»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'perl6: say 3;' or rakudo:, niecza:, std:, or /msg p6eval perl6: ... | irclog: irc.perl6.org/ | UTF-8 is our friend! | tinyurl.com/p6contest
Set by moritz_ on 28 December 2010.
dalek d: 4afb07c | larry++ | lib/Test.pm6:
catch up standard with s/done_testing/done/
 00:01
masak \o/ 00:02
00:03 payload joined
colomon masak: so, how many people have turned in solutions to all problems now? 00:04
masak colomon: that's classified until Friday.
colomon masak: is it bigger than a breadbox?
masak :)
TimToady depends on whether you're french 00:06
masak just assume, for the time being, that the competition is cause for everyone to be concerned. I'm not saying that's necessarily the case, just that it's a good assumption. :)
colomon masak: you just want me to keep on obsessing with p4...
TimToady: do the French really have French bread length bread boxes? They seem vaguely impractical at that size... 00:07
00:07 am0c joined
allbery_b afaik the French buy that bread fresh and finish it well before they'd need a breadbox 00:09
colomon that definitely seems like a better plan.
allbery_b you can get it on any street corner in most French cities
TimToady so a French bread box is not very big at all :)
the only question is how many of them fit on the head of a pin, and how that relates to the size of contest entries 00:10
dalek rixel: 24a5eac | diakopter++ | sprixel/ (6 files):
fixed all (heretofore) known regressions
 00:16
00:17 ashleydev left 00:20 flatwhatson_ joined
masak 'night, #perl6. 00:24
00:24 masak left
payload i wanted to port cairo today with zavolaj... but it is broken on current Star. but more likely parrot is broken. jnthn 00:24
00:25 QinGW joined 00:31 woosley joined 00:32 flatwhatson_ left
dalek albot: 9961391 | (Perl 6 Evalbot)++ | evalbot.pl:
pushed sprixel dir 1 level deeper in evalbot
 00:44
diakopter github was teh slow
wolverian stares the file named stdin 00:46
00:48 ashleydev joined
diakopter sigh 00:50
00:50 p6eval joined, ChanServ sets mode: +v p6eval
diakopter there we go. 00:52
perlesque: say(5)
p6eval perlesque: OUTPUT«5␤»
00:52 cdarroch left 01:00 justatheory left
sorear wolverian: it's just the austrian national anthem 01:00
dalek d: 0a2b4bd | larry++ | STD.pm6:
allow >) as infixstopper in regex

Code was probably confusing it with the )> token. sorear++		 01:03
01:06 plainhao left, woosley left 01:07 am0c left
dalek d: 84ca780 | larry++ | STD.pm6:
parse ...^ correctly

oops, was parsing it as ... and then prefix ^		 01:09
sorear wonders if &infix:<ff> has the macro-nature 01:10
TimToady std: 1,2 ...^ -> $a { $a > 10 }
p6eval std 625303c: OUTPUT«===SORRY!===␤Preceding context expects a term, but found infix > instead at /tmp/vhF7nyhV_7 line 1:␤------> 1,2 ...^ -⏏> $a { $a > 10 }␤Parse failed␤FAILED 00:01 120m␤»
sorear A ff B -> _flipflopop(state $, A, B) perhaps
TimToady if it's to follow P5 semantics, yes, it must be macroish 01:11
01:11 rgrau left
TimToady it can probably be written as a stateful ??!! somehow, in fact I think I wrote it out once 01:12
afk &
01:18 woosley joined 01:24 lue joined
lue OHAI! o/ 01:24
colomon \o\ 01:32
diakopter -o_ 01:40
01:43 gbacon left
snarkyboojum _o| 01:44
dalek rixel: 7f319c7 | diakopter++ | sprixel/ (8 files):
misc fixes/cleanup
 01:49
01:50 plobsing joined
jpr5 owowowow my eyes 01:52
diakopter ?
karma sorear 01:55
aloha sorear has karma of 671.
diakopter karma larry
aloha larry has karma of 8.
sorear buubot: karma sorear 01:56
buubot sorear: sorear has karma of 1625
diakopter buubot: karma larry
buubot diakopter: larry has karma of 9
diakopter buubot: karma diakopter 01:57
buubot diakopter: diakopter has karma of 621
diakopter buubot: karma TimToady
buubot diakopter: TimToady has karma of
diakopter karma TimToady
aloha TimToady has karma of 77.
diakopter sigh
flussence maybe aloha just wraps the int, while buubot uses a float and can't display +Inf. 01:58
01:58 cafesofie left
sorear neither bot has a reliable host, so they both only see a subset of deeds 02:01
02:01 noganex_ joined
sorear (I'm biased, I used to run the channel bot before somebody banned it) 02:01
cognominal_ karma gnole 02:02
aloha gnole has karma of 0.
colomon rakudo: say 1..10.map({; $_ => $_; }.perl; 02:03
p6eval rakudo 8681da: OUTPUT«===SORRY!===␤Unable to parse postcircumfix:sym<( )>, couldn't find final ')' at line 22␤»
colomon rakudo: say 1..10.map({; $_ => $_; }).perl;
p6eval rakudo 8681da: OUTPUT«␤»
colomon rakudo: say (1..10).map({; $_ => $_; }).perl; 02:04
p6eval rakudo 8681da: OUTPUT«(1 => 1, 2 => 2, 3 => 3, 4 => 4, 5 => 5, 6 => 6, 7 => 7, 8 => 8, 9 => 9, 10 => 10)␤»
colomon rakudo: say (1..10).map({; $_ => $_; }).grep( *.value ~~ 5..7 ).perl;
p6eval rakudo 8681da: OUTPUT«()␤»
colomon say (*.value ~~ 5..7).WHAT
02:04 noganex left
colomon rakudo: say (*.value ~~ 5..7).WHAT 02:05
p6eval rakudo 8681da: OUTPUT«Bool()␤»
cognominal_ karma gnole is a pun on en.wikipedia.org/wiki/Carmagnole, and gnole, a strong alcohol
aloha gnole is a pun on en.wikipedia.org/wiki/Carmagnole, and gnole, a strong alcohol has karma of 0.
