Perl 6 - the future is here, just unevenly distributed

IRC log for #perl6, 2014-12-11

Perl 6 | Reference Documentation | Rakudo

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

All times shown according to UTC.

Time Nick Message
00:16 |Tux| joined #perl6
00:20 dayangkun joined #perl6
00:20 robinsmidsrod joined #perl6
00:20 jnthn OK, advent post scheduled at a minute past midnight. :)
00:21 jnthn And I'm gonna sleep :)
00:21 jnthn &
00:21 psch night jnthn
00:38 * timotimo suggests someone puts a "Day 11 - " in front of the title
00:41 atroxaper joined #perl6
00:41 psch uh
00:42 psch scheduled for a minute past midnight..?
00:43 psch timotimo: i can do that, but i'm confused regarding when the post should be published at the moment...
00:44 psch because it doesn't look like it is yet, or is that just something on my end?
00:44 timotimo depends on the timezone jnthn has interpreted "midnight" in
00:45 psch hm, it says 12/11/2014 @ 00:19 in quick edit
00:47 psch but no TZ info
00:47 psch visible to me at least, but then i hardly know wordpress
00:48 timotimo ah
00:48 timotimo December 11, 2014, 12:48 am
00:48 timotimo this is what it shows in the stats page
00:48 timotimo that must be WP's local timezone
00:49 psch ...now there's 2 autosaves, edited by me, with just the "Day 11 - " added to the title o.o
00:49 timotimo it could very well be that wordpress interprets 12/11/2014 @ 00:19 to be "19 minutes past midnight on the 12th"
00:49 timotimo because, as we all know, 00:19 is 12:19am, which is part of the previous day
00:50 psch i don't follow
00:50 psch 12/11/2014 is american date format
00:50 lue joined #perl6
00:50 timotimo oh!
00:50 timotimo derp
00:50 timotimo how do americans cope?!
00:51 geekosaur by being bewildered at sane date formats
00:51 geekosaur >.>
00:51 psch heh
00:51 psch it's still possible to schedule the post for 12/11/2014 @ 00:01
00:51 timotimo or you could just post it immediately; WP thinks the new day has already begun
00:52 psch yeah, i guess that's safe
00:53 psch does it need some magic to appear on the landing page?
00:53 geekosaur the new day started 53 minutes ago, of course...
00:54 psch 'cause it's not there but is in the category 2014...
00:56 timotimo i see it
00:57 psch i can see it in the logged-in browser, but the other only sees it in the category
00:57 psch err, NOW
00:57 psch some caching
00:57 psch probably
00:58 timotimo seems likely
01:01 psch i also updated the ToC
01:01 psch not sure if the link from day 10 to day 11 happens automatically eventually or if that has to be added manually
01:01 psch 'cause that's not there either on my end atm
01:03 timotimo aye, only one of them shows up here; from 11 to 10
01:05 BenGoldberg joined #perl6
01:08 psch well, i don't know what to do about that link
01:08 psch i also have to go sleep &
01:09 timotimo gnite psch!
01:42 adu joined #perl6
01:46 kurahaupo joined #perl6
01:50 dalek specs: 23ef17e | TimToady++ | S0 (2 files):
01:50 dalek specs: Add List to 'OKness' types
01:50 dalek specs: review: https://github.com/perl6/specs/commit/23ef17e57a
01:56 * TimToady changes List.ACCEPTS to return self, to see what breaks
02:06 dalek specs: b40b20c | TimToady++ | S05-regex.pod:
02:06 dalek specs: another fake Match becomes List
02:06 dalek specs: review: https://github.com/perl6/specs/commit/b40b20c683
02:13 rmgk joined #perl6
02:13 TimToady Someone needs to entitle a talk: The following step can take a long time, please be patient.
02:25 japhb .oO( *** Implementing Perl 6 ... )
02:33 TimToady looks like it fails about 6 test files, but passes new TODOs in 5
02:33 TimToady so we almost broken even :)
02:34 TimToady *broke
02:34 * colomon thinks broken might be more appropriate here...
02:36 TimToady now I just need to figure out what word to name the old whatever-matching code
02:39 TimToady I could just move the ACCEPTS over to Parcel, but that would break as soon as we we get rid of Parcel
02:42 jimmy2 joined #perl6
02:45 TimToady what the hey, maybe it can be a transitional form while we decide what to do with parcel list matchers
02:45 TimToady whoops, I said 'make install' instead of 'make m-install'...
02:45 * TimToady goofs off, since "he's compiling"
02:51 japhb m: role NewDateish { multi method truncated-to(|) { nextsame } }; class NewDate is Date does NewDateish { }; my $nd = NewDate.today; say $nd.truncated-to('week');
02:51 camelia rakudo-moar 5f6494: OUTPUT«Nil␤»
02:51 japhb OK, what stupid mistake am I making there?  ^^
02:53 aborazmeh_ joined #perl6
02:56 lue TimToady: is there any written thing out there about this list refactoring you seem to be doing, or is it all kinda being improvised right now?
02:59 TimToady we talked about it some on irc :)
02:59 lue not surprised :)
03:00 chenryn joined #perl6
03:01 lue I have to say though, it's about time. I've always felt like I had to fiddle around a bit too much when it came to using lists in P6. Here's the last thing I ran into trouble with, from a while ago:
03:01 lue m: say 1,2 X+ 8,9; say [X+] (1,2;8,9);
03:01 camelia rakudo-moar 5f6494: OUTPUT«9 10 10 11␤2 2␤»
03:03 japhb TimToady: Can you see a mistake in what I'm doing above (the NewDate class)?
03:05 baest joined #perl6
03:06 araujo joined #perl6
03:07 TimToady m: say 1,2 X+ 8,9; say [X+](1,2;8,9);
03:07 camelia rakudo-moar 5f6494: OUTPUT«9 10 10 11␤9 10 10 11␤»
03:07 TimToady lue: you passed a single argument to [X+]
03:09 lue funny, I thought you weren't supposed to leave out the space there. Though I think I see why (spaceless makes it like &sub(1,2;8,9) , right?)
03:10 araujo joined #perl6
03:11 lue Doesn't seem to work when the LoL is in a variable, though.
03:12 noganex joined #perl6
03:15 jimmy_ joined #perl6
03:15 TimToady japhb: leave off the 'multi'
03:16 TimToady lue:  or were you asking about the GLR?  pmichaud++ has a blog on that...
03:17 lue GLR? I was just mentioning that [X+] @listoflists wasn't DWIMming; at the time I had to resort to @a[0] X+ @a[1] X+ @a[2] ...
03:17 TimToady okay
03:17 araujo joined #perl6
03:28 japhb TimToady: I actually had two multis in the role, so I could add a new truncation boundary ('quarter', in this case).  Am I forced to do it with an if or when in an only method in the role?
03:29 TimToady alternately, try a 'lastcall; nextsame'
03:31 atroxaper joined #perl6
03:32 TimToady the basic problem being that nextsame goes to the wrong dispatcher
03:34 TimToady so you have to terminate the multi dispatch to get to the next method dispatch
03:49 aborazmeh joined #perl6
03:49 aborazmeh joined #perl6
03:50 japhb I was afraid the workaround would be much uglier.  I'll give lastcall a try.
03:51 jack_rabbit joined #perl6
03:54 japhb Dang, no dice.  Is lastcall not implemented?
03:54 TimToady could be
03:54 JimmyZ_ m: lastcall
03:54 camelia rakudo-moar 5f6494: OUTPUT«lastcall is not in the dynamic scope of a dispatcher␤  in block  at src/gen/m-CORE.setting:744␤  in block <unit> at /tmp/HXHpxpZu3o:1␤␤»
03:55 TimToady hmm
03:55 JimmyZ_ m: nextsame
03:55 camelia rakudo-moar 5f6494: OUTPUT«nextsame is not in the dynamic scope of a dispatcher␤  in block  at src/gen/m-CORE.setting:734␤  in block <unit> at /tmp/6MaViaT6J3:1␤␤»
03:55 TimToady odd
03:56 * TimToady doesn't see how a role would interfere...
04:00 japhb m: class NewDate is Date { multi method truncated-to(|) { lastcall; nextsame } }; my $nd = NewDate.today; say $nd.truncated-to('week');
04:00 camelia rakudo-moar 5f6494: OUTPUT«Nil␤»
04:00 japhb Role isn't the problem; it's not getting to the right dispatcher it looks like
04:12 anaeem1_ joined #perl6
04:15 anaeem1 joined #perl6
04:17 pyrimidi_ joined #perl6
04:32 KCL_ joined #perl6
04:54 tinyblak joined #perl6
05:00 dalek rakudo/nom: 080b40f | TimToady++ | src/core/ (3 files):
05:00 dalek rakudo/nom: move List.ACCEPTS Parcel and Array for now
05:00 dalek rakudo/nom:
05:00 dalek rakudo/nom: List.ACCEPTS now indicates a success status.
05:00 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/080b40feab
05:00 tony-o m: 'something.min.js'.match(/ '.' .*? $ /).Str.say;
05:00 camelia rakudo-moar 5f6494: OUTPUT«.min.js␤»
05:02 kaare_ joined #perl6
05:03 KCL joined #perl6
05:08 tony-o how do i do [^\.] in a regex for perl6 ?
05:09 tony-o i'm not using the right words and all i'm finding is <!before
05:11 chenryn joined #perl6
05:14 flussence <-[.]> ?
05:17 tony-o ah.. missed the < >, thanks flussence
05:20 * TimToady wonders if -. would be a useful sugar
05:21 TimToady or something else a bit shorter than 6 chars
05:24 TimToady m: say "foo" ~~ m:g/./
05:24 camelia rakudo-moar 080b40: OUTPUT«「f」␤ 「o」␤ 「o」␤␤»
05:24 TimToady \o/
05:27 tony-o m: say "foo." ~~ m/-'.'/;
05:27 camelia rakudo-moar 080b40: OUTPUT«[31m===[0mSORRY![31m===[0m�Unrecognized regex metacharacter - (must be quoted to match literally)�at /tmp/m4R7psleri:1�------> [32msay "foo." ~~ m/[33m�[31m-'.'/;[0m�Unrecognized regex metacharacter ' (must be quoted to match literally)…»
05:27 tony-o m: say "foo." ~~ m/<-'.'>/;
05:27 camelia rakudo-moar 080b40: OUTPUT«[31m===[0mSORRY![31m===[0m�Unrecognized regex metacharacter < (must be quoted to match literally)�at /tmp/_pv56ufYIX:1�------> [32msay "foo." ~~ m/<-[33m�[31m'.'>/;[0m�Unrecognized regex metacharacter - (must be quoted to match literally)…»
05:27 tony-o m: say "foo." ~~ m/<-['.']>/;
05:27 camelia rakudo-moar 080b40: OUTPUT«Potential difficulties:â�¤    Repeated character (') unexpectedly found in character classâ�¤    at /tmp/CjtEtL88Cs:1â�¤    ------> [32msay "foo." ~~ m/<-[33mâ��[31m['.']>/;[0mâ�¤ï½¢fï½£â�¤â�¤Â»
05:31 rurban joined #perl6
05:39 ptc_p6 joined #perl6
05:42 adu joined #perl6
05:44 JimmyZ_ joined #perl6
05:52 telex joined #perl6
06:08 nige joined #perl6
06:10 tinyblak joined #perl6
06:24 kaare_ joined #perl6
06:32 ptc_p6 joined #perl6
06:36 tinyblak_ joined #perl6
06:48 TimToady also need to set $/ to a List if the match or subst returns a list, but some tests seem to think $/ is always a Match, surprise, surprise...
