:41 lizmat colomon: nothing with keys in this one
12:41 colomon nr: my %h = a => 1; my @a := %h.keys; @a[0].key = "boo!"; say @a.perl; say %h.perl
12:41 camelia niecza v24-51-g009f999: OUTPUT«Unhandled exception: Unable to resolve method key in type Str␤  at /tmp/anEbtj0Pz1 line 1 (mainline @ 6) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4331 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4332 (module-CORE @ 582) ␤  at /home/p6…
12:41 camelia ..rakudo 570d1d: OUTPUT«No such method 'key' for invocant of type 'Str'␤  in block  at /tmp/cYDvt8gVWv:1␤␤»
12:41 colomon lizmat: yes, that one is values
12:41 colomon same difference
12:42 colomon nr: my %h = a => 1; my @a := %h.keys; @a[0] = "boo!"; say @a.perl; say %h.perl
12:42 camelia niecza v24-51-g009f999: OUTPUT«Unhandled exception: Writing to readonly scalar␤  at /tmp/rWdVgUQand line 1 (mainline @ 5) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4331 (ANON @ 3) ␤  at /home/p6eval/niecza/lib/CORE.setting line 4332 (module-CORE @ 582) ␤  at /home/p6eval/niecza/li…
12:42 camelia ..rakudo 570d1d: OUTPUT«("boo!",).list␤("a" => 1).hash␤»
12:42 colomon okay, that's interesting.
12:43 colomon (and you'd already gotten it above)
12:43 lizmat p5eval: my %h=(a=>1); for $v (values %h) { $v=3 }; say "OK" if $h{a} == 3  # is what I expected
12:43 p5eval lizmat: OK1
12:45 [Coke]_ p5eval: 1;
12:45 p5eval [Coke]_: 1
12:45 moritz (the 1 in OK1 is the return value from say())
12:45 [Coke]_ wonder if we should make that ignore the return value, like p6eval.
12:46 lizmat p5eval: my %h=(a=>1); for $v (values %h) { $v=3 }; "OK" if $h{a} == 3  # just testing p5eval
12:46 p5eval lizmat: OK
12:46 colomon lizmat: presumably one of those is an error, but I don't have any notion which one it is.
12:47 lizmat I think it's more a matter of NYI, let's wait for jnthn/masak/pmichaud to become available
12:47 lizmat afk for a bit
13:01 * moritz wonders if he should write an apache module that bans IPs that requests URLs ending in signup.php
13:02 kaare_ joined #perl6
13:08 hoelzro moritz: just use fail2ban =)
13:08 hoelzro that's what I do
13:08 arnsholt Sounds like a good idea to me =)
13:09 moritz hoelzro: isn't fail2ban on the ssh level?
13:09 hoelzro moritz: it analyses any log file and takes actions based on patterns
13:09 hoelzro so I have a series of patterns for my web server log
13:09 hoelzro and an iptables rule is added for addresses that match
13:10 hoelzro I use it for all of my external-facing services
13:11 kivutar joined #perl6
13:13 moritz I am now experimenting with mod_evasive for http://irclog.perlgeek.de/
13:13 moritz the problem is that now and then, a crawler discovers the site, and follows all links, and causes immense load
13:14 mikemol joined #perl6
13:15 moritz but there is also a legitimate use case of causing many requests: the front page loads a .png for every channel
13:16 arnsholt Do you cache the images?
13:17 moritz arnsholt: they are static files
13:17 moritz but mod_evasive can't count request for only-dynamic files
13:18 arnsholt That counts. Just checking in case they were on-demand
13:20 moritz I'm not that crazy :-)
13:20 arnsholt Good, good =)
13:20 moritz I've now set the limit to 50 requests in 15 seconds
13:21 moritz that allows loading the front page and then a bit of browsing
13:21 bluescreen10 joined #perl6
13:21 moritz if you get many 403s on the irclog page, please tell me :-)
13:22 moritz we'll see if it actually prevents accidental crawler DOS
13:23 LoRe_ moritz: did bing make a visit to you?
13:23 SmokeMachine joined #perl6
13:23 moritz LoRe_: bing *and* msn
13:23 moritz LoRe_: I've added bing's /16 net to my blacklist
13:25 LoRe_ i've created a robots.txt: http://pb.rbfh.de/cRxXLL6pkubA that helped :)
13:26 btyler joined #perl6
13:26 moritz LoRe_: the problem is that once it starts hammering the site, it's too late to install a robots.txt
13:26 prevost joined #perl6
13:27 moritz because it has already fetched the robots.txt, and doesn't feel the need to re-fetch
13:27 LoRe_ yeah, for the current run i blocked them too und unblocked them 24h later, since then it works :)
13:39 skids joined #perl6
13:48 lizmat r: my %h{Any}; %h<a><b>=1  # could it be that the auto-vivication of object hashes is NYI ?
13:48 camelia rakudo 570d1d: OUTPUT«No such method 'postcircumfix:<{ }>' for invocant of type 'Mu'␤  in block  at /tmp/59VwgIfXOm:1␤␤»
13:49 lizmat r: my %h{Str}; %h<a><b>=1  # could it be that the auto-vivication of object hashes is NYI ?
13:49 camelia rakudo 570d1d: OUTPUT«No such method 'postcircumfix:<{ }>' for invocant of type 'Mu'␤  in block  at /tmp/6x0kSaQ9W0:1␤␤»
13:49 moritz looks like
13:50 lizmat any place in the code I could look at for this?
13:51 PacoAir joined #perl6
13:51 moritz src/core/Hash.pm search for TypedHash
13:52 lizmat thanks, will have a stab at it
14:01 arnsholt "Saying Perl is hard to use because of its symbols is like saying Greek is hard to learn because it has a different alphabet." =D
14:11 sxw joined #perl6
14:15 lizmat afk for a few hours
14:17 geekosaur txt proves people find language hard to use because of all the letters and punctuation. (u agree?)      :p
14:27 * TimToady always thought it was the phonology, morphology, semantics, and pragmatics that made it hard...
14:28 TimToady oh, and syntax
14:28 sxw left #perl6
14:30 colomon nah, it's just the speling
14:30 spider-mario joined #perl6
14:31 TimToady just have to learn speliology then
14:31 xilo joined #perl6
14:37 mtk joined #perl6
14:39 nwc10 these days decent version control makes speliology a lot less painful
14:39 TimToady I suspect that visitor-y methods like .values should be allowed to return containers if that's the fastest way to return things that already have containers
14:40 TimToady it's the later binding that tends to be readonly or rw
14:40 TimToady however, for arrays of native types, we can't require that .values return containers, since that would be a lot of extra overhead
14:41 TimToady in that case, there should probably be a .values-rw that forces containerization
14:42 TimToady afk &
14:58 mucker joined #perl6
14:58 patspam joined #perl6
14:59 patspam left #perl6
15:04 arlinius joined #perl6
15:08 kbenson joined #perl6
15:16 jnthn evening, #perl6 o/
15:16 colomon \o
15:16 PerlJam jnthn: greetings!