cognominal_ arf 02:06
colomon rakudo: say (4 < *.value < 10 ).WHAT 02:07
p6eval rakudo 8681da: OUTPUT«WhateverCode()␤»
colomon rakudo: say (1..10).map({; $_ => $_; }).grep( 4 < *.value < 8 ).perl;
p6eval rakudo 8681da: OUTPUT«(1 => 1, 2 => 2, 3 => 3, 4 => 4, 5 => 5, 6 => 6, 7 => 7, 8 => 8, 9 => 9, 10 => 10)␤»
colomon bug? 02:08
dalek href="https://modules.perl6.org:">modules.perl6.org: bee61f4 | snarkyboojum++ | web/fame-and-profit.html:
[fame-and-profit.html] Fix the camelia icon
diakopter rakudo: say (1..10).map({; $_ => $_; }).grep({ 4 < $_.value < 8 }).perl; 02:09
p6eval rakudo 8681da: OUTPUT«(5 => 5, 6 => 6, 7 => 7)␤»
diakopter :?
rakudo: say (1..10).map({; $_ => $_; }).grep({ 4 < $_.key < 8 }).perl; 02:10
p6eval rakudo 8681da: OUTPUT«(5 => 5, 6 => 6, 7 => 7)␤»
colomon I think maybe 4 < *.value < 8 is being silently translated to 4 < *.value && *.value < 8, or something like that? 02:11
I don't know if it's a bug or not, but it definitely kind of sucks that the first two ways I thought of writing that code didn't work. :\
flussence rakudo: say (1..10).map({; $_ => $_ }).grep(4 < $^a.value < 8).perl # ? 02:14
p6eval rakudo 8681da: OUTPUT«Null PMC access in find_method('value')␤ in main program body at line 22:/tmp/cSSI1mDUjp␤»
flussence argh
colomon needs parans.
er, braces.
flussence oh right, $^ doesn't do that on its own...
colomon rakudo: say (1..10).map({; $_ => $_ }).grep({4 < $^a.value < 8}).perl 02:15
p6eval rakudo 8681da: OUTPUT«(5 => 5, 6 => 6, 7 => 7)␤»
flussence rakudo: say (1..10).grep(4 < * < 8).perl
p6eval rakudo 8681da: OUTPUT«(1, 2, 3, 4, 5, 6, 7, 8, 9, 10)␤»
dalek href="https://modules.perl6.org:">modules.perl6.org: 513d1b0 | snarkyboojum++ | web/fame-and-profit.html:
Fix link to blurb about Camelia.
colomon rakudo: say (4 < * < 8)(5).perl
p6eval rakudo 8681da: OUTPUT«Bool::True␤»
colomon rakudo: say (4 < * < 8)(100).perl 02:16
p6eval rakudo 8681da: OUTPUT«Bool::True␤»
flussence yeah, I'd expect chained-comparison + Whatever to dwim...
colomon rakudo: say (4 < * < 8)("green").perl
p6eval rakudo 8681da: OUTPUT«Bool::True␤»
colomon I don't really have any idea how to explain that.
unless....
rakudo: say (* < 8)("green").perl
p6eval rakudo 8681da: OUTPUT«Bool::True␤»
colomon rakudo: say (* < 8)(100).perl
p6eval rakudo 8681da: OUTPUT«Bool::False␤»
flussence rakudo: say (Bool::False < 8 && Bool::True < 8).perl 02:17
p6eval rakudo 8681da: OUTPUT«Bool::True␤»
flussence I think it's being interpreted as (4 < *) < 8 02:18
dalek rixel: 13edce2 | diakopter++ | sprixel/sprixel.csproj:
default is Release mode build...
 02:20
02:30 VXZ left, VXZ joined 02:33 VXZ is now known as VX64Z 02:34 VX64Z is now known as VXZ 02:45 molaf left 02:54 justatheory joined 02:55 stkowski left 02:59 molaf joined 03:12 mtk left, justatheory left
payload is there a fast version of .match(*, :g) ? -.- insane 03:23
sorear niecza: my $st = "do99" x 1000; say +[ $st.comb(/\d+/) ]; 03:40
p6eval niecza v1-140-g7905557: OUTPUT«1000␤»
sorear niecza: my $st = "do99" x 10000; say +[ $st.comb(/\d+/) ];
p6eval niecza v1-140-g7905557: OUTPUT«10000␤»
sorear niecza: my $st = "do99" x 50000; say +[ $st.comb(/\d+/) ];
p6eval niecza v1-140-g7905557: OUTPUT«(timeout)»
payload i get a website "blog.fefe.de" and .match( /(.)(html<-[$0]>*/, :g ) 03:44
argh, maybe the $0 makes it slow 03:45
sorear <-[$0]> doesn't do what you think it does
payload ah, no... i use '"' hardcoded
no?
