»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg camelia perl6: ... | irclog: irc.perl6.org | UTF-8 is our friend!
Set by sorear on 25 June 2013.
00:10 Rounin left 00:16 |Tux| joined 00:18 TuxCM left 00:20 robinsmidsrod left, dayangkun joined, robinsmidsrod joined, dayangkun left
jnthn OK, advent post scheduled at a minute past midnight. :) 00:20
And I'm gonna sleep :) 00:21
&
psch night jnthn
00:28 gfldex left
timotimo suggests someone puts a "Day 11 - " in front of the title 00:38
00:41 atroxaper joined
psch uh 00:41
scheduled for a minute past midnight..? 00:42
timotimo: i can do that, but i'm confused regarding when the post should be published at the moment... 00:43
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 atroxaper left
psch hm, it says 12/11/2014 @ 00:19 in quick edit 00:45
but no TZ info 00:47
visible to me at least, but then i hardly know wordpress
timotimo ah 00:48
December 11, 2014, 12:48 am
this is what it shows in the stats page
that must be WP's local timezone
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"
because, as we all know, 00:19 is 12:19am, which is part of the previous day
psch i don't follow 00:50
12/11/2014 is american date format
00:50 lue joined
timotimo oh! 00:50
derp
how do americans cope?!
geekosaur by being bewildered at sane date formats 00:51
>.>
psch heh
it's still possible to schedule the post for 12/11/2014 @ 00:01
timotimo or you could just post it immediately; WP thinks the new day has already begun
psch yeah, i guess that's safe 00:52
does it need some magic to appear on the landing page? 00:53
geekosaur the new day started 53 minutes ago, of course...
psch 'cause it's not there but is in the category 2014... 00:54
timotimo i see it 00:56
psch i can see it in the logged-in browser, but the other only sees it in the category 00:57
err, NOW
some caching
probably
timotimo seems likely 00:58
psch i also updated the ToC 01:01
not sure if the link from day 10 to day 11 happens automatically eventually or if that has to be added manually
'cause that's not there either on my end atm
timotimo aye, only one of them shows up here; from 11 to 10 01:03
01:05 BenGoldberg joined
psch well, i don't know what to do about that link 01:08
i also have to go sleep &
timotimo gnite psch! 01:09
01:13 psch left 01:42 adu joined 01:46 kurahaupo joined
dalek ecs: 23ef17e | TimToady++ | S0 (2 files):
Add List to 'OKness' types
 01:50
TimToady changes List.ACCEPTS to return self, to see what breaks 01:56
dalek ecs: b40b20c | TimToady++ | S05-regex.pod:
another fake Match becomes List
 02:06
02:09 rmgk left 02:13 rmgk joined
TimToady Someone needs to entitle a talk: The following step can take a long time, please be patient. 02:13
japhb
.oO( *** Implementing Perl 6 ... )
 02:25
TimToady looks like it fails about 6 test files, but passes new TODOs in 5 02:33
so we almost broken even :)
*broke 02:34
colomon thinks broken might be more appropriate here... 02:35
TimToady now I just need to figure out what word to name the old whatever-matching code 02:36
I could just move the ACCEPTS over to Parcel, but that would break as soon as we we get rid of Parcel 02:39
02:42 jimmy2 joined 02:43 jimmy2 is now known as JimmyZ_
TimToady what the hey, maybe it can be a transitional form while we decide what to do with parcel list matchers 02:44
whoops, I said 'make install' instead of 'make m-install'... 02:45
TimToady goofs off, since "he's compiling"
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␤»
japhb OK, what stupid mistake am I making there? ^^
02:53 aborazmeh_ joined 02:55 aborazmeh_ left, aborazmeh left
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? 02:56
TimToady we talked about it some on irc :) 02:59
lue not surprised :)
03:00 chenryn joined, baest left
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
m: say 1,2 X+ 8,9; say [X+] (1,2;8,9);
camelia rakudo-moar 5f6494: OUTPUT«9 10 10 11␤2 2␤»
03:03 khisanth_ is now known as Khisanth
japhb TimToady: Can you see a mistake in what I'm doing above (the NewDate class)? 03:03
03:05 baest joined 03:06 araujo joined, araujo left, araujo joined
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␤»
TimToady lue: you passed a single argument to [X+] 03:08
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?) 03:09
03:10 araujo left, araujo joined
lue Doesn't seem to work when the LoL is in a variable, though. 03:11
03:12 noganex joined 03:13 grettis left 03:14 noganex_ left 03:15 jimmy_ joined
TimToady japhb: leave off the 'multi' 03:15
lue: or were you asking about the GLR? pmichaud++ has a blog on that... 03:16
03:16 araujo left
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 03:19 jimmy_ left
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? 03:28
TimToady alternately, try a 'lastcall; nextsame' 03:29
03:31 atroxaper joined
TimToady the basic problem being that nextsame goes to the wrong dispatcher 03:32
so you have to terminate the multi dispatch to get to the next method dispatch 03:34
03:38 adu left, atroxaper left 03:49 aborazmeh joined, aborazmeh left, aborazmeh joined
japhb I was afraid the workaround would be much uglier. I'll give lastcall a try. 03:50
03:51 jack_rabbit joined
japhb Dang, no dice. Is lastcall not implemented? 03:54
TimToady could be
JimmyZ_ m: lastcall
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␤␤»
TimToady hmm 03:55
JimmyZ_ m: nextsame
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␤␤»
TimToady odd
TimToady doesn't see how a role would interfere... 03:56
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␤»
japhb Role isn't the problem; it's not getting to the right dispatcher it looks like
04:12 anaeem1_ joined 04:14 anaeem1_ left 04:15 anaeem1 joined 04:17 pyrimidi_ joined 04:32 KCL_ joined 04:47 chenryn left 04:54 tinyblak joined
dalek kudo/nom: 080b40f | TimToady++ | src/core/ (3 files):
move List.ACCEPTS Parcel and Array for now

List.ACCEPTS now indicates a success status.		 05:00
tony-o m: 'something.min.js'.match(/ '.' .*? $ /).Str.say;
camelia rakudo-moar 5f6494: OUTPUT«.min.js␤»
05:02 kaare_ joined 05:03 KCL joined 05:05 KCL_ left
tony-o how do i do [^\.] in a regex for perl6 ? 05:08
i'm not using the right words and all i'm finding is <!before 05:09
05:11 chenryn joined
flussence <-[.]> ? 05:14
tony-o ah.. missed the < >, thanks flussence 05:17
TimToady wonders if -. would be a useful sugar 05:20
or something else a bit shorter than 6 chars 05:21
m: say "foo" ~~ m:g/./ 05:24
camelia rakudo-moar 080b40: OUTPUT«｢f｣␤ ｢o｣␤ ｢o｣␤␤»
TimToady \o/
tony-o m: say "foo." ~~ m/-'.'/; 05:27
camelia rakudo-moar 080b40: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter - (must be quoted to match literally)␤at /tmp/m4R7psleri:1␤------> say "foo." ~~ m/⏏-'.'/;␤Unrecognized regex metacharacter ' (must be quoted to match literally)…»
tony-o m: say "foo." ~~ m/<-'.'>/;
camelia rakudo-moar 080b40: OUTPUT«===SORRY!===␤Unrecognized regex metacharacter < (must be quoted to match literally)␤at /tmp/_pv56ufYIX:1␤------> say "foo." ~~ m/<-⏏'.'>/;␤Unrecognized regex metacharacter - (must be quoted to match literally)…»
tony-o m: say "foo." ~~ m/<-['.']>/;
camelia rakudo-moar 080b40: OUTPUT«Potential difficulties:␤ Repeated character (') unexpectedly found in character class␤ at /tmp/CjtEtL88Cs:1␤ ------> say "foo." ~~ m/<-⏏['.']>/;␤｢f｣␤␤»
05:30 JimmyZ_ left 05:31 rurban joined 05:35 kurahaupo left 05:39 ptc_p6 joined 05:42 adu joined 05:44 [Sno] left, JimmyZ_ joined 05:51 telex left, ptc_p6 left 05:52 telex joined 05:55 kaare_ left 06:08 nige joined 06:10 tinyblak left, tinyblak joined 06:24 kaare_ joined 06:32 ptc_p6 joined 06:33 mr-foobar left 06:36 tinyblak_ joined 06:40 tinyblak left 06:41 ptc_p6 left
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... 06:48
lue I would almost suggest that a list of matches should rather be in @/ :) . 06:52
06:54 atroxaper joined 06:56 kaleem joined 06:58 Isp-sec joined 07:00 atroxaper left
TimToady @() is sufficient for that 07:02
raydiak amusing typo of the day: "rakudobrew build most" 07:04
07:08 BenGoldberg left
TimToady feels as though that should be 'build moast' 07:09
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 07:12
TimToady so 'build niar' would be the opposite mistake... 07:14
07:15 bjz joined
raydiak for one of several interpretations of "opposite", absolutely 07:17
Woodi hallo today :) 07:18
raydiak: in Polish "most" is "bridge" so "build bridge" is quite correct :) 07:19
raydiak heh funny...hi Woodi 07:21
07:22 denis_boyun joined 07:23 darutoko joined 07:40 rurban_ joined 07:43 FROGGS joined
mathw wonders what happens if you rakudobrew build moat 07:45
raydiak fixed it before I hit enter, but iirc it tells you something like "building most is NYI; well volunteered!" 07:48
07:49 gtodd joined
TimToady most appropriate 07:50
07:51 vti joined
raydiak I volunteer to dig the moat (given transportation, room, and board) 07:51
07:56 blackbolt joined
TimToady
.oO(you'll never get rich, by digging a ditch...)
