»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!
Set by sorear on 4 February 2011.
quietfanatic that sounds like a syntax error though 00:02
don't know why
nebuchadnezzar arf, I just must not use multi
r: pastebin.com/raw.php?i=eJ7gF8Bc 00:03
p6eval rakudo 0eea68: OUTPUT«===SORRY!===␤Confused␤at /tmp/IoeSuTjAVl:1␤»
quietfanatic ah,,I can see why using multis would make disambiguation difficult
Yeah, rakudo's not parsing that for some reason
00:03 Khisanth joined
nebuchadnezzar ok, my mistake, role should not declare multi if we want to override them 00:04
geekosaur that should be throwing a different error; I believe that situation is illegal (but not a syntax error)
quietfanatic Can "Confused" mean something that's not syntax?
nebuchadnezzar perl book: 7.2.1: Otherwise, if the class has a method of the same name that is not declared as a multi, then the method in the class alone­as usual­will take precedence. 00:05
ok, so according to this, my case is a bug ;-) 00:06
geekosaur Confused may mean rakudo thinks it has a syntax error, when it;s actually a semantic error + a parsing bug in rakudo
nebuchadnezzar Cannot have a multi candidate for 'boo' when an only method is also in the package 'Boo'
geekosaur hmmmm
quietfanatic But that stlil shouldn't be a parsing issue 00:07
nebuchadnezzar geekosaur: even a working example make it confused
geekosaur n: pastebin.com/raw.php?i=eJ7gF8Bc
p6eval niecza v19-15-g051783d: OUTPUT«===SORRY!===␤␤Confused at /tmp/0pdnIHBeJQ line 1:␤------> http⏏://pastebin.com/raw.php?i=eJ7gF8Bc␤␤Undeclared routine:␤ 'http' used at line 1␤␤Parse failed␤␤»
geekosaur oh, that's not supported by niecza, I was hoping it was something the bot did for it :/
quietfanatic Oh, maybe that's Rakudo's problem too. 00:08
I know evalbot does github gists, but I don't know about pastebins
nebuchadnezzar hmm 00:09
I'll try gist
quietfanatic r: role boo1 { multi method boo () { say "Boo1"; } }; role boo2 { multi method boo () { say "Boo2"; } }; class Boo does boo1 does boo2 { method boo () { say "Boo"; } }; my $booable = Boo.new; $booable.boo
p6eval rakudo 0eea68: OUTPUT«===SORRY!===␤Cannot have a multi candidate for 'boo' when an only method is also in the package 'Boo'␤»
quietfanatic there you go
r: pastebin.com/L2p6aGwG 00:10
p6eval rakudo 0eea68: OUTPUT«===SORRY!===␤Confused␤at /tmp/C_MUEJsFM7:1␤»
quietfanatic nope
(tried it without the raw.php part)
nebuchadnezzar r: gist.github.com/3268320
p6eval rakudo 0eea68: OUTPUT«===SORRY!===␤Cannot have a multi candidate for 'boo' when an only method is also in the package 'Boo'␤»
quietfanatic yeah
nebuchadnezzar ok, either the perl book needs some update, either there is a bug 00:11
it only support parsing gist 00:12
quietfanatic how old is the book?
nebuchadnezzar 2012.05.23
quietfanatic oh 00:13
nebuchadnezzar github.com/perl6/book/downloads
quietfanatic that's pretty new. The spec shouldn't have changed that recently
nebuchadnezzar my rakudo is 0.1~2012.04.1-2 00:14
so older than the edition of that book
quietfanatic oh, then if anything, it'd have changed the other direction :) 00:15
nebuchadnezzar so it's not working on my old rakudo and on the new one 00:16
hmm
n: gist.github.com/3268320
p6eval niecza v19-15-g051783d: OUTPUT«Boo2␤»
nebuchadnezzar ok, this should have been "Boo"
00:18 sudokode joined
nebuchadnezzar earlier today (well in fact yesterday), masak was happy with a fun bug I found trying examples in the book, hope this one is a bug to make him happier ;-) 00:19
karma: nebuchadnezzar
erf
00:20 tokuhiro_ joined
[Coke] karma nebuchadnezzar 00:22
aloha nebuchadnezzar has karma of 0.
quietfanatic nebuchadnezzar++
00:24 fridim_ left
nebuchadnezzar hmm, the masak ++ was not taken... ;-) 00:25
night perl6 00:36
quietfanatic night 00:37
00:41 aindilis joined 00:57 benabik joined 01:00 cognominal left 01:04 jeffreykegler joined 01:11 jeffreykegler left 01:21 benabik left
sorear nebuchadnezzar: does the book say p6eval supports pastebin? 01:21
I can confirm the code is gist-only 01:22
01:24 benabik joined 01:29 vinian joined 01:41 sivoais left 01:42 sivoais joined 01:47 Chillance left 01:59 benabik left, benabik joined
quietfanatic I think he was referencing the book for role+multi behavior, not p6eval 02:02
02:16 HarryS joined 02:23 daniel-s__ is now known as daniel-s 02:37 Psyche^ joined 02:41 Patterner left, Psyche^ is now known as Patterner 02:49 fgomez left 02:53 orafu left, orafu joined 03:03 fgomez joined 03:06 sftp left 03:31 skids left 03:42 sivoais left 03:43 colomon left 03:44 colomon joined 03:45 sivoais joined 03:55 adu joined
diakopter masak: tell phenny I think my brain is running in a simulation in your brain. 03:57
04:04 tokuhiro_ left
quietfanatic diakopter: No, I'm pretty sure it's my brain. 04:08
diakopter um.
swap/masak/phenny/
quietfanatic lol 04:09
I didn't see that
diakopter yeah me neither
yeah no; yeah, no
yeah, no; no, yeah
phenny: tell masak caught you backlogging! 04:14
phenny diakopter: I'll pass that on when masak is around.
quietfanatic wait, but you didn't catch him backlogging if phenny tells him 04:16
diakopter d'oh. 04:18
quietfanatic: btw, I can get you 3 quintillion twitter followers *yesterday* 04:19
04:22 HarryS left, HarryS joined 04:24 crab2313 joined 04:26 sivoais left, sivoais joined 04:38 ingy left, ingy joined 04:46 sivoais left 04:48 fgomez left, sivoais joined 04:55 jdrab joined 05:01 Khisanth left
moritz \o 05:07
crab2313 o/ 05:08
05:09 crab2313 left 05:14 Khisanth joined 05:20 birdwindupbird joined
diakopter argh; the mars lander news site is ... down 05:24
benabik nasa.gov/ntv for a video feed 05:27
diakopter augh. dns on the iphone is failing for that domain. 05:28
stupid verizon.
05:28 cognominal joined 05:30 cognominal left
benabik "Touchdown confirmed. We are safe on Mars." 05:34
(Spoiler alert?) 05:35
diakopter hopefully it didn't land in a crevasse or on top of a boulder
benabik It was guided down, so it's supposed to be on a nice flat surface. 05:37
First picture back: 64x64 of a wheels on Mars
diakopter I'm not seeing it on www.nasa.gov/mission_pages/mars/ima..._mars.html
benabik Saw it on the JPL monitors on the live feed. Probably hasn't been sent to the web team yet. :-D 05:38
diakopter well, at least the little green men are smart enough to stay out of sight for now 05:42
hopefully nasa sent a chimp to Mars in secret. I mean, what's one more chimp with the tens of thousands killed in drug and surgery testing 05:43
benabik heh. Web page still doesn't have it, but twitter does: pic.twitter.com/894ouNJt 05:44
moritz \o/ marsfeed on #perl6 05:46
diakopter hopefully they also sent a nuke, to protect nasa's intellectual property and advanced technology lest it fall into enemy hands
05:47 thou left
benabik Well, it is powered by a RTG 05:48
(which means nuclear, for those not up on the lingo)
@MarsCuriosity: No photo or it didn't happen? Well lookee here, I'm casting a shadow on the ground in Mars' Gale crater #MSL t.co/cj1zFJty 05:49
sorear What would be the point of sending a chimp to mars? 05:50
benabik Animal cruelty?