06:52 lue I would almost suggest that a list of matches should rather be in @/ :) .
06:54 atroxaper joined #perl6
06:56 kaleem joined #perl6
06:58 Isp-sec joined #perl6
07:02 TimToady @() is sufficient for that
07:04 raydiak amusing typo of the day: "rakudobrew build most"
07:09 * TimToady feels as though that should be 'build moast'
07:12 raydiak well if it felt right it wouldn't be a typo :)  happened b/c my left hand was offset one key...amusing that the accident was more correct english than the intended string, yet still so related
07:14 TimToady so 'build niar' would be the opposite mistake...
07:15 bjz joined #perl6
07:17 raydiak for one of several interpretations of "opposite", absolutely
07:18 Woodi hallo today :)
07:19 Woodi raydiak: in Polish "most" is "bridge" so "build bridge" is quite correct :)
07:21 raydiak heh funny...hi Woodi
07:22 denis_boyun joined #perl6
07:23 darutoko joined #perl6
07:40 rurban_ joined #perl6
07:43 FROGGS joined #perl6
07:45 * mathw wonders what happens if you rakudobrew build moat
07:48 raydiak fixed it before I hit enter, but iirc it tells you something like "building most is NYI; well volunteered!"
07:49 gtodd joined #perl6
07:50 TimToady most appropriate
07:51 vti joined #perl6
07:51 raydiak I volunteer to dig the moat (given transportation, room, and board)
07:56 blackbolt joined #perl6
07:56 TimToady .oO(you'll never get rich, by digging a ditch...)
07:57 atroxaper joined #perl6
07:58 raydiak .oO(it is easier for a camel to go through the eye of a needle, than for a rich man to enter the kingdom of God)
08:03 FROGGS o/
08:03 raydiak in which case, I ought to slip right through those gates like a carbon nanotube :)\
08:03 raydiak \o
08:03 rindolf joined #perl6
08:04 mathw I always thought that quote just encouraged rich people to be really nasty, since they're doomed anyway they might as well be arrogant whatsits
08:06 raydiak He goes on to say "With people this is impossible, but with God all things are possible."
08:10 zakharyas joined #perl6
08:17 TimToady well, most people will take any excuse to be nasty, really :)
08:18 TimToady speaking of which, the current implementation of s/// is
08:21 TimToady it returns the current value of $_ rather than the success of the substitution
08:23 mathw Hmm. Seems unhelpful.
08:29 nige joined #perl6
08:31 * raydiak finds himself contracted to build a 1-page website for a pole-dancing fitness gym
08:33 * Woodi is dreaming about web pages making and "normal" programming be a plain, artisan job for few centuries from now ;)
08:36 Woodi btw. internet is in very infancy state at the moment... today I wanted to know prices for wood paint in our small town shops and it's imposible to find this on internet ! ;)
08:36 Woodi even futuristic to wish such simple information :)
08:38 raydiak all you had to do was call around and put the prices in comments on their google pages and you could have fixed the internet for us! :)
08:38 pecastro joined #perl6
08:39 andreoss left #perl6
08:41 Woodi raydiak: no! becouse price can change any second and information will be out of data ! all that shops should have some actualization stream in place to update prices and stock !
08:41 Woodi err... do bakerise should do that also ?
08:41 nwc10 I'm failing to find this in the documentation - what's the Perl 6 way of writing cat?
08:41 nwc10 ie perl -e 'print while <>'
08:43 TimToady m: slurp.print
08:43 camelia rakudo-moar 080b40: OUTPUT«Céad slán ag sléibhte maorga Chontae Dhún na nGall␤Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤I mo dhiaidh bhí gleanntáin ghlas’ G…»
08:43 TimToady m: .say for lines
08:43 camelia rakudo-moar 080b40: OUTPUT«Céad slán ag sléibhte maorga Chontae Dhún na nGall␤Agus dhá chéad slán ag an Eireagal ard ina stua os cionn caor is coll;␤Nuair a ghluais mise thart le Loch Dhún Lúich’ go ciúin sa ghleann ina luí␤I mo dhiaidh bhí gleanntáin ghlas’ G…»
08:43 nwc10 thanks
08:48 dalek rakudo/nom: 6b429b9 | TimToady++ | src/core/Str.pm:
08:48 dalek rakudo/nom: set $/ on listy match too
08:48 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/6b429b95d9
08:49 [Sno] joined #perl6
08:50 raydiak alright enough of my silliness for now, this was one horrific day and I'm thankful it's over...g'night #perl6
08:51 FROGGS gnight raydiak
08:52 [Tux] joined #perl6
08:56 blackbolt can someone help me with substitution. for (@test){ s/\'//:g;}. it remove only first \'
08:57 brrt joined #perl6
08:58 TimToady us s:g/\'// instead
08:58 TimToady *use
08:59 blackbolt ohhh thanks TimToady
09:00 TimToady and you don't need the parens around @test, just spaces
09:01 ptc_p6 joined #perl6
09:02 mathw that would require shops to have decent IT systems
09:02 mathw (delayed, sorry, forgot I'd scrolled up)
09:04 mathw The lack of parens there is still one of my favourite syntactic things in Perl 6. It lightens up the code so much
09:04 TimToady scrolling down requires decent IT systems...
09:04 mathw it would help if I had drunk more than a few sips of this delicious hot chocolate already
09:04 mathw then I might be marginally more awake
09:06 ptc_p6 joined #perl6
09:09 dalek roast: bd41b86 | TimToady++ | S05-modifier/ (2 files):
09:09 dalek roast: m:g// can now work under smartmatch, returns List
09:09 dalek roast: review: https://github.com/perl6/roast/commit/bd41b863d8
09:09 dalek rakudo/nom: 1eb7fad | TimToady++ | src/core/List.pm:
09:09 dalek rakudo/nom: fake up .from/.to on List for lists of matches
09:09 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/1eb7fada84
09:10 * TimToady leaves the s/// work for tomorrow, assuming he doesn't blow away in this storm...
09:10 mathw My advice is to avoid going kite flying
09:10 mathw Also stay indoors if you're water-soluble
09:12 * TimToady hopes staying in bed is adequate to stay dry for the next 8 hours or so
09:12 TimToady zzz &
09:15 mathw Depends how sturdy your house is I suppose. Sleep well!
09:23 tinyblak joined #perl6
09:32 fhelmberger joined #perl6
09:32 andreoss joined #perl6
09:32 fhelmberger joined #perl6
09:32 woolfy left #perl6
09:33 dalek roast: 5a1239a | (Elizabeth Mattijsen)++ | S05-modifier/perl5_0.t:
09:33 dalek roast: Unfudge passing TODO, TimToady++
09:33 dalek roast: review: https://github.com/perl6/roast/commit/5a1239a325
09:34 lizmat afk&
09:39 kurahaupo joined #perl6
09:39 robinsmidsrod joined #perl6
09:43 Rounin joined #perl6
09:43 woolfy joined #perl6
09:49 rurban1 joined #perl6
09:53 itz_ joined #perl6
10:06 mvuets joined #perl6
10:09 mvuets hello, #perl6
10:09 aborazmeh joined #perl6
10:13 mathw hello
10:20 aborazmeh joined #perl6
10:22 andreoss I got "Unhandled exception: const_iX NYI" with huge amount of iterations, but with fewer iterations code runs just fine
10:23 andreoss what is not yet implemented in this case?
10:24 tinyblak joined #perl6
10:31 masak antenoon, #perl6
10:32 masak m: sub f { my &b = { return }; return &b }; my &c = f(); &c()
10:32 camelia rakudo-moar 1eb7fa: OUTPUT«Attempt to return outside of any Routine␤  in block  at src/gen/m-CORE.setting:629␤  in block <unit> at /tmp/fyt7jTUWyS:1␤␤»
10:33 masak m: say X::ControlFLow::Return.new.message
10:33 camelia rakudo-moar 1eb7fa: OUTPUT«Could not find symbol '&Return'␤  in method <anon> at src/gen/m-CORE.setting:13810␤  in any find_method_fallback at src/gen/m-Metamodel.nqp:2725␤  in any find_method at src/gen/m-Metamodel.nqp:988␤  in block <unit> at /tmp/JfDtDJj6ni:1␤␤»
10:33 masak m: say X::ControlFlow::Return.new.message
10:33 camelia rakudo-moar 1eb7fa: OUTPUT«Attempt to return outside of any Routine␤»
10:37 mvuets lizmat: re http://paste.scsys.co.uk/451709, it is tricky to reproduce, i filed a ticket with details: #123406
10:37 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=123406
10:37 mvuets lizmat: i think hoelzro made a fix last night
10:38 mvuets yup, hoelzro++ https://github.com/MoarVM/MoarVM/pull/162
10:46 Alula joined #perl6
10:53 rindolf joined #perl6
10:54 tinyblak joined #perl6
10:58 masak today's autopun: "Those who believe in telekinesis, raise my hand."
11:02 brrt left #perl6
11:03 tadzik :D
11:09 Alula joined #perl6
11:12 psch joined #perl6
11:13 psch hi #perl6
11:15 robinsmidsrod joined #perl6
11:20 Grrrrrrr joined #perl6
11:23 chenryn joined #perl6
11:26 bjz joined #perl6
11:32 anaeem1_ joined #perl6
11:55 tinyblak joined #perl6
12:00 dalek nqp: ca209b6 | Carlin++ | 3rdparty/libtommath/ (2 files):
12:00 dalek nqp: use arc4random on platforms that support it
12:00 dalek nqp:
12:00 dalek nqp: this merges in patches from MoarVM's libtommath
12:00 dalek nqp: review: https://github.com/perl6/nqp/commit/ca209b6f32
12:00 dalek nqp: 241c631 | (Tobias Leich)++ | 3rdparty/libtommath/ (2 files):
12:00 dalek nqp: Merge pull request #206 from carbin/arc4random
12:00 dalek nqp:
12:00 dalek nqp: use arc4random on platforms that support it
12:00 dalek nqp: review: https://github.com/perl6/nqp/commit/241c63160f
12:00 Grrrrrrr left #perl6
12:02 Alula joined #perl6
12:09 brrt joined #perl6
12:12 Util joined #perl6
12:20 anaeem1_ joined #perl6
12:33 myp joined #perl6
12:38 andreoss map and grep don't work within promises?