 07:56
07:57 atroxaper joined
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)
 07:58
08:02 atroxaper left, adu left
FROGGS o/ 08:03
raydiak in which case, I ought to slip right through those gates like a carbon nanotube :)\
\o
08:03 rindolf joined
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 08:04
raydiak He goes on to say "With people this is impossible, but with God all things are possible." 08:06
08:10 zakharyas joined
TimToady well, most people will take any excuse to be nasty, really :) 08:17
speaking of which, the current implementation of s/// is 08:18
it returns the current value of $_ rather than the success of the substitution 08:21
mathw Hmm. Seems unhelpful. 08:23
08:24 nige left 08:29 nige joined
raydiak finds himself contracted to build a 1-page website for a pole-dancing fitness gym 08:31
Woodi is dreaming about web pages making and "normal" programming be a plain, artisan job for few centuries from now ;) 08:33
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
08:36 pecastro left
Woodi even futuristic to wish such simple information :) 08:36
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
08:38 pecastro joined 08:39 andreoss left
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
err... do bakerise should do that also ?
nwc10 I'm failing to find this in the documentation - what's the Perl 6 way of writing cat?
ie perl -e 'print while <>'
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…»
TimToady m: .say for lines
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…»
nwc10 thanks
08:44 Isp-sec left
dalek kudo/nom: 6b429b9 | TimToady++ | src/core/Str.pm:
set $/ on listy match too
 08:48
08:49 [Sno] joined
raydiak alright enough of my silliness for now, this was one horrific day and I'm thankful it's over...g'night #perl6 08:50
FROGGS gnight raydiak 08:51
08:52 [Tux] joined 08:53 rindolf left
blackbolt can someone help me with substitution. for (@test){ s/\'//:g;}. it remove only first \' 08:56
08:57 brrt joined
TimToady us s:g/\'// instead 08:58
*use
blackbolt ohhh thanks TimToady 08:59
TimToady and you don't need the parens around @test, just spaces 09:00
09:01 KCL left, ptc_p6 joined
mathw that would require shops to have decent IT systems 09:02
(delayed, sorry, forgot I'd scrolled up)
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...
mathw it would help if I had drunk more than a few sips of this delicious hot chocolate already
then I might be marginally more awake
09:06 ptc_p6 left, ptc_p6 joined
dalek ast: bd41b86 | TimToady++ | S05-modifier/ (2 files):
m:g// can now work under smartmatch, returns List
 09:09
kudo/nom: 1eb7fad | TimToady++ | src/core/List.pm:
fake up .from/.to on List for lists of matches
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
Also stay indoors if you're water-soluble
TimToady hopes staying in bed is adequate to stay dry for the next 8 hours or so 09:12
zzz &
mathw Depends how sturdy your house is I suppose. Sleep well! 09:15
09:20 nige left 09:23 tinyblak_ left, tinyblak joined, _4d47 left 09:27 tinyblak left 09:32 fhelmberger joined, andreoss joined, fhelmberger left, fhelmberger joined, woolfy left
dalek ast: 5a1239a | (Elizabeth Mattijsen)++ | S05-modifier/perl5_0.t:
Unfudge passing TODO, TimToady++
 09:33
lizmat afk& 09:34
09:37 robinsmidsrod left 09:39 kurahaupo joined, robinsmidsrod joined 09:43 Rounin joined, woolfy joined 09:49 rurban1 joined 09:53 itz_ joined 09:55 rurban1 left 10:05 aborazmeh left 10:06 mvuets joined
mvuets hello, #perl6 10:09
10:09 aborazmeh joined
mathw hello 10:13
10:14 aborazmeh left 10:17 blackbolt left 10:19 JimmyZ_ left 10:20 aborazmeh joined
andreoss I got "Unhandled exception: const_iX NYI" with huge amount of iterations, but with fewer iterations code runs just fine 10:22
what is not yet implemented in this case? 10:23
10:24 tinyblak joined 10:25 tinyblak left
masak antenoon, #perl6 10:31
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␤␤»
10:32 Alula left
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␤␤»
masak m: say X::ControlFlow::Return.new.message
camelia rakudo-moar 1eb7fa: OUTPUT«Attempt to return outside of any Routine␤»
mvuets lizmat: re paste.scsys.co.uk/451709, it is tricky to reproduce, i filed a ticket with details: #123406 10:37
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=123406
mvuets lizmat: i think hoelzro made a fix last night
yup, hoelzro++ github.com/MoarVM/MoarVM/pull/162 10:38
10:46 Alula joined 10:53 rindolf joined 10:54 tinyblak joined
masak today's autopun: "Those who believe in telekinesis, raise my hand." 10:58
10:59 tinyblak left 11:02 brrt left
tadzik :D 11:03
11:04 Alula left 11:09 Alula joined 11:10 kurahaupo left 11:12 chenryn left, psch joined 11:13 robinsmidsrod left
psch hi #perl6 11:13
11:15 robinsmidsrod joined 11:20 Grrrrrrr joined 11:23 bjz left, chenryn joined 11:26 kaleem left, bjz joined 11:32 anaeem1 left, anaeem1_ joined 11:33 anaeem1_ left 11:35 rindolf left 11:54 pecastro left 11:55 tinyblak joined 11:59 tinyblak left
dalek p: ca209b6 | Carlin++ | 3rdparty/libtommath/ (2 files):
use arc4random on platforms that support it

this merges in patches from MoarVM's libtommath		 12:00
p: 241c631 | (Tobias Leich)++ | 3rdparty/libtommath/ (2 files):
Merge pull request #206 from carbin/arc4random

use arc4random on platforms that support it
12:00 Grrrrrrr left, bcode left 12:02 Alula left, Alula joined 12:09 brrt joined 12:12 Util joined 12:15 eMBee is now known as no-one, no-one is now known as eMBee 12:20 anaeem1_ joined 12:33 myp joined
andreoss map and grep don't work within promises? 12:38
my $p = start { grep { is-prime $_ }, 1..100 }; 12:39
FROGGS m: my $p = start { grep { is-prime $_ }, 1..100 }; say $p
camelia rakudo-moar 1eb7fa: OUTPUT«Promise.new(scheduler => ThreadPoolScheduler.new(initial_threads => 0, max_threads => 16, uncaught_handler => Callable), status => PromiseStatus::Kept)␤»
andreoss say await $p; # empty 12:40
12:42 rurban1 joined 12:43 jack_rabbit left
FROGGS m: my $p = start { grep { is-prime $_ }, 1..100 }; say await $p 12:44
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
FROGGS m: my $p = start { grep { is-prime $_ }, 1..100 }; say (await $p).perl
camelia rakudo-moar 1eb7fa: OUTPUT«().list␤»
brrt is camelia allowed to start threads?
