»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
00:00 pierre_ joined, kurahaupo left, obfusk joined
viki context: advent theme :) 00:01
AlexDaniel viki: is there any way to add santa camelia logo? 00:03
00:07 effbiai left 00:10 canopus left
viki Oh, right, forgot about that 00:12
She ruins the sexy! 00:13
AlexDaniel looks great
viki Yeah?
OK then :) 00:14
00:16 espadrine_ left 00:19 canopus joined 00:27 adu joined
AlexDaniel viki: to be honest, #222 looks better than #555 :P 00:27
viki We can't change stuff like that 00:28
AlexDaniel :'( 00:29
u: CRY 00:32
unicodable6 AlexDaniel, U+1A7F TAI THAM COMBINING CRYPTOGRAMMIC DOT [Mn] (◌᩿)
AlexDaniel, U+2CB6 COPTIC CAPITAL LETTER CRYPTOGRAMMIC EIE [Lu] (Ⲷ)
AlexDaniel, U+2CB7 COPTIC SMALL LETTER CRYPTOGRAMMIC EIE [Ll] (ⲷ)
AlexDaniel u: CRY FACE 00:33
unicodable6 AlexDaniel, gist.github.com/1eeff9ae31c3692fef...a4dc12f1ec
AlexDaniel, U+1F622 CRYING FACE [So] (😢)
AlexDaniel, U+1F62D LOUDLY CRYING FACE [So] (😭)
AlexDaniel, U+1F63F CRYING CAT FACE [So] (😿)
00:38 wamba left 00:40 adu left 00:42 vendethiel left 00:43 seatek joined, vendethiel joined 00:46 rindolf left
yoleaux AlexDaniel: submit weird <--Inf++Inf\i> edge cases as roast tests 00:49
AlexDaniel alright
viki :) 00:50
00:51 rindolf joined
viki huggable: advent 00:52
huggable viki, github.com/perl6/mu/blob/master/mi...6/schedule
AlexDaniel viki: it looks a bit too empty 00:54
00:55 rindolf left
samcv why is it called Rakudo 'blead' why spelled this way? does it mean something special? 00:55
viki mst told me once, but I forgot 00:56
16:14 mst iBakeCake: because that's been the perl tradition forever 00:57
16:15 I believe it's because that branch is the bleading edge
TimToady at some point "the bleeding edge" was misspelled in Perl 5 culture
and it stuck
viki heh
samcv that's not how you spell bleeding tho
geekosaur 2016 Oct 21 16:14:57 <mst>iBakeCake: because that's been the perl tradition forever
2016 Oct 21 16:15:19 <mst>I believe it's because that branch is the bleading edge
samcv hah i see TimToady
geekosaur right, but bleeding edge is itself a pun on leading edge
and I see viki found their log first, o well 00:58
samcv i was pretty sure it was refering to blades, like the bleeding edge the edge that is sharp and causes bleeding etc 00:59
geekosaur that is where the original pun came from
samcv but i guess. the perl version could be leading?
geekosaur the leading edge of development is sharp and can cut you...
so this is kinda trending back toward the original without losing the pun completely 01:00
samcv well i mean there's also cutting edge as well. 01:01
dalek : 8977b19 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Claim Dec 7

For "Subset In Your Ways: How To Make, Use, And Abuse Perl 6 Subsets"
AlexDaniel viki: wait, um
viki ?
AlexDaniel m: dd <0-1i>
camelia rakudo-moar 843a6b: OUTPUT«<0-1i>␤»
AlexDaniel m: dd <-0-1i>
camelia rakudo-moar 843a6b: OUTPUT«<-0-1i>␤»
samcv “Cutting edge” actually preceded “leading edge,” first appearing in a literal sense (the sharp edge of a knife or other cutting implement) in the early 19th century. The modern figurative use appeared in the mid-1800s, and is defined by the Oxford English Dictionary (OED) as “A dynamic, invigorating, or incisive factor or quality, especially one that delivers a decisive advantage.
AlexDaniel is it supposed to be this way?
samcv what would be the bleating edge ;P ( bleating like sheep) 01:02
MasterDuke m: sub a($bbb) { $bbb = 1 }; a(2) 01:03
yoleaux 29 Nov 2016 21:35Z <AlexDaniel> MasterDuke: tests?
camelia rakudo-moar 843a6b: OUTPUT«Cannot assign to a readonly variable or a value␤ in sub a at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
01:03 aborazmeh joined, aborazmeh left, aborazmeh joined
viki AlexDaniel: well, Complex's parts are nums and nums can do -0... 01:03
AlexDaniel oh, well… okay then?