15:18 FROGGS o/
15:23 patspam joined #perl6
15:37 btyler joined #perl6
15:49 bluescreen100 joined #perl6
15:55 cognominal joined #perl6
15:56 denisboyun joined #perl6
15:59 gtodd hmm long ago when IRC was young there was regexp creation race channel
16:00 gtodd if I recall ....
16:00 gtodd I guess mad regexp skills are so pre y2k ...
16:11 FROGGS[mobile] joined #perl6
16:23 SmokeMachine joined #perl6
16:27 daxim rakudo: say Set.WHAT
16:27 camelia rakudo 570d1d: OUTPUT«(Set)␤»
16:28 daxim rakudo: say Set.methods
16:28 camelia rakudo 570d1d: OUTPUT«No such method 'methods' for invocant of type 'Set'␤  in block  at /tmp/8yAiPFQ8jD:1␤␤»
16:28 daxim rakudo: say Set.methods()
16:28 camelia rakudo 570d1d: OUTPUT«No such method 'methods' for invocant of type 'Set'␤  in block  at /tmp/FqfWSr9GdJ:1␤␤»
16:28 daxim rakudo: say Set.^methods
16:28 camelia rakudo 570d1d: OUTPUT«default keys values elems exists Bool Numeric Real hash at_key exists_key new BUILD postcircumfix:<( )> iterator list pick roll of Str gist perl elems infinite item fmt Int Num chrs Numeric Str␤»
16:32 dalek rakudo/jvm-support: 8e54191 | jonathan++ | src/core/ (2 files):
16:32 dalek rakudo/jvm-support: Unbust the Parrot build after JVM tweaks.
16:32 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/8e54191b3c
16:35 dalek rakudo/jvm-support: c8d0895 | (Donald Hunter)++ | / (3 files):
16:35 dalek rakudo/jvm-support: Added a simple perl6 runner on JVM.
16:35 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/c8d0895bb1
16:35 dalek rakudo/jvm-support: 7e1b46d | (Donald Hunter)++ | tools/build/Makefile-JVM.in:
16:35 dalek rakudo/jvm-support: Add perl6.class to clean.
16:35 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/7e1b46d154
16:35 dalek rakudo/jvm-support: aaea644 | jonathan++ | / (3 files):
16:35 dalek rakudo/jvm-support: Merge pull request #140 from donaldh/jvm-support
16:35 dalek rakudo/jvm-support:
16:35 dalek rakudo/jvm-support: Added a simple perl6 runner on JVM.
16:35 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/aaea644367
16:37 mucker joined #perl6
16:38 daxim where's the documentation for the Test module?
16:42 AndChat|356841 joined #perl6
16:43 FROGGS joined #perl6
16:43 domidumont joined #perl6
16:46 [Coke]_ daxim: http://perlcabal.org/syn/S24.html is the closest I see.
16:46 daxim that suffices, thanks
16:47 [Coke] (probably out of date compared to rakudo's Test.pm - https://github.com/rakudo/rakudo/blob/nom/lib/Test.pm
16:48 mucker perl6: sub ∑ { return @_ }; ∑(1,2,3);
16:48 camelia niecza v24-51-g009f999: OUTPUT«[31m===[0mSORRY![31m===[0m��Malformed block at /tmp/bvyxHxRmmz line 1:�------> [32msub [33m�[31m∑ { return @_ }; ∑(1,2,3);[0m��Parse failed��»
16:48 camelia ..rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Missing blockâ�¤at /tmp/N3WMiDs_Lh:1â�¤------> [32msub [33mâ��[31m∑ { return @_ }; ∑(1,2,3);[0mâ�¤    expecting any of:â�¤        statement listâ�¤        prefix or termâ�¤        prefix or meta-prefixâ�¤        new name to be definedâ�¤Â»â€¦
16:49 donaldh jnthn: did you test the runner on Windows?
16:52 * [Coke] wonders if anyone would like a hangout in place of a #phaser occasionally.
16:52 donaldh as in G+ ?
16:53 jnthn donaldh: Not yet, got a build running of the JVM stuff now
16:53 donaldh :)
16:53 jnthn Was just unbusting the Parrot build first.
16:53 jnthn But feeling hungry, so may go and hunt down some curry and test it after :)
16:53 donaldh good call
16:54 [Coke] donaldh: they just revamped hangouts, so "yes, sort of"
16:54 donaldh jnthn: what would be most useful to work on next?
16:55 lichtkind joined #perl6
16:56 jnthn donaldh: In 08b9fff I stubbed some nqp:: ops that are not yet implemented in Ops.java. Same for 029e3ad5.
16:56 jnthn (sha1s both in nqp repo)
16:57 jnthn Alternatively, could work on getting the file/line info incorporated into the backtrace printed by the ExceptionHandling class.
16:57 jnthn It's probably a little fun as it needs correlating frames from the Java stack trace with the caller chain of CallFrame.
16:58 donaldh yikes
16:58 jnthn ;)
16:59 jnthn Well, we have the MethodHandle reachable from each CallFrame, which should help.
16:59 jnthn The other thing that would be a startup time improvement would be to look at the code-gen in P6opaque and see when backing classes could be shared.
16:59 jnthn Though not sure how often that will be...
16:59 jnthn The alternative is generating them on first instantiation I guess.
17:00 donaldh Have you thought about generating more readable class and method names?
17:00 donaldh On the java side.
17:01 jnthn We could but (a) each compilation unit is a single Java class, so it's hard to know what to call it and we need to make sure they end up unique, and (b) the method names would need mangling 'cus there's various chars you may not have in a Java method name.
17:01 jnthn Not to mention the uniqueness thing.
17:01 jnthn Is there no annotation for "name to show in the debugger", ooc?
17:01 [Coke] pony6code
17:01 donaldh Sure, was just thinking about incorporating aspects of the compunit filename and method names to aid reading java stack traces.
17:02 donaldh I'll investigate.
17:02 jnthn Yeah. There's a reason ExceptionHandling tries to do something better than the Java stack trace :)
17:02 jnthn But I guess for folks wanting to use tools, that ain't such a good story.
17:03 jnthn It'd be nice if the Indy frames and similar could be hidden also. I dunno what the options are there.
17:03 donaldh I suspect it requires a debugger extension to munge the stacktrace view. Also require an Eclipse extension for source file breakpoints.
17:03 jnthn Mmmm...an Eclipse Perl 6 extension... :)
17:04 donaldh Might be something that can be added to E.P.I.C.
17:05 jnthn That'd be an E.P.I.C. win...
17:05 donaldh Anyhoo, time to go and 'relax' at a rock climbing wall.
17:05 jnthn Have fun
17:05 donaldh Might have some hacking calories later.
17:05 * jnthn too
17:05 donaldh bye o/
17:05 jnthn o/
17:05 jnthn dinner &
17:08 dalek v5: e966f41 | (Tobias Leich)++ | Makefile:
17:08 dalek v5: add the source files as a dep for their target
17:08 dalek v5: review: https://github.com/rakudo-p5/v5/commit/e966f412b8
17:08 dalek v5: a8aff27 | (Tobias Leich)++ | lib/Perl5/ (3 files):
17:08 dalek v5: treat $] like a normal variable, so we can export it from Perl5::Terms
17:08 dalek v5: review: https://github.com/rakudo-p5/v5/commit/a8aff2756e
17:08 dalek v5: fc0825f | (Tobias Leich)++ | lib/Perl5/ (2 files):
17:08 dalek v5: make qr// and m// work
17:08 dalek v5:
17:08 dalek v5: The modifiers are not handled yet.