FROGGS it at least looks like
psch m: await start { say "i think so" }
camelia rakudo-moar 1eb7fa: OUTPUT«i think so␤»
FROGGS m: my $p = start { 1..100 }; say await $p 12:45
camelia rakudo-moar 1eb7fa: OUTPUT«1..100␤»
FROGGS m: my $p = start { (^100).map: 'ö' }; say await $p
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
12:45 blackbolt joined
psch m: await start { say grep .is-prime, [1, 2, 3, 4] } 12:45
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
FROGGS m: say await start { (^100).map: 'ö' } 12:46
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
FROGGS j: say await start { (^100).map: 'ö' }
camelia rakudo-jvm 1eb7fa: OUTPUT«Can't open perl script "/home/camelia/jvm-rakudo/eval-client.pl": No such file or directory␤»
FROGGS bah
m: say await start { [~] 1, 2, 3 }
camelia rakudo-moar 1eb7fa: OUTPUT«123␤»
FROGGS m: say await start { (^3).join: '' } 12:47
camelia rakudo-moar 1eb7fa: OUTPUT«012␤»
FROGGS okay, it really looks like map and grep are fishy here
andreoss also 12:48
psch m: say await start { my $x; $x ~= $_ for 1..10 } 12:49
camelia rakudo-moar 1eb7fa: OUTPUT«Nil␤»
psch for goes through MapIter as well i think?
hm, am i actually returning $x there..?
m: say await start { my $x; $x ~= $_ for 1..10; $x }
camelia rakudo-moar 1eb7fa: OUTPUT«12345678910␤»
psch right, nevermind :)
andreoss m: my $p = start { gather for 1..1000 { take $_ if $_.is-prime } };say await $p; 12:50
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
FROGGS m: say await start { do for 1..10 { $_ } }
camelia rakudo-moar 1eb7fa: OUTPUT«1 2 3 4 5 6 7 8 9 10␤»
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 } }
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…»
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]
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
colomon m: say grep .is-prime, 1, 2, 3, 4
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
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'␤»
colomon m: say (1, 2, 3, 4).grep(*.is-prime)
camelia rakudo-moar 1eb7fa: OUTPUT«2 3␤»
12:55 chenryn left
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'␤»
FROGGS ohh, that's actually correct
m: say await start { (^100).map: { 'ö' }; 1 } 12:56
andreoss does "do" mean something special for promises?
camelia rakudo-moar 1eb7fa: OUTPUT«1␤»
12:56 tinyblak joined
FROGGS andreoss: no 12:56
'do' puts a loop in statement context, which means it will produce results
dalek href="https://modules.perl6.org:">modules.perl6.org: 427af07 | moritz++ | web/lib/P6Project.pm:
Try to make UTF-8 handling more robust
 12:57
FROGGS so you can use 'do' to indicate that a while/until/for loop should either produce values or throw them away
m: my $x = 64; say do while $x++ < 85 { .chr } 12:58
camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
FROGGS huh
12:58 rurban left
moritz it returns the last evaluated condition 12:59
m: say do while 0 { 42 }
camelia rakudo-moar 1eb7fa: OUTPUT«0␤»
FROGGS m: my $x = 64; say do while $x++ < 85 { $x.chr }
camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
FROGGS weird
that's not what I expect 13:00
13:00 tinyblak left
FROGGS m: my $x = 64; say do while $x++ < 85 { $x.chr; last } 13:00
camelia rakudo-moar 1eb7fa: OUTPUT«-> (|) { #`(Block|45850752) ... }␤»
moritz and it's not up to spec, iirc
13:04 Woodi left, Woodi joined 13:05 pecastro joined
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" 13:06
i did successfully build a BOOTArray, but i didn't figure out how to hllize that properly either :/ 13:08
hoelzro morning #perl6 13:09
moritz modules.p6c.org/ now running
colomon o/ 13:12
dalek : d70722e | moritz++ | misc/svn-search/ (3 files):
remove outdated svn-search
 13:15
13:17 rindolf joined 13:20 atroxaper joined
dalek : 9c8f365 | moritz++ | util/update-design.perl6.org.sh:
script for generating design.perl6.org HTML (untested)
 13:27
: 461f574 | moritz++ | misc/mothballed/yap6/ (29 files):
remove stuff that has been mothballed for 5 years
 13:28
sergot m:say <a b c d>.splice: 0, 2; 13:32
m: say <a b c d>.splice: 0, 2;
camelia rakudo-moar 1eb7fa: OUTPUT«No such method 'splice' for invocant of type 'Parcel'␤ in block <unit> at /tmp/TEsxO2hsDZ:1␤␤»
sergot do we have splice for arrays?
psch m: say [1, 2, 4].splice: 2, 3
camelia rakudo-moar 1eb7fa: OUTPUT«4␤»
sergot psch++ 13:33
psch m: say @(<a b c>).splice: 0, 2;
camelia rakudo-moar 1eb7fa: OUTPUT«a b␤»
psch sergot: the latter one actually relates, i guess :)
sergot m: say <a b c d>.list.splice: 0, 2; 13:34
camelia rakudo-moar 1eb7fa: OUTPUT«a b␤»
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
this way, at the end of the year, we'll know what to write about :) 13:35
13:36 pecastro left
andreoss how do i declare thread-shared variable? 13:37
for promises
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
andreoss: but it's better to avoid that, and use some reactive pattern instead
(though I'm not an expert on concurrency) 13:40
13:43 kaleem joined, kaleem left 13:44 kaleem joined
colomon tadzik: Typed::Subroutines failed overnight smoke 13:52
tadzik oh noes
I'll check it out, thanks
psch andreoss: jnthn.net/papers/2014-nlpw-reactive.pdf and jnthn.net/papers/2014-yapceu-async.pdf might be of interest to you 13:53
13:56 [Sno]_ joined, tinyblak joined 13:57 [Sno] left, [Sno]_ is now known as [Sno], atroxaper left
TimToady tadzik: if the tests rely on the old List semantics under smartmatch, could be my changes 13:57
13:58 aborazmeh left
nwc10 TimToady: are you still dry? 13:58
tadzik TimToady: the code uses @types ~~ $s.signature.params.map(*.type) 13:59
so, might be
13:59 onebitboy left
tadzik TimToady: what's changed about the list smartwatching? 14:01
erm, smartmatching
14:01 tinyblak left
TimToady nwc10: the heavy rain doesn't start for another 10 minutes or so :) 14:21
List is now a distinguished result type for smartmatching, just like Bool, Match, and Nil 14:22
so you need to cast it to Parcel or Array to get the same effect currently
but that's probably going away eventually too, in favor of a specific match-this-list keyword of some sort 14:23
overloading the base types for this is kinda wrongish 14:24
m: say "foo" ~~ m:g/./ 14:25
camelia rakudo-moar 1eb7fa: OUTPUT«｢f｣␤ ｢o｣␤ ｢o｣␤␤»
TimToady the List changes were to make ^^ work
psch m: "foo" ~~ m:g/./; $/.perl.say 14:26
camelia rakudo-moar 1eb7fa: OUTPUT«(Match.new(orig => "foo", from => 0, to => 1, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo", from => 1, to => 2, ast => Any, list => ().list, hash => EnumMap.new()), Match.new(orig => "foo", from => 2, to => 3, ast => Any, li…»
TimToady it's a list of match
moritz TimToady: I thought the desired solution was to return a list-ish Match object from m:g//