I find it somewhat depressing that these photos are appearing on Twitter faster than on nasa.gov
sorear That's not a point
diakopter to test whether the life support systems would work for that long 05:51
05:51 sivoais left, sivoais joined
diakopter and keep the animal in some kind of stasis for months, presumably 05:52
flexing its muscles for it..
sorear MSL is not part of the manned Mars programme 05:53
moritz OH HAI MARS. HUMANITY SENT ME AND I AM NUKULAR!!! 05:54
diakopter sorear: I'm sorry it wasn't obvious I was initially joking but then playing along 05:55
05:56 arnsholt_ is now known as arnsholt
adu mars? 05:57
the planet?
05:57 jeffreykegler joined
benabik Mystery solved. moritz is a robot. Specifically, the Curiosity Rover. :-D 05:57
05:59 jeffreykegler left 06:01 xdbr joined
moritz benabik: actually the clone that was left behind on earth 06:01
TCP doesn't work out well between Mars and freenode
benabik Yeah, the ping times via the DSN are ridiculous. 06:02
tadzik good morning #perl6 06:03
moritz \o tadzik
tadzik . o O ( TCP over hailing frequencies )
benabik Has to be more efficient than RFC1149 06:05
tadzik I don't even have to look it up to know what it is:) 06:06
benabik IPoAC: high latency, but high theoretical bandwidth 06:07
sorear DSN doesn't have great bandwidth either, the noise floor is just too high 06:08
apparently the MSL has a 10W transmitter driver 06:09
i'm actually suprised by this
benabik MSL primarily just has to communicate to orbit... There are three orbiters that are doing repeater duty,
06:11 nodmonkey left, sivoais left 06:12 sivoais joined 06:14 SamuraiJack joined 06:19 adu left
sorear diakopter: sorry 06:19
06:22 xinming left 06:23 leprevost left 06:27 fgomez joined 06:35 cognominal joined 06:36 sivoais left, sivoais joined 06:37 cognominal left 06:45 wtw joined 06:46 xinming joined 06:49 kaleem joined 07:10 brrt joined 07:16 GlitchMr joined 07:24 adu joined
arnsholt Bah. It's annoying that make doesn't support several wildcards 07:32
07:33 cognominal joined
moritz it doesn't? 07:34
07:34 cognominal left, cognominal joined
moritz # Run a single test 07:34
t/*.t t/*/*.t t/*/*/*.t: all spectest_checkout @$(HARNESS_WITH_FUDGE) --verbosity=1 $@
that's from rakudo's Makefile
arnsholt There is that 07:36
But my rule wants to be "sv-train-o%-p%.conll: ..." (which isn't called wildcards, I recall now)
Suffix rule? Something like that
moritz arnsholt: write your own make replacement :- 07:41
)
with regex support in rule 07:42
s
m/sv-train-o(\w+)-p(\w+).conll/: # use $0 and $1 here
arnsholt Yeah. I've thought about it more than once =)
But more seriously, I've considered trying redo 07:43
moritz and while you are at it, please include proper support for commands that produce more than one output file :-)
07:43 xinming left
arnsholt Yeah, that one's annoying as well 07:45
Although "file1 file2:\n\tcommand" gets close enough for me
07:47 GlitchMr left 07:48 xinming joined 07:49 PacoAir left 07:52 fhelmberger joined 08:00 GlitchMr joined 08:04 hoelzro|away is now known as hoelzro
jnthn morning, #perl6 08:09
moritz mrnng, jnthn 08:11
tadzik hello hello 08:15
08:21 nodmonkey joined 08:22 GlitchMr left
sorear hello hello jnthn moritz & tadzik 08:30
masak morning, #Perl6 08:34
phenny masak: 04:14Z <diakopter> tell masak caught you backlogging!
masak how deep :) 08:35
sorear masak! 08:39
masak \o/
sorear fff. I accidentally let MSL preempt trip planning. hopefully I still have enough time
masak so, I hear we put some robot on Mars. that's nice. 08:40
what kind of software does it run?
moritz do US citizens need a visa for visiting .de?
sorear moritz: no, that was the first thing I checked actually :)
moritz that's wise, because it's what takes the longest time 08:41
sorear well, we do if the trip is longer than 90 days or we intend to become employed there
I also already got my passport and made room sharing plans with masak 08:42
I think the most important thing I'm missing now is plane tickets
08:42 GlitchMr joined
moritz sorear: are you 18 years or older? 08:43
sorear moritz: yes
moritz good
then you're legally of age in Germany, which makes stuff much easier
mhasch Is there an alternative for feather.perl6.nl:3000 using some standard port, 80, say? 08:44
moritz mhasch: do you need that because of firewalling?
mhasch yep, (though I could set up an ssh tunnel) 08:45
sorear moritz: apparently Liz wants me in Luxembourg, and mberends has been saying ... something I haven't quite fully understood ... about visiting .uk, .nl, and/or masak in .se
moritz sorear: or you could visit me after YAPC::EU, it's just 200km from Frankfurt :-) 08:46
(iirc mberends has a house in .nl, but currently does much of his work in .uk, so that would make sense)
08:47 mucker joined, mucker left
sorear it's only like a week before the PRS, this may be a bit too close 08:47
08:48 mucker joined, xinming left
mhasch moritz: Erlangen is nice 08:48
sorear mm, Erlangen
I should correct my low level of knowlege about that someday 08:49
"I've heard of the Erlangen Program, that's something right?" 08:50
08:50 xinming joined
moritz it's a town with 100k inhabitants, so not too big 08:50
just special because it happens to be my home town :-)
and has hosted the German Perl Workshop in 2008 and 2012
sorear sounds pretty big by EU standards :D 08:51
moritz sorear: you do have some knowledge updating to do :-)
mhasch did not know there was an EU standard for Perl workshops :-)
moritz r: say 230.5 / 31.6 # population density of DE/USA 08:53
p6eval rakudo 0eea68: OUTPUT«7.294304␤»
jnthn has a talk accepted at YAPC::Asia! o/
tadzik \o/
brrt completely unrelated to anything here, but did anyone else notice javascript developers effectively all started programming in CPS? 08:54
(and call it 'asynchronous') 08:55
moritz did :-)
tadzik :>
sorear my home town has 2300k.
brrt funny, eh? :-)
tadzik mine's 1,802,398, 3,326/km2 08:56
brrt my city has about has about 200k
2300k is alot
masak brrt: hyperboleandahalf.blogspot.se/2010/...thing.html
tadzik 3,054,678 counting "Metropolitan area", * that means
08:56 adu left
jnthn heh, where I grew up had 3k :) 08:56
masak where I grew up had 0.2k 08:57
sorear in the US, 200k is a medium town. we have a strange way of doing things
brrt lol
sorear I'm not certain Frankfurt would be called a city
jnthn masak: Ö RLY?
sorear (in colloquial usage, that is. legally a city is anything with a city charter, and you could probably make a city with 1 inhabitant if you lived out in the countryside and were really bored.) 08:58
masak :P
08:59 fgomez left
masak sorear: in .cn, 1000k seems to be a smallish town. 08:59
moritz I've asked a Russian friend of mine where she came from. She answered "a small town you've never heard of". 09:01
turned out to have 1.5 Mio inhabitants or so
and indeed I've never heard of it before :-)
moritz can't solve the majority of recaptha challenges 09:03
tadzik moritz: I like that blog post alot 09:04
brrt masak: thanks for that link 09:05
bonsaikitten sorear: in luxemburg there's a city called Rumelingen/Rumelange that still has city charter with about, err, 5k inhabitants iirc?