12:39 andreoss my $p = start { grep { is-prime $_ },  1..100  };
12:39 FROGGS m: my $p = start { grep { is-prime $_ },  1..100  }; say $p
12:39 camelia rakudo-moar 1eb7fa: OUTPUT«Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Kept)␤»
12:40 andreoss say await $p; # empty
12:42 rurban1 joined #perl6
12:44 FROGGS m: my $p = start { grep { is-prime $_ },  1..100  }; say await $p
12:44 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:44 FROGGS m: my $p = start { grep { is-prime $_ },  1..100  }; say (await $p).perl
12:44 camelia rakudo-moar 1eb7fa: OUTPUT«().list␤»
12:44 brrt is camelia allowed to start threads?
12:44 FROGGS it at least looks like
12:44 psch m: await start { say "i think so" }
12:44 camelia rakudo-moar 1eb7fa: OUTPUT«i think so␤»
12:45 FROGGS m: my $p = start { 1..100  }; say await $p
12:45 camelia rakudo-moar 1eb7fa: OUTPUT«1..100␤»
12:45 FROGGS m: my $p = start { (^100).map: 'ö'  }; say await $p
12:45 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:45 blackbolt joined #perl6
12:45 psch m: await start { say grep .is-prime, [1, 2, 3, 4] }
12:45 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:45 FROGGS m: say await start { (^100).map: 'ö' }
12:45 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:46 FROGGS j: say await start { (^100).map: 'ö' }
12:46 camelia rakudo-jvm 1eb7fa: OUTPUT«Can't open perl script "/home/camelia/jvm-rakudo/eval-client.pl": No such file or directory␤»
12:46 FROGGS bah
12:46 FROGGS m: say await start { [~] 1, 2, 3 }
12:46 camelia rakudo-moar 1eb7fa: OUTPUT«123␤»
12:47 FROGGS m: say await start { (^3).join: '' }
12:47 camelia rakudo-moar 1eb7fa: OUTPUT«012␤»
12:47 FROGGS okay, it really looks like map and grep are fishy here
12:48 andreoss also
12:49 psch m: say await start { my $x; $x ~= $_ for 1..10 }
12:49 camelia rakudo-moar 1eb7fa: OUTPUT«Nil␤»
12:49 psch for goes through MapIter as well i think?
12:49 psch hm, am i actually returning $x there..?
12:49 psch m: say await start { my $x; $x ~= $_ for 1..10; $x }
12:49 camelia rakudo-moar 1eb7fa: OUTPUT«12345678910␤»
12:49 psch right, nevermind :)
12:50 andreoss m: my $p = start { gather for 1..1000 { take $_ if $_.is-prime } };say await $p;
12:50 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:50 FROGGS m: say await start { do for 1..10 { $_ } }
12:50 camelia rakudo-moar 1eb7fa: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
12:51 andreoss gather/take actually gives some results, but not the whole list of primes below 1000
12:51 psch m: say await start { do for 1..1000 { .is-prime ?? $_ !! Nil } }
12:51 camelia rakudo-moar 1eb7fa: OUTPUT«2 3 5 7 11 13 17 19 23 29 31 37 41 43 47 53 59 61 67 71 73 79 83 89 97 101 103 107 109 113 127 131 137 139 149 151 157 163 167 173 179 181 191 193 197 199 211 223 227 229 233 239 241 251 257 263 269 271 277 281 283 293 307 311 313 317 331 337 347 349 353 3…»
12:52 * psch is distracting himself from figuring out list/array boxing from java to rakudo
12:52 colomon m: say grep .is-prime, [1, 2, 3, 4]
12:52 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:52 colomon m: say grep .is-prime, 1, 2, 3, 4
12:52 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:53 FROGGS m: say await start { (^100).map: 'ö'; 1 } # lol
12:53 camelia rakudo-moar 1eb7fa: OUTPUT«===SORRY!===␤No such method 'count' for invocant of type 'Str'␤»
12:53 colomon m: say (1, 2, 3, 4).grep(*.is-prime)
12:53 camelia rakudo-moar 1eb7fa: OUTPUT«2 3␤»
12:55 FROGGS m: say await start { (^100).map: 'ö'; 1 } # <--- that happens in MapIter.reify at 'my $argc   := $!block.count;'
12:55 camelia rakudo-moar 1eb7fa: OUTPUT«===SORRY!===␤No such method 'count' for invocant of type 'Str'␤»
12:55 FROGGS ohh, that's actually correct
12:56 FROGGS m: say await start { (^100).map: { 'ö' }; 1 }
12:56 andreoss does "do" mean something special for promises?
12:56 camelia rakudo-moar 1eb7fa: OUTPUT«1␤»
12:56 tinyblak joined #perl6
12:56 FROGGS andreoss: no
12:56 FROGGS 'do' puts a loop in statement context, which means it will produce results
12:57 dalek modules.perl6.org: 427af07 | moritz++ | web/lib/P6Project.pm:
12:57 dalek modules.perl6.org: Try to make UTF-8 handling more robust
12:57 dalek modules.perl6.org: review: https://github.com/perl6/modul​es.perl6.org/commit/427af075eb
12:57 FROGGS so you can use 'do' to indicate that a while/until/for loop should either produce values or throw them away
12:58 FROGGS m: my $x = 64; say do while $x++ < 85 { .chr }
12:58 camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
12:58 FROGGS huh
12:59 moritz it returns the last evaluated condition
12:59 moritz m: say do while 0 { 42 }
12:59 camelia rakudo-moar 1eb7fa: OUTPUT«0␤»
12:59 FROGGS m: my $x = 64; say do while $x++ < 85 { $x.chr }
12:59 camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
12:59 FROGGS weird
13:00 FROGGS that's not what I expect
13:00 FROGGS m: my $x = 64; say do while $x++ < 85 { $x.chr; last }
13:00 camelia rakudo-moar 1eb7fa: OUTPUT«-> (|) { #`(Block|45850752) ... }␤»
13:00 moritz and it's not up to spec, iirc
13:04 Woodi joined #perl6
13:05 pecastro joined #perl6
13:06 psch so what do i have to do to create hll array in the vm level?  i'm allocating what i get from Ops.hlllist(tc), but that REPR "does not support bind_pos_boxed"
13:08 psch i did successfully build a BOOTArray, but i didn't figure out how to hllize that properly either :/
13:09 hoelzro morning #perl6
13:09 moritz http://modules.p6c.org/ now running
13:12 colomon o/
13:15 dalek mu: d70722e | moritz++ | misc/svn-search/ (3 files):
13:15 dalek mu: remove outdated svn-search
13:15 dalek mu: review: https://github.com/perl6/mu/commit/d70722ef23
13:17 rindolf joined #perl6
13:20 atroxaper joined #perl6
13:27 dalek mu: 9c8f365 | moritz++ | util/update-design.perl6.org.sh:
13:27 dalek mu: script for generating design.perl6.org HTML (untested)
13:27 dalek mu: review: https://github.com/perl6/mu/commit/9c8f365a64
13:28 dalek mu: 461f574 | moritz++ | misc/mothballed/yap6/ (29 files):
13:28 dalek mu: remove stuff that has been mothballed for 5 years
13:28 dalek mu: review: https://github.com/perl6/mu/commit/461f574827
13:32 sergot m:say <a b c d>.splice: 0, 2;
13:32 sergot m: say <a b c d>.splice: 0, 2;
13:32 camelia rakudo-moar 1eb7fa: OUTPUT«No such method 'splice' for invocant of type 'Parcel'␤  in block <unit> at /tmp/TEsxO2hsDZ:1␤␤»
13:32 sergot do we have splice for arrays?
13:32 psch m: say [1, 2, 4].splice: 2, 3
13:32 camelia rakudo-moar 1eb7fa: OUTPUT«4␤»
13:33 sergot psch++
13:33 psch m: say @(<a b c>).splice: 0, 2;
13:33 camelia rakudo-moar 1eb7fa: OUTPUT«a b␤»
13:33 psch sergot: the latter one actually relates, i guess :)
13:34 sergot m: say <a b c d>.list.splice: 0, 2;
13:34 camelia rakudo-moar 1eb7fa: OUTPUT«a b␤»
13:34 tadzik you know what we should do, starting next year? Say something like #adventcalendar on the IRC whenever something historical happens, or a question becomes a FAQ
13:35 tadzik this way, at the end of the year, we'll know what to write about :)
13:37 andreoss how do i declare thread-shared variable?
13:37 andreoss for promises
13:39 andreoss usual my $var in global scope causes segmentation faults
13:39 moritz andreoss: iirc you can use it, you just have to lock-protect it
13:39 moritz andreoss: but it's better to avoid that, and use some reactive pattern instead
13:40 moritz (though I'm not an expert on concurrency)
13:43 kaleem joined #perl6
13:43 kaleem joined #perl6
13:52 colomon tadzik: Typed::Subroutines failed overnight smoke
13:52 tadzik oh noes
13:52 tadzik I'll check it out, thanks
13:53 psch andreoss: http://jnthn.net/papers/2014-nlpw-reactive.pdf and http://jnthn.net/papers/2014-yapceu-async.pdf might be of interest to you
13:56 [Sno]_ joined #perl6
13:56 tinyblak joined #perl6
13:57 TimToady tadzik: if the tests rely on the old List semantics under smartmatch, could be my changes
13:58 nwc10 TimToady: are you still dry?
13:59 tadzik TimToady: the code uses @types ~~ $s.signature.params.map(*.type)
13:59 tadzik so, might be
14:01 tadzik TimToady: what's changed about the list smartwatching?