TimToady the other night we put our heads together and decided it was better to make List a result type than to fake up a Match
the boolean semantics of List work out very nicely this way, and you don't have to .list the result to use it 14:27
tadzik colomon: fixed, thanks
andreoss psch: for me it's buggy, i've tried to make my naive solution to be thread-aware, but it seems a with more than two iterations the program segfaults 14:28
psch m: "abc" ~~ m:g/$<foo=.>/; say $()>><foo>
camelia rakudo-moar 1eb7fa: OUTPUT«No such method 'ast' for invocant of type 'List'␤ in block <unit> at /tmp/jGNuDOnNWr:1␤␤»
14:28 novice666 joined
andreoss fpaste.org/158752/18307922/ 14:28
psch m: "abc" ~~ m:g/$<foo=.>/; say @$/>><foo>
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
psch hm
m: "abc" ~~ m:g/$<foo=.>/; say $<foo> # this seems not-so-great
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␤␤»
novice666 m: say eval("abs(-10)"); 14:30
camelia rakudo-moar 1eb7fa: OUTPUT«===SORRY!=== Error while compiling /tmp/nWbC9_1wUV␤Undeclared routine:␤ eval used at line 1␤␤»
TimToady it's EVAL now
FROGGS m: say EVAL("abs(-10)");
camelia rakudo-moar 1eb7fa: OUTPUT«10␤»
novice666 thanks
psch andreoss: i can't really help with concurrency, i just remembered that these two talks by jnthn++ exist 14:31
14:31 smls joined
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
not sure what to put in there instead, though, considering smartmatch for subst and match checks $/ anyway
TimToady m: "abc" ~~ m:g/$<foo=.>/; say $/.map: *.<foo>
camelia rakudo-moar 1eb7fa: OUTPUT«␤»
TimToady m: "abc" ~~ m:g/$<foo=.>/; .say for @$/ 14:33
camelia ( no output )
TimToady oh, it's malformed
m: "abc" ~~ m:g/<foo=.>/; .say for @$/ 14:34
camelia rakudo-moar 1eb7fa: OUTPUT«===SORRY!=== Error while compiling /tmp/gHQf7ZgDTz␤Unable to parse expression in metachar:sym<assert>; couldn't find final '>' ␤at /tmp/gHQf7ZgDTz:1␤------> "abc" ~~ m:g/<foo=.⏏>/; .say for @$/␤ expecting a…»
TimToady m: "abc" ~~ m:g/$<foo=>./; .say for @$/
camelia ( no output )
TimToady m: "abc" ~~ m:g/$<foo>./; .say for @$/
camelia ( no output )
smls TimToady: What about incremental matching, like we could do with which m//g and m//gc in Perl 5? Is there no replacement planned for that?
TimToady m: "abc" ~~ m:g/$<foo>=./; .say for @$/
camelia rakudo-moar 1eb7fa: OUTPUT«｢a｣␤ foo => ｢a｣␤␤｢b｣␤ foo => ｢b｣␤␤｢c｣␤ foo => ｢c｣␤␤»
TimToady there we go
m:g is lazy 14:35
smls you mean it returns a lazy list?
TimToady yes
smls ok
TimToady supposed to, anyway
psch do we need Cats for that? 14:36
TimToady no
smls And will it be possible to teach the GC to free elements of lazy lists that have already been iterated over? :P
TimToady depends on how they're stored
that's one of the things we talked about in Salzburg 14:37
14:37 anaeem1_ left
TimToady Cats might be useful for incremental output of s/// 14:37
psch ah, right
TimToady m: "abc" ~~ m:g/$<foo>=./; say $/»<foo> 14:38
camelia rakudo-moar 1eb7fa: OUTPUT«｢a｣␤ ｢b｣␤ ｢c｣␤␤»
TimToady works if you actually match the pattern
smls Btw when building moar with rakudobrew, I see a bunch of lines like «gcc -c -fPIC -O1 -DNDEBUG -D_REENTRANT -D_FILE_OFFSET_BITS=64 ...» 14:40
Why "-O1"?
TimToady m: "abc" ~~ m:g/$<foo>=./; say $1 14:41
camelia rakudo-moar 1eb7fa: OUTPUT«｢b｣␤ foo => ｢b｣␤␤»
14:41 kaare_ left
TimToady -O2 was getting indigestion on something some number of months ago 14:42
14:42 ptc_p6 left
psch hm, i guess that's ok. having $0, $1 etc reference the different 'instances' of :g Match objects makes sense to me at least 14:42
smls Interesting; I thought -O2 did only "safe" optimizations? 14:43
TimToady some things are not so safe for large interpreters playing with goto 14:44
which may or may not have been the problem at the time 14:45
brrt TimToady - already awake :-) 14:46
TimToady yes, alas 14:47
14:47 tinyblak joined
TimToady jnthn,brrt: btw, when trying (and failing) to fix up s/// last night, I was getting some: Spesh: failed to fix up handlers (-1, 556, 556) 14:51
what does that mean?
14:51 tinyblak left
brrt iirc, handlers are basically pointers into the bytecode, and they are made 'symbolic' during the bytecode->spesh graph transformation, and back to pointers in the spesh graph->bytecode transformation 14:52
it looks to me as if some blocks were lost, but to be sure i'd have to see the spesh graph 14:53
TimToady wonders if something is cheating on s/// in there
course it was in the setting, so things can get weird in any case 14:54
14:54 andreoss left, blackbolt left
TimToady .subst is a bit weird insofar as the proto is not a bare {*} 14:54
probably has something to do with that... 14:55
together with futzing around with OUTER::<$/> 14:59
brrt hmmm
possible, but that would still be an error on moars part imho
does it persist when MVM_SPESH_DISABLE_INLINE=1
TimToady didn't try that yet
I'll have to reproduce my changes, having backed them out in the interests of finishing up m:g// 15:01
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 15:06
TimToady treating blobs and bufs as strings to the extent it makes sense is part of the intended design 15:12
the Stringy role is kind of a down payment on that
you probably won't get grapheme semantics, or even Unicode codepoints 15:13
smls nor would you probably want it there :) 15:14
psch oh, someone fixed the Day 10 -> Day 11 link 15:16
or maybe it does happen automatically?
FROGGS we should be able to get rid of OUTER::<$/> by passing the $/ around, starting from the actions, at least for s/// and friends 15:17
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
subst still sets $/ internally for its replacement, but it's setting a private $/ in the proto 15:19
if I read it aright
15:21 rurban1 left
masak TimToady: github.com/perl6/specs/commit/23ef...3193c847a4 seems to do more than add List to 'OKness' types... was that deliberate? 15:22
TimToady no, but I think it is probably closer to reality wrt PEG rules 15:23
so we can let it stand 15:24
15:26 telex left 15:28 atroxaper joined, telex joined 15:32 c0nspiracy joined 15:34 atroxaper left 15:36 tinyblak joined
masak oki. 15:45
15:45 andreoss joined
masak quick poll: do people prefer `die MyException.new` or `MyException.new.throw` ? 15:45
15:46 kaleem left
tadzik die 15:46
you know what's going on from the first 3 characters 15:47
masak ah, end-weight.
yes, I was thinking the same thing.
15:49 kaare_ joined
masak what's the standard way to throw a control exception? 15:50
`warn MyException.new` ?
smls TimToady: How would we express something like «read($fh, $buf, 2); my ($format) = unpack("n", $buf); if ($format == 42) { ...read...unpack...etc. } else { ... }» in a grammar though? 15:53
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.