sorear it matches any character except $ or 0
payload whee 03:46
well, '"' hardcoded is still slow 03:47
03:54 cafesofie joined
colomon rakudo: say ~([\+] (1..10)) 03:56
p6eval rakudo 8681da: OUTPUT«1 3 6 10 15 21 28 36 45 55␤»
colomon huh. 03:57
TimToady double useless use of parens :) 03:59
colomon rakudo: say ~[\+] 1..10 04:00
p6eval rakudo 8681da: OUTPUT«1 3 6 10 15 21 28 36 45 55␤»
colomon \o
\o/
payload there is a need for a regression test. doent print anything on current Star: perlgeek.de/blog-en/perl-6/starry-obfu.html
04:06 ashleydev left 04:23 Su-Shee_ joined 04:26 ashleydev joined 04:27 Su-Shee left 04:35 Eevee left
TimToady payload: I believe the semantics of * changed since then, from * + * meaning { $^a + $^a } to meaning { $^a + $^b } 04:35
04:43 wormphlegm joined 04:49 orafu left, orafu joined
payload ah, okay 04:49
04:50 wormphlegm left 05:07 satyavvd joined
lue rakudo: say 0,10, *+* ... 100; 05:09
p6eval rakudo 8681da:
..OUTPUT«(timeout)508013021034055089014402330377061009870159702584041810676501094601771102865704636807502501213930196418031781105142290832040013462690217830903524578057028870922746501493035202415781703908816906324598601023341550165580141026791429604334944370701408733011349031700183631190302…
lue that's... unexpected. afk! \o 05:10
05:22 typepeter joined, payload left 05:42 starcoder left 05:43 dayangkun joined 05:44 starcoder joined 05:47 ashleydev left, dukeleto left, dukeleto joined 05:55 kaare_ joined 06:02 Bzek joined 06:07 cjk101010 joined 06:09 ashleydev joined 06:22 typepeter left 06:31 Bzek_ joined
dalek ecza: 6c69124 | sorear++ | v6/harness:
[v6] Fix order of operations in Alt optimization
 06:32
ecza: 4135121 | sorear++ | v6/ (10 files):
[v6] Mergeback
ecza: 3a6eeb2 | sorear++ | v6/NieczaCompiler.pm6:
[v6] Cache loaded units between compiler runs
06:33 ashleydev left 06:34 Bzek left 06:38 Bzek_ left
diakopter perlesque: my $sw = Diagnostics::StopWatch.StartNew(); my $a = 10_000_000; sub b() { if (--$a) { b() } }; b(); say($sw.Elapsed) 06:46
p6eval perlesque: OUTPUT«Could not allocate 16432 bytes␤Stacktrace:␤␤ at sprixel.FrameBase.Run () <0x000fe>␤ at (wrapper runtime-invoke) <Module>.runtime_invoke_object__this__ (object,intptr,intptr,intptr) <0xffffffff>␤ at (wrapper managed-to-native) System.Reflection.MonoMethod.InternalInvoke
..(System.R…
diakopter bah
perlesque: my $sw = Diagnostics::StopWatch.StartNew(); my $a = 5_000_000; sub b() { if (--$a) { b() } }; b(); say($sw.Elapsed) 06:48
p6eval perlesque: OUTPUT«(timeout)9204037␤»
diakopter weird
it was doing so much better earlier
perlesque: my $sw = Diagnostics::StopWatch.StartNew(); my $a = 2_000_000; sub b() { if (--$a) { b() } }; b(); say($sw.Elapsed) 06:49
p6eval perlesque: OUTPUT«00:00:02.8491494␤»
diakopter frowny frown
06:49 wtw joined 06:51 mberends joined 06:54 cafesofie left, cafesofie joined 07:06 Alias left, Alias joined 07:16 cognominal_ left 07:28 mberends left 07:38 nsh- left, dalek joined, ChanServ sets mode: +v dalek, nsh_ joined 07:39 HarryS left 07:42 Woody4286 left 07:43 araujo left 07:47 dalek left 07:48 Woody4286 joined 07:49 araujo joined 07:51 p6eval left, dalek joined, ChanServ sets mode: +v dalek 07:54 p6eval joined, ChanServ sets mode: +v p6eval 07:58 typepeter joined 08:02 dalek left, Eevee joined 08:04 dalek joined, ChanServ sets mode: +v dalek 08:26 Su-Shee_ is now known as Su-Shee 08:31 typepeter left 08:34 daxim joined 08:37 [particle] joined 08:45 Bzek joined, cognominal joined 08:46 HarryS joined 09:14 noganex_ is now known as noganex, kensanata joined 09:25 dakkar joined 09:31 fhelmberger joined, bacek left 09:43 nadim left 09:44 nadim joined 10:05 woosley left 10:09 QinGW left 10:10 dayangkun left
sorear niecza-compiled-niecza seems to work fine 10:45
but (niecza-compiled-niecza)-compiled-niecza is unusable
it can't even handle my $x; $x 10:46
it thinks $x is undeclared
arnsholt The joys of bootstrapping? 10:49
moritz_ it acts as a harsh test suite with not-so-good diagnostics
sorear aye 10:50
mathw so clearly niecza-compiled-niecza is not behaving exactly the same as it should
recursively compiling yourself is a damned good test though :)
sorear this would be inconceivable without a proper test suite to catch 90% of the compilation problems beforehand
moritz_ time to use spectest suite :-)
you can stub eval_{lives,dies}_ok if you don't have eval yet 10:51
sorear alternatively, I can find some way to diff 7MB of whitespace-free JSON 10:52
moritz_ first find a pretty printer
shouldn't be too hard
then diff the pretty-printed JSON
sorear I've been using one for debugging, but I need to find one that produces deterministic output 10:53
10:53 wtw left
mathw might have to stop and write one 10:53
which would be a pain
sorear ha
moritz_ can't you do a whitespace insensitive diff? 10:55
mathw sorting 10:57
moritz_ ah 10:58
crap
sorear well, also there are spurious differences
moritz_ what's the biggest blocker to using roast?
sorear tuits :) 10:59
moritz_ ok :-)
sorear for idiotic historical reasons, the Perl 5 version of Niecza includes a copy of the full pathname in every position annotation node
moritz_ would an update script + test harness help?
sorear already have both of those
pmurias++ has been tinkering with fudges 11:00
moritz_ so it's just a matter of fudging tests and adding them to the list of passing test files?