14:01 tadzik erm, smartmatching
14:21 TimToady nwc10: the heavy rain doesn't start for another 10 minutes or so :)
14:22 TimToady List is now a distinguished result type for smartmatching, just like Bool, Match, and Nil
14:22 TimToady so you need to cast it to Parcel or Array to get the same effect currently
14:23 TimToady but that's probably going away eventually too, in favor of a specific match-this-list keyword of some sort
14:24 TimToady overloading the base types for this is kinda wrongish
14:25 TimToady m: say "foo" ~~ m:g/./
14:25 camelia rakudo-moar 1eb7fa: OUTPUT«「f」␤ 「o」␤ 「o」␤␤»
14:25 TimToady the List changes were to make ^^ work
14:26 psch m: "foo" ~~ m:g/./; $/.perl.say
14:26 camelia rakudo-moar 1eb7fa: OUTPUT«(Match.new(orig => "foo", from => 0, to => 1, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo", from => 1, to => 2, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo", from => 2, to => 3, ast => Any, li…»
14:26 TimToady it's a list of match
14:26 moritz TimToady: I thought the desired solution was to return a list-ish Match object from m:g//
14:26 TimToady the other night we put our heads together and decided it was better to make List a result type than to fake up a Match
14:27 TimToady the boolean semantics of List work out very nicely this way, and you don't have to .list the result to use it
14:27 tadzik colomon: fixed, thanks
14:28 andreoss psch: for me it's buggy, i've tried to make my naive solution to be thread-aware, but it seems a with more than two iterations the program segfaults
14:28 psch m: "abc" ~~ m:g/$<foo=.>/; say $()>><foo>
14:28 camelia rakudo-moar 1eb7fa: OUTPUT«No such method 'ast' for invocant of type 'List'␤  in block <unit> at /tmp/jGNuDOnNWr:1␤␤»
14:28 novice666 joined #perl6
14:28 andreoss http://fpaste.org/158752/18307922/
14:28 psch m: "abc" ~~ m:g/$<foo=.>/; say @$/>><foo>
14:28 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
14:28 psch hm
14:28 psch m: "abc" ~~ m:g/$<foo=.>/; say $<foo> # this seems not-so-great
14:28 camelia rakudo-moar 1eb7fa: OUTPUT«postcircumfix:<{ }> not defined for type List␤  in method gist at src/gen/m-CORE.setting:13806␤  in sub say at src/gen/m-CORE.setting:16560␤  in block <unit> at /tmp/orqBWKkj13:1␤␤»
14:30 novice666 m: say eval("abs(-10)");
14:30 camelia rakudo-moar 1eb7fa: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/nWbC9_1wUVâ�¤Undeclared routine:â�¤    eval used at line 1â�¤â�¤Â»
14:30 TimToady it's EVAL now
14:30 FROGGS m: say EVAL("abs(-10)");
14:30 camelia rakudo-moar 1eb7fa: OUTPUT«10␤»
14:30 novice666 thanks
14:31 psch andreoss: i can't really help with concurrency, i just remembered that these two talks by jnthn++ exist
14:31 smls joined #perl6
14:32 psch TimToady: i like the List change, but i wonder about the consistency of $/ semantics.  having to pay attention if the last match was :g seems unwieldy
14:32 psch not sure what to put in there instead, though, considering smartmatch for subst and match checks $/ anyway
14:32 TimToady m: "abc" ~~ m:g/$<foo=.>/; say $/.map: *.<foo>
14:32 camelia rakudo-moar 1eb7fa: OUTPUT«␤»
14:33 TimToady m: "abc" ~~ m:g/$<foo=.>/; .say for @$/
14:33 camelia rakudo-moar 1eb7fa: ( no output )
14:33 TimToady oh, it's malformed
14:34 TimToady m: "abc" ~~ m:g/<foo=.>/; .say for @$/
14:34 camelia rakudo-moar 1eb7fa: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/gHQf7ZgDTzâ�¤Unable to parse expression in metachar:sym<assert>; couldn't find final '>' â�¤at /tmp/gHQf7ZgDTz:1â�¤------> [32m"abc" ~~ m:g/<foo=.[33mâ��[31m>/; .say for @$/[0mâ�¤    expecting a…»
14:34 TimToady m: "abc" ~~ m:g/$<foo=>./; .say for @$/
14:34 camelia rakudo-moar 1eb7fa: ( no output )
14:34 TimToady m: "abc" ~~ m:g/$<foo>./; .say for @$/
14:34 camelia rakudo-moar 1eb7fa: ( no output )
14:34 smls TimToady: What about incremental matching, like we could do with which m//g and m//gc in Perl 5? Is there no replacement planned for that?
14:34 TimToady m: "abc" ~~ m:g/$<foo>=./; .say for @$/
14:34 camelia rakudo-moar 1eb7fa: OUTPUT«「a」␤ foo => 「a」␤␤「b」␤ foo => 「b」␤␤「c」␤ foo => 「c」␤␤»
14:34 TimToady there we go
14:35 TimToady m:g is lazy
14:35 smls you mean it returns a lazy list?
14:35 TimToady yes
14:35 smls ok
14:35 TimToady supposed to, anyway
14:36 psch do we need Cats for that?
14:36 TimToady no
14:36 smls And will it be possible to teach the GC to free elements of lazy lists that have already been iterated over? :P
14:36 TimToady depends on how they're stored
14:37 TimToady that's one of the things we talked about in Salzburg
14:37 TimToady Cats might be useful for incremental output of s///
14:37 psch ah, right
14:38 TimToady m: "abc" ~~ m:g/$<foo>=./; say $/»<foo>
14:38 camelia rakudo-moar 1eb7fa: OUTPUT«「a」␤ 「b」␤ 「c」␤␤»
14:38 TimToady works if you actually match the pattern
14:40 smls Btw when building moar with rakudobrew, I see a bunch of lines like «gcc -c -fPIC -O1 -DNDEBUG  -D_REENTRANT -D_FILE_OFFSET_BITS=64 ...»
14:40 smls Why "-O1"?
14:41 TimToady m: "abc" ~~ m:g/$<foo>=./; say $1
14:41 camelia rakudo-moar 1eb7fa: OUTPUT«「b」␤ foo => 「b」␤␤»
14:42 TimToady -O2 was getting indigestion on something some number of months ago
14:42 psch hm, i guess that's ok.  having $0, $1 etc reference the different 'instances' of :g Match objects makes sense to me at least
14:43 smls Interesting; I thought -O2 did only "safe" optimizations?
14:44 TimToady some things are not so safe for large interpreters playing with goto
14:45 TimToady which may or may not have been the problem at the time
14:46 brrt TimToady - already awake :-)
14:47 TimToady yes, alas
14:47 tinyblak joined #perl6
14:51 TimToady jnthn,brrt: btw, when trying (and failing) to fix up s/// last night, I was getting some: Spesh: failed to fix up handlers (-1, 556, 556)
14:51 TimToady what does that mean?
14:52 brrt iirc, handlers are basically pointers into the bytecode, and they are made 'symbolic' during the bytecode->spesh graph transformation, and back to pointers in the spesh graph->bytecode transformation
14:53 brrt it looks to me as if some blocks were lost, but to be sure i'd have to see the spesh graph
14:53 * TimToady wonders if something is cheating on s/// in there
14:53 TimToady course it was in the setting, so things can get weird in any case
14:54 TimToady .subst is a bit weird insofar as the proto is not a bare {*}
14:55 TimToady probably has something to do with that...
14:59 TimToady together with futzing around with OUTER::<$/>
14:59 brrt hmmm
14:59 brrt possible, but that would still be an error on moars part imho
14:59 brrt does it persist when MVM_SPESH_DISABLE_INLINE=1
14:59 TimToady didn't try that yet
15:01 TimToady I'll have to reproduce my changes, having backed them out in the interests of finishing up m:g//
15:06 smls TimToady: Is parsing binary blobs part of the long-term vision for Perl 6 grammars? Custom while+unpack+if/else spaghetti code for parsing binary formats, doesn't seem very fun when you've been spoiled with grammars for parsing strings-based formats... :P
15:12 TimToady treating blobs and bufs as strings to the extent it makes sense is part of the intended design
15:12 TimToady the Stringy role is kind of a down payment on that
15:13 TimToady you probably won't get grapheme semantics, or even Unicode codepoints
15:14 smls nor would you probably want it there :)
15:16 psch oh, someone fixed the Day 10 -> Day 11 link
15:16 psch or maybe it does happen automatically?
15:17 FROGGS we should be able to get rid of OUTER::<$/> by passing the $/ around, starting from the actions, at least for s/// and friends
15:18 FROGGS not so for the method forms... but this maybe means that the method forms won't have side effects anymore, as they perhaps shouldn't have anyway
15:18 TimToady I'm guessing we'll end up with a different method anyway
15:19 TimToady subst still sets $/ internally for its replacement, but it's setting a private $/ in the proto
15:19 TimToady if I read it aright
15:22 masak TimToady: https://github.com/perl6/specs/commit/2​3ef17e57a03ff7aaf59dff986d90d3193c847a4 seems to do more than add List to 'OKness' types... was that deliberate?
15:23 TimToady no, but I think it is probably closer to reality wrt PEG rules
15:24 TimToady so we can let it stand
15:28 atroxaper joined #perl6
15:28 telex joined #perl6
15:32 c0nspiracy joined #perl6
15:36 tinyblak joined #perl6
15:45 masak oki.
15:45 andreoss joined #perl6
15:45 masak quick poll: do people prefer `die MyException.new` or `MyException.new.throw` ?
15:46 tadzik die
15:46 tadzik you know what's going on from the first 3 characters
15:47 masak ah, end-weight.
15:47 masak yes, I was thinking the same thing.
15:49 kaare_ joined #perl6
15:50 masak what's the standard way to throw a control exception?
15:50 masak `warn MyException.new` ?
15:53 smls TimToady: How would we express something like   «read($fh, $buf, 2); my ($format) = unpack("n", $buf); if ($format == 42) { ...read...unpack...etc. } else { ... }»   in a grammar though?
15:53 smls There would need to be a way to tell the grammar: Gobble up 2 bytes if, interpeted as an unsigned short in network order, they hold the value 42.
15:53 masak haha -- you know you're doing something funky with your program when it stops and says "Trying to unwind over wrong handler". :)
15:54 [Sno]_ joined #perl6
15:57 masak m: class C is Exception {}; { die C.new; CATCH { default { say .WHAT } } }
15:57 camelia rakudo-moar 1eb7fa: OUTPUT«(C)␤»
15:58 masak m: class C is Exception {}; { warn C.new; CONTROL { default { say .WHAT } } }
15:58 camelia rakudo-moar 1eb7fa: OUTPUT«(X::AdHoc)␤»
15:58 moritz smls: (..) [ <?{ unpack('n', $0) == 42 }> ... | ...]
15:58 masak why do I get the exception itself when I `die` with it, but `X::AdHoc` when I `warn` with it?
15:58 moritz (if the grammar is in byte mode, if that exists)
15:58 masak should I use another base class for control exceptions?
15:58 kaleem joined #perl6
15:59 moritz masak: because nobody really worked on warnings yet
15:59 [Sno]_ joined #perl6
15:59 masak hm. maybe I'd better stay away from control exceptions, then.
16:01 TimToady smls: nothin' sez we can't have subrules like <unpack('n')>
16:01 TimToady they're just methods, after all
16:04 smls Interesting; Where can I read more on subrules with parameters?
16:09 TimToady well, the calling forms are mentioned at S05:1426
16:09 synopsebot Link: http://perlcabal.org/syn/S05.html#line_1426
16:10 TimToady on the recieving end, any regex/token/rule declaration can take an ordinary signature, so it really doesn't need much additional documentation
16:11 TimToady if you write an ordinary method as part of a grammar, you have to know how to deal with cursors correctly, but other than that, it's an ordinary method
16:12 TimToady basically, self in such a method is a cursor, and you're expected to return a new cursor saying how much you matched
16:13 smls Also, if the grammar matched  <format=unpack('n')>  then how would we get at the unpacked integer?  $<format>  would just hold a Buf with the two original bytes, right?