MasterDuke huh, i thought camelia got new commits pretty quickly? 01:04
AlexDaniel eval: sub a($bbb) { $bbb = 1 }; a(2)
evalable6 AlexDaniel, rakudo-moar 3800e99: OUTPUT«(exit code 1) Cannot assign to a readonly variable ($bbb) or a value␤ in sub a at /tmp/YfAjRms_iG line 1␤ in block <unit> at /tmp/YfAjRms_iG line 1␤»
AlexDaniel whateverable has a 10 minute delay tops
MasterDuke eval: use Test; throws-like {sub a($bbb){$bbb=1};a(2)}, X::AdHoc, message => /bbb/
evalable6 MasterDuke, rakudo-moar 3800e99: OUTPUT«(exit code 1) 1..3␤ ok 1 - code dies␤ ok 2 - right exception type (X::AdHoc)␤ not…»
MasterDuke, Full output: gist.github.com/c9185a4f3e6aef9b6a...d4396b10c8
lucs Jack Reeves (C++ guy) started using "(B)Leading Edge" (pun intended) in the mid 90s. 01:05
AlexDaniel hmmm seems like camelia is stuck on that commit for some reason
MasterDuke the error i get from the throws-like is from src/6model/containers.c in MoarVM
viki m: dd -0-1i 01:06
camelia rakudo-moar 843a6b: OUTPUT«<0-1i>␤»
MasterDuke eval: use MONKEY; EVAL '{sub a($bbb){$bbb=1};a(2)}'
evalable6 MasterDuke, rakudo-moar 3800e99: OUTPUT«(exit code 1) Cannot assign to an immutable value␤ in block <unit> at EVAL_0 line 1␤ in block <unit> at /tmp/ppYC9W505M line 1␤»
viki m: dd -0e0-1i
camelia rakudo-moar 843a6b: OUTPUT«<-0-1i>␤»
viki AlexDaniel: yeah, it's OK. 01:07
AlexDaniel: FWIW, I have a separate file in roast for all the negative zero stuff
like... I'm still hunting/fixing cases where we don't handle it well
m: dd "\x[2212]0".Num
camelia rakudo-moar 843a6b: OUTPUT«-0e0␤»
viki m: dd sprintf "%f", -0e0 01:08
camelia rakudo-moar 843a6b: OUTPUT«"-0.000000"␤»
viki m: dd "\x[2212]0".parse-base: 10 01:09
camelia rakudo-moar 843a6b: OUTPUT«0␤»
MasterDuke locally i get 'Cannot assign to a readonly variable ($bbb) or a value', when running that EVAL
viki m: dd "\x[2212]0e0".parse-base: 10
camelia rakudo-moar 843a6b: OUTPUT«Failure.new(exception => X::Syntax::Number::InvalidCharacter.new(radix => 10, str => "0e0", filename => Any, pos => 2, line => Any, column => Any, modules => [], is-compile-time => Bool::False, pre => Any, post => Any, highexpect => []), backtrace => Backt…»
viki hm 01:10
01:10 zakharyas joined
MasterDuke oh wait 01:10
eval: use MONKEY; EVAL q|sub a($b){$b=1};a(2);CATCH{say .WHAT.gist}|
evalable6 MasterDuke, rakudo-moar 3800e99: OUTPUT«(exit code 1) (AdHoc)␤Cannot assign to a readonly variable ($b) or a value␤ in sub a at EVAL_0 line 1␤ in block <unit> at EVAL_0 line 1␤ in block <unit> at /tmp/FuPkvclSSs line 1␤»
viki AlexDaniel: I have another gift :) 01:12
AlexDaniel oh no…
viki m: say :10('0e0')
camelia rakudo-moar 843a6b: OUTPUT«0␤»
viki m: say :10<0e0>
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Couldn't process entire number: 1/3 int chars, 2/-1 fractional chars␤at <tmp>:1␤------> 3say :10<0e0>7⏏5<EOL>»
MasterDuke eval: use Test; throws-like {sub a($bbb){$bbb=1};a(2);CATCH{}}, X::AdHoc, message => /bbb/
viki :)
evalable6 MasterDuke, rakudo-moar 3800e99: OUTPUT« 1..3␤ ok 1 - code dies␤ ok 2 - right exception type (X::AdHoc)␤ ok 3 - .message matches /bbb/␤ok 1 - did we throws-like X::AdHoc?»