sometimes those things are quite absurd
moritz tadzik: did you mean masak?
tadzik moritz: oh, yes, pardon :) 09:06
masak: ^
sorear bonsaikitten: those names are de/fr? 09:08
sorear looks up "Mio". interesting but unsuprising. 09:09
bonsaikitten sorear: yes, and the first one should have a diaresis on the u
moritz sorear: million 09:10
sorear I was talking money with an American acquaintance not long ago and the figure of '3 MM $' came up. my response, "WTF? I think you mistyped that. Also, double prefixes have been discouraged since the 60s, you're supposed to use T$ now" 09:12
masak tadzik: I like it alot too. 09:15
sorear I am amused that alot has made it here
masak alot makes it here ;) 09:16
tadzik it's funny when someone really makes a typo, and everyone assumes that it's a deliberate, evil error
moritz you just have to earn a reputation for typo-ing a lot :-)
sorear tadzik: apparently MM=million is a legit US thing, insane as it is
tadzik sorear: hm, I've never seen "MM" before 09:17
masak neither
sorear also apparently I'm the only person in my generation who remembers kilomegacycles and micromicrofarads
tadzik and speaking of typos, I meant "alot" :P
sorear too much time spent with my grandfather :P
moritz used Mm (mega meters) in one physics test, and was called out for it
apparently you aren't supposed to prefix m with anything higher than k :( 09:18
masak shoulda used attoparsecs :P
or beard seconds.
09:19 fgomez joined
sorear moritz: huh, I've never heard of people having a problem with megametres. 09:21
bonsaikitten attoparsec <-- superb unit 09:22
moritz iirc that's about 20cm, no?
(or similar order of magnitude)
masak no, about 3 cm.
tadzik www.wolframalpha.com/input/?i=attoparsec 09:23
moritz ok, just one order of magnitude off :-)
that's acceptable for a theoretician :-)
masak and a beard second is somewhere between 5 and 10 nanometers.
so, one Angstrom :)
moritz is that how far a beard growth in a second?
s/far/long/
masak oh, sorry. 100 Angstrom. 09:24
moritz: yes.
jnthn Doesn't that vary from person to person? :)
masak jnthn: yeah, that's why it's 5..10 nm.
moritz average of 1cm/month iirc
jnthn
.oO( nm? My beard is nanotechnology! )
 09:25
tadzik nanobots building beards
masak r: say 0.01 #`(m) / (30 #`(days/month) * 24 #`(hours/day) * 3600 #`(seconds/hour)) 09:26
p6eval rakudo 0eea68: OUTPUT«0.0000000039␤»
masak r: say 3.9e-9
p6eval rakudo 0eea68: OUTPUT«3.9e-09␤»
09:26 TBA joined
TBA morning #p6 09:26
tadzik hello TBA
masak be announced, TBA.
moritz r: say 0.0000000039.Num 09:27
p6eval rakudo 0eea68: OUTPUT«3.9e-09␤»
masak \o/
moritz++ # was thinking how to do that
TBA r: say "f" ~~ "f"|"g"|"h";
p6eval rakudo 0eea68: OUTPUT«True␤»
moritz r: say 0.0000000039.fmt('%f')
TBA r: say "f" ~~ [any]<f g h>;
p6eval rakudo 0eea68: OUTPUT«0.000000␤»
rakudo 0eea68: OUTPUT«False␤»
moritz r: say 0.0000000039.fmt('%e')
p6eval rakudo 0eea68: OUTPUT«3.900000e-09␤»
moritz r: say "f" ~~ [any] <f g h>; 09:28
p6eval rakudo 0eea68: OUTPUT«===SORRY!===␤Confused␤at /tmp/cQfgl8dlK3:1␤»
TBA should they work the same?
moritz TBA: any is not an infix operator
TBA: so you can't reduce it with []
TBA ahh, makes sense, thx!
moritz it actually parsed as [ any() ]
just say
r: say "f" ~~ any <f g h>
p6eval rakudo 0eea68: OUTPUT«True␤»
moritz (and the <...> was a postcircufmix to the array, just like a hash key) 09:29
sorear sleep& 09:32
mucker hi #perl6, what is the official way to embed perl5 in perl6 ... like __asm__ in C ?? 09:34
sorear: sweet dreams :)
moritz mucker: use v5; in an inner scope. But, as you said, "sweet dreams"
(ie nobody supports that yet) 09:35
mucker moritz: I'm working on embedding perl5 into perlito
moritz there's also eval($string, :lang<perl5>)
mucker: \o/
mucker i've succeded, but with the tokens PERL text __PERL__ 09:36
now i need to make it match the spec :)
09:38 nsh joined 09:41 GlitchMr left 09:46 xinming_ joined
masak r: role B1 { multi method x($m) {} }; role B2 { multi method x { say "OH NOES" } }; class C does B1 does B2 { method x { say "OH HAI" } }; C.new.x 09:47
p6eval rakudo 0eea68: OUTPUT«===SORRY!===␤Cannot have a multi candidate for 'x' when an only method is also in the package 'C'␤»
masak well, I think I actually agree with Rakudo there. 09:48
n: role B1 { multi method x($m) {} }; role B2 { multi method x { say "OH NOES" } }; class C does B1 does B2 { method x { say "OH HAI" } }; C.new.x
p6eval niecza v19-15-g051783d: OUTPUT«Potential difficulties:␤ $m is declared but not used at /tmp/7HuFNgPFXj line 1:␤------> role B1 { multi method x(⏏$m) {} }; role B2 { multi method x { say␤␤Unhandled exception: Cannot call x; none of these signatures match:␤ Any, Any…
jnthn Me too
masak nebuchadnezzar: so, no bug AFAICS. 09:49
09:49 xinming left
jnthn Well, possibly bookbug... 09:49
masak there is that. 09:53
moritz ok, suppose I want to compose two roles, one has an only method, one has a multi method, both of the same name 09:54
what do I do?
jnthn You can use "proto" to force multiness on all 09:57
Then the only will just be added to the candidate list.
moritz r: role R1 { method a() { 1 } }; role R2 { multi method a($) { 2 } }; class A does R1 does R2 { proto method a(|$) {*} }; say A.new.a, A.new.a(1) 09:59
p6eval rakudo 0eea68: OUTPUT«Cannot call 'a'; none of these signatures match:␤:(A , , Mu *%_)␤␤ in method a at /tmp/iXk_8pXFBb:1␤ in block at /tmp/iXk_8pXFBb:1␤␤»
moritz r: role R1 { method a() { 1 } }; role R2 { multi method a($) { 2 } }; class A does R1 does R2 { proto method a(|$) {*} }; say A.new.a(1)
p6eval rakudo 0eea68: OUTPUT«2␤»
moritz that seems to lose the only method
I guess that's acceptable
jnthn Hmm...that's surprising. 10:00
moritz not to me
jnthn OK :)
moritz just like role R { method a() { } }; class A { method a() {} }
also loses the method A from R
jnthn True 10:01
masak is there a way to call R1.a from A?
moritz it's only subs where a proto forces all others of the sam name to be multis
masak: self.R1::a
jnthn Should just be a qualified method call
masak then I think I'm fine with it too.
though it's tricky to see all the consequences of this one. 10:02
moritz I kinda think of an only-method as a proto with just one candidate
which is backwards (but not wrongly) from the usual definition, where a proto is an only method that internally redispatches to its candidates
masak aye. 10:03
an only is like a singleton multi.