17:08 dalek v5: review: https://github.com/rakudo-p5/v5/commit/fc0825f401
17:08 dalek v5: 6545f9d | (Tobias Leich)++ | lib/Perl5/ (4 files):
17:16 daxim I have a module under development.  it sits in lib, tests in t.  I run prove -e perl6 for TDD.  it does not find the module in lib.  what do I have to do?  prove -l does not help, obviously
17:18 daxim man perl6 doesn't say anything about include dirs
17:20 arnsholt PERL6LIB=lib prove
17:20 arnsholt Or use ufo and make test =)
17:21 daxim tell me more about ufo
17:24 daxim I found its readme.  do NOT tell me more about ufo
17:25 arnsholt It's a weird README
17:25 daxim I'm going to be the poopyhead now and say that playfulness in the documentation is over the top for my taste
17:25 arnsholt But it autocreates the Makefile for you
17:27 daxim pfffff    Failed to get the directory contents of 'Subdir': readdir failed: No such file or directory
17:30 dmol joined #perl6
17:35 daxim rakudo:  method Int::is-perfect { say 'fnord' }; 6.is-perfect;
17:35 camelia rakudo 570d1d: OUTPUT«Useless declaration of a has-scoped method in mainline␤No such method 'is-perfect' for invocant of type 'Int'␤  in block  at /tmp/ZJka2Du4Ou:1␤␤»
17:36 daxim how do I add a method to an existing class?
17:37 moritz r: use MONKEY_TYPING; augment class Int { method foolish() { 42 } }; say Int.foolish
17:37 camelia rakudo 570d1d: OUTPUT«42␤»
17:37 kaleem joined #perl6
17:40 SamuraiJack joined #perl6
17:40 daxim rakudo:  use MONKEY_TYPING; augment class Set { method have-some-more(Str $foo) { say $foo } };  set <fnord quux>.have-some-more('stuff');
17:40 camelia rakudo 570d1d: OUTPUT«===SORRY!===␤Class 'Set' already has a Parrot v-table override for 'invoke'␤»
17:40 daxim this error message lacks a remedy
17:44 moritz and it looks weird
17:44 moritz you didn't try to override invoke
17:47 SamuraiJack_ joined #perl6
17:49 pmichaud good morning, #perl6
17:51 pmichaud also, it seems to me that .have-some-more('stuff') would end up working on the list ("fnord", "quux") and not the Set.
17:51 pmichaud rakudo: <fnord quux>.have-some-more('stuff');
17:51 camelia rakudo 570d1d: OUTPUT«No such method 'have-some-more' for invocant of type 'Parcel'␤  in block  at /tmp/eF8EbJ8OQH:1␤␤»
17:51 pmichaud rakudo: set <fnord quux>.have-some-more('stuff');
17:51 camelia rakudo 570d1d: OUTPUT«No such method 'have-some-more' for invocant of type 'Parcel'␤  in block  at /tmp/dZiKI59Efs:1␤␤»
17:52 pmichaud ...but I guess it never even got that far.
17:57 Chillance joined #perl6
17:59 berekuk joined #perl6
18:01 dalek perl6-roast-data: a6c04d7 | coke++ | / (4 files):
18:01 dalek perl6-roast-data: today (automated commit)
18:01 dalek perl6-roast-data: review: https://github.com/coke/perl6-roast-data/commit/a6c04d70ee
18:04 [Coke] niecza has been dirty for 303 days. pugs has been clean for 3 days. rakudo has been dirty for 12 days.
18:05 rjbs What do dirty and clean mean, here?  Passing believed-to-pass tests?
18:05 [Coke] running against roast.
18:05 [Coke] so rakudo has been failing roast spec tests for 12 days in a row. (but, as noted by FROGGS, et al., only on feather, it seems)
18:06 [Coke] I would be happy to use better words than dirty/clean if someone can suggest them. ;)
18:07 btyler joined #perl6
18:07 pmichaud "dirty" sounds about right to me.
18:07 pmichaud it's definitely not clean.  :)
18:07 pmichaud are the feather failures logged somewhere?
18:09 [Coke] https://raw.github.com/coke/perl6-roast-data/master/rakudo_summary.out is the daily log. there's an RT for this particular failure.
18:09 [Coke] https://rt.perl.org/rt3/Ticket/Display.html?id=117957
18:10 [Coke] that first URL looks wrong. https://raw.github.com/coke/perl6-roast-data/master/rakudo_summary.out
18:10 [Coke] second is the ticket for that failure with more details.
18:11 [Coke] ah. and that failure has masked other failures. rakudo is now failing 37 tests, up from 36 yesterday, up from 7, up from the 3 I knew about.
18:12 [Coke] S32-hash/delete-adverb.rakudo S29-os/system.t S09-hashes/objecthash.rakudo
18:13 on joined #perl6
18:14 pmichaud those are useful links... I'll want to review them when I have p6-tuits again.
18:14 gtodd p5eval:  perl -E 'foreach (keys %ENV) { say  if /PERL*/ } '
18:14 p5eval gtodd: ERROR: syntax error at (eval 7) line 1, near "E 'foreach (keys %ENV) { say  if /PERL*/ } '"
18:14 [Coke] pmichaud: also just added this one:
18:14 pmichaud loos like lots of lhf in the summary output
18:14 pmichaud *looks
18:14 [Coke] https://github.com/coke/rakudo-star-daily
18:15 pmichaud \o/  Coke++
18:15 [Coke] (more helpfully: https://github.com/coke/rakudo-star-daily/blob/master/log/test-modules.log_
18:15 [Coke] (more helpfully: https://github.com/coke/rakudo-star-daily/blob/master/log/test-modules.log)
18:15 pmichaud "Rakudo Star Daily" sounds like a news service.  :)
18:15 [Coke] once I have a few days worth of data, I'll see about whipping up a summary file.
18:15 [Coke] heh!
18:16 pmichaud I guess I should wander towards my gate so as not to miss my flight.
18:16 pmichaud bbl
18:16 gtodd arrgh sorry bout hat
18:17 lizmat [Coke]: S32-hash/delete-adverb.rakudo would be mine, probably
18:18 Ulti http://mmb.pcb.ub.es/~carlesfe/funnypapers/#PerlComparison
18:20 [Coke] lizmat: is the test failing for you?
18:20 gdey joined #perl6
18:20 lizmat no, it's not
18:21 [Coke] let me do a run of just that test using the last built version.
18:21 lizmat apart from the warnings, is there something failing there?
18:22 [Coke] http://feather.perl6.nl/~coke/lizmat.out
18:22 [Coke] Yes.
18:22 [Coke] You sure you pushed everything?