11:01 wamba joined
sorear "just" 11:01
moritz_ well, I've done that for rakudo before 11:02
and I might have some tuits tonight (ie in about 6..8 hours) 11:03
dalek href="https://modules.perl6.org:">modules.perl6.org: cefbeea | snarkyboojum++ | web/ (2 files):
Align headings and link them to the homepage
arnsholt @ means does Positional, right? 11:18
sorear maybe
it did at one point, but then we decided my @foo is Bag should work
it's been reslushified 11:19
arnsholt Oh
moritz_ and Bag !~~ Positional?
sorear correct
bags are unordered
like hashes
moritz_ my state is still that @ means "restrict to Positional and default to Array"
shouldn't Bag be % sigiled then? 11:20
arnsholt In S32 Array isn't Positional either, it seems
Array does List, and List does Container 11:21
sorear Container is nonsense
arnsholt Oh, great =) 11:22
moritz_ but List does Positional 11:23
the specs are very vague in that area
11:26 cafesofie left
arnsholt Or, to ask a different question: When $/.ast is an Array, why does for $/.ast simply iterate once, with $_ being the whole array? 11:28
Or, why does @stuff = $/.ast become [[stuff]]? 11:29
moritz_ arnsholt: it probably is not an Array, but an Array in a Scalar
try @stuff = $/.ast.list
iirc is make() specced to be item context, or something
arnsholt Aha 11:30
That might be it. Although $/.ast.WHAT being Array() in that case does seem weird
moritz_ the scalarness is a property of the container, whereas .WHAT returns the type of the contents 11:31
sorear well, consider my @foo = ([1],[2]);
@foo[0].WHAT is Array()
moritz_ rakudo: say [].PARROT
p6eval rakudo 8681da: OUTPUT«ObjectRef->Array␤»
moritz_ rakudo: say (my @a).PARROT
p6eval rakudo 8681da: OUTPUT«Array␤»
arnsholt Ah, I see
moritz_ this whole flatten/non-flatten bit is going to be an FAQ 11:32
and we should make it easy to inspect
arnsholt Yeah, I was about to say pretty much the same thing =)
dalek href="https://modules.perl6.org:">modules.perl6.org: bff1a08 | snarkyboojum++ | web/ (2 files):
Make pages validate and fix link styles
arnsholt (About the FAQ thing)
moritz_ -> lunch
arnsholt But yay. My dead-stupid p3 solution works =D 11:33
sorear note that it works completely differently under the hood in niecza
11:35 DarkWolf84 joined
DarkWolf84 hi :) 11:36
11:37 wtw joined 11:40 hanekomu joined
sorear hello and welcome, DarkWolf84 11:41
DarkWolf84 I can't compile the last rakudo :) 11:45
:(
it freezes my system
It never happened before 11:46
coldhead what is your system?
DarkWolf84 amd64 512m ram
I know the ram is not too much
i'm on archlinux 11:47
coldhead how much swap space do you have?
DarkWolf84 the x86_64 version
about 1G
free says 1004056 11:49
coldhead do you know how far the compile gets? 11:50
DarkWolf84 last time more than 12 hours 11:51
11:51 masak joined
masak hi, #perl6. 11:51
snarkyboojum o/
mathw hi masak
DarkWolf84 hi, masak
:)
masak hi, snarkyboojum. hi DarkWolf84 -- long time, no see. 11:52
sorear I've found a lovely misparse that I can't effectively golf
it seems to be context sensitive in some way 11:53
11:53 satyavvd left
sorear $*DECLARAND = $curstash.{$name} = $declaring; is getting compiled as left-associative in STD but not from the command line 11:53
sorear out 11:54
masak intriguing.
mathw ...run
...fun*
or maybe just run screaming for the hills, and don't come back until you've reformulated Perl 6 using S-expressions which are at least easy to parse correctly 11:55
DarkWolf84 why compiling rakudo takes so much memory
sorear masak: niecza-compiled-niecza seems to work fine, (niecza-compiled-niecza)-compiled-niecza is having... issues
'my $x; $x' -> $x not predeclared
masak DarkWolf84: there are some memory optimizations that require more effort than people have put in so far. we more or less know about them, but they'll require some work. 11:56
sorear: I love the smell of strange loops in the morning.
mathw rakudo: ([+] (16..100)) * 4 12:00
p6eval rakudo 8681da: ( no output )
mathw rakudo: say ([+] (16..100)) * 4
p6eval rakudo 8681da: OUTPUT«19720␤»
snarkyboojum rakudo: say [+] 16..100 * 4; say ([+] 16..100) * 4 12:11
p6eval rakudo 8681da: OUTPUT«80080␤19720␤»
snarkyboojum rakudo: say [+] 16..400 12:12
p6eval rakudo 8681da: OUTPUT«80080␤»
masak huh.
masak thought [+] was a listop 12:13
oh wait.
never mind.
100 * 4 == 400
yes, it does make sense as-is.