MasterDuke interesting, just sticking that CATCH{} made it the right error 01:13
01:13 pierre_ left
MasterDuke * just sticking that CATCH{} in there 01:13
AlexDaniel viki: hmmmmmmmmm
this looks familiar
viki: rt.perl.org/Ticket/Display.html?id=128804
:P 01:14
m: say :35<lizmat>
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Couldn't process entire number: 2/6 int chars, 2/-1 fractional chars␤at <tmp>:1␤------> 3say :35<lizmat>7⏏5<EOL>␤»
01:14 dataf3l_ left
viki AlexDaniel: but how come it works in :10('0e0') form? I thought the two were equivalent 01:14
AlexDaniel ah, wtf
O_o
viki :) 01:15
AlexDaniel m: say :35('lizmat')
camelia rakudo-moar 843a6b: OUTPUT«Cannot convert string to number: malformed ':35' style radix number, expecting '>' after the body in '3:35<li⏏5zmat>' (indicated by ⏏)␤ in block <unit> at <tmp> line 1␤»
01:15 dataf4l left
viki hah 01:15
AlexDaniel well, that's a nice error message
seatek if error messages get sufficiently good, there's little reason they shouldn't correct the error for you :) 01:16
viki seatek: hah, I often think that about some of the errors.
geekosaur thinks back to watfiv and shudders
AlexDaniel which is a very common idea actually
and the typical response is “GREAT! Write a module!” 01:17
seatek it would be interesting trying to determine where to draw the lines, at assuming you know better :)
might have to implement a danger scale on run 01:18
dalek : cf89606 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Fix name
 01:19
01:19 pierre_ joined
seatek much of that thinking must go on with perl anyway, as it parses with its many ways of figuring out what people may have meant 01:20
so at what point does it actually become an error
i hadn't really thought of that before
TimToady The problem with a compiler that guesses right most of the time is what happens when it doesn't. :) 01:23
01:23 pierre_ left
geekosaur hence my shudder about WATFIV 01:24
first commercial compiler to do such on the fly correction. when it worked it was great, but when it was wrong, oh was it ever WRONG
...and this in a language that was simple enough that such guessing was much easier than it would be in perl (any version) 01:25
AlexDaniel well, it does not have to be fully autonomous 01:27
seatek that's what the danger slider is for :)
AlexDaniel just a simple question “Want me to change it to ｢………｣ [Y/n]” will do
01:27 dontbe^2 joined
seatek maybe we could send all uncertain assumptions to google in real time and let deep learing answers come back, like those spooky dream pictures 01:29
might be interesting on global financial market data 01:30
01:36 pierre_ joined
AlexDaniel .in 1d go through clog and see if there are any other < > cases that are obviously wrong 01:38
yoleaux AlexDaniel: I'll remind you on 1 Dec 2016 01:38Z
01:38 Vynce left 01:40 cdg left
TimToady
.oO("Hello Dave, you appear to be terminating the wrong entity; fix it for you? [Y/n]")
 02:09
.oO("Hello Dave, you appear to be terminating the wrong entity; fix it for you? [Y/y]")
 02:10
AlexDaniel Y 02:15
Y 02:18
02:19 pierre_ left, kalkin- left, kalkin-_ joined 02:20 geekosaur left 02:22 geekosaur joined 02:25 pierre_ joined 02:29 pierre_ left, pierre_ joined 02:38 pierre_ left 02:41 leego left 02:44 dataf3l_ joined, dataf4l joined, pierre_ joined 02:45 ilbot3 left 02:47 ilbot3 joined 02:48 pierre_ left 02:50 leego joined 02:55 BenGoldberg joined 03:07 pierre_ joined 03:09 rpburkholder left, Guest42590 joined 03:13 Guest42590 left 03:15 Vynce joined 03:22 noganex joined 03:25 noganex_ left 03:33 aborazmeh left 03:40 avalokite joined 03:44 chris2 left 03:47 chris2 joined 03:57 avalokite left 04:05 pierre_ left 04:11 pierre_ joined 04:15 Vynce left 04:16 pierre_ left 04:17 pierre_ joined, b7j0c joined
b7j0c so I recently stumbled upon the "is pure" annotation for subroutines...how meaningful is this given perl6's strong support for powerful exceptions? 04:18