18:23 lizmat some of my stuff is still in a pull request  :-(
18:23 [Coke] my copy of roast is shown here: https://github.com/coke/perl6-roast-data/blob/master/perl6_pass_rates#L2  f039dde
18:24 [Coke] lizmat: you should be able to commit directly to roast, no?
18:24 lizmat I guess I forgot to mark some of these tests as todo
18:24 lizmat yes
18:24 [Coke] I thought rakudo proper was the only thing that we needed CLAs for.
18:26 lizmat afaik, that's true, I already have a commit bit for e.g. nqp
18:27 moritz lizmat: you even committed to roast before :-)
18:27 moritz roast eq t/spec/
18:30 dalek roast: e3e2560 | (Elizabeth Mattijsen)++ | S32-hash/delete-adverb.t:
18:30 dalek roast: Fudging in delete-adverb tests needed until my pull request is merged
18:30 dalek roast:
18:30 dalek roast: Spotted by [Coke]++
18:30 dalek roast: review: https://github.com/perl6/roast/commit/e3e2560f94
18:30 lizmat moritz: yes, and quite a few already  :-)
18:30 SmokeMachine joined #perl6
18:30 lizmat that's what's caused some of the smoke ;-(
18:31 lizmat [Coke]: I *think* it is fine now
18:31 lizmat grrrs…  hold it
18:32 [Coke] ooooh, I thought you meant a pull request to ROAST was holding things up.
18:32 [Coke] still six failures.
18:34 dalek roast: f83718b | (Elizabeth Mattijsen)++ | S32-hash/delete-adverb.t:
18:34 dalek roast: Further fudging of delete-adverb tests
18:34 dalek roast: review: https://github.com/perl6/roast/commit/f83718bc92
18:35 [Coke] lizmat++
18:36 lizmat [Coke]: no, it is a pull request in nom:  I hope pmichaud finds the time soon to have a look at it
18:36 [Coke] S09-hashes/objecthash.t - dies after test 25
18:36 [Coke] might need a skip for new tests?
18:37 lizmat I don't think I touched that
18:37 lizmat fwiw, that doesn't fail for me
18:37 [Coke] checking.
18:38 [Coke] 5c7b4be9 (Elizabeth Mattijsen 2013-05-14 16:40:50 +0200 75)
18:38 lizmat hmmm.. I changed the plan?
18:38 lizmat good thing there is git blame
18:38 lizmat I'll fudge them
18:41 lizmat is the number in the fudge the number of lines, or the number of tests?
18:41 jnthn tests
18:42 lizmat no, it's not, it's lines
18:42 jnthn Oh?
18:42 jnthn I've always treated it as tests :P
18:42 lizmat uncommented lines, actually
18:43 lizmat the test dies on "my %h{Mu};"
18:43 jnthn r: my %h{Mu}
18:43 camelia rakudo 570d1d:  ( no output )
18:43 lizmat hmmm….
18:44 jnthn .tell donaldh the runner seems to work fine on Windows
18:44 yoleaux jnthn: I'll pass your message to donaldh.
18:46 lizmat r: my %h{Mu}; %h{Mu}=2
18:46 camelia rakudo 570d1d: OUTPUT«Cannot call 'postcircumfix:<{ }>'; none of these signatures match:␤:(: Mu *%_)␤:(: :p(:$p)!, Mu *%_)␤:(: :k(:$k)!, Mu *%_)␤:(: :kv(:$kv)!, Mu *%_)␤:(: :v(:$v)!, Mu *%_)␤:(: :BIND(:$BIND)!, Mu *%_)␤:(\SELF: $key, Mu *%_)␤:(\SELF: $key, Mu \$BIND, Mu *%_)␤:(\SELF: $k…
18:46 lizmat r: my %h{Mu}; say %h{*}
18:46 camelia rakudo 570d1d: OUTPUT«␤»
18:47 lizmat r: my %h{Mu}; %h<a>=1; say %h{*}
18:47 camelia rakudo 570d1d: OUTPUT«1␤»
18:49 dalek roast: e53a16e | (Elizabeth Mattijsen)++ | S09-hashes/objecthash.t:
18:49 dalek roast: Still need to fudge %h{Mu} related tests
18:49 dalek roast:
18:49 dalek roast: Also pending my pull request to nom.  [Coke]++ for spotting
18:49 dalek roast: review: https://github.com/perl6/roast/commit/e53a16e882
18:49 lizmat [Coke]: good chance it is ok now
18:53 [Coke] lizmat++
18:54 lizmat sorry for the smoke.   Please blame this on getting the workflow going
18:55 [Coke] no worries.
18:55 [Coke] more concerned about the one that's been failing for 2 weeks. :|
18:59 lizmat fwiw, S02-types/bool.t succeeds with me
19:00 lizmat also S29-os/system.t
19:09 Guest12382 joined #perl6
19:12 lizmat r: my %h{Mu}; say %h.WHAT   # how am I supposed to find out that "Mu" is what the hash was typed with?
19:12 camelia rakudo 570d1d: OUTPUT«(Hash+{TypedHash})␤»
19:14 jnthn Good question. For the value you use the .of method on the hash.
19:14 jnthn er, value type that is.
19:15 lizmat hmm… that's in Cool or higher, as I can't find it in ^methods
19:15 lizmat unless I use my eyes of course, duh  :-(
19:16 lizmat thanks, jnthn++
19:16 lizmat r: my %h of Mu; say %h.of   # makes it much clearer
19:16 camelia rakudo 570d1d: OUTPUT«(Mu)␤»
19:17 zby_home_ joined #perl6
19:19 lizmat r: my %h1 of Mu; my %h2{%h1.of}  # how to create a hash with the same type as another hash
19:19 camelia rakudo 570d1d: OUTPUT«===SORRY!===␤Invalid hash shape; type expected␤»
19:19 PerlJam lizmat++ (asking the question I've been wondering about but never got around to asking)
19:19 lizmat r: my %h1 of Mu; my %h2 of %h1.of
19:19 camelia rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Confusedâ�¤at /tmp/7u7AETCHfg:1â�¤------> [32mmy %h1 of Mu; my %h2 of [33mâ��[31m%h1.of[0mâ�¤    expecting any of:â�¤        scoped declaratorâ�¤Â»
19:20 lizmat std: my %h1 of Mu; my %h2 of %h1.of
19:20 camelia std 9906f18: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Invalid type name at /tmp/o9paKfWPNN line 1:â�¤------> [32mmy %h1 of Mu; my %h2 of [33mâ��[31m%h1.of[0mâ�¤    expecting typenameâ�¤Parse failedâ�¤FAILED 00:00 42mâ�¤Â»
19:20 dalek rakudo/jvm-support: 67feb5a | jonathan++ | src/ (2 files):
19:20 dalek rakudo/jvm-support: Fix 'say 42' (and probably loads more).
19:20 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/67feb5a042
19:22 * masak waves tiredly in greeting
19:22 lizmat masak!
19:22 tadzik hey masak!
19:23 labster masak o/
19:23 * lizmat congratulates jnthn++ and masak++ for bringing home the bacon for yet another day!
19:23 jnthn o⁄ masak
19:23 labster mmm bacon
19:24 masak hey, thanks.
19:24 masak yeah, today felt good.