10:11 crab2313 joined 10:13 nodmonkey left 10:16 vinian left 10:19 spider-mario joined 10:24 crab2313 left 10:34 nodmonkey joined 10:46 xinming_ left 10:47 xinming joined 11:08 tokuhiro_ joined
mhasch can panda be configured to use a HTTP proxy? 11:22
masak never heard of it being able to. what would it take to make it do so? 11:23
tadzik I think there's a guide in the README file, no? 11:24
"If you are behind a proxy, you neet to configure git and configure wget to use the proxy." 11:25
there are links to guides as well
Siddhant++ for adding those
mhasch Ah, should have looked there, but thanks anyway. So it is hard-coded to use wget (which gives us our loop-hole). 11:26
11:27 daxim joined
tadzik yeah :/ 11:27
I don't have any better ideas for it though, patches welcome
daxim please review stackoverflow.com/questions/1182545...6#11826626 for accuracy
tadzik I think I once wrote a .pir script (brr) that could be used instead of wget 11:28
jnthn tadzik: How hard would it be to just use Rakudo's socket support instead? Or ship a very minimal module (like we do with JSON)? 11:29
"just" a matter of tuits?
masak daxim: looks good.
rjbs someone mistakenly contacted perladvent.pm.org about perl6advent; to whom should she be directed? 11:31
moritz me, for example 11:33
[email@hidden.address]
rjbs thanks
masak moritz++ rjbs++
tadzik jnthn: I'd just go for using LWP::Simple I think
moritz tadzik: cosimo has talked about abandoning LWP::Simple in favor of HTTP::Client 11:34
afaict there's no decision yet
but HTTP::Client is worth investigating 11:35
tadzik well, it's mostly the matter of adding one more dependency to panda
no big deal, *but*( 11:36
we already have problems with bootstraping all this stuff :/
especially on windows, where nothing seems to work anyway
so, what do you think about merging offline-bootstrap while we're talking about it?
since neither branch appears to be working on windows
jnthn I'm pretty sure not depending on wget will help on the windows side.
tadzik: You'll be at YAPC?
tadzik jnthn: yes 11:37
heh, true that
jnthn We could try and find some time to hack together on Windows support for Panda.
tadzik sure
I can take my external HDD and you'll show me how to setup a working env there
or maybe an .msi will be enough? 11:38
moritz panda is already shipped with R*, no?
tadzik it is 11:39
mhasch For bootstrapping, I would be content with a route involving perl5. This should work on Windows. 11:40
moritz p5 doesn't help in this case 11:41
mhasch (talking about bootstrapping panda, not perl6 of course) 11:42
moritz yes
the prolem is more to fiddle with the include paths and doing stuff in the right order etc.
11:47 JimmyZ joined
JimmyZ What's the difference between panda and ufo? 11:47
11:48 xinming_ joined
jnthn The panda eats shoots and leaves. The ufo just shoots and leaves. 11:48
mhasch That's neat :-) 11:49
jnthn I suspect the answer I was meant to give is, panda is a module installer, ufo makes makefiles. 11:50
11:51 xinming left
masak jnthn++ # shoots and leaves 11:53
ufo is a developer-only Makefile laziness tool.
panda is a module installer for Perl 6 module users.
nebuchadnezzar masak: thanks for the answer, so 7.2.1 in the perlbook 2012.05.23.a4.pdf is wrong ;-)
masak what does it say? 11:54
11:56 MayDaniel joined
nebuchadnezzar perl book: 7.2.1: Otherwise, if the class has a method of the same name that is not declared as a multi, then the method in the class alone­as usual­will take precedence. 11:57
flussence regarding the wget-in-panda thing: the rakudo makefile also has curl hardcoded, maybe we should change one so they're both using the same 12:01
masak nebuchadnezzar: that... makes a certain amount of sense, too.
masak wonders what S12 and S14 have to say about this
nebuchadnezzar is looking S14 12:02
r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x () { say "Hay!"; }; }; C1.new.x(1); 12:04
p6eval rakudo 0eea68: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤ in method x at /tmp/7Ttor_bzCu:1␤ in block at /tmp/7Ttor_bzCu:1␤␤»
flussence checks what distros tend to use by default, seems to be wget...
nebuchadnezzar hmm, using proto in the class to force multi in roles should works according to S14 12:05
flussence rn: my $len = 6; my $str = ('a'..'z').join; my $substr = rx/ ^ . ** {0..$len} /; say $str ~~ $substr 12:06
p6eval niecza v19-15-g051783d: OUTPUT«#<match from(0) to(6) text(abcdef) pos([].list) named({}.hash)>␤»
..rakudo 0eea68: OUTPUT«===SORRY!===␤Quantifier quantifies nothing at line 2, near " {0..$len}"␤»
mhasch flussence: what does rakudo use curl for? It worked for me w/o curl being present.
cognominal the ufo is more like "Mars attack"?
flussence mhasch: automatic spectesting
moritz no, it uses git for fetching the spectests
flussence moritz: it uses curl for uploading the tar.gz 12:07
mhasch but maybe curl for reporting success.
moritz ah, but only for 'spectest_smolder'
12:07 gongyiliao left 12:09 t0rn left
nebuchadnezzar r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x () { say "Hay!"; }; }; C1.new.x; 12:09
p6eval rakudo 0eea68: OUTPUT«Hay!␤»
nebuchadnezzar r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x () { say "Hay!"; }; }; C1.new.x("Hou");
p6eval rakudo 0eea68: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤ in method x at /tmp/XTEs5ouXCc:1␤ in block at /tmp/XTEs5ouXCc:1␤␤»
12:12 tokuhiro_ left 12:13 tokuhiro_ joined 12:17 tokuhiro_ left 12:21 cognominal left 12:23 cognominal joined
nebuchadnezzar erf, I knew it will happened lh3.googleusercontent.com/-hQwvriY...7/mars.jpg 12:32
12:33 mtk left 12:34 mtk joined
[Coke] (backscroll) I find I see MM a lot in the financial industry, and in every case, we have a big UK presence as well. 12:35
12:37 rafl left 12:38 rafl joined
masak yes, but... *how* is "MM" short for "millions"? 12:39
huf a thousand thousand?
M*M
jnthn a thousand thousand 12:40
huf roman imperialist culture!
isnt pp short for pages though? so maybe it's just "to get a plural, double the abbreviation for the singular"
flussence Metric Million? :)
moritz huf: nono, you're confusing that with vim, where doubling a character applies the operation to the whole line :-) 12:41
huf oh, so that's my opinion then 12:42
thanks :)
arnsholt pp is short for pages, yes. But I think the convention is to double the last page. For example the plural of ms (for manuscript) is mss 12:43
huf yes, so M is for Million, MM is for Millions. works out
arnsholt There's also § for paragraph and §§ for paragraphs, seq for following and seqq for more than one following 12:44
rjbs the OED has an entry for "08:31 <@mst> new server announcement 12:45
08:39 >@mst< rjbs: we name machines after authors
08:40 >@mst< rjbs: and announce new servers by picking an epigraph worthy quote from one of their books
"er, oops
@#$#@ OED.app
The OED has an entry for "milliardaire"
masak r: sub kilo{1e3};sub mega{1e6};sub giga{1e9};sub postfix:<second>($unit){appropriate-unit($unit)};sub appropriate-unit($s){my ($q, $u)=$s,"seconds";for <minutes 60 hours 60 days 24 years 365.24> ->$nu,$d {last if $q/$d <1;$q/=$d;$u=$nu};"$q $u"};say kilo\second;say mega\second;say giga\second 12:46
p6eval rakudo 0eea68: OUTPUT«16.6666666666667 minutes␤11.5740740740741 days␤31.6889554103441 years␤»
masak \o/
r: sub kilo{1e3};sub mega{1e6};sub giga{1e9};sub postfix:<second>($s){my ($q, $u)=$s,"seconds";for <minutes 60 hours 60 days 24 years 365.24> ->$nu,$d {last if $q/$d <1;$q/=$d;$u=$nu};"$q $u"};say kilo\second;say mega\second;say giga\second 12:47
p6eval rakudo 0eea68: OUTPUT«16.6666666666667 minutes␤11.5740740740741 days␤31.6889554103441 years␤»
arnsholt I read a book earlier this summer which used Msec and so on as standard units of time 12:53
Good book too 12:54
12:54 leprevost joined, xinming_ left 12:55 leprevost left
masak arnsholt: "Accelerando"? 12:57
arnsholt A deepness in the sky, actually 12:58
masak ooh
must try that one, then.