I realize "is pure" is probably just ignored now...but it seems an odd claim to make in perl6 that subs can be side-effect free 04:19
04:21 pierre_ left 04:22 pierre_ joined
ZzZombo why" 04:25
?*
sub x is pure {put 'x'}
04:28 dataf4l left, dataf3l_ left
b7j0c but thats just an expression. if you take a parameter, you have to assume it can carry an exception 04:29
perl6 doesn't complain when i attach "is pure" to functions that throw exceptions, for example 04:30
how can a sub be "pure" if it doesn't even return a value of its signature? 04:31
04:32 pierre_ left
b7j0c almost seems to be begging a "noexcept" annotation a la c++ 04:33
04:34 adu joined 04:36 dataf4l joined, dataf3l joined 04:41 pierre_ joined 04:46 pierre_ left 05:02 khw left 05:06 adu left
b7j0c actually, even "put 'x'" is not "pure" 05:07
TimToady the primary purpose of 'is pure' is to tell the compiler that it is allowed to do constant folding with it 05:13
it is an assertion that any given set of arguments will always produce the same result 05:16
secondarily, we can report "Useless use" of any pure function in sink context, since the only reason to call any function in sink context is for side effects 05:17
b7j0c so should "is pure" cause the runtime to fail if I do something like get the time-of-day in a sub and do something with it? 05:18
TimToady 'is pure' is not for adding any kind of type checking 05:19
it is not an instruction to check anything
b7j0c I guess I can't figure out if it is a promise I am supposed to be making to the runtime or the runtime to me..or what are the implications of lying 05:20
TimToady it is merely an assertion of something you already know to be true, and if you put it on non-pure functions, it's your own fault :)
it's merely an assertion to the compiler that there are no side effects, nor is there any internal state that would cause the function to be non-idempotent 05:21
hence it is safe to assume we can call it once at compile time to do constant folding when we know the arguments are known at compile time 05:22
b7j0c feels more like "can memoize" than "is pure"
if i ask the runtime to memoize a function result and then I end up blowing my own foot off, its on me
TimToady memoizing has additional run-time ramifications, especially for types that are not value types 05:24
but yes, pure is a kind of subset of that, as we use it
b7j0c cool, thanks for the clarification
TimToady another slight difference is that it's pretty much always a win to do constant folding on a pure function, but the benefit of memoizing depends critically on usage patterns, since a cache does you no good if you never repeat a calculation 05:28
and if you limit the size of the cache, you lose the benefit if it does not repeat often enough 05:29
which is why we've had a separate 'is cached' that we've played with from time to time 05:30
05:32 skids left
seatek well, tonight's rabbit hole has been junctions. going to map a new unicode character as a momento 05:42
05:46 mohae_ left 05:53 kalkin-_ left, kalkin- joined 05:57 japhb left 05:59 japhb joined, rindolf joined 06:02 b7j0c left, Cabanossi left 06:03 Cabanossi joined, effbiai joined, effbiai left, effbiai joined 06:12 eisen74 joined 06:18 pierre_ joined 06:23 japhb left 06:25 BenGoldberg left, japhb joined 06:29 canopus left 06:32 canopus joined 06:34 _28_ria left, girafe left 06:37 wamba joined, domidumont joined 06:39 _28_ria joined 06:40 domidumont left 06:44 CIAvash joined 06:49 RabidGravy joined 06:51 nadim joined 06:53 japhb left 06:55 japhb joined, domidumont joined 07:00 CIAvash left, CIAvash joined 07:01 domidumont left 07:03 domidumont joined 07:09 Tonik joined 07:10 AlexDaniel left 07:12 nightfrog left 07:15 CIAvash left 07:20 japhb left 07:21 japhb joined 07:22 eisen74 left, bjz joined 07:24 RabidGravy left 07:25 darutoko joined
moritz viki++ # advent theme, nicely readable 07:27
yoleaux 29 Nov 2016 23:59Z <viki> moritz: I checked a few and they sucked. This one is sexy, so I went with it: perl6advent.wordpress.com/2015/12/18/