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 15:55 [Sno] left, [Sno]_ is now known as [Sno]
masak m: class C is Exception {}; { die C.new; CATCH { default { say .WHAT } } } 15:57
camelia rakudo-moar 1eb7fa: OUTPUT«(C)␤»
masak m: class C is Exception {}; { warn C.new; CONTROL { default { say .WHAT } } } 15:58
camelia rakudo-moar 1eb7fa: OUTPUT«(X::AdHoc)␤»
moritz smls: (..) [ <?{ unpack('n', $0) == 42 }> ... | ...]
masak why do I get the exception itself when I `die` with it, but `X::AdHoc` when I `warn` with it?
moritz (if the grammar is in byte mode, if that exists)
masak should I use another base class for control exceptions?
15:58 kaleem joined
moritz masak: because nobody really worked on warnings yet 15:59
15:59 [Sno]_ joined
masak hm. maybe I'd better stay away from control exceptions, then. 15:59
16:00 [Sno] left, [Sno]_ is now known as [Sno], kaleem left
TimToady smls: nothin' sez we can't have subrules like <unpack('n')> 16:01
they're just methods, after all
smls Interesting; Where can I read more on subrules with parameters? 16:04
16:05 abraxxa left
TimToady well, the calling forms are mentioned at S05:1426 16:09
synopsebot Link: perlcabal.org/syn/S05.html#line_1426
TimToady on the recieving end, any regex/token/rule declaration can take an ordinary signature, so it really doesn't need much additional documentation 16:10
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 16:11
basically, self in such a method is a cursor, and you're expected to return a new cursor saying how much you matched 16:12
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
16:13 denis_boyun left
TimToady would depend on how the method processed it and where it put it 16:13
the new cursor carries all the match info 16:14
as well as any ast you're constructing
the standard grammar contains many examples 16:15
though most of the ast making is hidden off in the actions modules instead of the grammar, at least for rakudo 16:16
smls So, you'd let the unpack subrule implicitly call `make` to make the unpacked value available as part of the ast? 16:19
Also, does <unpack('n', :test(42))> look sensible? 16:21
andreoss why does perl6 run git sometimes? 16:22
smls Or maybe even <ushort(42)> for maximum sugar... 16:23
tadzik andreoss: the installation process might, Configure.pl to be exact 16:25
or 'make spectest'. Perl 6 itself does not run gin :) 16:26
or git
16:27 mr-foobar joined 16:29 lucas_ joined
lucas_ Hello 16:29
Having a easter egg like 42 won't make my code less correct-ish if some funny client pass that exact string as input?
hoelzro what's special about 42, again? 16:30
iirc, 42.WHY is special
m: 42.WHY
camelia ( no output )
hoelzro m: say 42.WHY
camelia rakudo-moar 1eb7fa: OUTPUT«(Any)␤»
hoelzro huh, nvm
masak m: "Life, the Universe, and Everything".WHY
camelia ( no output )
lucas_ its WHAT
masak m: say "Life, the Universe, and Everything".WHY
camelia rakudo-moar 1eb7fa: OUTPUT«42␤»
lucas_ *it's
psch m: say 42.WHAT
camelia rakudo-moar 1eb7fa: OUTPUT«(Int)␤»
andreoss tadzik: i dont think so, i run some script with git repo, and as its seen in tmux title, perl6 run git
*running 16:31
16:31 brrt left
TimToady "some script" can do anything it pleases 16:31
16:32 atroxaper joined
TimToady the perl6 process itself does not run git for anything 16:33
andreoss but it definitely doesn't have system("git") inside
TimToady m: system()
camelia rakudo-moar 1eb7fa: OUTPUT«===SORRY!=== Error while compiling /tmp/BU7M_5gCvo␤Undeclared routine:␤ system used at line 1␤␤»
TimToady there is no 'system' in Perl 6
there are lots of other ways to run a subprocess 16:34
including using a module that does so
16:34 zakharyas left
TimToady doing stuff with panda, for instance 16:35
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? 16:36
so it'd look like «use Grammar::Binary; grammar BMP does Grammar::Binary { ... <unpack(...)> ... }» 16:37
16:37 atroxaper left 16:39 FROGGS left
hoelzro smls: I had a similar thought 16:39
andreoss my bad, it was bash who ran git, nevermind
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... 16:40
Oh, nevermind, I'm making the confusion again about WHAT vs. WHY... I've never used WHY... 16:43
psch .WHY is documentation 16:44
lucas_ psch: Like Pod? Oh, thanks!
TimToady and 42 is cool with pretending to be a string 16:45
hoelzro m: #|(marine) sub yellow {}; say &yellow.WHY
camelia rakudo-moar 1eb7fa: OUTPUT«marine␤»
psch i was about to look up how it works... :) 16:47
hoelzro++
lucas_ hoelzro: :D
TimToady m: #| WHY NOT!?!␤ sub philosopher {}; say &philosopher.WHY
camelia rakudo-moar 1eb7fa: OUTPUT«WHY NOT!?!␤»
psch m: #|[cut me some slack] sub genius {}; say &genius.WHY 16:48
camelia rakudo-moar 1eb7fa: OUTPUT«cut me some slack␤»
psch back to blog-writing
raydiak worries for TimToady who, at twice my age, seems to get half as much sleep at night 16:49
TimToady only last night, and I plan to take a geezer nap later 16:50
raydiak worries somewhat less :)
mvuets #| is even better in combination with 'multi sub MAIN' → paste.scsys.co.uk/451928
raydiak oh, yeah! mvuets++ ...I should add that to some of my p6 work, kinda forgot about it 16:52
hoelzro mvuets++ # p⬢ post? 16:53
mvuets hoelzro: on my backlog already (-:
hoelzro =)
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?
TimToady yup
method arguments too 16:56
in either case, as long as they're declared multi
lucas_ TimToady: nice, thank you
16:57 sporus joined
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 :) 16:57
TimToady remembers being delighted when he realized that multi would work for MAIN 16:58
smls d
[sorry, wrong window] 16:59
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
mvuets lucas_: check out the rock-paper-scissors example here raw.githubusercontent.com/perl6/bo...spatch.pod 17:02
lucas_ mvuets: Thank you! And nice example the 'multi sub MAIN'... 17:03
mvuets → home 17:04
17:05 mvuets left, atroxaper joined, fhelmberger left 17:06 Isp-sec joined, c0nspiracy left
psch btw, i've sent my CLA a few days back. who do i talk to to confirm arrival? 17:08
17:12 lucas_ left, kurahaupo joined 17:15 treehug88 joined
[Coke] jnthn: minor nit: "It's job" -> "Its job" 17:15
17:16 sporus left
[Coke] tadzi++ #hashtag the future. 17:21
jnthn++ #moarvm blog post 17:24
sergot Does perl6/doc is responsible for generating doc.perl6.org?
what 17:25
does is?
I wanted to write somthing else
is perl6/doc ~ anyway
17:26 raiph joined
smls sergot: Yes, it's generated fro that git repo 17:26
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 17:27
sergot smls++ thanks 17:29
moritz psch: I think pmichaud++ has access to the CLA lists 17:31
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 127.0.0.1:3000 17:32
sergot smls++ thank you!