19:24 masak teaching JavaScript is weirdly satisfactory.
19:25 FROGGS I want bacon, naow
19:25 FROGGS and eggs
19:25 FROGGS :/
19:25 jnthn FROGGS: So grab the frying pan? :)
19:25 bloonix joined #perl6
19:25 on joined #perl6
19:25 FROGGS I dont have bacon ó.ò
19:25 nwc10 oooh, numbers
19:26 * FROGGS takes a walk to the kitchen
19:26 lizmat you're *that* skinny ?
19:26 masak if I backlog #perl6, will I find discussion about the schwern thing?
19:26 lizmat no
19:26 jnthn The Schwern thing?
19:26 PerlJam masak: I hope now.  Mostly that's been on #yapc
19:26 PerlJam s/now/not
19:26 lizmat this was the first mention, afaik
19:27 flussence for some reason that name rings a bell, but not from here
19:27 lizmat and I hope this will be the last: schwern thing
19:28 flussence but it sounds like something I don't want to be reminded of anyway
19:28 masak the schwern thing: http://blog.schwern.net/2013/05/15/yapcna-2013-withdrawal/
19:28 lizmat *sigh*
19:28 gtodd perl -E 'foreach (keys %ENV) { say if /PERL/ }'
19:29 gtodd sigh
19:29 flussence oh, I must've been thinking of a different name.
19:29 tadzik oh ffs
19:29 masak tadzik: that's about the size of it.
19:29 gtodd EMOUSECUTSEL
19:30 tadzik well, I'm still going to go there and be nice to everyone, CoC or not
19:30 sorear let's not have a spillover
19:31 skids What's wrong with Corrosion of Conformity :-)
19:31 lizmat tadzik: that's an autopun?
19:31 jac50 joined #perl6
19:32 masak I'm sorry. spillover was the last thing I wanted.
19:32 tadzik lizmat: if it is, I didn't notice :)
19:32 masak I just think that matters of community maintenance are automatically on-topic here.
19:32 * skids wonders if TLA saturation follows moore's law.
19:32 masak (since it's one of the reasons Perl 6 was created.)
19:33 * lizmat is repressing the urge to comment
19:34 domidumont joined #perl6
19:34 lizmat r: my %h of Mu; %h.of ~~ m/\w+/; say ~$/  # instead point out some more NYI's / bugs / nags
19:34 camelia rakudo 570d1d: OUTPUT«No such method 'match' for invocant of type 'Mu'␤  in block  at /tmp/_RaQdyyO__:1␤␤»
19:36 lizmat r: my %h of Mu; say ~%h.of
19:36 camelia rakudo 570d1d: OUTPUT«Nominal type check failed for parameter ''; expected Any but got Mu instead␤  in sub prefix:<~> at src/gen/CORE.setting:1271␤  in block  at /tmp/rR0sj51D6R:1␤␤»
19:37 * lizmat is trying to figure out a way to create a hash with the same type as another *reliably*
19:38 lizmat r: my %h{Int}; say %h.of
19:38 camelia rakudo 570d1d: OUTPUT«(Mu)␤»
19:38 * lizmat giving up for now
19:39 flussence just to be clear: are you looking for type constraints on the keys or the values?
19:39 masak lizmat: .clone ?
19:39 moritz lizmat: %h{Int} types the keys, .of gives the type constraint of the values
19:39 flussence (because "of" doesn't seem to do anything on my end...)
19:40 moritz TypedHash doesn't seem to expose the types
19:40 lizmat moritz: aha!
19:41 lizmat masak: .clone also copies the contents
19:41 * lizmat has a look at the clone code
19:42 masak ah, troo.
19:43 masak I definitely agree that getting a new container with the same .of seems a reasonable thing to do.
19:43 masak even before I have a concrete use case.
19:46 nwc10 jnthn: for Rakudo on the JVM, in the ./perl6 for *nix, you need "" round $@, ie exec java ... perl6 "$@"
19:47 nwc10 and I have to admit that I prefer the startup speed of NQP
19:47 jnthn .new should do it. If it doesn't, that'd be a bug.
19:47 nwc10 I couldn't work out where that ./perl6 comes from
19:47 jnthn nwc10: The startup speed of both of them on the JVM sucks at the moment.
19:48 nwc10 parrot has to be an order of magniture better for somethign :-)
19:50 masak lizmat: what jnthn++ said. I shoulda thought of that.
19:50 dalek rakudo/jvm-support: c52bedd | jonathan++ | tools/build/create-jvm-runner.pl:
19:50 dalek rakudo/jvm-support: Fix runner for *nix; nwc10++.
19:50 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/c52bedd7cf
19:51 lizmat r: my %h{Mu}; say %h.WHAT; my %h2=%h.new; say %h2.WHAT;
19:51 camelia rakudo 570d1d: OUTPUT«(Hash+{TypedHash})␤(Hash)␤»
19:51 lizmat the .new loses the Type information  :-(
19:52 lizmat seems it is using the one from Hash: method new(*@args) { @args.hash }
19:52 jnthn lizmat: Yeah, so I notice...
19:53 fgomez joined #perl6
19:53 lizmat simply putting a method new in the TypedHash would not solve the pb, I think, as it probably lies deeper, no?
19:53 jnthn Probably we need to look at the way Hash.new works.
19:54 lizmat so what *is* the way to get at the Type?  ".of" is about constraining the values, as moritz pointed out
19:55 jnthn lizmat: there isn't one implemented. S09 may have one spec'd, but I don't recall it.
19:56 nwc10 Bother:
19:56 nwc10 nick@raspberrypi:~/test$ ./perl6 -e 'say "Hello portable world"'
19:56 nwc10 Unhandled exception: java.lang.BootstrapMethodError: call site initialization exception in <anon> in <anon>
19:56 nwc10 so much for Java bytecode being portable
19:57 jnthn Oddnes.
19:58 nwc10 yes.
19:58 jnthn What does java -version show?
19:58 nwc10 Although IIRC that was the exception thrown when I tried to build nqp-jvm on the raspberry pi
19:58 nwc10 java version "1.7.0_03"
19:58 nwc10 OpenJDK Runtime Environment (IcedTea7 2.1.7) (7u3-2.1.7-1)
19:58 nwc10 OpenJDK Zero VM (build 22.0-b10, mixed mode)
19:59 nwc10 it was built by this Java:
19:59 nwc10 java version "1.7.0_19"
19:59 nwc10 OpenJDK Runtime Environment (rhel-2.3.9.1.el6_4-x86_64)
19:59 nwc10 OpenJDK 64-Bit Server VM (build 23.7-b01, mixed mode)
19:59 jnthn I actually managed to get the nqp-jvm build to segfault one of the earlier ones in the 7 serious.
19:59 jnthn *series
19:59 nwc10 pwnage!
19:59 jnthn Segfaulting the JVM felt like an, uh, achievement...
19:59 jnthn But it was already fixed in a later version.