07:30 awwaiid left, awwaiid joined 07:37 _28_ria left, bjz left 07:39 _28_ria joined 07:42 dataf4l left, dataf3l left 07:45 lizmat left 07:52 bjz joined, dataf3l joined 07:54 _28_ria left 07:55 dataf4l joined, _28_ria joined 07:58 Actualeyes left 08:01 dataf4l left, dataf3l left 08:03 CIAvash joined 08:04 Wiertek joined 08:11 ufobat joined 08:13 Wiertek left
nadim viki: very nice indeed. if I may, can we get more than 50% advent post? the calendar column on the left is more or less noise and takes too much room IMO. also, the links in a color that doesn't look like errors. 08:15
08:17 avalokite joined 08:19 Actualeyes joined 08:20 firstdayonthejob joined
moritz if the screen is too narrow, you don't see the calendar column, so it's not a problem 08:38
and on wide screens I don't really mind 08:39
08:42 brillenfux joined 08:47 pierre_ left
nadim It's not that there is too little room, just too little text, and too short sentences. i.imgur.com/nFUoMb5.png 08:54
08:56 bjz_ joined 08:57 pierre_ joined, bjz left 09:12 LeCamarade joined
LeCamarade I know this channel is much too civil to poke fun at other languages, even though it has richly earned the right to. However, the recent article by Zed Shaw on Python 3's failures is proof-positive of why Perl 6 was doing the right things the right way all along. 09:14
Even if it takes 15 years, let it take 15 years and a serious and humble attempt to get things right and composable, rather than the mad rush that has ended up at Python 3.
moritz LeCamarade: I'd like that to be to true, but I don't see the evidence
LeCamarade learnpythonthehardway.org/book/nopython3.html 09:15
moritz the usage numbers for python 3 are higher than Perl 6, both in relative and absolute numbers
and the damage done to pyhon 2 and perl 5 is hard to quantify
LeCamarade Yes, but Perl 6 is a hundred-year language. Python 3 is a feeble attempt at Python 1.0.
09:16 CIAvash left
LeCamarade My point is that Py3 is a terrible mess. I have been deploying production stuff on the latest Pythons for many years now. I have much less confidence in Py3 than I would have in PHP 5. No jokes. 09:16
Perl 6 got difficult problems right; it is a genuine improvement on Perl 5.20! Python 3, on the other hand, clearly did not have a humble linguist agonising over the features and how they work together. 09:18
timotimo the very early days of py3 were extremely bumpy, but isn't it the clear winner over py2 nowadays at least? 09:20
LeCamarade Py3 assumed improvements to a language at this point would be easy. Wrong. TimToady knew right. 09:21
timotimo i have stopped using python for my stuff when i joined the p6 community, but back then py3 was painful to use (no numpy or scipy or something)
moritz at $work, we use py3 for new projects, and it works fine
LeCamarade timotimo No. It is actually worse now. For example: we know about the Unicode implementation in Perl 6, and it is extremely far removed from where Py3 is _headed_. See that? 09:22
moritz Yes, I use Py3 too, and that is why I know to hate it! :-D I was hoping to convert all my Py2.7 projects on GAE to Py3 (having earlier moved them from Py2.5), and so I started messing with it. Dear God.
seatek Only time I've ever messed with Python is when I've had to go fix or tweak little things here or there. Reminds me of playing with plastic Lego blocks. 09:24
moritz I guess the experience for a conversion is quite different than for a green-field project
LeCamarade SourceBaby: fine, I stopped looking for a new scripting language ever since Perl 6 came out, but before that time, I was probably one of the first to deploy Ruby 2.0 (called 1.9 at the time). I have a website running on Rakudo Star. I do bleeding edge. And Python 3 is haemorrhaging edge. 09:25
seatek oh my gosh, "once" is not fooling around. it even survives multiple different class instantiations.