psch moritz: thanks, i've been contacted already
sergot smls: I want to add syntax highlighting :)
smls cool :) 17:33
17:35 anaeem1_ joined
smls btw, is this Pod I wrote really malformatted: github.com/perl6/doc/blob/master/l...pe/Bag.pod ...or is github just bad at parsing Pod? 17:35
(Pod::To::HTML converts it just fine) 17:36
PerlJam I think github is using Perl 5 POD parsing
jnthn evening, #perl6 17:38
[Coke] pmichaud & I both have access to the CLA lists. 17:39
vendethiel o/, #perl6 17:41
17:42 Rounin left 17:46 atroxaper left, atroxaper joined, atroxaper left
[Coke] (bad pod) you can nest =begin/=end in p6? 17:46
m: 3.say 17:47
camelia rakudo-moar 1eb7fa: OUTPUT«3␤»
[Coke] m: =begin pod␤=begin code␤=end code␤=end pod␤say "alive";
camelia rakudo-moar 1eb7fa: OUTPUT«alive␤»
[Coke] m: =begin pod␤=begin code␤=end pod␤=end code␤say "alive";
camelia rakudo-moar 1eb7fa: OUTPUT«===SORRY!=== Error while compiling /tmp/yf_YiCSuRL␤'=begin' not terminated by matching '=end pod'␤at /tmp/yf_YiCSuRL:5␤------> say "alive";⏏<EOL>␤ expecting any of:␤ whitespace␤ horizontal…»
[Coke] that error message is slightly LTA. Still, nifty. 17:48
17:50 ptc_p6 joined 17:52 kurahaupo1 joined
jnthn [Coke]: Fixed the nit - thank you! 17:56
I hope the post wasn't too boring/weird...I was tireder than expected when I came to work on it. 17:57
dalek osystem: c528917 | moritz++ | server/updatelist.pl:
[server] reduce dependencies
 17:58
osystem: 20fa4a7 | moritz++ | server/updatelist.pl:
[server] make output location configurable
moritz jnthn: I liked it
17:59 kurahaupo1 left
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
jnthn OK. I guess it's beefy enough that one core busy on that won't cause a big issue... 18:00
moritz right
andreoss how do i limit amount of threads perl6 spawns?
jnthn andreoss: Set up $*SCHEDULER to a ThreadPoolScheduler with the max threads configured as you like it.
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 :) 18:03
dalek osystem: f8b11ba | moritz++ | server/updatelist.pl:
[server] avoid temporary file
 18:06
psch perl6advent.wordpress.com/?p=2560&a...eview=true feedback much apreciated
andreoss jnthn: thanks a lot! 18:08
tadzik moritz++ 18:10
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
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).
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
tadzik moritz: awesome 18:12
jnthn moritz: What port will it run on? Can we do it in a normal HTTP one nowadays?
(Sorry if you already did that and I missed it)
18:12 KCL_ joined
jnthn Just thinking of the folks behind firewall nastiness. 18:13
moritz jnthn: sure, will be port 80 18:14
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++
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
jnthn In 2014, I'd hope so :) 18:16
andreoss jnthn: why it runs extremply slow and segfaults with out specifing max_threads?
is it trying to do some special scheduling than? 18:19
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. :/ 18:24
The slow is because the scheduler isn't especially smart yet, and it mostly worries about avoding deadlock rather than optimizing throughput. 18:26
For the first issue, Rakudo on JVM can offer a decent bit more stability at the moment.
It's also less fun to develop on though, due to high startup time. 18:28
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
(and p6 generated)
BreakDancer could be employed :)
moritz tadzik: do you have any plans to make it more dynamic?
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␤»
moritz tadzik: if not, I might do a quick p5 rewrite to generate static files
and save me the trouble of deploying something dynamic
tadzik moritz: it's untouched for about two years, I think it's likely to stay that way for a while :
:)
andreoss jnthn: i've marked all optimizations which prevent segfaulting paste.fedoraproject.org/158853/41832260
naive solution can be found in perl6-examples prob027 18:31
18:31 mvuets joined
jnthn andreoss: Thanks; it's nice to have examples that expose problems. 18:32
Well, it's nicer not to have problems, of course, but... :) 18:33
18:33 FROGGS joined 18:36 bcode joined 18:41 bcode left 18:50 ptc_p6 left
dalek osystem/static: 724a0f1 | moritz++ | server/ (18 files):
Generate JSON files statically

also generate an errors.json file with error messages per URL		 18:53
moritz shortstats for the commit above: 18 files changed, 34 insertions(+), 652 deletions(-)
ok, most of the deletions were for automatically generated boilerplate; still feels nice :-)
FROGGS yeah :o) 18:55
I also love to delete code
jnthn yeah, but I don't really like regressins :) 18:56
oh wait...mischannel
lol
FROGGS: Stop being in all the discussions about patches removing code when I'm tired and easily confused :P
FROGGS jnthn: wait... you are about to blug, no? 18:57
jnthn: :P
moritz FROGGS: do you want to move cpandatatesters to the new machine(s)?
jnthn FROGGS: I did my advent post last night?
FROGGS jnthn: just distract me with a blog
ohh
psch fwiw, i've scheduled my post about 45 minutes ago and was asking for feedback... :P
FROGGS moritz: I think so... because I am not that happy with the performance of my server
moritz FROGGS: what's the bottleneck? CPU? 18:58
FROGGS jnthn: see, I am still confused by your 'Ja' answer :o)
moritz: probably... it is a virtual server
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
Stage parse : 28.386
FROGGS ohh nice!
moritz on hack.p6c.org for raudo-moar
FROGGS my box takes about 70s me thinks 19:00
19:02 ab5tract joined
moritz FROGGS: so, first step would be an account for you, I think :-) 19:02
FROGGS :o) 19:03
that'd be sweet
19:04 bcode joined 19:08 rindolf left 19:09 FROGGS_ joined
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 19:09
including dinner& 19:10
19:13 FROGGS left
ab5tract psch: is a "longname" a formal thing or a typo? :) 19:14
psch ab5tract: it's in STD.pm6 i've been told, and i think it's also in Perl6/Grammar.nqp 19:16
so i'd consider it formal :)
<name> {} [ <?before ':' <+alpha+[\< \[ \« ]>> <colonpair> ]* # that's a longname
FROGGS_ a longname is a thing, yeah 19:17
"infix:<&&>" is a longname for example
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. 19:19
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...
ab5tract psch: in the case of it being formal, a small word or link to it might be cool! 19:28
org.perl6.nqp.runtime.BootJavaInterop
the line with that on it is sad :(
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? :)
ab5tract (missing a comma there)
19:31 [Sno] left
ab5tract well, could do, could do ;) 19:31
19:31 atroxaper joined
japhb jnthn: Did you see my example from last night about 'lastcall; nextsame' not seeming to do the right thing? 19:31
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␤»
ab5tract honestly i think it could do with more inline code samples
19:32 grettis joined
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!) 19:33
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
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."
japhb: So you throw away the rest of the candidate list and then do nextsame, which needs the candidate list? :)
psch ab5tract: the problem with inline code samples is that it's much code, and hard to condense :/
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
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.) 19:37
19:38 atroxaper left
jnthn Ah 19:40
I'd be tempted just to fully qualify the call :P
smls ab5tract: For Java snippets, [code language="java"] ... [/code] should work on wordpress.com for syntax highlighting 19:42
It doesn't support Perl 6 though 19:43
japhb jnthn: As in: multi method truncated-to(|c) { self.Date::truncated-to(|c) } ? 19:44
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
japhb What is the right way to jump dispatchers that way?
smls (using these convenience scripts for the first and last step: github.com/smls/undef/tree/master/scripts)
ab5tract .hug smls
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 :) 19:46
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)) 19:48
bartolin hi, #perl6 19:49
ab5tract smls: in my opinion, you are a hero. we need a better way :/
psch ab5tract: that's the point though, with my changes you can just call "$crc.update($myint)" :)
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.
tadzik psch: show it as example, definitely 19:50
I read the article, didn't see any example at the end and thought "oh, so this is just a plan then"
psch well, yes, i probably should
i mean, the "that's the point" kind of made it clear to myself as well :D
ab5tract psch: yes, but surely there is something more fun than a CRC in the default Java SDK? :)
psch tadzik++ ab5tract++ # great feedback 19:52
ab5tract: i'll have to look, i hardly know the JDK
bartolin FROGGS_: I pushed some changes for rakudo PR 336. to me those changes look like improvements ;-) 19:53
moritz tadzik: could you please test panda with ecosystem-api.p6c.org/ as the API URL? 19:54
19:54 Sqirrel joined
FROGGS_ bartolin: will take a look soonish :o) 19:54
bartolin ++FROGGS_
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>␤»
smls psch: ^^ is there maybe a «.list» missing there?