wouldn't some Perl 6 implementor like to put in the optimization [+] $A..$B ==> ($B - $A + 1) * ($A + $B) / 2? :) 12:16
(that only holds when $B >= $A, of course) 12:17
rakudo: my $A = 16; my $B = 400; say ($B - $A + 1) * ($A + $B) / 2
p6eval rakudo 8681da: OUTPUT«80080␤»
masak \o/
snarkyboojum _o/ 12:18
masak really looks forward to toying with optimizations in Yapsi 12:21
12:24 pmurias joined 12:36 mtk joined 12:37 helv12 left 12:40 wamba left, helv12 joined 12:41 wamba joined 12:50 pmurias left, zby_home_ joined
arnsholt masak: That optimisation only holds if there's no funny business going on with the .. op and so on, though =) 12:51
12:53 WonTu joined, WonTu left
jnthn Thankfully, you can statically know about multi overrides. :) 12:53
masak yay
moritz_ masak: I wanted to do in rakudo
12:53 bacek joined
moritz_ masak: but it turns out that if I install one [+] multi, the rest isn't autogenerated 12:53
and it needs to check if the boundaries are excluded 12:54
erm, s/ex/in/ 12:55
12:57 DarkWolf84 left
masak of course. 12:58
13:02 dave_miles joined
takadonet morning all 13:07
moritz_ \o 13:09
13:09 DarkWolf84 joined, jedai left
tadzik o/ 13:11
masak takadonet: \o
13:15 pmurias joined
pmurias moritz_: niecza using roast see t/run_spectests 13:16
13:16 takadonet left
moritz_ pmurias: thanks 13:17
sadly my tuits just got destroyed
(related only to real life being mean)
pmurias roast tests are a mixture of easy and (very) hard test cases so it takes a lot of fuding to pass a spec test 13:19
moritz_ I know
13:20 MayDaniel joined 13:24 takadonet joined, takadonet left, takadonet joined
takadonet yo 13:24
colomon \o 13:25
moritz_ it seems that my parenthood is approaching faster than anticipated 13:26
new ETA: tomorrow 13:27
moritz_ panics
masak hugs moritz_
pmurias: do you think it would be worth it to segregate easy and very hard test cases? would such a division be the same across implementations?
moritz_ in general it would not 13:28
but there are some things that are generally considered harder than others
colomon hugs moritz_
moritz_ ie multi harder than single dispatch
moritz_ feels hugged. Thanks masak and colomon
colomon moritz_: how early is this?
jnthn moritz_: *hug* Good luck, hope all goes well! 13:29
moritz_ colomon: original ETA was Feb 1st
13:29 satyavvd joined
moritz_ so it's about 2.5 weeks earlier 13:29
colomon not bad.
mux they should have progress bars on pregnant women bellies 13:30
moritz_ not too dramatic
masak sounds relatively OK.
mux: I doubt that would've helped in this case.
moritz_ mux: they have, kindof. But sometimes there are IO errors near the end
mux heh.
moritz_ they make regular size measurements with supersonic imagining
mathw moritz_: oooh good luck 13:31
mux moritz_: congrats btw
moritz_ thanks everyone
and sorry for distracting from masak's and pmurias' p6 discussion :-)
takadonet mortiz_ looks like I be in the same boat as you in a few months
masak moritz_: np. I didn't feel distracted from any discussion. 13:32
moritz_ takadonet: \o/ congratulations, and my well wishes to the whole (small) family
colomon takadonet: congratulations as well! 13:38
13:38 drbean joined, ashleydev joined
coldhead i hope perl6 doesn't get me pregnant too 13:39
13:40 MayDaniel left
masak coldhead: not if you use Perl 6 in the intended way. 13:44
jnthn
.oO( Always use Safe; )
daxim p3rl.org/Sex 13:46
13:46 coldhead left
masak coldhead: no! don't leave :) 13:49
jnthn "Scared of getting pregnant" would be a new reason for not using Perl 6... 13:50
masak I'll take my chances. I've been a Perl 6 user thus far, and nothing like that has happened to me. 13:53
I may have gained a little weight, but...
jnthn Me too, and it's Perl 6's fault. I turn beer into code and consuming beer in order to produce code for the Perl 6 project made me fatter. :) 13:55
takadonet thanks guys! 13:57
14:03 takadonet left 14:05 hanekomu left
gfldex jnthn: did you try to code while riding a bicycle? 14:05
14:06 hanekomu joined
masak gfldex: I find I can do routine tasks, like passing tests, on a bicycle. deeper design-related coding issues, however, no. 14:07
either the code comes out not very good, or I run off the road into a ditch.
I also find the whole keyboard-and-steering thing complicated.
gfldex if we had to power our computers via the dynamo, we would all be much healthier :) 14:08
PerlJam good * #perl6
masak: They make chorded keyboards that are handle-bar-sized. You just have to get used to them :) 14:09
14:09 satyavvd_ joined
masak PerlJam: ooh 14:09
PerlJam In fact, I think the original models were specifically for bicyclers.
14:11 satyavvd left, satyavvd_ is now known as satyavvd 14:12 plobsing left 14:13 takadonet joined 14:14 ggoebel left
pmurias masak: re segregating test i think it would be hard 14:16
masak: some implemenation might have crufy syntax implemented others might have multimethods 14:17
14:19 ggoebel joined
pmurias masak: what could be helpfull would be to have a convienient fuding tool 14:19
masak pmurias: I think no-one would object if you were to put forward more convenient tooling for fudging tests. 14:21
14:24 ggoebel left 14:29 ggoebel joined
[Coke] doesn't moritz_ have something for auto-fudging and un-fudging for rakudo? 14:34
14:35 plobsing joined
moritz_ just unfudging 14:35
14:35 REPLeffect left
[Coke] ponders the sanity of writing a fudger in perl6. 14:36
pmurias [Coke]: the fudger would be very complex 14:38
* woudln't
* wouldn't
14:38 lamstyle left 14:40 BaggioKwok joined
masak [Coke]: it wouldn't be insane at all to do it in Perl 6. but I would consider featureset to be vastly more important than language choice in this case. 14:40
14:41 satyavvd left 14:43 BaggioKwok left 14:48 cjk101010 left, REPLeffect joined, GinoMan left 14:51 jedai joined
tadzik what's a fudger? 14:57
takadonet tadzik: someone who makes fudge :)
PerlJam A more polite expletive? :)
masak tadzik: a filter that selectively disables tests for implementations that don't support them yet.
tadzik thanks
masak tadzik: so that all implementations can (in theory) use the same spectest suite.
takadonet masak: my explanation is better
masak takadonet: oh, no doubt. 14:58
daxim the interesting question then is: why not let e.g. pugs just fail instead of distorting the outcome?