arnsholt: and you must try "Accelerando" :)
arnsholt Although Accelerando is a book I'll probably read soon-ish too
Charlie Stross is a recent fave of mine
Though I might read rule 34 and the other one whose name escapes me first 12:59
colomon <3s Deepness in the Sky and Accelerando. 13:01
13:01 mucker left
arnsholt Yeah, Deepness was fantastic 13:01
A couple of the climaxes towards the end gave me more of a rush than most exciting movies =D 13:02
A fire upon the deep was nifty as well
13:02 lorn_ joined
arnsholt My most recent book was The apocalypse codex. The laundry files are pretty cool 13:02
13:03 mucker joined
colomon can never remember which laundry books he's read. I mean, the first one and the James Bond one, sure. And one other? And random short stories? 13:03
13:03 lorn_ is now known as lorn
colomon is not awake enough to talk about himself in the first person. 13:03
arnsholt Well, if you've read the first two and one more, you've read all of them except the one that came out last month, probably 13:04
There are two short ones according to Wikipedia, as well 13:05
colomon Ah. I definitely have not read one in the last month.
arnsholt The best one is the one where he goes into the Laundry's insane asylum
The best clue I can give to whether you've read the third one is that it has a pretty important plot point about Angleton 13:06
colomon oh, right. 13:07
13:07 bluescreen10 joined
arnsholt The latest one is pretty cool too. A few nifty parallels between dominionist christian sects and Cthulhu worksip, among others 13:07
13:11 xinming joined
colomon tries to remember what he has read in the last month. Range of Ghosts (E. Bear), The Lost Hero and Neptune's Son (Riordan), Soldier of Sidon (Wolfe), It Happened On A Train (Barnett), The Devil's Art (Cooper and Nyborg), finished the Fantine section of Les Miserables (Hugo), and is working on A Knight of Ghosts and Shadows (Anderson) 13:13
arnsholt Speaking of Perl 6, are any other sixers considering submitting to the Italian Perl Workshop? 13:14
moritz colomon: I envy you for reading this much
arnsholt I think colomon's days have more hours than mine ^_^
moritz There are only 30 hours a day -- HPMoR 13:15
colomon Well, I did spend the week before last in fishing camp. And I may have gone back a little further than a month. And three of those are kids books. :)
But I've definitely been reading more than usual lately. 13:16
It's also been a particularly good run of books, IMO. 13:17
13:17 atrodo joined
[Coke] I'm pretty sure MM is "mille mille" 13:20
masak oh, that's horrible. 13:21
[Coke] ah, not just financial, but also oil/gas, which would explain it at Enron, also.
masak: "industry standard". <shrug> 13:22
masak: see the similarly fun CPM. 13:25
dalek c: 5922a03 | moritz++ | htmlify.pl:
[htmlify] embed type graph images

requires the very latest Pod::To::HTML
moritz doc.perl6.org/type/Array # for example
masak Wikipedia gives me 502 Bad Gateway. :/ 13:26
moritz++ # nice!
moritz japhb++ did most of the work 13:27
[Coke] moritz++ japhb++ 13:29
13:31 skids joined 13:34 GlitchMr joined 13:37 brrt left 13:38 brrt joined 13:54 TBA left
mhasch whow, almost 34 hours of talks (not counting lightning) already accepted to yapc::eu. at last, including four p6 talks. 14:00
14:01 GlitchMr left 14:10 thou joined
moritz science. It's hard. 14:11
14:12 thou left
moritz has just falsified a peer's conference talk 14:12
mhasch moritz: which one? 14:13
moritz mhasch: something about flat-top gain phase sensitive all-optical amplifiers 14:14
14:14 kaleem left 14:15 fhelmberger left
mhasch moritz: cool. Now there can be two talks. Did he/she like your findings? 14:20
moritz mhasch: I haven't informed the author (and not sure I will, I'll finish/abort my academic career in a month) 14:22
mhasch moritz: What a pity you should leave just when you were onto something. 14:29
moritz mhasch: "onto something" is a bit overstated. It took me about 2 hours (with lots of procrastination in between while the simulations were running) to destroy a PhD-students 2 months of work 14:30
bonsaikitten amusing 14:31
moritz: you should turn that into a career 14:32
mhasch moritz: out of curiousity, does your career change mean doing more or less cool perl stuff?
bonsaikitten when I was involved in P2P stuff I had a feeling that at least half the papers were using cooked data, but I never found the time for a systematic disassembly of things
moritz mhasch: more perl5 stuff
bonsaikitten: well, it only worked because I had very similar simulation already running 14:33
bonsaikitten: not applicable to many papers out there
14:33 kaare_ joined
bonsaikitten moritz: in p2p stuff most things I saw were smelly ... oh you claim higher throughput, but in your simulation upload speed is unbounded? might that not shift things badly, hmm? 14:34
masak well, falsification may hurt and feel harsh, but it's *very* important. moritz++
moritz bonsaikitten: I know that kind of paper 14:35
14:38 GlitchMr joined
[Coke] if you know their thesis is wrong, and you let them get a doctorate, doesn't that make you evil? ;) 14:39
14:39 kaleem joined
mhasch moritz: two similar setups producing antipodal results should at least account for a closer look on both setups. 14:39
14:45 Chillance joined 14:53 GlitchMr left, birdwindupbird left, _sri left, pnu_ left, _fall0ut_ left
mhasch [Coke]: There can be wrong, as in "I rigged my data to support some nonsense", or wrong as in "I started out with a hypothesis that supported my initial data, but new data forced me to revise it". 14:54
14:55 GlitchMr joined, birdwindupbird joined, _sri joined, pnu_ joined, _fall0ut_ joined, shachaf joined, baest joined
GlitchMr perl6 -e'loop {say @*ARGS||"y"}' 14:57
I guess I have fun with this
14:58 sftp joined
mhasch GlitchMr: /usr/bin/yes is faster :-) 14:59
15:07 orafu left, kaare__ joined, thou joined
[Coke] yes. 15:09
15:10 birdwindupbird left
[Coke] ww. 15:10
... but excellent timing.
15:10 kaare_ left
masak yes. 15:11
PerlJam
.oO( #perl6 is full of yes-men? )
masak phenny: en sv "yes-man"? 15:12
phenny masak: "Ja -man" (en to sv, translate.google.com)
masak :P
15:16 kaleem left 15:18 JimmyZ left 15:22 Tedd1^ joined, Tedd1 left
[Coke] finds it slightly nifty that github.com/coke/perl6-roast-data/c...1b1214bafa shows rakudo fixing a bug and that niecza didn't change. 15:29
15:33 atrodo left 15:38 atrodo joined
masak does anyone know what programming language(s) the software of the Curiosity rover is written in? 15:38
brrt java, ada, or fortran, probably :-) 15:39
15:39 GlitchMr left
brrt ... probably not java 15:39
hoelzro ada is present
en.wikipedia.org/wiki/Vxworks
holy crap, Java *is* on there
PerlJam hoelzro: why would this be a surprise?
hoelzro shrugs 15:40
PerlJam now ... if there were some Perl 6 on there .... ;)
huf is there some sh in it?
everything needs a bit of sh fudging
brrt java is famous for coming with statements that it should not be used for mission-critical software 15:41
i.e., nuclear reactors and stuff
[Coke] that just means "you can't blame us if your code explodes"
brrt i'd thought NASA to have been more conservative
masak likes Ovid's old post about how Ada feels less powerful and less type safe than Perl 6 15:42
huf heh.