for 'Hello, Java'.encode('UTF-8').list { .say } 19:57
m: for 'Hello, Java'.encode('UTF-8').list { .say }
camelia rakudo-moar 1eb7fa: OUTPUT«72␤101␤108␤108␤111␤44␤32␤74␤97␤118␤97␤»
psch smls: actually, it's complicated
smls: the example that jnthn posted doesn't work anymore nowadays, for two reason
marshaling of arrays back to perl6 is broken
and the method descriptor doesn't exist 19:58
i supposed the latter happened because a [ got lost somewhere
...which means i should probably correct the example
tadzik moritz: sure
psch anyway, the method that would work has the desc "method/update/([I)V"
and that takes an int[]
...iirc 19:59
i'd have to look it up, but yes, the example has to be corrected in any case
20:01 rurban joined
moritz also there is now ecosystem-api.p6c.org/errors.json for a list of errors while fetching / processing META.info files 20:03
tadzik Bytecode validation error at offset 4, instruction 2: register operand index 54 out of range 0..15 20:05
that's perfect
jnthn what the...
psch fwiw, there's now a heavily condensed version of the fallback method in the draft 20:07
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
20:08 kurahaupo left
tadzik jnthn: how do I get more details out of it, provided I can reproduce it? 20:08
20:08 pecastro joined
jnthn tadzik: Well, --ll-exception may give a stack trace that tells us where it comes from... 20:09
What are you doing, ooc? 20:10
20:11 darutoko left
tadzik 'panda install .' 20:11
well, now it worked >_<
jnthn ^Hgrmbl
I hope this isn't going to be The Next Heisenbug...
psch righto, time to see if i can find something fun and interesting in the jdk 20:13
20:13 andreoss left
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
tadzik nope 20:14
FROGGS_ no
moritz tadzik: it's a name-based virtual host, so that's the bare minimum
FROGGS_ it does http 1.0 without that header
tadzik $s = IO::Socket::INET.new(:host<ecosystem-api.p6c.org>, :port(80));
$s.send("GET /projects.json HTTP/1.0\n\n");
FROGGS_ but it is easily tweakable, no?
tadzik it's just this
yep
FROGGS_ give it a whirl then :D
dalek : 201d297 | moritz++ | misc/dalek- (2 files):
Remove dalek polling config and instruction

it has not been running for several weeks, and nobody has complained so far.
Also it seems that the poller was never adapted to github URL changes, so it was pretty much broken		 20:16
20:18 raiph left
jnthn psch: "no method multiple dispatch (MMD)" 20:18
psch: I had it as meaning "multi-mmethod dispatch" 20:19
uh, without the spello :)
psch oh
that's probably because i had "multiple dispatch" from somewhere else
i'll fix that too
20:20 bjz left
jnthn psch: "except we are supplying a longname () consisting of the module name and the colonpair" 20:20
I'm guessing something is going on with the empty parens?
20:20 bjz joined
psch no, i think the parens have to go because they're empty :) 20:20
dalek ecs: f6ea401 | duff++ | S99-glossary.pod:
[S99] Add long name, short name, and multiple dispatch
 20:21
moritz my $s = IO::Socket::INET.new(:host<ecosystem-api.p6c.org>, :port(80));
$s.send("GET /projects.json HTTP/1.0\nHost: ecosystem-api.p6c.org\n\n");
say $s.recv;
PerlJam psch: I added some of those term to our glossary in case you want to link to them.
moritz that seems to work
psch PerlJam: thanks, will do
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
ab5tract tadzik: OS X? 20:24
tadzik ab5tract: nah, linux 20:25
now it worked
moritz: yeah, now it work
s*
moritz: do you plan any downtime, or can I commit changes to panda? 20:27
moritz tadzik: no downtimes planned
tadzik okay
dalek nda: 9fe5154 | tadzik++ | lib/Panda/Ecosystem.pm:
Switch ecosytem API host to p6c.org
 20:29
tadzik please test
FROGGS_ why not just ecosystem.p6c.org? 20:31
(I don't like 'API' in any name tbh)
moritz FROGGS_: because it's not an HTML site 20:32
PerlJam FROGGS_: even if it's ecosystem.p6c.org/api ?
FROGGS_ PerlJam: even then :o)
moritz: okay, I see
moritz FROGGS_: and if there is both modules.perl6.org and ecosystem.perl6.org, the difference is not obvious
FROGGS_ true 20:33
moritz FROGGS_: (p6c.org is only meant as a temporary name until perl6.org DNS switches over)
jnthn psch: I guess "*insert fun example here that works*" is something you know is still to do? :)
psch++ # nice post :) 20:34
psch jnthn: yes, i'm working on inserting the fun, working example 20:35
dalek kudo/nom: 8b85116 | usev6++ | src/core/operators.pm:
Allow chained sequences with closures
kudo/nom: da8b73d | usev6++ | src/core/operators.pm:
Optimization of chained sequences with closures

still passes tests from RT #123329
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=123329
jnthn \o/
dalek kudo/nom: 3fa1bdf | (Tobias Leich)++ | src/core/operators.pm:
Merge pull request #336 from usev6/nom

Allow chained sequences with closures
FROGGS_ bartolin++
dalek nda/ecosystem-api: 26f1164 | moritz++ | bin/redpanda:
Start to fill move HTTP communication into a separate module
 20:38
bartolin oh, great. thanks for reviewing!
20:38 Rounin joined
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?
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 :-)
tadzik it looks like I wanted to have a cpanm for perl6 :P
psch smls: thanks, i'll do
tadzik I'm tempted to leave it like that to see if anyone notices and reports it
this way I'll know if there's a single user of it 20:41
psch *that
20:41 kurahaupo joined
psch i think my test class is good enough - i'm having trouble finding something that'd actually work 20:43
because i realized there's another thing still not quite working yet, that is overloaded constructors 20:44
20:44 Alula left
tadzik good thing to note too :) 20:47
psch oh yes, definitely 20:48
if had that i could pull out an AWT or Swing example as "look what works now", alas...
psch .oO( THERE'S STILL TIME!!1 )
ab5tract :D
psch ooh 20:49
yeah
i don't think i'll try that, it'd take more than two hours 20:50
moritz psch: and don't pressure yourself too much; it's just an advent post :-) 20:53
20:57 [Sno] joined
psch moritz: i'm fine, thanks for the reminder though :) 20:58
20:58 Alula joined 21:01 FROGGS_ is now known as FROGGS
FROGGS m: say $_, $/, $! # TimToady: all of them should be Any? 21:02
camelia rakudo-moar 1eb7fa: OUTPUT«NilNilNil␤»
FROGGS I've got quite some spectest fallout about $/, but need to investigate 21:03
psch alright, i'm done i think 21:06
[Coke] is skipping daily run today. will see about moving to the new box before the next run. 21:07
FROGGS TimToady: S02-types/nil.t explicitly checks that these three are Nil...