16:13 TimToady would depend on how the method processed it and where it put it
16:14 TimToady the new cursor carries all the match info
16:14 TimToady as well as any ast you're constructing
16:15 TimToady the standard grammar contains many examples
16:16 TimToady though most of the ast making is hidden off in the actions modules instead of the grammar, at least for rakudo
16:19 smls So, you'd let the unpack subrule implicitly call `make` to make the unpacked value available as part of the ast?
16:21 smls Also, does  <unpack('n', :test(42))>  look sensible?
16:22 andreoss why does perl6 run git sometimes?
16:23 smls Or maybe even <ushort(42)>  for maximum sugar...
16:25 tadzik andreoss: the installation process might, Configure.pl to be exact
16:26 tadzik or 'make spectest'. Perl 6 itself does not run gin :)
16:26 tadzik or git
16:27 mr-foobar joined #perl6
16:29 lucas_ joined #perl6
16:29 lucas_ Hello
16:29 lucas_ Having a easter egg like 42 won't make my code less correct-ish if some funny client pass that exact string as input?
16:29 hoelzro what's special about 42, again?
16:30 hoelzro iirc, 42.WHY is special
16:30 hoelzro m: 42.WHY
16:30 camelia rakudo-moar 1eb7fa: ( no output )
16:30 hoelzro m: say 42.WHY
16:30 camelia rakudo-moar 1eb7fa: OUTPUT«(Any)␤»
16:30 hoelzro huh, nvm
16:30 masak m: "Life, the Universe, and Everything".WHY
16:30 camelia rakudo-moar 1eb7fa: ( no output )
16:30 lucas_ its WHAT
16:30 masak m: say "Life, the Universe, and Everything".WHY
16:30 camelia rakudo-moar 1eb7fa: OUTPUT«42␤»
16:30 lucas_ *it's
16:30 psch m: say 42.WHAT
16:30 camelia rakudo-moar 1eb7fa: OUTPUT«(Int)␤»
16:30 andreoss tadzik: i dont think so, i run some script with git repo, and as its seen in tmux title, perl6 run git
16:31 andreoss *running
16:31 TimToady "some script" can do anything it pleases
16:32 atroxaper joined #perl6
16:32 TimToady the perl6 process itself does not run git for anything
16:33 andreoss but it definitely doesn't have system("git") inside
16:33 TimToady m: system()
16:33 camelia rakudo-moar 1eb7fa: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/BU7M_5gCvoâ�¤Undeclared routine:â�¤    system used at line 1â�¤â�¤Â»
16:33 TimToady there is no 'system' in Perl 6
16:33 TimToady there are lots of other ways to run a subprocess
16:34 TimToady including using a module that does so
16:35 TimToady doing stuff with panda, for instance
16:36 smls TimToady: I understood correctly, the "only" missing puzzle piece for binary grammars that needs to be provided by Perl6 itself is the support for matching against Bufs and other non-Str's - and then modules could provide all the sugar (like the unpack subroule), right?
16:37 smls so it'd look like  «use Grammar::Binary; grammar BMP does Grammar::Binary { ... <unpack(...)> ... }»
16:39 hoelzro smls: I had a similar thought
16:39 andreoss my bad, it was bash who ran git, nevermind
16:40 lucas_ Well, if the programmer expects a Str typeobject and gets 42 as the answer will be fun anyway. Maybe I should grep 'WHY' in the module ecosystem...
16:43 lucas_ Oh, nevermind, I'm making the confusion again about WHAT vs. WHY... I've never used WHY...
16:44 psch .WHY is documentation
16:44 lucas_ psch: Like Pod? Oh, thanks!
16:45 TimToady and 42 is cool with pretending to be a string
16:45 hoelzro m: #|(marine) sub yellow {}; say &yellow.WHY
16:45 camelia rakudo-moar 1eb7fa: OUTPUT«marine␤»
16:46 psch i was about to look up how it works... :)
16:47 psch hoelzro++
16:47 lucas_ hoelzro: :D
16:47 TimToady m: #| WHY NOT!?!␤ sub philosopher {}; say &philosopher.WHY
16:47 camelia rakudo-moar 1eb7fa: OUTPUT«WHY NOT!?!␤»
16:48 psch m: #|[cut me some slack] sub genius {}; say &genius.WHY
16:48 camelia rakudo-moar 1eb7fa: OUTPUT«cut me some slack␤»
16:48 * psch back to blog-writing
16:49 * raydiak worries for TimToady who, at twice my age, seems to get half as much sleep at night
16:50 TimToady only last night, and I plan to take a geezer nap later
16:50 * raydiak worries somewhat less :)
16:50 mvuets #| is even better in combination with 'multi sub MAIN' → http://paste.scsys.co.uk/451928
16:52 raydiak oh, yeah! mvuets++ ...I should add that to some of my p6 work, kinda forgot about it
16:53 hoelzro mvuets++ # p⬢ post?
16:53 mvuets hoelzro: on my backlog already (-:
16:53 hoelzro =)
16:54 raydiak well, off to the doctor with me, just stopped in to say g'mornin, so...have a good morning, #perl6 \o
16:55 lucas_ So P6 has FP pattern matching in subroutine arguments?
16:55 TimToady yup
16:56 TimToady method arguments too
16:56 TimToady in either case, as long as they're declared multi
16:56 lucas_ TimToady: nice, thank you
16:57 sporus joined #perl6
16:57 ugexe ive been playing with `use`ing multiple modules with multi MAIN (including #| usage) for dynamic plugin type command line arguments. i love multi MAIN more than anything else :)
16:58 * TimToady remembers being delighted when he realized that multi would work for MAIN
16:58 smls d
16:59 smls [sorry, wrong window]
17:00 PerlJam TimToady: I was delighted too until I realized that subcommands wouldn't quite work like I wanted
17:00 TimToady well, we've had theories on that subject
17:02 mvuets lucas_: check out the rock-paper-scissors example here https://raw.githubusercontent.com/pe​rl6/book/397fd9c7334ff59ad8bc27ab178​2045a272056c8/src/multi-dispatch.pod
17:03 lucas_ mvuets: Thank you! And nice example the 'multi sub MAIN'...
17:04 * mvuets → home
17:05 atroxaper joined #perl6
17:06 Isp-sec joined #perl6
17:08 psch btw, i've sent my CLA a few days back.  who do i talk to to confirm arrival?
17:12 kurahaupo joined #perl6
17:15 treehug88 joined #perl6
17:15 [Coke] jnthn: minor nit: "It's job" -> "Its job"
17:16 sporus left #perl6
17:21 [Coke] tadzi++ #hashtag the future.
17:24 [Coke] jnthn++ #moarvm blog post
17:24 sergot Does perl6/doc is responsible for generating doc.perl6.org?
17:25 sergot what
17:25 sergot does is?
17:25 sergot I wanted to write somthing else
17:25 sergot is perl6/doc ~ anyway
17:26 raiph joined #perl6
17:26 smls sergot: Yes, it's generated fro that git repo
17:27 smls The htmlify.p6 script in the top-level of that repo generates the html (and related) files from the Pod files in the lib/ folder
17:29 sergot smls++ thanks
17:31 moritz psch: I think pmichaud++ has access to the CLA lists
17:32 smls sergot: If you want to view it locally, the simplest way is to pull the perl6/doc repo, run «./htmlify.p6», and then run  «perl ./app.pl daemon» to start a temporary mini-webserver (which will handle the link paths right etc.) that will let you view the generated docs at http://127.0.0.1:3000
17:32 sergot smls++ thank you!
17:32 psch moritz: thanks, i've been contacted already
17:32 sergot smls: I want to add syntax highlighting :)
17:33 smls cool :)
17:35 anaeem1_ joined #perl6
17:35 smls btw, is this Pod I wrote really malformatted: https://github.com/perl6/doc/​blob/master/lib/Type/Bag.pod  ...or is github just bad at parsing Pod?
17:35 smls (Pod::To::HTML converts it just fine)
17:36 PerlJam I think github is using Perl 5 POD parsing
17:38 jnthn evening, #perl6
17:38 [Coke] pmichaud & I both have access to the CLA lists.
17:41 vendethiel o/, #perl6
17:46 atroxaper joined #perl6
17:46 [Coke] (bad pod) you can nest =begin/=end in p6?
17:47 [Coke] m: 3.say
17:47 camelia rakudo-moar 1eb7fa: OUTPUT«3␤»
17:47 [Coke] m: =begin pod␤=begin code␤=end code␤=end pod␤say "alive";
17:47 camelia rakudo-moar 1eb7fa: OUTPUT«alive␤»
17:47 [Coke] m: =begin pod␤=begin code␤=end pod␤=end code␤say "alive";
17:47 camelia rakudo-moar 1eb7fa: OUTPUT«[31m===[0mSORRY![31m===[0m Error while compiling /tmp/yf_YiCSuRLâ�¤'=begin' not terminated by matching '=end pod'â�¤at /tmp/yf_YiCSuRL:5â�¤------> [32msay "alive";[33mâ��[31m<EOL>[0mâ�¤    expecting any of:â�¤        whitespaceâ�¤        horizontal…»
17:48 [Coke] that error message is slightly LTA. Still, nifty.
17:50 ptc_p6 joined #perl6
17:52 kurahaupo1 joined #perl6
17:56 jnthn [Coke]: Fixed the nit - thank you!
17:57 jnthn I hope the post wasn't too boring/weird...I was tireder than expected when I came to work on it.
17:58 dalek ecosystem: c528917 | moritz++ | server/updatelist.pl:
17:58 dalek ecosystem: [server] reduce dependencies
17:58 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/c528917991
17:58 dalek ecosystem: 20fa4a7 | moritz++ | server/updatelist.pl:
17:58 dalek ecosystem: [server] make output location configurable
17:58 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/20fa4a7fa2
17:58 moritz jnthn: I liked it
17:59 kurahaupo1 left #perl6
17:59 jnthn moritz: If somebody wanted to set up daily perl6-bench runs on the server, would that be appropriate use?
17:59 moritz jnthn: by all means
18:00 jnthn OK. I guess it's beefy enough that one core busy on that won't cause a big issue...
18:00 moritz right
18:00 andreoss how do i limit amount of threads perl6 spawns?
18:00 jnthn andreoss: Set up $*SCHEDULER to a ThreadPoolScheduler with the max threads configured as you like it.
18:02 jnthn Of course, some code could go and start a thread directly, or create its own scheduler, but the first is discouraged and the second is kinda anti-social. I guess your question is for keeping your own code bounded rather than enforcing a limit for arbitary code, though :)
18:06 dalek ecosystem: f8b11ba | moritz++ | server/updatelist.pl:
18:06 dalek ecosystem: [server] avoid temporary file
18:06 dalek ecosystem: review: https://github.com/perl6/e​cosystem/commit/f8b11bad86
18:06 psch http://perl6advent.wordpress.​com/?p=2560&amp;preview=true feedback much apreciated
18:08 andreoss jnthn: thanks a lot!
18:10 tadzik moritz++
18:11 * psch notices the missing "Day 12 - " in the headline...