PerlJam masak: What's this "(in theory)" stuff? It seems to have worked in practice so far :)
moritz_ daxim: because we also use the spectests as regression tests for the implementation 14:59
masak daxim: Pugs is not a good example, because no-one's developing it right now. but for an implementation that's being developed, you'd want to see only the relevant tests.
PerlJam: mostly what pmurias said. he seems not altogether happy with the current fudger.
PerlJam: I don't know what that's about more exactly, but I don't think pmurias would say something like that without reason. 15:00
moritz_ sorear: what's your preferred way of receiving patches for niecza? 15:04
15:06 lamstyle joined
pmurias moritz_: fork it on github 15:09
pyrimidine takadonet: Congrats! I have a 4-month old now :)
Babies everywhere 15:10
(and belated congrats to moritz_ as well!)
takadonet pyrimidine: thanks
PerlJam I have two 60-month olds as of Jan 10 ;-) 15:12
masak twins?
15:12 skangas left
PerlJam aye 15:12
masak nice.
colomon my guy is 28-months.
mathw my dad's a twin
none in my generation of the family though
colomon and just "Super Chief, meet the Little Blue Engine." (He's playing trains.)
mathw his dad was a twin too 15:13
15:14 KyleHa joined
KyleHa Howdy #perl6. 15:14
moritz_ is abusing rakudo's update_passing_test_data.pl for niecza
\o/ it's KyleHa
masak KyleHa! \o/
PerlJam mathw: I didn't know there were any twins in my family until I had two myself. Suddenly I found out that my mother's father was a twin as were a majority of his cousins. (My mother's parents died when I was young and we moved out of state, so I didn't get to really know that part of my family)
KyleHa Rakudo testers: I ran spectest on a system with no icu and discovered two other files that failed because of that (they passed after I installed icu). They are dash-e.t and ignorecase.t. Those don't seem all that Unicode-related. Should they be marked #icu in the spectest.data, or should we do something about the tests in them? 15:16
moritz_ good question 15:17
PerlJam maybe a little of both.
masak the interesting question is *why* they currently depend on icu. 15:18
PerlJam dash-e.t uses \c[LATIN SMALL LETTER A WITH DOT ABOVE] which seems a little hard to come by without some knowledge of unicode 15:19
KyleHa As I recall, it's this test in ignorecase: ok('Δ' ~~ m:i/δ/, ':i with greek chars');
mathw PerlJam: Wht I never understood is why my grandma insisted on dressing Dad and his brother exactly the same whenever possible when they were boys. 15:22
PerlJam maybe split ignorecase.t into two parts? ignorecase-ascii.t and ignorecase-unicode.t ?
mathw PerlJam: in all the photos, even they don't know which is which.
15:23 ashleydev left
masak mathw: they're *twins*. that's the way it works. 15:23
15:23 MayDaniel joined
masak it's an unwritten rule or something. 15:23
mathw maybe that explains why they're so different now
a backlash :)
PerlJam My twins are boy-girl, so it would be really weird to dress them alike ;)
mathw ah fraternal twins 15:24
much more interesting
KyleHa I saw twins on the train this morning. The only difference was one had her pants over her boots and the other hand her pants bunched up above her boots.
15:29 wtw left 15:30 wtw joined, wtw left, skangas joined, skangas is now known as kangas
Su-Shee (I will rephrase in the next article accordingly.. "if vim gets you laid these days - Perl 6 gets you pregnant.." :) 15:31
takadonet wow...
masak heh
15:31 kangas is now known as skangas
masak Perl 6 got me a job. that's enough for now. 15:31
Su-Shee masak: wait and see.. ;) 15:32
15:32 gbacon joined
TimToady rakudo: say [+](1..100) + 0 # is there a ticket for this misparse? 15:32
p6eval rakudo 8681da: OUTPUT«100␤»
takadonet looks at masak 15:33
15:33 orafu left
masak std: say [+](1..100) + 0 15:33
p6eval std 625303c: OUTPUT«ok 00:01 121m␤»
15:34 risou joined
masak TimToady: oh, I see now. 15:34
not sure there's a ticket for that.
TimToady point is, it's supposed to work like a list op
masak there are similar tickets, for sure.
masak submits rakudobug, Justin Case
moritz_ fwiw it got me a job too, but I rejected
masak moritz_: s/rejected/declined/
moritz_ erm, yes :-) 15:35
TimToady or ejected, if it was to be a fighter pilot
colomon Wait, is it a misparse?
15:36 orafu joined
moritz_ niecza: say +"23" 15:36
p6eval niecza v1-143-g3a6eeb2: OUTPUT«Unhandled exception: Unable to resolve method Numeric in class Str␤ at line 0 (ExitRunloop @ 0)␤ at (eval) line 1 (MAIN mainline @ 1)␤ at /home/p6eval/niecza/lib/SAFE.setting line 1070 (SAFE C378_ANON @ 1)␤ at /home/p6eval/niecza/lib/SAFE.setting line 1070 (SAFE
..module-SAFE @ …
colomon rakudo: say (1..100) + 0
p6eval rakudo 8681da: OUTPUT«100␤»
15:37 DarkWolf84 left
TimToady rakudo: say ([+] 1..100) + 0 # should get this 15:37
p6eval rakudo 8681da: OUTPUT«5050␤»
TimToady a listop with immediate () is a function call
15:41 kensanata left 15:42 Kodi joined, plobsing left
TimToady rakudo: say [+](1..100) + 0 # instead of this 15:42
p6eval rakudo 8681da: OUTPUT«100␤»
15:42 MayDaniel left
TimToady anyway, shouldn't need extra parens around [+]() to make it come out right 15:43
masak nod
masak adds this to the ticket
dalek ecza: 6b2a40e | pmurias++ | cl-backend/backend.lisp:
started working on classes in the common lisp backend
 15:45
pmurias hmm, what syntax should embedding common lisp in a Perl 6 program use?