psch the fun and working example is rather contrived, but it makes the point i guess 21:08
26 revisions heh 21:09
bartolin FROGGS: what kind of spectest fallout do you see? I've run a spectest two hours ago and saw unexpected results in S05-modifier/overlapping.rakudo.moar (You planned 22 tests but ran 38) plus one passed TODO in S10-packages/precompilation.rakudo.moar 21:12
FROGGS: that spectest was on moar only
FROGGS bartolin: the S05-modifier will go away once you pull
but my local patch for RT #123175 makes regexes very unhappy 21:13
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=123175
bartolin aha, I misunderstood. so it's your local spectest fallout 21:15
FROGGS correct 21:16
m: say "\x[846D]" ~~ m/^<:!L>$/
camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
FROGGS that, for example, comes out true here O.o
m: $/ = Any; say "\x[846D]" ~~ m/^<:!L>$/ 21:17
camelia rakudo-moar 1eb7fa: OUTPUT«False␤»
21:20 konsolebox joined 21:22 telex left
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
21:23 bjz left
moritz naddiseo: you don't 21:23
21:24 telex joined
naddiseo So how would I do that kind of grouping? 21:24
PerlJam naddiseo: why exactly? 21:25
naddiseo a.b.c -> attr( attr('a', 'b'), 'c' )
moritz naddiseo: you simply parse with <term>+ % ',' 21:26
naddiseo: so you get a flat list
naddiseo: and then do the grouping outside the grammar
21:26 kaare_ left, kaare_ joined
naddiseo Okay, I'll give that a try. Thanks. 21:29
FROGGS m: my @a = <a b c d e f g>; my $r = shift @a; for @a -> $x { $r = [$r, $x] }; say $r.perl 21:35
camelia rakudo-moar 3fa1bd: OUTPUT«[[[[[["a", "b"], "c"], "d"], "e"], "f"], "g"]␤»
PerlJam looks like a reduce operation to me :) 21:38
smls m: my @a = <a b c d e f g>; @a.reduce({ [$^a, $^b] }).perl
camelia ( no output )
smls m: my @a = <a b c d e f g>; @a.reduce({ [$^a, $^b] }).perl.say
camelia rakudo-moar 3fa1bd: OUTPUT«[[[[[["a", "b"], "c"], "d"], "e"], "f"], "g"]␤»
PerlJam apparently to smls too :)
smls yep 21:39
jnthn detrain; bbiab 21:41
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"]␤»
21:42 grettis left
FROGGS bartolin: I might have been wrong about this one: 21:43
t/spec/S05-modifier/overlapping.rakudo.moar ................... Dubious, test returned 255 (wstat 65280, 0xff00)
could it be a regression caused by your patch?
because I am not sure anymore what revision was clean :/ 21:44
21:45 kaare_ left 21:46 ab5tract left, dalek2 joined
masak 'night, #perl6 21:46
21:47 d4l3k2_ joined, d4l3k2_ left, d4l3k2_ joined
bartolin FROGGS: I'm also still looking at it. 21:48
21:48 d4l3k2_ left
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 21:52
moritz overlapping.t was last touched on 2014-07-05, so at least we know it's not a test change 21:53
TimToady my patch caused more tests to be run, I think 21:54
bartolin that would explain it:
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.
moritz oh, then it's a bad test file
TimToady people shouldn't really write tests that run only if a previous test succeeds
FROGGS it might not skip correctly... 21:55
moritz the number of tests shouldn't depend on the correctness of the feature being tested
bartolin if I change the plan to 38 all is well (some passed TODO)
moritz gist.github.com/moritz/f37058d81010819692d6 # my TODO for the new server; if you miss any service that you think should be migrated, please let me know
21:56 vti left 21:57 mvuets left 21:59 anaeem1_ left 22:00 ab5tract joined, rurban_ left
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
bartolin is waiting 22:03
(especially for rakudo.jvm and rakudo.parrot to build)
TimToady if you're compiling, you're supposed to be goofing off now
that's what compiling is for 22:04
bartolin ;-) 22:05
22:05 rurban left, rurban joined
dalek ast: e3b109e | (Tobias Leich)++ | S05-modifier/overlapping.t:
fix skip count via #?DOES

The skipped tests skipped the wrong number of tests because some of them are in a loop, and we dont get the loop right either.		 22:06
japhb moritz: You and jnthn discussed a daily perl6-bench run; also, will [Coke]'s daily smoke test be moved there? 22:08
moritz japhb: it will if somebody moves it 22:09
bartolin FROGGS++
moritz japhb: I'm trying to not do the things that regular users can do themselves
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
I'll mebbe have a look into setting up the perl6-bench thingy at the weekend... 22:11
japhb Cool beans, jnthn. 22:17
Understood, moritz. And yes, TODOs for others makes sense.
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 :) 22:19
japhb Sounds like an *excellent* day. 22:22
japhb is just watching the flooding slowly rise outside 22:23
jnthn uh-oh
japhb May flood the garage, but I think the house is safe.
TimToady just so's it's slow...
timotimo you reap what you so. 22:24
japhb
.oO( waitpid() should have been called reap() ... }
moritz root@hack:~# grep -iE 'invalid user|failed password' /var/log/auth.log|wc -l 22:25
TimToady he said grimly...
moritz 2815
it's a sad internet we live in
dalek ast: 1bcc540 | usev6++ | S03-sequence/misc.t:
Unskip passing tests for RT #123329
synopsebot Link: rt.perl.org/rt3//Public/Bug/Displa...?id=123329
moritz that host has only been running for two days
timotimo yeah, that's par for the course
moritz time for sleep 22:26
hoelzro moritz: =(
moritz good * everybody
hoelzro fail2ban to the rescue? 22:27
o/ moritz
bartolin o/
Ulti moritz do you know abut fail2ban?
japhb good * moritz
Ulti oh hoelzro beat me to it :D
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
22:31 FROGGS left 22:32 perturbation joined
TimToady wrote his own one of those 22:32
just keeps its own private set of subrules in the firewall 22:33
22:38 raiph joined 22:41 rurban left
ab5tract psch: the java code snippets are cut short on the RHS 22:50
22:51 smls left
ab5tract the advent theme breaks on code snippets in general, but usually with overflows (which are jarring but readable) 22:51
psch ab5tract: aren't they scrollable for you? 22:52
ab5tract ah! indeed they are :) 22:53
does anyone mind if i polish up the blogroll section?
it still links to planetsix, and is missing 6guts / strangeconsistent / perlgeek / likely more 22:54
*strangelyconsistent
[Coke] ab5tract: go for it
ab5tract oh, and perl6weekly
cool
jnthn Hm. Seems the build got slower (NQP and Rakudo) since last time I built on my desktop. :( 22:56
TimToady my fastest times seem about the same, but it also seems more variable 22:57
jnthn It's from about 60s to 66s on Windows for the full Rakudo build. 22:58
10% is quite a lot. 22:59
psch i also found a bunch of html entities...
ab5tract done :)
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> ); }" 23:02
TimToady why not just loop {} ? 23:04
23:04 baest left
japhb And if you're using Bag.roll, why do 'xx 60'? 23:04
TimToady more spaces
japhb Damn precedence.
23:05 Isp-sec left
ab5tract japhb: that's the weight of the snowfall :) 23:05
japhb Or its inverse. :-) 23:06
ab5tract well, the snow falls upside down, doesn't it? ;)
jnthn .u ❆❄❅
yoleaux U+2744 SNOWFLAKE [So] (❄)
U+2745 TIGHT TRIFOLIATE SNOWFLAKE [So] (❅)
U+2746 HEAVY CHEVRON SNOWFLAKE [So] (❆)
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
TIL... Thanks Unicode :)
ab5tract everyone has infinite loops, but infinite lists are a rarer breed :) 23:09
23:10 baest joined
japhb Exchange u+274[456] with u+262[0123] and you get the sci-fi apocalypse 23:10
psch .u ☡ 23:11
yoleaux U+2621 CAUTION SIGN [So] (☡)
TimToady falling cuneiform would be cool
23:12 atroxaper joined
TimToady 'course, it would mostly look like it was raining fish 23:12
japhb fish bones, at least ...
23:13 rurban joined
TimToady wonders if there are any fish swimming in his street... 23:13
23:13 colomon left, colomon_ joined 23:16 oetiker left
ab5tract the golfed matrix.. now i'm curious 23:17
23:17 atroxaper left
TimToady well, but you have to flip your screen left-to-right without flipping it vertically for that, since the characters are all backwards 23:18
or you need to flip it vertically too, but put each character back right-side up, but backwards 23:20
I think you need the green pill for that...
ab5tract lol
23:21 naddiseo left 23:23 raiph left 23:27 gfldex left 23:34 Rounin left, treehug88 left 23:50 aborazmeh joined 23:58 rurban left

Powered by Cro and the Raku® Programming Language.

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

Thank you!