20:00 nwc10 Sadly the properly sarcastic headline isn't in the URL: http://www.theregister.co.uk/2013/05/15/oracle_new_versioning_scheme/ -- "Oracle updates Java versioning to allow more security fixes"
20:01 jnthn heh :)
20:05 breinbaas joined #perl6
20:16 dorlamm joined #perl6
20:21 estrabd joined #perl6
20:38 spider-mario_ joined #perl6
20:39 betterwo1ld joined #perl6
20:40 lizmat r: https://gist.github.com/lizmat/5594924
20:40 camelia rakudo 570d1d: OUTPUT«("Foo::Bar" => {"perl6" => {}}).hash␤Nominal type check failed for parameter 'lhs'; expected Any but got Mu instead␤  in sub METAOP_TEST_ASSIGN:<//> at src/gen/CORE.setting:13798␤  in sub module2stash at /tmp/_BTgCQ8SZ_:15␤  in block  at /tmp/_BTgCQ8SZ_:22␤␤»…
20:40 lizmat I'm stumped, haven't been able to simplify this down further  :-(
20:40 masak '= Any' better written '?'
20:41 lizmat that changes things
20:42 masak ah.
20:42 lizmat no it doesn't  :-(   must put quotes around ?
20:42 Timbus|Away joined #perl6
20:42 lizmat ah, ok, gotcha now
20:42 lizmat no changes
20:44 dalek v5: 5980153 | (Tobias Leich)++ | lib/Perl5/Grammar.nqp:
20:44 dalek v5: support ${foo} variable name syntax
20:44 dalek v5: review: https://github.com/rakudo-p5/v5/commit/59801533b8
20:44 dalek v5: 5f94a6c | (Tobias Leich)++ | / (4 files):
20:44 dalek v5: move undef to Perl5::Terms
20:44 dalek v5: review: https://github.com/rakudo-p5/v5/commit/5f94a6cf4a
20:44 lizmat the lines with return Proxy:new can be removed, it still creates the same error
20:46 masak is the sub necessary?
20:47 lizmat masak: do you see anything else wrong with the code ?
20:47 lizmat the sub?
20:47 masak I mean, does the failing code have to be in a sub?
20:47 masak or can you just set the parameters as variables?
20:47 lizmat for testing, yes...
20:47 lizmat ultimately it has to become a sub
20:47 FROGGS r: say "abc" ~~ m?a?g
20:48 camelia rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Regex not terminatedâ�¤at /tmp/9hP6aNAAnu:1â�¤------> [32msay "abc" ~~ m?a?g[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infixâ�¤        prefix or termâ�¤        prefix or meta-prefix…
20:48 masak lizmat: for the purposes of golfing.
20:50 lizmat r: https://gist.github.com/lizmat/5594924   # golfed
20:50 camelia rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Two terms in a rowâ�¤at /tmp/W2J319ZHpe:1â�¤------> [32mhttps[33mâ��[31m://gist.github.com/lizmat/5594924   # go[0mâ�¤    expecting any of:â�¤        argument listâ�¤        postfixâ�¤        infix stopperâ�¤        infix or meta-infixâ�¤        st…
20:50 yakshavr joined #perl6
20:51 lizmat r: https://gist.github.com/lizmat/5594924
20:51 camelia rakudo 570d1d: OUTPUT«("Foo::Bar" => {"perl6" => {}}).hash␤Nominal type check failed for parameter 'lhs'; expected Any but got Mu instead␤  in sub METAOP_TEST_ASSIGN:<//> at src/gen/CORE.setting:13798␤  in block  at /tmp/FVbfGg38LX:11␤␤»
20:51 lizmat grrsss…  when using a URL, you can't comment
20:52 lizmat so: no sub
20:54 lizmat masak: you agree that this *should* work?
20:58 [Coke] joined #perl6
20:59 clkao joined #perl6
21:00 masak lizmat: yes, definitely.
21:00 lizmat I'll try to golf it further down
21:01 masak r: my $auths = do { my %h{Any} of Hash;  %h }; my $versions = $auths{Any}; say "alive"
21:01 camelia rakudo 570d1d: OUTPUT«alive␤»
21:01 masak hm.
21:04 lizmat down to 5 lines, it is related to %h{Any} being stored as a value in a hash
21:07 masak r: (my %h{Any} of Hash){Any} //= my %
21:07 camelia rakudo 570d1d: OUTPUT«Nominal type check failed for parameter 'lhs'; expected Any but got Mu instead␤  in sub METAOP_TEST_ASSIGN:<//> at src/gen/CORE.setting:13798␤  in block  at /tmp/QD7khTjRjR:1␤␤»
21:07 masak this is as golf'd as I can get it.
21:08 lizmat r: my $froms= do { my %h{Any} of Hash;  %h }; $froms<a> //= do { my %h{Any} of Hash;  %h };  # my golf
21:08 camelia rakudo 570d1d: OUTPUT«Nominal type check failed for parameter 'lhs'; expected Any but got Mu instead␤  in sub METAOP_TEST_ASSIGN:<//> at src/gen/CORE.setting:13798␤  in block  at /tmp/773hQw7_T5:1␤␤»
21:08 masak I think that makes it fairly clear that something is Wrong, at least.
21:08 * jnthn is a little surprised to discover that "use v6;" actually depends on *a gather block*!
21:08 jnthn o.O
21:09 lizmat why would this involve a META op ?
21:09 jnthn The foo= form is a meta-op
21:09 timotimo will we be getting a jakudo evalbot thingie for the channel?
21:10 jnthn timotimo: Probably, when it's worth having :)
21:10 lizmat ah, of course, yes
21:11 dalek nqp/rak-jvm-support: 0440860 | jonathan++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
21:11 dalek nqp/rak-jvm-support: Implement nqp::getlexcaller.
21:11 dalek nqp/rak-jvm-support: review: https://github.com/perl6/nqp/commit/04408607fe
21:11 dalek nqp/rak-jvm-support: 1baee5d | jonathan++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java:
21:11 dalek nqp/rak-jvm-support: Some more decont, tweaks.
21:11 dalek nqp/rak-jvm-support: review: https://github.com/perl6/nqp/commit/1baee5dc61
21:12 lizmat seems to me the code generated isn't getting the left hand side passed to it if the destination is a non-existent hash value on the left
21:13 lizmat r: my $froms= do { my %h{Any} of Hash;  %h }; $froms<a>=Any; $froms<a> //= do { my %h{Any} of Hash;  %h }; say $froms;  # works
21:13 camelia rakudo 570d1d: OUTPUT«("a" => {}).hash␤»
21:14 lizmat adding a $h<key>=Any apparently creates a container that //= can find
21:16 lizmat r: my %h{Any}; %h<a> //= "a"   #golfed shortest
21:16 camelia rakudo 570d1d: OUTPUT«Nominal type check failed for parameter 'lhs'; expected Any but got Mu instead␤  in sub METAOP_TEST_ASSIGN:<//> at src/gen/CORE.setting:13798␤  in block  at /tmp/Q3e_ebVOMA:1␤␤»
21:16 lizmat r: my %h{Any}; %h<a>=Any; %h<a> //= "a"   #golfed shortest
21:16 camelia rakudo 570d1d:  ( no output )
21:16 dalek rakudo/jvm-support: b47f7a3 | jonathan++ | src/vm/jvm/runtime/org/perl6/rakudo/Ops.java:
21:16 dalek rakudo/jvm-support: Implement nqp::p6recont_ro.