LeCamarade (I meant to write "So", not "SourceBaby".)
seatek m: class D { method m() { once say 'this'; }}; my $c = D.new; $c.m; my $d = D.new; $d.m; my $e = D.new; $e.m; 09:26
camelia rakudo-moar 843a6b: OUTPUT«this␤»
seatek anyone know of a way to make it be... once... until i tell you again that you haven't done it? :) 09:27
LeCamarade I just want to say something to this channel, and I hope it makes it to TimToady somehow: the slow, deliberate, careful approach to going beyong Perl 5 was exactly the right thing to do. Otherwise, Perl 6 would have been the kind of trivial (probably illusory) improvement on the former.
seatek I think the people here are really wonderful people LeCamarade 09:28
LeCamarade Even Python 3 was more than 15 years in the making (as ovid or someone once pointed out in an e-mail spat about why Perl 6 was taking forever). I for one had been waiting for Perl 6 for 10 of those years.
seatek Yeah, I find that they are nice. But there is a time for peace and a time for war. I am more on the latter end. ;-) The thing is, there is a time to say that "That sucks!" and I can never trust these people to say it so directly. Now, Perl 6 rocks to high heaven, because a man who was called to do languages for God (he was going to be a Bible translator) worked on it as his life's work. Py3 sucks, because they did not understand that this matter 09:30
timotimo seatek: build your own state variable and give the user access to resetting it by making it named 09:31
seatek seatek: yeah that's what i was going to do. either that, or beg for a onceish function 09:32
i'm amazed it survives through that
LeCamarade: it sounds stressful
i say let them play with their legos :) 09:33
LeCamarade seatek Not if you feel called to do it; not if you feel incomplete otherwise. Isn't all this geekery and math stressful to you? ;-) 09:34
seatek naw i kinda go zen on it. but arguing with people is is stressful to me. the mentalities are very different. python people seem to believe in One Way that is correct. and so it's very hard to help them see other things and ways 09:36
at least that's what i've found
i think it's good to have the good fights out there though :) 09:37
timotimo we've seen enough good snark about python having more than one way, and in some cases all ways suck :(
LeCamarade Fine, but they are failing in their own methodology because they did not have someone sit down and figure out the one right way to do the one-right-way language. Py3 emphatically isn't it.
seatek i hear ya :) 09:38
LeCamarade On the other hand, Perl 6 clearly understands the philosophy behind it so well that I can enforce a pythonic worldview in it--python done properly, even provably, in perlistic terms.
09:39 wamba left
LeCamarade You see, Py3 thought that "all we need is to tweak this here." Wrong. This deep into the 21st century, if you are doing a language++, and especially if you break backward compatibility, you had better have had someone sit and think. We are not interested in yet another new syntax. Honestly, it is relieving that Perl 6 is so well-done that I do not expect to look around ever again. A pity I still can't get it to run on my Raspberry Pi! :-( 09:41
seatek yeah one of the first things that struck me when starting into Perl6 was that somehow it felt WAY more structured, yet still felt almost completely free at the same time. Very odd.
09:42 dakkar joined, noganex left
seatek ok who needs once?? not me 09:42
LeCamarade Now I use only two languages: Haskell and Perl 6. It used to be Haskell and Ruby, but Ruby was only for programs less than 301 lines long. Perl 6, which I write almost entirely statically-typed, can sustain thousands of lines. Py3 has not even began to be more-useful than Forth in that regard. It's an unfortunate joke.
09:42 noganex joined
seatek i'm pretty well committed to Perl 6 for life now. which may be quite short, if i shoot myself because of the size/compiling stuff continuing to grow on this crazy thing 09:44
LeCamarade What makes it all the more poignant for me is that, when Perl 6 was "delayed", the noise was (as is typical for us) all about how long it was taking. When I saw the result, I knew it was God who had commissioned TimToady. The result is what should have mattered. And this thing is insanely brilliant. I have been a language fanatic from my childhood, and very early on I decided I wanted a "perfect" language. I am sure I meant a "hundred-year language". P 09:45
09:46 pmurias joined
LeCamarade seatek Good on you! :-) That makes two of us. I literally do not bother with alternatives anymore. If you have an alternative for "system software", preferably Haskell, you should be fine. Lucky you, though, because I, on the other hand, have not yet successfully compiled Rakudo Star on my computer! :-( 09:46
(I mean my domestic-usage computer; it is a Raspberry Pi. So whenever I do something there, it is in Haskell, or Perl 5 because I can now tolerate that weirdness if I know I am not bound to it for the long haul.) 09:47
seatek LeCamarade: maybe you've been a very naughty person and you're being punished. while i, on the other hand, am practically perfect... ;)
09:48 M_o_C joined, M_o_C left
LeCamarade seatek I am sure I have been naughty. However, punishment should hurt some more than this ... :-D 09:48
seatek i think you'll make a great holy crusader LeCamarade :) 09:49
i'm far too self absorbed i think .. :) 09:50
pmurias LeCamarade: re the anti python 3 rant, Perl 6 also avoids hidden binary data to unicode conversions
LeCamarade seatek Thanks! I wish you knew how right you are ... :-D (Hint: sacralist Reformed Presbyterians still exist, and they do believe in armed revolution, and, if they are me, are actively working to that end.) 09:51
09:51 CIAvash joined
LeCamarade pmurias Yes, but Perl 6 is way more orthogonal: in regular expressions (using Unicode classes, while Py3 goes "Wha'?"), and in allowing, say, Unicode quotes that work as they should (where Py3 just dies). 09:53
09:54 labster left
LeCamarade pmurias Actually, when it comes to Unicode, at present, no man should mention Perl 6 and a joke like Py3 in the same breath. World apart. I write _all_ my strings with proper quotes. _All_ of them. I cannot even type them because this client does not do Unicode as well as my Perl 6 does. 09:54
“Quotes”
Yay!