* embedding common lisp code
masak pmurias: 'use Common::Lisp;', perhaps. 15:46
colomon does that mean that [+](1..100) + 0 is different from [+] (1..100) + 0 ? # note extra space between ] & (
TimToady yes 15:47
same as for any other listop
rakudo: say (1+2) + 3; say(1+2) + 3
p6eval rakudo 8681da: OUTPUT«6␤3␤»
colomon makes sense, I guess 15:48
masak colomon: compare it to the path Perl 5 has chosen (explained at the beginning of 'perldoc perlfunc') 15:50
KyleHa rakudo: use Test;is [+](1..10) + 0, ([+] 1..10) +0, "RT 82210"
p6eval rakudo 8681da: OUTPUT«get_attr_str() not implemented in class 'Sub'␤ in main program body at line 1:/tmp/FZvL0tOl9E␤»
KyleHa C'est what? 15:51
masak I'm tempted to submit another rakudobug for that :)
colomon you should, it seems 15:52
masak submits rakudobug
TimToady not that the first [+] is slurping is's args
due to the bug in question
masak troo
TimToady *note
masak KyleHa: try with more parentheses.
colomon btw, has anyone looked at all the weird behavior turned up last night on the list? 15:53
TimToady particularly around the first arg to is
colomon no longer remembers the details, there is a lack of sleep thing going on here.
KyleHa Oh, that makes sense. Masak: I worked around it as "ok … == … " instead.
TimToady that will still misparse if you're not careful
colomon I mean irclog.perlgeek.de/perl6/2011-01-13#i_3181402 and further on 15:54
WhateverClosures misbehaving with respect to common idioms for looking at ranges of numbers.
TimToady we've said that ~~ doesn't autocurry, but maybe it doesn't autocurry only on the right arg 15:57
and chained ops are hard to autocurry correctly, but it will presumably get done someday 15:58
dalek ast: c5240b2 | (Kyle Hasselbacher)++ | S03-metaops/reduce.t:
[reduce] Test for RT 82210: Misparse of metaop with immediate ()
 15:59
ast: 8be7057 | (Kyle Hasselbacher)++ | S32-hash/push.t:
Merge branch 'master' of github.com:perl6/roast
pmurias masak: {use Common::Lisp;(format t "Hello from Common Lisp~%")}? 16:00
16:01 pmurias left
TimToady a chained op has to be considered a function of all its args to curry right: 4 < *.value < 7 really means chain'em(4, &[<], *.value, &[<], 7) or some such 16:01
16:01 plobsing joined, KyleHa left
masak pmurias: right. 16:02
dalek kudo: a350ff4 | kyleha++ | t/spectest.data:
[spectest.data] Mark dash-e and ignorecase as depending on icu (because they do)
 16:03
[Coke] masak: (perl 6 got you a job) pics or it didn't happen. (well, ok, blog.) 16:09
TimToady he didn't say it was a paying job 16:10
masak [Coke]: twitter.com/Edument/status/24580314319822849
[Coke] sadly, cold fusion is the only language that has gotten me a job. (twice, even.)
masak [Coke]: I didn't say it was a Perl 6 job. 16:11
[Coke] masak++
masak: I understand.
(one of the jobs I got with CF ended up being java mostly anyway. ;) 16:12
16:13 plobsing left
masak [Coke]: incidentally, it's the same employer as jnthn's, making that company probably the most p6-dense one in the world. 16:14
hm, unfortunate choice of words. "p6-concentrated" :) 16:15
16:18 Trashlord left 16:19 plobsing joined 16:20 cjk101010 joined 16:27 risou_ joined
colomon masak: well, my company has one employee... ;) 16:27
masak that's a bit of a corner case, methinks. 16:28
16:29 risou left
masak at least you're on good terms with the boss ;) 16:29
16:29 plobsing left
moritz_ 2 perl 6 hackers in one company sounds like a higher density than 1 in one company :-) 16:30
masak depends whether you count per company or per employees in the company, I guess. 16:31
colomon yeah, but as long as your company has more than two employees, the percentage win goes to mine. :) 16:32
16:33 avuserow left, pyrimidine left 16:35 avuserow joined 16:40 kjeldahl joined
masak jnthn: colomon is challenging our role as the most p6-packed company! :) 16:41
jnthn: we need to lay a few people off, I think. :P
16:42 plainhao joined
jnthn masak: :P 16:45
masak: Well, at least we're safe in such a layoff. :P
16:47 pyrimidine joined
jnthn
.oO( We could always teach our boss Perl 6... )
 16:48
16:50 MayDaniel joined
colomon you could always give me the funds to hire a couple more employees... ;) 16:56
16:57 Trashlord joined 17:03 cjk101010 left 17:04 ashleydev joined, daxim left 17:08 mtk left 17:09 mtk joined 17:10 spq1 joined 17:11 wamba left 17:14 plobsing joined 17:16 Vlavv` joined 17:17 Vlavv_ left 17:22 envi left, justatheory joined 17:24 Kodi left 17:26 timbunce joined 17:37 ashleydev left 17:53 hanekomu left 17:56 estrabd_afk is now known as estrabd 17:58 ggoebel left, MayDaniel left, dakkar left 18:00 rgrau joined 18:02 ggoebel joined 18:12 snearch joined 18:13 kanishka joined 18:18 masak left 18:29 jaldhar left 18:31 jferrero joined 18:35 risou_ left 18:41 timbunce left 18:51 kanishka left 18:55 molaf left 19:10 cdarroch joined, cdarroch left, cdarroch joined 19:13 _twitch joined 19:17 vmspb joined 19:34 mberends joined 19:42 orafu left 19:43 ggoebel left 19:44 orafu joined 19:48 ggoebel joined 19:49 alester joined 19:53 hercynium joined 19:54 mtk left, _twitch left
TimToady was it something I said? 19:55
sorear Hi! 19:57
good * #perl6
TimToady likewise 19:58
colomon \o
takadonet sorear: yo 19:59
moritz_ hi sorear. Pull requests waits for you :-) 20:01
20:06 coldhead joined 20:07 KatrinaTheLamia left 20:09 dave_miles left 20:23 mtk joined 20:27 shi joined 20:44 plainhao left 20:49 MayDaniel joined 20:52 masak joined
masak ahoy! \o/ 20:52
moritz_ lolitsmasak! 20:53
tadzik o/
jnthn yayitsmasak!