21:16 dalek rakudo/jvm-support: review: https://github.com/rakudo/rakudo/commit/b47f7a3c5f
21:16 lizmat masak: rakudobug?
21:17 * masak submits
21:18 lizmat masak++
21:18 lizmat BTW, same goes for ||= and &&=
21:19 lizmat also note that there is no problem if the hash is not typed
21:19 jnthn Possible patch (please somebody else try it) for the lizmat++ bug: https://gist.github.com/jnthn/5595193
21:20 [Coke] jnthn: is there anyone aside from you who can wrangle GC bugs that pass through sixmodel's PMC's destroy() vtable?
21:20 lizmat jnthn: I'll try that
21:21 jnthn [Coke]: Somewhere in the world, probably... :)
21:21 adu joined #perl6
21:21 jnthn [Coke]: is it a double free happening inside there?
21:21 adu hi jnthn
21:21 jnthn o/ adu
21:22 adu how goes?
21:22 [Coke] jnthn: aye.
21:22 jnthn Tiredly...
21:22 [Coke] er... at one point it presented as a double free, anyway.
21:22 jnthn [Coke]: A while back I tried and failed to hunt one of those. I got it down to being related to mixins and bigints interacting interestingly.
21:22 adu o right, Europe time
21:23 [Coke] current error was: *** glibc detected *** ./perl6: free(): invalid next size (fast):
21:23 jnthn [Coke]: As in, those seemed to be the ingredients. I never managed to get down to the exact issue.
21:23 [Coke] https://rt.perl.org/rt3/Ticket/Display.html?id=117957 is the ticket if, you know, you have spare time. :)
21:23 jnthn ;)
21:24 [Coke] I don't think there are any bigints near this one.
21:26 [Coke] ~~
21:32 swarley left #perl6
21:33 kurahaupo joined #perl6
21:35 gdey joined #perl6
21:37 sdo joined #perl6
21:37 sdo hello
21:37 lizmat hi sdo!
21:38 kurahaupo sdo: hi
21:38 sdo lizmat: hi kurahaupo: I am reading at this time UsingPerl6
21:38 masak hi, sdo.
21:38 sdo hi masak:
21:40 lizmat jnthn: alas, no change
21:40 sdo and I was wondering is Perl6 can be considered as a program that can allow constraint programing as can do ada the NG?
21:40 sorear .oO( solar observation spacecraft studying perl 6 )
21:40 sdo NG: Next Generation
21:40 jnthn lizmat: Hmm...OK, then it's not that I thought it was. That patch feels needed even if not sufficient, mind...
21:42 lizmat ok
21:42 sdo the book UsingPerl6 is good but hard to assimilate
21:42 adu sdo: Ada95?
21:42 sdo no ADA 2012
21:42 adu I'm not familiar with that
21:43 sdo as far as I know Ada 95 is full object
21:44 adu sdo: what do you mean constraint programing?
21:44 lizmat jnthn: did you catch me saying that this happens if the key is not in the hash yet?  If you do %h<key>=Any first, it doesn't happen
21:44 sdo sub circle-radius-from-area(Real $area where { $area >= 0 }) {($area / pi).sqrt} say circle-radius-from-area(3); # OK
21:45 lizmat fwiw: now that I know what the problem is, I can work around it
21:45 jnthn lizmat: Yes, which led me to suspect busted auto-viv things, which I thought the patch I suggested may fix. But it seems not.
21:45 sdo I mean to certify that an object respect constraint
21:45 lizmat jnthn: but doesn't at_key return Mu if the key doesn't exist?
21:46 jnthn lizmat: It should return a con...oh...wait a moment...
21:46 FROGGS sdo: I believe you can do that with perl6
21:46 sdo that's perl 6
21:46 FROGGS jnthn: dary?
21:46 adu sdo:  there is something called "as"
21:46 lizmat r: my %h{Mu}; say %h.at_key("a")   # typed hash returns Mu, not Any
21:46 camelia rakudo 570d1d: OUTPUT«(Mu)␤»
21:46 FROGGS what is a condary?
21:47 jnthn lizmat: no, I though it may be missing "is rw" on those at_pos methods...but it ain't...
21:48 adu sdo: wait, apparently there is a "where" also
21:48 jnthn FROGGS: What?
21:48 sdo I don't know about as because I am at page 42
21:48 FROGGS jnthn: just a joke (legen... dary)
21:48 FROGGS r: sub a ( Int $a where $a > 5 ) { say $a }; a( 42 )
21:48 lizmat r: my %h; say %h.at_key("a")   # untyped hash returns Any
21:48 camelia rakudo 570d1d: OUTPUT«42␤»
21:48 camelia rakudo 570d1d: OUTPUT«(Any)␤»
21:48 FROGGS r: sub a ( Int $a where $a > 5 ) { say $a }; a( 2 )
21:48 camelia rakudo 570d1d: OUTPUT«Constraint type check failed for parameter '$a'␤  in sub a at /tmp/rhXNBbSHrO:1␤  in block  at /tmp/rhXNBbSHrO:1␤␤»
21:48 jnthn r: my %h{Mu}; say %h.at_key("a")
21:48 camelia rakudo 570d1d: OUTPUT«(Mu)␤»
21:48 jnthn r: my %h; say %h.at_key("a")
21:48 camelia rakudo 570d1d: OUTPUT«(Any)␤»
21:49 FROGGS sdo: see? works
21:49 jnthn lizmat: ah...sorry, I'm really slow tonight :)
21:49 lizmat I'm very glad that you're here at all!  :-)   jnthn++
21:49 jnthn r: my %h{Mu} of Any; say %h.at_key("a")
21:49 camelia rakudo 570d1d: OUTPUT«(Mu)␤»
21:49 jnthn r: my Any %h{Mu}; say %h.at_key("a")
21:49 camelia rakudo 570d1d: OUTPUT«(Any)␤»
21:49 jnthn ahya
21:49 jnthn hmmmm.
21:50 jnthn lizmat: Take two: https://gist.github.com/jnthn/5595395
21:51 masak 'night, #perl6
21:51 adu omg, this syntax just blew up my mind: $obj = { pair 1, 2, 3, 4, 5, 6 }
21:51 lizmat night masak!
21:51 FROGGS I think I've got a fix for: "abc" ~~ m?abc?