18:11 moritz tadzik: my plan is to set up ecosystem-api.p6c.org from it
18:11 jnthn I saw a quesiton about using start with a grep inside of it earlier also. While I don't know why you don't get any data back, I do know thta the Promise will be kept almost immediately, because grep is lazy. If the grep should actually be done as part of the execution of the promise, then make it eager (stick an eager before the grep works I guess).
18:11 moritz tadzik: and once we do the DNS switch of perl6.org to the new server, it will be available as ecosystem-api.perl6.org as well
18:12 onebitboy joined #perl6
18:12 tadzik moritz: awesome
18:12 jnthn moritz: What port will it run on? Can we do it in a normal HTTP one nowadays?
18:12 jnthn (Sorry if you already did that and I missed it)
18:12 KCL_ joined #perl6
18:13 jnthn Just thinking of the folks behind firewall nastiness.
18:14 moritz jnthn: sure, will be port 80
18:15 tony-o does anyone have good docs or code showing a TLS handshake and the actual structure of a [Client/Server]Hello message?
18:15 jnthn moritz++
18:15 moritz jnthn: if necessary, it could get its own IP address; but if the clients all support name-based virtual hsots, we'll be fine without it
18:16 jnthn In 2014, I'd hope so :)
18:16 andreoss jnthn: why it runs extremply slow and segfaults with out specifing max_threads?
18:19 andreoss is it trying to do some special scheduling than?
18:24 jnthn andreoss: The segfaults are mostly due to insufficient reliability engineering. There's quite a backlog of known issues, and way too few people to work on them. By and large, it's mostly me who fixes them, and I've been horribly bogged down with other things in the last couple of months. :/
18:26 jnthn The slow is because the scheduler isn't especially smart yet, and it mostly worries about avoding deadlock rather than optimizing throughput.
18:26 jnthn For the first issue, Rakudo on JVM can offer a decent bit more stability at the moment.
18:28 jnthn It's also less fun to develop on though, due to high startup time.
18:29 moritz tadzik: does the API actually need to run any code? I mean, it seems to support '/list' and '/module/:name', both of which could easily be static files
18:29 tadzik moritz: yep, it could be all static
18:29 tadzik (and p6 generated)
18:29 tadzik BreakDancer could be employed :)
18:29 moritz tadzik: do you have any plans to make it more dynamic?
18:30 smls m: say await start { eager grep *.is-prime, 1, 2, 3, 4 };   # jnthn is right, eager makes it work
18:30 camelia rakudo-moar 1eb7fa: OUTPUT«2 3␤»
18:30 moritz tadzik: if not, I might do a quick p5 rewrite to generate static files
18:30 moritz and save me the trouble of deploying something dynamic
18:30 tadzik moritz: it's untouched for about two years, I think it's likely to stay that way for a while :
18:30 tadzik :)
18:30 andreoss jnthn: i've marked all optimizations which prevent segfaulting http://paste.fedoraproject.org/158853/41832260
18:31 andreoss naive solution can be found in perl6-examples prob027
18:31 mvuets joined #perl6
18:32 jnthn andreoss: Thanks; it's nice to have examples that expose problems.
18:33 jnthn Well, it's nicer not to have problems, of course, but... :)
18:33 FROGGS joined #perl6
18:36 bcode joined #perl6
18:53 dalek ecosystem/static: 724a0f1 | moritz++ | server/ (18 files):
18:53 dalek ecosystem/static: Generate JSON files statically
18:53 dalek ecosystem/static:
18:53 dalek ecosystem/static: also generate an errors.json file with error messages per URL
18:53 dalek ecosystem/static: review: https://github.com/perl6/e​cosystem/commit/724a0f106f
18:53 moritz shortstats for the commit above: 18 files changed, 34 insertions(+), 652 deletions(-)
18:53 moritz ok, most of the deletions were for automatically generated boilerplate; still feels nice :-)
18:55 FROGGS yeah :o)
18:55 FROGGS I also love to delete code
18:56 jnthn yeah, but I don't really like regressins :)
18:56 jnthn oh wait...mischannel
18:56 jnthn lol
18:56 jnthn FROGGS: Stop being in all the discussions about patches removing code when I'm tired and easily confused :P
18:57 FROGGS jnthn: wait... you are about to blug, no?
18:57 FROGGS jnthn: :P
18:57 moritz FROGGS: do you want to move cpandatatesters to the new machine(s)?
18:57 jnthn FROGGS: I did my advent post last night?
18:57 FROGGS jnthn: just distract me with a blog
18:57 FROGGS ohh
18:57 psch fwiw, i've scheduled my post about 45 minutes ago and was asking for feedback... :P
18:57 FROGGS moritz: I think so... because I am not that happy with the performance of my server
18:58 moritz FROGGS: what's the bottleneck? CPU?
18:58 FROGGS jnthn: see, I am still confused by your 'Ja' answer :o)
18:58 FROGGS moritz: probably... it is a virtual server
18:59 FROGGS so I have four cores, but it does not feel like four cores
18:59 moritz FROGGS: well, {hack,www}.p6c.org are also virtual - but the hypervisor isn't resource-starved
18:59 moritz Stage parse      :  28.386
18:59 FROGGS ohh nice!
18:59 moritz on hack.p6c.org for raudo-moar
19:00 FROGGS my box takes about 70s me thinks
19:02 ab5tract joined #perl6
19:02 moritz FROGGS: so, first step would be an account for you, I think :-)
19:03 FROGGS :o)
19:03 FROGGS that'd be sweet
19:04 bcode joined #perl6
19:09 FROGGS_ joined #perl6
19:09 moritz psch: fwiw I've quickly read through your post, and liked it. But I'm afraid I have too much other stuff going on right now to do a thorough read-through
19:10 moritz including dinner&
19:14 ab5tract psch: is a "longname" a formal thing or a typo? :)
19:16 psch ab5tract: it's in STD.pm6 i've been told, and i think it's also in Perl6/Grammar.nqp
19:16 psch so i'd consider it formal :)
19:16 psch <name> {} [ <?before ':' <+alpha+[\< \[ \« ]>> <colonpair> ]* # that's a longname
19:17 FROGGS_ a longname is a thing, yeah
19:17 FROGGS_ "infix:<&&>" is a longname for example
19:19 japhb jnthn, moritz: Let me know if you find you need any perl6-bench functionality to make daily runs easier to manage or get useful info from.  I will attempt to prioritize those requests.
19:20 japhb You're also both collaborators on the GitHub project, so you're welcome to just commit anything you need and hack up yourself.
19:20 psch ab5tract: the post is littered with other typos though, as i'm realizing during the 10th proof reading...
19:28 ab5tract psch: in the case of it being formal, a small word or link to it might be cool!
19:28 ab5tract org.perl6.nqp.runtime.BootJavaInterop
19:28 ab5tract the line with that on it is sad :(
19:30 ab5tract also, i know your pain.. i was still proofreading and revising until late writing until ~2am or so
19:30 psch that should also link to the class to encourage following along i guess? :)
19:30 ab5tract (missing a comma there)
19:31 ab5tract well, could do, could do ;)
19:31 atroxaper joined #perl6
19:31 japhb jnthn: Did you see my example from last night about 'lastcall; nextsame' not seeming to do the right thing?
19:32 japhb m: class NewDate is Date { multi method truncated-to(|) { lastcall; nextsame } }; my $nd = NewDate.today; say $nd.truncated-to('week');
19:32 camelia rakudo-moar 1eb7fa: OUTPUT«Nil␤»
19:32 ab5tract honestly i think it could do with more inline code samples
19:32 grettis joined #perl6
19:33 ab5tract unless they are really complicated, i think being able to read along with the code would help my understanding of the story you are telling (of which i am quite interested, by the way!)
19:34 jnthn japhb: Uh...what's the lastcall doing there?
19:34 ab5tract i'm not sure if smls has posted his method of getting such nice code snippets, but that would be icing on the cake, so to speak
19:35 nwc10 jnthn++ # bloggage
19:35 jnthn japhb: From S12: "It is also possible to trim the candidate list so that the current call is considered the final candidate."
19:35 jnthn japhb: So you throw away the rest of the candidate list and then do nextsame, which needs the candidate list? :)
19:35 psch ab5tract: the problem with inline code samples is that it's much code, and hard to condense :/
19:36 japhb jnthn: The lastcall was a workaround suggested by TimToady for the problem that since this truncated-to is a multi, and Date's is not, you have to end the current dispatcher and fall back to the other one.  But that idea didn't seem to work.
19:36 psch although maybe i can get at least a heavily cut final version in there somewhere
19:37 gfldex joined #perl6
19:37 japhb (Yes, I want a multi in the derived class, there's multiple truncated-to() methods, but this is the fallback method to just use the base class's functionality.)
19:40 jnthn Ah
19:40 jnthn I'd be tempted just to fully qualify the call :P
19:42 smls ab5tract: For Java snippets,  [code language="java"] ... [/code]  should work on wordpress.com for syntax highlighting
19:43 smls It doesn't support Perl 6 though
19:44 japhb jnthn: As in: multi method truncated-to(|c) { self.Date::truncated-to(|c) }   ?
19:45 smls ab5tract: So I used pygmentize on the code snippets in my advent post, then hand-edited the result to fix parts where the hgihlighter hadn't done a good job, and then inlined the CSS styles so it would work on wordpress
19:45 jnthn japhb: Yeah
19:45 japhb What is the right way to jump dispatchers that way?
19:45 smls (using these convenience scripts for the first and last step: https://github.com/smls/undef/tree/master/scripts)
19:45 ab5tract .hug smls
19:46 jnthn japhb: Not sure, tbh. Maybe what TimToady suggested is meant to work...but it implies we have some nested model of dispatchers, which is a little tricky to set up :)
19:48 ab5tract psch: then maybe a few examples of the new syntax in use (for JVM neophytes, we don't even know what to call once we do checkout your changes))
19:49 bartolin hi, #perl6
19:49 ab5tract smls: in my opinion, you are a hero. we need a better way :/
19:49 psch ab5tract: that's the point though, with my changes you can just call "$crc.update($myint)" :)
19:49 japhb jnthn: ISTR there are indeed supposed to be multiple levels of dispatchers, but it's been a while since I read through that whole section.
19:50 tadzik psch: show it as example, definitely
19:50 tadzik I read the article, didn't see any example at the end and thought "oh, so this is just a plan then"
19:50 psch well, yes, i probably should
19:50 psch i mean, the "that's the point" kind of made it clear to myself as well :D
19:50 ab5tract psch: yes, but surely there is something more fun than a CRC in the default Java SDK? :)
19:52 psch tadzik++ ab5tract++ # great feedback
19:52 psch ab5tract: i'll have to look, i hardly know the JDK
19:53 bartolin FROGGS_: I pushed some changes for rakudo PR 336. to me those changes look like improvements ;-)
19:54 moritz tadzik: could you please test panda with http://ecosystem-api.p6c.org/ as the API URL?