[Coke] if you are you relying on your own tests, that's less of a concern, no?
hoelzro masak: link, please?
huf that article comes to mind where the author presents ada as a secret project to collapse the soviet union...
15:43 kaare__ left
masak hoelzro: tricky, because use.perl.org is no longer operational. 15:43
hoelzro: I don't know if archive.org will retreive old use.perl posts.
oh, seems it does. 15:44
15:45 kaare_ joined
masak web.archive.org/web/20081027005144/...rnal/37699 15:46
archive++
hoelzro masak: thanks!
mhasch masak: I tried to install Game::Crypt with panda but the test suite failed. 15:47
t/crypt.t .. 1/? Could not find symbol 'X::Adventure::&NoExitThere' in block <anon> at t/crypt.t:120
masak mhasch: yes, because of module precompilation fucking up symbols. 15:49
I complained very loudly about it, and jnthn++ told me to figure out a way to increase the bus number. 15:50
jnthn Well, the other option is to be patient while I figure out how to fix it. :)
Well, figure out why it's wrong probably comes first.
masak no rush but hurry dammit
jnthn :P
masak :P
15:51 thelazydeveloper joined
mhasch I don't know about bus numbers but will be patient for now :-) 15:51
15:52 orafu joined, hoelzro left, hoelzro joined
masak jnthn: I don't think I ever submitted that one, due to a strong similarity with moritz++' ticket. 15:52
jnthn: would a golfing of it and a separate RT ticket help? 15:53
15:53 hoelzro is now known as hoelzro|away
jnthn masak: I suspect it's the same issue moritz++ submitted. Any further analysis/golfing is welcome. 15:54
Comparing contents of packages (which it's easy to do by looking into .WHO) could help.
15:54 crab2313 joined
masak gotcha. 15:54
will make a few experiments after $work. 15:55
[Coke] bus number: the number of people that can be hit by a bus before a team loses the ability to know how to do something. 15:58
for many things on a team this size, it's usually 1.
mhasch Thanks [Coke]++, I was thinking of something much more technical for a moment :-) 16:00
masak in terms of evolutionary bottlenecks, it's still very risky to have a low bus number.
huf with all the code and all the conversations about the code being available online 16:01
i dont think your bus number can really go as low as 1
but it would take a considerable effort for a new team to figure it all out and pick up where you left off, in case of bus
so don't bus.
PerlJam huf: It can go lower than 1 if you're not careful (Pugs was that way for a while) 16:02
huf heh :)
16:02 brrt left
jnthn will try to fly and train instead of bus :) 16:03
huf :)
train yourself to fly
mhasch huf: half baked code can be a dead end no matter how much you put it online. jnthn++: the trick with flying is to miss the ground when you are about to hit it 16:04
huf but that's not a trick to flying 16:05
that's falling.
[Coke] huf douglas adams reference.
huf yes, i think so too 16:06
:)
16:06 imarcusthis left
masak huf: there's a wide gap between "could be picked up" and "someone picks it up". 16:06
huf masak: yep.
[Coke] "Do not listen to what anybody says to you at this point because they are unlikely to say anything helpful.
huf so i suppose bus number is like portability. it's all relative.
16:09 imarcusthis joined
arnsholt masak: programmers.stackexchange.com/quest...e-built-in (via HN) 16:16
masak \o/ 16:17
arnsholt++
dalek kudo/nom: d9c09e8 | (Geoffrey Broadwell)++ | src/core/Main.pm:
First step of val() work: replace grammar CLIVal in hack-val() with direct use of Str.Numeric()
jnthn the nilpotent page on wikipedia is odd :) 16:19
er, wrong window
masak .oO( try viagra ) 16:21
[Coke] I cialis you did there? 16:23
16:23 kaare_ left
jnthn
.oO( Can we raise the quality of the humor? )
 16:24
masak so, Curiosity runs on 2.5..3.5 megalines of C, some of it generated probably by Python, but 1 megaline hand-written.
flussence 1 MMLL? 16:25
masak aaaurghhh
flussence it sounds like a lot, but most of it's going to be recycled from older missions... 16:26
masak aye. 16:28
16:30 kaare_ joined
geekosaur under strong impression there is a lot of custom hand written code in use for this stuff, the core of which goes back many years but which is tweaked for every single mission. 16:32
there is not a lot of actual software design/engineering involved 16:33
16:35 fridim_ joined
PerlJam except where they have to change from english units to metric units ;) 16:39
masak (oops) 16:41
16:45 GlitchMr joined 16:50 Chillance left, Chillance joined 16:53 fgomez left 16:59 wtw left
masak ha ha ha, oh right Perl 5 requires parentheses around all the conditional expressions, how hilarious. *sob* 17:17
moritz except for statement modifiers
masak I turned this one into a statement modifier. incidentally. 17:18
by the way, I feel much more comfortable about 'unless' these days. though I only ever use it in its statement modifier form.
Juerd For some reason, I still prefer "if not" 17:28
I have no idea what that reason could be, though. 17:29
[Coke] if not unless then what? 17:31
masak for me, the determiner isn't so much "will I avoid an extra negation here?", but "is the statement an assertion that is best specified in the positive?" 17:32
17:32 fridim_ left
masak 'next unless <some condition holds>;' 17:32
Juerd I tend to write that as <some condition holds> or next 17:33
masak yeah. I don't do that at all.
Juerd In non-trivial code, I put the "or next" on the next line, indented.
geekosaur that's kinda where I am; I see some sense to the claims against "unless" but at the same time the code often *reads* better if I phrase it that way
masak I know the boolean ops are powerful enough to do that, but I only use 'or' and 'and' for side-effectful expressions, like 'open' or 'system'.
I guess I'm using Damian's rule there of what's the most important thing to emphasize. 17:34
with a conditional return, the 'return' is most important, so I use 'unless' rather than 'or'. 17:35
with an open that might fail, the 'open' is most important, so I use 'or' rather than 'unless'.
I don't expect that argument will sway anyone, but that's how my thinking goes :) 17:36
Juerd masak: I tend not to think in terms of what's most important, but of patterns. 17:38
masak that's a very vague thing to say :)
Juerd To me, "die", "next", and "return" all feel the same.
Once I have "or die ..." in code, it feels wrong to write "die ... if" or even "next ... if". 17:39
So after a while I decided to just go with "or die", "or next" and "or return".
masak ah. interesting. 17:40
Juerd Trying to pick the "most important" of the condition and conditional is strange. They're equally important.
masak that's where I disagree, and I think I have good reasons. 17:41
Juerd The condition is most important because without it, the code would break.
The conditional is most important because without it, the code would break.
huf put them both on the left.
masak the 'open' is what I want to do. the 'die' is off to the side because then I'm already outside of the main flow. that's why 'autodie' is possible for that case; we shouldn't really have to care about 'or die' for open. 17:42
Juerd huf: That's sort of what I do.
huf clearly code needs a 3rd dimension so you can write in a layer above another...
Juerd masak: I do have a lot of "open or next" in my code.
masak in the case of 'next unless', I'm still in control and basically defending my block-or-whatever from failed preconditions using control flow.
Juerd If opening the file didn't work, I usually don't care for that and will just ignore it.
masak Juerd: ok. I've never done that one. feels a bit risky, to be honest. but maybe in your use cases it's the right thing. 17:43
Juerd It is risky if you have nothing else to detect that something's off.
But in these use cases, simply doing a chmod -r to skip files is very useful :) 17:44
I'll often use directories as queues. The completed jobs are moved to another directory. 17:46
masak oh, I do that too occasionally. 17:48
hadn't thought of the chmod -r thing, though.
dalek rl6-roast-data: bd4703a | coke++ | p (2 files):
today
[Coke] seen pmichaud ?
aloha pmichaud was last seen in 15558 days 17 hours ago .