21:51 lizmat jnthn: testing
21:52 lizmat r: my $obj = { pair 1, 2, 3, 4, 5, 6 }; say $obj
21:52 camelia rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Undeclared routine:â�¤    pair used at line 1. Did you mean '&pairs'?â�¤â�¤Â»
21:53 dalek perl6-bench: 00d48f6 | (Geoffrey Broadwell)++ | analyze:
21:53 dalek perl6-bench: Untabify analyze
21:53 dalek perl6-bench: review: https://github.com/japhb/perl6-bench/commit/00d48f6701
21:53 dalek perl6-bench: 3d439fb | (Geoffrey Broadwell)++ | analyze:
21:53 dalek perl6-bench: Allow individual Perls to be clicked on and off using legend in plots
21:53 dalek perl6-bench: review: https://github.com/japhb/perl6-bench/commit/3d439fb14c
21:53 dalek perl6-bench: 77647e8 | (Geoffrey Broadwell)++ | README:
21:53 dalek perl6-bench: Update dates in README a bit
21:53 dalek perl6-bench: review: https://github.com/japhb/perl6-bench/commit/77647e8aa8
21:53 dalek perl6-bench: 323bde2 | (Geoffrey Broadwell)++ | README:
21:53 dalek perl6-bench: Small tweaks to README to make it work better as Markdown using: `pandoc -s --toc -f markdown -o README.html README`
21:53 dalek perl6-bench: review: https://github.com/japhb/perl6-bench/commit/323bde2697
21:59 labster If I have a statement like `my Cool $foo = 3;`, how do I get the container type from $foo (i.e. Cool instead of Int)?
22:00 FROGGS r: my Cool $foo = 3; say $foo.^WHAT
22:00 camelia rakudo 570d1d: OUTPUT«[31m===[0mSORRY![31m===[0mâ�¤Cannot use .^ on a non-identifier method callâ�¤at /tmp/j0QRyDKbzc:1â�¤------> [32mmy Cool $foo = 3; say $foo.^WHAT[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        method argumentsâ�¤Â»
22:00 FROGGS r: my Cool $foo = 3; say $foo.WHAT
22:00 camelia rakudo 570d1d: OUTPUT«(Int)␤»
22:00 jnthn r: my Cool $foo = 3; say $foo.VAR.of
22:00 camelia rakudo 570d1d: OUTPUT«No such method 'of' for invocant of type 'Scalar'␤  in block  at /tmp/MObM9NlnUX:1␤␤»
22:00 jnthn r: my Cool $foo = 3; say $foo.VAR.type
22:00 camelia rakudo 570d1d: OUTPUT«No such method 'type' for invocant of type 'Scalar'␤  in block  at /tmp/BEGDxZB9oO:1␤␤»
22:00 jnthn hmm :)
22:00 FROGGS r: my Cool $foo = 3; say $foo.VAR.^methods
22:00 camelia rakudo 570d1d: OUTPUT«Int Num Rat FatRat abs Bridge chr sqrt base expmod is-prime floor round ceiling sign conj rand sin asin cos acos tan atan atan2 sec asec cosec acosec cotan acotan sinh asinh cosh acosh tanh atanh sech asech cosech acosech cotanh acotanh unpolar cis Complex log exp …
22:00 FROGGS r: my Cool $foo = 3; say $foo.VAR.^attr
22:00 camelia rakudo 570d1d: OUTPUT«No such method 'attr' for invocant of type 'Perl6::Metamodel::ClassHOW'␤  in block  at /tmp/DYucNTlMtc:1␤␤»
22:00 jnthn r: my Cool $foo = 3; say $foo.VAR.name
22:00 camelia rakudo 570d1d: OUTPUT«$foo␤»
22:01 jnthn well, that's something... :)
22:01 FROGGS hehe
22:01 chayin_ joined #perl6
22:01 FROGGS r: my Cool $foo = 3; say $foo.VAR.^attributes
22:01 camelia rakudo 570d1d: OUTPUT«No such method 'gist' for invocant of type 'BOOTSTRAPATTR'␤  in method gist at src/gen/CORE.setting:5259␤  in method gist at src/gen/CORE.setting:893␤  in sub say at src/gen/CORE.setting:10920␤  in block  at /tmp/4GXaMOXpbC:1␤␤»
22:01 FROGGS sad
22:02 lizmat Success, jnthn++
22:02 FROGGS r: my Cool $foo = 3; print $foo.VAR.^attributes
22:02 camelia rakudo 570d1d: OUTPUT«Nominal type check failed for parameter '$got'; expected Any but got BOOTSTRAPATTR instead␤  in method REIFY at src/gen/CORE.setting:6571␤  in method reify at src/gen/CORE.setting:5650␤  in method gimme at src/gen/CORE.setting:6058␤  in method print at src/gen/CORE…
22:02 labster This is related to fixing RT#93980 , assigning Nil to that would set it to Cool
22:02 lizmat jnthn: will run a spectest now to see if it doesn't break anything
22:02 lizmat will let you know tomorrow, ok?
22:03 jnthn lizmat: ok
22:03 jnthn FROGGS, labster: https://gist.github.com/jnthn/5595470
22:03 jnthn Sleep time for me... 'night
22:03 lizmat night jnthn, sleep tight!
22:03 jnthn hope so :)
22:03 jnthn o/
22:03 labster night, thanks
22:05 * lizmat also calling it a night, cu all tomorrow!
22:08 donaldh joined #perl6
22:31 konundra joined #perl6
22:34 v5DsX joined #perl6
22:38 adu omg Ada2012 has a video
22:38 adu we should make a video for Perl6
22:39 adu just have a bunch of people send me interviews with themselves, then I could splice it together and add soundtrack
22:41 timotimo ... interviews? why not ... space stations and guitars?
22:41 FROGGS I sold my guitar ages ago :/
22:41 timotimo a friend of mine recently introduced me to the erlang movie and its "sequel"
22:43 flussence if you want to impress lots of people with a programming language, give them something made in it that runs at 60fps.
22:46 timotimo i don't quite see that happen with perl6 soon :|
22:46 flussence probably not, but it's something to aim for
22:46 FROGGS not at 60fps, no
22:46 flussence (took a while before Javascript could do that, too)
22:47 timotimo and now there's asm.js and we have unreal engine stuff running in browsers, that's pretty crazy
22:48 avuserow joined #perl6
22:49 avuserow good *, #perl6 :)
22:49 adu timotimo: have you seen this? http://bellard.org/jslinux/
22:49 labster hello avuserow
22:51 avuserow hey labster, how's it going?
22:52 labster pretty good: jnthn++ gave me some code I needed to solve a problem, and I'm going to YAPC:NA.  How about you?
22:54 avuserow not much here, wrapping up $dayjob soon. I've been enjoying watching jnthn++'s progress on the JVM port too.
22:54 avuserow and I am planning on going to YAPC::NA. Unfortunately I have to leave wednesday evening but I'll be around starting on Saturday, assuming I get my hotel plans figured out
22:56 UTrGm joined #perl6
22:57 adu flussence: you mean a first person shooter? or frames per second?
22:57 skids joined #perl6
22:58 adu flussence: also, it depends on the setup, because povray is a very nice language, but it usually runs at around 0.00000000001 fps
22:59 FROGGS frames per seconds I think
23:00 adu FROGGS: I think first person shooter would also fit the theme of the statement
23:06 adu is moritz here?
23:06 FROGGS adu: already sleeping I suppose
23:07 jerome__ joined #perl6
23:25 labster what does 'is pure' mean in rakudo?
23:27 shinobicl left #perl6
23:32 FROGGS I believe "having no side effects"
23:39 DNQcj joined #perl6
23:41 xilo joined #perl6
23:45 gdey joined #perl6

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

Perl 6 | Reference Documentation | Rakudo