19:54 Sqirrel joined #perl6
19:54 FROGGS_ bartolin: will take a look soonish :o)
19:54 bartolin ++FROGGS_
19:56 smls m: for 'Hello, Java'.encode('UTF-8') { .say }
19:56 camelia rakudo-moar 1eb7fa: OUTPUT«Buf:0x<48 65 6c 6c 6f 2c 20 4a 61 76 61>␤»
19:56 smls psch: ^^ is there maybe a «.list» missing there?
19:57 smls for 'Hello, Java'.encode('UTF-8').list { .say }
19:57 smls m: for 'Hello, Java'.encode('UTF-8').list { .say }
19:57 camelia rakudo-moar 1eb7fa: OUTPUT«72␤101␤108␤108␤111␤44␤32␤74␤97␤118␤97␤»
19:57 psch smls: actually, it's complicated
19:57 psch smls: the example that jnthn posted doesn't work anymore nowadays, for two reason
19:57 psch marshaling of arrays back to perl6 is broken
19:58 psch and the method descriptor doesn't exist
19:58 psch i supposed the latter happened because a [ got lost somewhere
19:58 psch ...which means i should probably correct the example
19:58 tadzik moritz: sure
19:58 psch anyway, the method that would work has the desc "method/update/([I)V"
19:58 psch and that takes an int[]
19:59 psch ...iirc
19:59 psch i'd have to look it up, but yes, the example has to be corrected in any case
20:01 rurban joined #perl6
20:03 moritz also there is now http://ecosystem-api.p6c.org/errors.json for a list of errors while fetching / processing META.info files
20:05 tadzik Bytecode validation error at offset 4, instruction 2: register operand index 54 out of range 0..15
20:05 tadzik that's perfect
20:05 jnthn what the...
20:07 psch fwiw, there's now a heavily condensed version of the fallback method in the draft
20:08 psch i'm not sure it really adds something, i think i'll also link to the actual method as it's implemented atm
20:08 tadzik jnthn: how do I get more details out of it, provided I can reproduce it?
20:08 pecastro joined #perl6
20:09 jnthn tadzik: Well, --ll-exception may give a stack trace that tells us where it comes from...
20:10 jnthn What are you doing, ooc?
20:11 tadzik 'panda install .'
20:11 tadzik well, now it worked >_<
20:11 jnthn grmbl
20:11 jnthn I hope this isn't going to be The Next Heisenbug...
20:13 psch righto, time to see if i can find something fun and interesting in the jdk
20:13 tadzik moritz: it doesn't seem like the improvised http client that panda uses is good enough for the new server
20:13 moritz tadzik: does it send the Hostname: header?
20:14 naddiseo joined #perl6
20:14 tadzik nope
20:14 FROGGS_ no
20:14 moritz tadzik: it's a name-based virtual host, so that's the bare minimum
20:14 FROGGS_ it does http 1.0 without that header
20:14 tadzik $s = IO::Socket::INET.new(:host<ecosystem-api.p6c.org>, :port(80));
20:14 tadzik $s.send("GET /projects.json HTTP/1.0\n\n");
20:14 FROGGS_ but it is easily tweakable, no?
20:14 tadzik it's just this
20:14 tadzik yep
20:14 FROGGS_ give it a whirl then :D
20:16 dalek mu: 201d297 | moritz++ | misc/dalek- (2 files):
20:16 dalek mu: Remove dalek polling config and instruction
20:16 dalek mu:
20:16 dalek mu: it has not been running for several weeks, and nobody has complained so far.
20:16 dalek mu:
20:16 dalek mu: Also it seems that the poller was never adapted to github URL changes,
20:16 dalek mu: so it was pretty much broken
20:16 dalek mu: review: https://github.com/perl6/mu/commit/201d2978b1
20:18 jnthn psch: "no method multiple dispatch (MMD)"
20:19 jnthn psch: I had it as meaning "multi-mmethod dispatch"
20:19 jnthn uh, without the spello :)
20:19 psch oh
20:19 psch that's probably because i had "multiple dispatch" from somewhere else
20:19 psch i'll fix that too
20:20 jnthn psch: "except we are supplying a longname () consisting of the module name and the colonpair"
20:20 jnthn I'm guessing something is going on with the empty parens?
20:20 bjz joined #perl6
20:20 psch no, i think the parens have to go because they're empty :)
20:21 dalek specs: f6ea401 | duff++ | S99-glossary.pod:
20:21 dalek specs: [S99] Add long name, short name, and multiple dispatch
20:21 dalek specs: review: https://github.com/perl6/specs/commit/f6ea401c87
20:21 moritz my $s = IO::Socket::INET.new(:host<ecosystem-api.p6c.org>, :port(80));
20:21 moritz $s.send("GET /projects.json HTTP/1.0\nHost: ecosystem-api.p6c.org\n\n");
20:21 moritz say $s.recv;
20:21 PerlJam psch: I added some of those term to our glossary in case you want to link to them.
20:21 moritz that seems to work
20:21 psch PerlJam: thanks, will do
20:22 PerlJam And everyone else look at the definitions I just added and update them if they seem a little off.   I just did this off the top of my head :)
20:22 tadzik oh, now I got a segfault
20:24 ab5tract tadzik: OS X?
20:25 tadzik ab5tract: nah, linux
20:25 tadzik now it worked
20:25 tadzik moritz: yeah, now it work
20:25 tadzik s*
20:27 tadzik moritz: do you plan any downtime, or can I commit changes to panda?
20:27 moritz tadzik: no downtimes planned
20:27 tadzik okay
20:29 dalek panda: 9fe5154 | tadzik++ | lib/Panda/Ecosystem.pm:
20:29 dalek panda: Switch ecosytem API host to p6c.org
20:29 dalek panda: review: https://github.com/tadzik/panda/commit/9fe51545a1
20:29 tadzik please test
20:31 FROGGS_ why not just ecosystem.p6c.org?
20:31 FROGGS_ (I don't like 'API' in any name tbh)
20:32 moritz FROGGS_: because it's not an HTML site
20:32 PerlJam FROGGS_: even if it's ecosystem.p6c.org/api ?
20:32 FROGGS_ PerlJam: even then :o)
20:32 FROGGS_ moritz: okay, I see
20:32 moritz FROGGS_: and if there is both modules.perl6.org and ecosystem.perl6.org, the difference is not obvious
20:33 FROGGS_ true
20:33 moritz FROGGS_: (p6c.org is only meant as a temporary name until perl6.org DNS switches over)
20:33 jnthn psch: I guess "*insert fun example here that works*" is something you know is still to do? :)
20:34 jnthn psch++ # nice post :)
20:35 psch jnthn: yes, i'm working on inserting the fun, working example
20:35 dalek rakudo/nom: 8b85116 | usev6++ | src/core/operators.pm:
20:35 dalek rakudo/nom: Allow chained sequences with closures
20:35 dalek rakudo/nom: and more than one element after last infix:<...> operator
20:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/8b85116017
20:35 dalek rakudo/nom: da8b73d | usev6++ | src/core/operators.pm:
20:35 dalek rakudo/nom: Optimization of chained sequences with closures
20:35 dalek rakudo/nom: FROGSS++; this is closer to the original version;
20:35 dalek rakudo/nom: still passes tests from RT #123329
20:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/da8b73ddcd
20:35 jnthn \o/
20:35 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=123329
20:35 dalek rakudo/nom: 3fa1bdf | (Tobias Leich)++ | src/core/operators.pm:
20:35 dalek rakudo/nom: Merge pull request #336 from usev6/nom
20:35 dalek rakudo/nom:
20:35 dalek rakudo/nom: Allow chained sequences with closures
20:35 dalek rakudo/nom: review: https://github.com/rakudo/rakudo/commit/3fa1bdfe23
20:35 FROGGS_ bartolin++
20:38 dalek panda/ecosystem-api: 26f1164 | moritz++ | bin/redpanda:
20:38 dalek panda/ecosystem-api: Start to fill move HTTP communication into a separate module
20:38 dalek panda/ecosystem-api: review: https://github.com/tadzik/panda/commit/26f11646b7
20:38 bartolin oh, great. thanks for reviewing!
20:38 Rounin joined #perl6
20:39 smls psch: I just tested it on a private wordpress.com blog, and using   [code language="java"]...[/code]   instead of   <pre>...</pre>  not only adds syntax highlighting, but also line numbers (and a horizontal scrollbar if the code is too wide). Might be nice for the two long-ish Java code snippets in your post.
20:39 moritz tadzik: was it intentional that you didn't update bin/redpanda?
20:40 tadzik moritz: now, I merely forgot about it. Honestly, I'd be surprised if anyone used it, maybe I should just kill it
20:40 moritz tadzik: I don't even know what it does :-)
20:40 tadzik it looks like I wanted to have a cpanm for perl6 :P
20:40 psch smls: thanks, i'll do
20:40 tadzik I'm tempted to leave it like that to see if anyone notices and reports it
20:41 tadzik this way I'll know if there's a single user of it
20:41 psch *that
20:41 kurahaupo joined #perl6
20:43 psch i think my test class is good enough - i'm having trouble finding something that'd actually work
20:44 psch because i realized there's another thing still not quite working yet, that is overloaded constructors
20:47 tadzik good thing to note too :)
20:48 psch oh yes, definitely
20:48 psch if had that i could pull out an AWT or Swing example as "look what works now", alas...
20:48 * psch .oO( THERE'S STILL TIME!!1 )
20:48 ab5tract :D
20:49 psch ooh
20:49 psch yeah
20:50 psch i don't think i'll try that, it'd take more than two hours
20:53 moritz psch: and don't pressure yourself too much; it's just an advent post :-)
20:57 [Sno] joined #perl6
20:58 psch moritz: i'm fine, thanks for the reminder though :)
20:58 Alula joined #perl6
21:02 FROGGS m: say $_, $/, $! # TimToady: all of them should be Any?
21:02 camelia rakudo-moar 1eb7fa: OUTPUT«NilNilNil␤»
21:03 FROGGS I've got quite some spectest fallout about $/, but need to investigate
21:06 psch alright, i'm done i think
21:07 * [Coke] is skipping daily run today. will see about moving to the new box before the next run.
21:07 FROGGS TimToady: S02-types/nil.t explicitly checks that these three are Nil...
21:08 psch the fun and working example is rather contrived, but it makes the point i guess
21:09 psch 26 revisions heh
21:12 bartolin FROGGS: what kind of spectest fallout do you see? I've run a spectest two hours ago and saw unexpected results in S05-modifier/overlapping.rakudo.moar (You planned 22 tests but ran 38) plus one passed TODO in S10-packages/precompilation.rakudo.moar
21:12 bartolin FROGGS: that spectest was on moar only
21:12 FROGGS bartolin: the S05-modifier will go away once you pull
21:13 FROGGS but my local patch for RT #123175 makes regexes very unhappy
21:13 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=123175
21:15 bartolin aha, I misunderstood. so it's your local spectest fallout
21:16 FROGGS correct
21:16 FROGGS m: say "\x[846D]" ~~ m/^<:!L>$/
21:16 camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
21:16 FROGGS that, for example, comes out true here O.o
21:17 FROGGS m: $/ = Any; say "\x[846D]" ~~ m/^<:!L>$/
21:17 camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
21:20 konsolebox joined #perl6
21:23 naddiseo How does one do a left recursive rule, such that "1,2,3" is parsed as "(1, 2), 3" instead of "1, (2, 3)" ?