So, yeah: in _every_ case, in my code, I writes strings either as “strings” or as ‘strings’, depending on the need. 09:55
And that is just one problem! I cannot live without that, by the way, because I insist on it as much as I insist on the commas, for which you can see evidence in my chats (since, unlike the quotes, I was already sure they would work …). 09:56
timotimo it does have the benefit of giving a clear start and end quote, for nesting and such
though our smart quote stuff might be problematic for that? giving you multiple possible end-quotes for the same start-quote?
m: say ‚foo’; say ‘bar’
camelia rakudo-moar 843a6b: OUTPUT«foo␤bar␤»
timotimo m: say ‚foo‘nested?’bar’
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Two terms in a row␤at <tmp>:1␤------> 3say ‚foo‘7⏏5nested?’bar’␤ expecting any of:␤ infix␤ infix stopper␤ postfix␤ statement end␤ statement…»
LeCamarade I also already use things like → and the like in my Haskell. So I expected it to work. In Perl 6, it does.
timotimo okay, not nested with those kinds of quotes
m: say ‘foo‘nested?’bar’ 09:57
camelia rakudo-moar 843a6b: OUTPUT«foo‘nested?’bar␤»
timotimo with this kind, though, it works
yeah, that makes sense
you don't want to have a quote character that you don't realize is a valid start-quote give you nesting within what you know to be your start-and-end marks
LeCamarade timotimo And that is just one simple thing that a grammar nazi like myself just cannot otherwise tolerate not having. Only one language gets it right. And that is one simple _source code_ matter! There are many millions of others. You still have stupid regexen in Py3. And they took as long at it as TimToady took with Perl 6! It’s an unfortunate joke. 09:58
You still have an “interpreted” language, and you still have to use “lex and yacc” to get a string parse. Not funny. 10:00
timotimo yeah, put the heaviest of quotations on the word ‘interpreted’ 10:01
LeCamarade For ranges, I use the proper diæreses: … 10:02
timotimo that term has long lost most of its usefulness
10:02 AlexDaniel joined
LeCamarade Unless it really means nothing now, you should be able to tell your interpreted-language parser to just adopt _these other rules_ while it goes over _that other line_. We think this is a big deal because our bar has been set way too low by things like Py3. 10:03
AlexDaniel LeCamarade: hello!
LeCamarade AlexDaniel Hi! :-D 10:04
AlexDaniel I was reading the log… and seems like somebody really likes the quotes :)
so I came here to say hi, because I really do too
LeCamarade Log?
AlexDaniel irclog.perlgeek.de/perl6/2016-11-30
LeCamarade: ah ya, by the way: gist.github.com/AlexDaniel/c89bd2786f9b63f31e4c 10:05
m: say ‚foo‚nested?’bar’ 10:06
camelia rakudo-moar 843a6b: OUTPUT«foo‚nested?’bar␤»
LeCamarade AlexDaniel Cool resource!