[Coke] seen pmichaud?
aloha pmichaud was last seen in #perl6 3 days ago saying "anyway, I'm afk for a bit".
[Coke] aloha--
aloha [Coke]: Pbbbbtt! 17:49
benabik karma aloha
aloha aloha has karma of -70.
masak aloha-- 17:50
aloha masak: Pbbbbtt!
masak aloha: Pbbbbtt! 17:55
arnsholt Having modules so that A uses B uses A will just cause Rakudo to loop endlessly, right?
masak ideally, no.
does it?
arnsholt I suspect so
Could be something else that's broken in my code
Yeah, it hangs. 17:57
Tried with A and B being just the uses and -e 'use A; say "alive"'
17:59 fgomez joined
masak arnsholt: care to submit a rakudobug? 18:00
moritz maybe the "module A is loaded now" is set *after* A has finished loading 18:01
masak it wasn't always like that, though. I'm pretty sure this used to work. 18:02
18:03 GlitchMr left 18:04 fgomez left
arnsholt masak: Will do 18:04
18:05 adu joined
jnthn Wait, what? 18:07
arnsholt Oh, it's a jnthn =) 18:08
Yeah, it breaks here
jnthn If you need the declarations from A in order to parse B, and you need the declarations from B in order to parse A, how is that gonna work?
TimToady poorly 18:09
arnsholt struct A has members of type struct B which has members of type A (it's Zavolaj stuff, just to make it extra fun)
But yeah, that's a good point
jnthn Within a file such circularities are fine. 18:10
Just stub.
arnsholt It just seems that class A is repr('CStruct') { ... } doesn't quite work
jnthn Hm
That's surprising...it shouldn't be at all REPR-dependent.
benabik Doe the repr need to be defined when it's stubbed or when it's fleshed out?
jnthn When it's stubbed.
That's the point the type object comes into existence. 18:11
arnsholt Yeah, it complains that "The following packages were stubbed but not defined"
And then my classes 18:12
jnthn Did you define them? 18:13
That is, in the same compilation unit?
r: class A is repr('CStruct') { ... }; class A { }
p6eval rakudo d9c09e: ( no output )
jnthn r: class A is repr('CStruct') { ... };
p6eval rakudo d9c09e: OUTPUT«===SORRY!===␤The following packages were stubbed but not defined:␤ A␤at /tmp/wx6sZnnz_K:1␤»
arnsholt r: use NativeCall; class A is repr('CStruct') { ... }; class A { } 18:14
p6eval rakudo d9c09e: OUTPUT«===SORRY!===␤Could not find NativeCall in any of: /home/p6eval/.perl6/lib, /home/p6eval/nom-inst/lib/parrot/4.4.0-devel/languages/perl6/lib␤»
arnsholt rs: use NativeCall; class A is repr('CStruct') { ... }; class A { }
jnthn "star"
arnsholt star: use NativeCall; class A is repr('CStruct') { ... }; class A { }
p6eval star 2012.07: ( no output )
arnsholt Thanks =)
Huh. That's odd
jnthn np :)
arnsholt star: module Lib; use NativeCall; class Lib::A is repr("CStruct") { ... }; class Lib::A { } 18:17
p6eval star 2012.07: OUTPUT«===SORRY!===␤The following packages were stubbed but not defined:␤ Lib::A␤at /tmp/cOtsTnOyyC:1␤»
arnsholt There, relatively golfed
benabik module Lib; class Lib::A ? Wouldn't that produce a class Lib::Lib::A ? 18:18
arnsholt Not sure. "module Lib; class Lib::A {}" let's me instantiate a Lib::A in the using code, at least
jnthn That one does indeed look buggy. 18:19
arnsholt But killing the class Lib:: does fix it
jnthn star: module Lib; use NativeCall; module Lib { class A is repr("CStruct") { ... }; class A { } } 18:20
p6eval star 2012.07: ( no output )
jnthn there's that workaround.
masak heh. "Quantum Mechanics obeys the Law of Conservation of Weirdness" :)
jnthn But the bug deserves a ticket.
18:20 lorn left, lorn joined 18:21 UncleFester6 joined
UncleFester6 perl6: sub foo(Str $s is rw) { $s ~= "C"; say $s }; my Str $s = "AB"; foo($s); foo(Str.new()) 18:22
p6eval rakudo d9c09e: OUTPUT«ABC␤Cannot assign to a non-container␤ in block at src/gen/CORE.setting:11682␤ in sub foo at /tmp/3_iqJdzwkS:1␤ in block at /tmp/3_iqJdzwkS:1␤␤»
..niecza v19-15-g051783d: OUTPUT«ABC␤Unhandled exception: Binding '$s' in 'foo', cannot bind read-only value to is rw parameter␤ at /tmp/yidbCTkxEP line 0 (foo @ 1) ␤ at /tmp/yidbCTkxEP line 1 (mainline @ 8) ␤ at /home/p6eval/niecza/lib/CORE.setting line 3929 (ANON @ 3) ␤ at /home/p6e…
arnsholt Yeah, there's some oddness here. Now I have "module Lib; class A is repr('CStruct') { ... }; class B is repr('CStruct') { ... }; class A {}; class B{};" but say Lib::A.new outputs "A.new()" 18:23
Right. Another module Lib inside 18:26
Nope. Still strips the Lib prefix
PerlJam How does one inspect the execution environment? Like to find out what modules have been declared/defined? or subroutines or classes or whatever? 18:29
jnthn r: use Test; say ::.keys
p6eval rakudo d9c09e: OUTPUT«&isa_ok &ok &dies_ok &nok &lives_ok &is GLOBALish &eval_dies_ok &isnt EXPORT &eval_lives_ok &is_approx $?PACKAGE &is_deeply &todo ::?PACKAGE &done_testing &skip $_ &done $/ &skip_rest $=pod $! &diag !UNIT_MARKER &plan &flunk &pass␤»
jnthn r: use Test; say GLOBAL::.keys
p6eval rakudo d9c09e: OUTPUT«Test␤»
jnthn Like that
18:30 Coleoid_aside joined, Coleoid_aside left
PerlJam ah 18:30
18:30 avarab joined 18:31 BinGOs_ joined, TimToady_ joined
UncleFester6 Just to be clear my second call to foo(Str.new()) was legit and should work ... 18:32
18:33 BinGOs left, sivoais_ joined
jnthn UncleFester6: No, it should not work. 18:34
There's no scalar container to assign it into
Did you want "is copy"?
PerlJam UncleFester6: you didn't allocate storage.
18:35 Ulti_ joined
UncleFester6 OK - I think I see - sorry, bit of an out of date C/Perl perspective I guess 18:36
18:36 HarryS_ joined, HarryS is now known as Guest16167, Celelibi left 18:37 BinGOs_ is now known as BinGOs, BinGOs left, BinGOs joined
UncleFester6 perl6: sub foo(Str $s is rw) { $s ~= "C"; say $s }; my Str $s = "AB"; foo($s); foo(my $t = Str.new()) 18:37
p6eval niecza v19-15-g051783d: OUTPUT«Potential difficulties:␤ $t is declared but not used at /tmp/rCHpZ8YcA8 line 1:␤------> $s }; my Str $s = "AB"; foo($s); foo(my ⏏$t = Str.new())␤␤ABC␤C␤»
..rakudo d9c09e: OUTPUT«ABC␤C␤»
18:38 nsh- joined, jdrab_ joined 18:39 nsh left, Guest16167 left, Ulti left, TimToady left, sivoais left, avar left, jdrab left, nsh- is now known as nsh, nsh left, nsh joined, Celelibi joined 18:40 avarab is now known as avar 18:41 Coleoid left 18:43 daxim left
UncleFester6 perl6: sub foo(Str $s is rw) { $s ~= "C"; say $s }; sub baz { my Str $s = "AB"; return $s }; foo(baz()) 18:46
p6eval niecza v19-15-g051783d: OUTPUT«ABC␤»
..rakudo d9c09e: OUTPUT«Cannot assign to a readonly variable or a value␤ in block at src/gen/CORE.setting:11682␤ in sub foo at /tmp/ckBIrx4IZf:1␤ in block at /tmp/ckBIrx4IZf:1␤␤»
UncleFester6 Anyone else think that niecza is better here? 18:49
jnthn return is meant to strip containers 18:53
perl6: sub foo(Str $s is rw) { $s ~= "C"; say $s }; sub baz { my Str $s = "AB"; return-rw $s }; foo(baz())
p6eval rakudo d9c09e, niecza v19-15-g051783d: OUTPUT«ABC␤»
benabik Does return-rw still exist? 18:54
jnthn Yes
benabik Oh. Bah, you just used it.