masak yayitsallyouguys!
jnthn It soll them?
masak I... don't know what that means. 20:54
jnthn Me either. :P
masak halp, jnthn is making English up as he goes along :)
that's cheating. I've worked hard for the level of English I have at this point. you can't go extending it now! :) 20:55
jnthn
.oO( I can imagine people saying that about Perl 6... )
20:55 kfo_ joined
masak yeah, but you didn't predeclare your unorthodox use of the English language :) 20:55
jnthn :P 20:56
moritz_ declares jnthn as a punster
masak I've created a punster! 20:57
TimToady you can think of the 'jnthn' on the front as a kind of meta-sigil that declares the following slang
moritz_ ... to be parsed with .* 20:58
20:58 kfo left
jnthn TimToady: Talking of slangs... Can I sorta think of them as a little piece of grammar that we need to compile into something role-ish immediately after parsing it, and mix into the current cursor? Or do you "enter" / consume a slang in some other explicit way? 20:59
perigrin s'all is a perfectly crumulous word
jnthn I'm guessing it's not quite as simple as being just applied to the current compilation unit or you couldn't export them.
21:01 thepler left 21:02 plobsing left
TimToady mixins are kinda the general approach; we do this writ small when we say Q:qq:c// and such; it just needs to be writ a bit bigger for importing 21:03
21:03 snearch left
jnthn TimToady: *nod* 21:04
TimToady: I want to implement it soonish.
TimToady: So just poking for more concrete. :) 21:05
sorear wonders what the syntax for implementing Cobol.pm6 is
jnthn My use case is related to meta-model stuff and introducing new package declarators, fwiw.
21:06 thepler joined
jnthn EXPORT SYNTACTIC EXTENSION TO CALLER NAMED "COBOL" 21:06
BEGIN GRAMMAR DIVISION
... :)
perigrin START PROGRAMMATIC UNIT OF CODE
gotta make sure it's < 80 columns though 21:07
or else it'll never fit on a card.
sorear jnthn: um, no
jnthn: bootstrapping can't be the first step :D
arnsholt masak: For p4, should our solutions do something when the game is over? 21:08
masak arnsholt: anything beyond base-test is optional/bonus stuff. that said, do something that you yourself believe in. 21:09
arnsholt: I will quite literally judge the code on the bases of small things like that, along with more sweeping criteria.
moritz_: jnthn and I discussed the return value of s/// a bit more. here's the solution we like best: a successful s/// returns the (whole) result string, an unsuccessful one returns Str. that way, it always returns something of type Str, and if you're interested in whether it matched, you can do &defined on it. 21:10
arnsholt masak: Right, that makes sense, in a Perl 6 kind of way =)
Thanks!
sorear masak: will all entrants be published? I'd love to see a cross section of p6 styles
arnsholt Yeah, that might be cool
masak sorear: yes, all submissions (even overridden ones) will be published. that's part of the contest, so to speak. 21:11
21:13 MayDaniel left
masak which means people will have the same basis as I have for judging a winner. I hope the increased transparency won't create any problems. :) 21:14
that's also why I ask people to send in exceptional solutions. makes it easier to pick a winner. :P
21:16 mberends left 21:28 plobsing joined 21:29 mtk left 21:30 masak left 21:37 fhelmberger left
Tene Hmm... so I could just send in anything at all, and he'd publish it with his solutions? 21:41
>.> 21:42
sjohnson <.< 21:46
[Coke] &.! 21:48
21:54 Patterner left 21:56 Psyche^ joined, Psyche^ is now known as Patterner 22:03 aindilis left
jnthn Tene: Mwahaha. :) 22:06
Tene obviously I need to write some TimToady / GVR slashfic to submit. 22:07
huf do it 22:09
colomon wait, overridden solutions will be published?!?!!! Oh noez! 22:16
22:25 colomon left 22:26 shi left 22:35 am0c joined 22:37 Flibble3657 joined 22:39 alester left 22:41 Flibble3657 left 22:43 colomon joined 22:44 pmurias joined
pmurias sorear: hi 22:44
22:44 shi joined
pmurias why do you want to implement cobol? 22:44
Khisanth because there is still a lot of cobol code around? :p 22:54
22:55 plobsing left
pmurias has never seen a cobol program 22:55
22:59 kaare_ left, Rodgey joined, Rodgey left
flussence they usually look VERY LOUD AND SERIOUS AND ONLY HAVE ONE PUNCTUATION SYMBOL. 22:59
23:00 plobsing joined
flussence by that metric I suppose they're the exact opposite of p6... 23:01
23:10 kjeldahl left 23:13 plobsing left 23:14 zby_home_ left, pmurias left 23:22 hercynium left 23:25 estrabd is now known as estrabd_afc, estrabd_afc is now known as estrabd_amf
mdxi cobol isn't as horrid as most people (who have also never read any) make it out to be. 23:37
it's just very verbose and unimaginative 23:38
a 5-10 page cobol program can be rewritten as a 10-15 line RPG program. not that that means anything to anyone, either :) 23:39
23:56 plobsing joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!