21:23 moritz naddiseo: you don't
21:24 telex joined #perl6
21:24 naddiseo So how would I do that kind of grouping?
21:25 PerlJam naddiseo: why exactly?
21:25 naddiseo a.b.c -> attr( attr('a', 'b'), 'c' )
21:26 moritz naddiseo: you simply parse with  <term>+ % ','
21:26 moritz naddiseo: so you get a flat list
21:26 moritz naddiseo: and then do the grouping outside the grammar
21:26 kaare_ joined #perl6
21:29 naddiseo Okay, I'll give that a try. Thanks.
21:35 FROGGS m: my @a = <a b c d e f g>; my $r = shift @a; for @a -> $x { $r = [$r, $x] }; say $r.perl
21:35 camelia rakudo-moar 3fa1bd: OUTPUT«[[[[[["a", "b"], "c"], "d"], "e"], "f"], "g"]␤»
21:38 PerlJam looks like a reduce operation to me  :)
21:38 smls m: my @a = <a b c d e f g>; @a.reduce({ [$^a, $^b] }).perl
21:38 camelia rakudo-moar 3fa1bd: ( no output )
21:38 smls m: my @a = <a b c d e f g>; @a.reduce({ [$^a, $^b] }).perl.say
21:38 camelia rakudo-moar 3fa1bd: OUTPUT«[[[[[["a", "b"], "c"], "d"], "e"], "f"], "g"]␤»
21:38 PerlJam apparently to smls too :)
21:39 smls yep
21:41 jnthn detrain; bbiab
21:42 moritz m: my @a = <a b c d e f g>; say @a.reduce(&circumfix:<[ ]>).perl
21:42 camelia rakudo-moar 3fa1bd: OUTPUT«[[[[[["a", "b"], "c"], "d"], "e"], "f"], "g"]␤»
21:43 FROGGS bartolin: I might have been wrong about this one:
21:43 FROGGS t/spec/S05-modifier/overlapping.rakudo.moar ................... Dubious, test returned 255 (wstat 65280, 0xff00)
21:43 FROGGS could it be a regression caused by your patch?
21:44 FROGGS because I am not sure anymore what revision was clean :/
21:46 dalek2 joined #perl6
21:46 masak 'night, #perl6
21:47 d4l3k2_ joined #perl6
21:47 d4l3k2_ joined #perl6
21:48 bartolin FROGGS: I'm also still looking at it.
21:52 bartolin FROGGS: I had clean spectests yesterday. maybe it has something to do with TimToady's commits earlier today. Actually I'm not sure 22 is the correct number of tests for S05-modifier/overlapping.t
21:53 moritz overlapping.t was last touched on 2014-07-05, so at least we know it's not a test change
21:54 TimToady my patch caused more tests to be run, I think
21:54 bartolin that would explain it:
21:54 bartolin t/spec/S05-modifier/overlapping.rakudo.moar (Wstat: 65280 Tests: 38 Failed: 16) Failed tests:  23-38 TODO passed:   1-3, 11-14, 22 Non-zero exit status: 255 Parse errors: Bad plan.  You planned 22 tests but ran 38.
21:54 moritz oh, then it's a bad test file
21:54 TimToady people shouldn't really write tests that run only if a previous test succeeds
21:55 FROGGS it might not skip correctly...
21:55 moritz the number of tests shouldn't depend on the correctness of the feature being tested
21:55 bartolin if I change the plan to 38 all is well (some passed TODO)
21:55 moritz https://gist.github.com/mo​ritz/f37058d81010819692d6 # my TODO for the new server; if you miss any service that you think should be migrated, please let me know
22:00 ab5tract joined #perl6
22:02 bartolin well, I'll take a look at the test file. if I'm unable to fix it properly, I'll change the plan to 38 for now and open a ticket.
22:02 FROGGS bartolin: wait
22:03 * bartolin is waiting
22:03 bartolin (especially for rakudo.jvm and rakudo.parrot to build)
22:03 TimToady if you're compiling, you're supposed to be goofing off now
22:04 TimToady that's what compiling is for
22:05 bartolin ;-)
22:05 rurban joined #perl6
22:06 dalek roast: e3b109e | (Tobias Leich)++ | S05-modifier/overlapping.t:
22:06 dalek roast: fix skip count via #?DOES
22:06 dalek roast:
22:06 dalek roast: The skipped tests skipped the wrong number of tests because some of them are in a loop, and
22:06 dalek roast: we dont get the loop right either.
22:06 dalek roast: review: https://github.com/perl6/roast/commit/e3b109e1f2
22:08 japhb moritz: You and jnthn discussed a daily perl6-bench run; also, will [Coke]'s daily smoke test be moved there?
22:09 moritz japhb: it will if somebody moves it
22:09 bartolin FROGGS++
22:09 moritz japhb: I'm trying to not do the things that regular users can do themselves
22:10 moritz japhb: but might be worth mentioning as TODOs for others :-)
22:10 jnthn When it comes to servers, I'm a pretty irregular user... :P
22:11 jnthn I'll mebbe have a look into setting up the perl6-bench thingy at the weekend...
22:17 japhb Cool beans, jnthn.
22:17 japhb Understood, moritz.  And yes, TODOs for others makes sense.
22:19 jnthn Tomorrow I finally have a day when I can (a) sleep in, (b) do Perl 6 things, and (c) not have to care about much else :)
22:22 japhb Sounds like an *excellent* day.
22:23 * japhb is just watching the flooding slowly rise outside
22:23 jnthn uh-oh
22:23 japhb May flood the garage, but I think the house is safe.
22:23 TimToady just so's it's slow...
22:24 timotimo you reap what you so.
22:24 japhb .oO( waitpid() should have been called reap() ... }
22:25 moritz root@hack:~# grep -iE 'invalid user|failed password' /var/log/auth.log|wc -l
22:25 TimToady he said grimly...
22:25 moritz 2815
22:25 moritz it's a sad internet we live in
22:25 dalek roast: 1bcc540 | usev6++ | S03-sequence/misc.t:
22:25 dalek roast: Unskip passing tests for RT #123329
22:25 dalek roast: review: https://github.com/perl6/roast/commit/1bcc540815
22:25 synopsebot Link: https://rt.perl.org/rt3//Publ​ic/Bug/Display.html?id=123329
22:25 moritz that host has only been running for two days
22:25 timotimo yeah, that's par for the course
22:26 moritz time for sleep
22:26 hoelzro moritz: =(
22:26 moritz good * everybody
22:27 hoelzro fail2ban to the rescue?
22:27 hoelzro o/ moritz
22:27 bartolin o/
22:27 Ulti moritz do you know abut fail2ban?
22:27 japhb good * moritz
22:27 Ulti oh hoelzro beat me to it :D
22:28 hoelzro Ulti: that's just more endorsement =)
22:28 Ulti yeah I used to use DenyHosts but its not well supported anymore... been running fail2ban for a few months on a new server and it works well
22:32 perturbation joined #perl6
22:32 * TimToady wrote his own one of those
22:33 TimToady just keeps its own private set of subrules in the firewall
22:38 raiph joined #perl6
22:50 ab5tract psch: the java code snippets are cut short on the RHS
22:51 ab5tract the advent theme breaks on code snippets in general, but usually with overflows (which are jarring but readable)
22:52 psch ab5tract: aren't they scrollable for you?
22:53 ab5tract ah! indeed they are :)
22:53 ab5tract does anyone mind if i polish up the blogroll section?
22:54 ab5tract it still links to planetsix, and is missing 6guts / strangeconsistent / perlgeek / likely more
22:54 ab5tract *strangelyconsistent
22:54 [Coke] ab5tract: go for it
22:54 ab5tract oh, and perl6weekly
22:54 ab5tract cool
22:56 jnthn Hm. Seems the build got slower (NQP and Rakudo) since last time I built on my desktop. :(
22:57 TimToady my fastest times seem about the same, but it also seems more variable
22:58 jnthn It's from about 60s to 66s on Windows for the full Rakudo build.
22:59 jnthn 10% is quite a lot.
22:59 psch i also found a bunch of html entities...
22:59 ab5tract done :)
23:02 ab5tract in other news, snow golf can now automagically use your console width: COLUMNS=$(tput cols) perl6 -e "for 0..Inf { say [~] [ '❆ ','❄ ', '❅ ', ' ' xx 60 ].Bag.roll( %*ENV<COLUMNS> ); }"
23:04 TimToady why not just loop {} ?
23:04 japhb And if you're using Bag.roll, why do 'xx 60'?
23:04 TimToady more spaces
23:04 japhb Damn precedence.
23:05 ab5tract japhb: that's the weight of the snowfall :)
23:06 japhb Or its inverse.  :-)
23:06 ab5tract well, the snow falls upside down, doesn't it?  ;)
23:06 jnthn .u ❆❄❅
23:06 yoleaux U+2744 SNOWFLAKE [So] (❄)
23:06 yoleaux U+2745 TIGHT TRIFOLIATE SNOWFLAKE [So] (❅)
23:06 yoleaux U+2746 HEAVY CHEVRON SNOWFLAKE [So] (❆)
23:07 ab5tract TimToady: i guess for `0..Inf` just feels way more epic
23:07 jnthn I didn't know a snowflake could be tight :P
23:07 jnthn TIL... Thanks Unicode :)
23:09 ab5tract everyone has infinite loops, but infinite lists are a rarer breed :)
23:10 baest joined #perl6
23:10 japhb Exchange u+274[456] with u+262[0123] and you get the sci-fi apocalypse
23:11 psch .u ☡
23:11 yoleaux U+2621 CAUTION SIGN [So] (☡)
23:11 TimToady falling cuneiform would be cool
23:12 atroxaper joined #perl6
23:12 TimToady 'course, it would mostly look like it was raining fish
23:12 japhb fish bones, at least ...
23:13 rurban joined #perl6
23:13 * TimToady wonders if there are any fish swimming in his street...
23:13 colomon_ joined #perl6
23:17 ab5tract the golfed matrix.. now i'm curious
23:18 TimToady well, but you have to flip your screen left-to-right without flipping it vertically for that, since the characters are all backwards
23:20 TimToady or you need to flip it vertically too, but put each character back right-side up, but backwards
23:20 TimToady I think you need the green pill for that...
23:20 ab5tract lol
23:50 aborazmeh joined #perl6

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

Perl 6 | Reference Documentation | Rakudo