jnthn Well, you coulda meant "still exist in the spec" :) 18:55
dinner, bbiab
UncleFester6 Hmm ... I was taking a buffer read from an IO::Socket::INET and passing it to a routine that tried to append to the rw parm and blew up. Why would the Socket people think of using return-rw? 18:57
nebuchadnezzar blup! 19:01
should I open somebug somewhere for something which seems not compliant with S14? 19:02
r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x () { say "Hay!"; }; }; C1.new.x("Hou");
p6eval rakudo d9c09e: OUTPUT«Too many positional parameters passed; got 2 but expected 1␤ in method x at /tmp/snCczrwA9b:1␤ in block at /tmp/snCczrwA9b:1␤␤»
19:03 crab2313 left
masak nebuchadnezzar: that looks right to me. 19:07
nebuchadnezzar: consider giving the proto a wider signature.
sorear good * #perl6
masak sorear! \o/
colomon o/ 19:12
nebuchadnezzar masak: a wider signature? I want to force roles methods to be multi by defining a proto in my class, I make my proto less specific because I do not want to use it, or at a last ressort, like explained in perlcabal.org/syn/S14.html#line_207 19:13
japhb moritz, I say your commit adding the thumbnails to the doc HTML, and needing blead of Pod::To::HTML to process it. Does this also mean linking the thumbs to the full-size SVGs is possible soon? 19:15
phenny japhb: 05 Aug 16:35Z <TimToady> tell japhb I think niecza's design of val() seems to work pretty well, as far as I've used it, though I think there should be method form
japhb TimToady_, duly noted, and makes sense to me too. 19:16
nebuchadnezzar Is there a way to find out where, in the build process, the compilation fails? To figure out what part of the compiler does not like my code? 19:17
masak nebuchadnezzar: as far as I know, the proto must have a signature that is at least the union of the signatures you want to intercept in your multis.
nebuchadnezzar r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x (Any $) { say "Hay!"; }; }; C1.new.x("Hou"); 19:21
p6eval rakudo d9c09e: OUTPUT«Hay!␤»
masak r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x (Any $) { {*} }; }; C1.new.x("Hou");
p6eval rakudo d9c09e: OUTPUT«===SORRY!===␤{*} may only appear in proto at line 2, near " }; }; C1."␤»
masak it *appears* in a proto! >:( 19:22
nebuchadnezzar r: role R1 { method x (Str $m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x (Any $) { say "Hay!"; }; }; C1.new.x("Hou");
p6eval rakudo d9c09e: OUTPUT«Hay!␤»
nebuchadnezzar erf, I must have missed something
SIGSTOMACK 19:24
I wondered why my brain does not work very well... 19:25
19:25 Chillance left
jnthn r: role R1 { method x ($m) { say "MSG: $m"; } }; role R2 { method x (Int $i) {say "Int: $i"; } }; class C1 does R1 does R2 { proto method x (Any $) { * }; }; C1.new.x("Hou"); 19:25
p6eval rakudo d9c09e: OUTPUT«Cannot call ''; none of these signatures match:␤␤ in method x at /tmp/akGSRZ9UAE:1␤ in block at /tmp/akGSRZ9UAE:1␤␤»
dalek p: 3abd848 | (Arne Skjærholt)++ | src/ (2 files):
Fix segfault bug when write-barriering CStruct REPRd objects.

child_objs is now allocated with mem_sys_allocate_zeroed, and a child object is ignored if child_objs[i] is NULL. This fixes a bug where write-barriering would trip over a bogus pointer when write-barriering a struct returned from C and a field was unread before passing it to another C function.		 19:26
19:26 Chillance joined
moritz japhb: where "soon" = "when somebody comes up with a good way to do it, and implements it" 19:27
japhb moritz, Hmmm. I guess that means a bit of design discussion about how we want it to look, followed by one of us doing the needed hacking. And I'm going to be tuit-constrained this week. Can we for now just have an explicit link next to the thumbnail to the larger (and navigable) version? 19:35
moritz japhb: +1 19:36
19:45 nsh left 19:50 nsh joined 20:07 birdwindupbird joined 20:17 PacoAir joined 20:33 SamuraiJack left 20:36 p6rd left 20:37 benabik left 20:40 benabik joined 20:50 kaare_ left
moritz $ panda install Net::IRC::Bot 20:55
Mu()
resolve stage failed for Net::IRC::Bot: Project Net::IRC::Bot not found in the ecosystem
20:55 atrodo left
moritz what's wrong? 20:55
according to modules.perl6.org/ the name is correct 20:56
jnthn moritz: Do other module installs work?
moritz jnthn: yes, installing URI worked
20:57 popl joined
jnthn Oddness 20:57
21:02 nsh left 21:03 dju left, ingy left, ingy joined 21:04 dju joined
dalek c: 1b3ffe4 | moritz++ | htmlify.pl:
link to SVG version of type graph images
 21:11
c: 9596fde | moritz++ | htmlify.pl:
show progress while writing type graph images
21:17 birdwindupbird left 21:18 skids left 21:36 MayDaniel left 21:37 adu left 21:39 mtk left 22:04 sivoais_ left, sivoais_ joined 22:14 colomon left 22:17 sivoais_ is now known as sivoais 22:18 bluescreen10 left 22:20 colomon joined 22:43 sivoais left 22:44 sivoais joined 22:50 Chillance left, Chillance joined
masak '♞, #perl6 22:51
colomon o/
shachaf o⎠
colomon thinks "'black knight, #perl6" is an odd thing to say. ;) 22:52
sorear o/
shachaf I wish Unicode has BLACK versions of all the WHITE characters.
22:53 fgomez joined, sivoais is now known as sivoais_, sivoais_ is now known as sivoais
geekosaur observes that it's a white knight here... 22:54
shachaf is annoyed that his IRC client doesn't support non-BMP characters.
sorear
.oO( My IRC client system mostly uses TTF characters )
 22:55
shachaf What, s/haracter/odepoint/? 22:56
23:00 sivoais left, sivoais joined
popl My font is 12pt so it's more like an irregular white blob here. :D 23:09
[Coke] yawns. 23:12
23:13 popl left, popl joined
ggoebel p5: print 'ok' 23:40
p5eval: print 'ok'
hmm...
masak mentioned "Perl 5 requires parentheses around all the conditional expressions"
what about... print 'ok' if 1==1
benabik eval print 'ok' 23:41
buubot_backup benabik: ok1
ggoebel thx
eval print 'ok' if 1==1 23:42
buubot_backup ggoebel: ok1
benabik he did also say "except for statement modifiers"
eval if 1 == 1 { print 'ok' }
buubot_backup benabik: ERROR: syntax error at (eval 20) line 1, near "if 1"
benabik eval if (1 == 1) { print 'ok' }
buubot_backup benabik: ok1
ggoebel oh right, moritz says that in the next line down...
23:47 skids joined 23:56 mucker 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!