AlexDaniel timotimo: works fine if you do it properly :)
timotimo well, yeah 10:07
i was talking about having different pairs
LeCamarade Sometimes I am coding in a francophone context, and I use 􏿽xC2 cette sorte ». 10:08
10:08 synopsebot6 left, M_o_C joined, synopsebot6 joined
LeCamarade « quotes! » 10:08
10:08 M_o_C left 10:09 M_o_C joined, M_o_C left
AlexDaniel LeCamarade: what about ∞, 2², ½ ? Do you use that stuff too? 10:09
u: MINUS SIGN
unicodable6 AlexDaniel, U+00B1 PLUS-MINUS SIGN [Sm] (±)
AlexDaniel, U+02D7 MODIFIER LETTER MINUS SIGN [Sk] (˗)
AlexDaniel, U+0320 COMBINING MINUS SIGN BELOW [Mn] (◌̠)
LeCamarade AlexDaniel Oh you bet! I get every excuse to!
unicodable6 AlexDaniel, gist.github.com/7fce0f40666a0d60db...5ce4b6fecd
AlexDaniel LeCamarade: what about −?
LeCamarade I use … in ranges, but … what is that, the em-dash? — ? 10:10
En-dash. I don’t know how to type it.
:-(
timotimo but ⟨⟩ and «» don't have regular quoting semantics
m: say «foo bar baz».perl 10:11
camelia rakudo-moar 843a6b: OUTPUT«("foo", "bar", "baz")␤»
LeCamarade AlexDaniel Is that the en-dash? I would assume it work like a mere -?
AlexDaniel u: - —
unicodable6 AlexDaniel, U+002D HYPHEN-MINUS [Pd] (-)
AlexDaniel, U+0020 SPACE [Zs] ( )
AlexDaniel, U+2014 EM DASH [Pd] (—)
AlexDaniel uhh that's not it
u: -−– 10:12
unicodable6 AlexDaniel, U+2212 MINUS SIGN [Sm] (−)
AlexDaniel, U+002D HYPHEN-MINUS [Pd] (-)
AlexDaniel, U+2013 EN DASH [Pd] (–)
LeCamarade timotimo Yes, actually. I use those for lists, I remember. There was something else I was demoing for my less-enlightened francophone friends. But, yes, I use them for qw// lists.
AlexDaniel LeCamarade: the issue with - is that it is used both as a hyphen and a minus
LeCamarade: so unicode actually has a separate character for a minus
LeCamarade: which normally looks identical, but still…
and Perl 6 supports it too 10:13
m: say −42
camelia rakudo-moar 843a6b: OUTPUT«-42␤»
LeCamarade AlexDaniel Woah. Now I am going to have learn how to type that with X’ compos.
u: -
unicodable6 LeCamarade, U+002D HYPHEN-MINUS [Pd] (-)
LeCamarade u: —
unicodable6 LeCamarade, U+2014 EM DASH [Pd] (—)
AlexDaniel LeCamarade: do you write your ‘’ “” ｢｣ quotes using compose key also? 10:14
LeCamarade AlexDaniel Yes.
AlexDaniel that's kind of inefficient, I just added them all to my keyboard layout
(so that the time to type " is identical to the time to type “) 10:15
and … is one key press instead of 3 :)
10:16 andrzeju_ joined
LeCamarade AlexDaniel Hmm. I wanted something I could google … 10:16
Actually, … is _four_ keypresses!
10:16 dogbert2 joined
LeCamarade Compose + . + . + .. Absurd. :-( 10:16
AlexDaniel yea, well… Compose is good for rare characters, but you should get common stuff right into your keyboard layout, I believe
10:17 noganex_ joined 10:18 zakharyas left, zakharyas joined
LeCamarade Compose is terrible, but I can only fix those things once I move away from using stock computers. The only piece I do not have, alas, is Rakudo Star on ARM64. I have even started thinking of replacing Pi with something like the JaguarBoard, so that I can move on. I do not want to configure crap; I want to build awesome. 10:18
10:18 perigrin_ joined, domm1 joined, krakan_ joined 10:19 baest_ joined, Spot__ left, ribasushi left, zacts left, setty1 left, perigrin left, domm left, dakkar left, apejens left, mr-foobar left, ilmari[m] left, PerlJam left, ShoaibJahejo left, gfldex left, ilbelkyr left, timotimo left, noganex left, brillenfux left, dogbert17 left, andrzejku left, Khisanth left, timeless left, huf left
LeCamarade AlexDaniel Which keyboard layout do you use? I use Dvorak. With Compose on caps. 10:19
10:19 saki_ left 10:20 jonadab joined, mr-foobar joined, PerlJam joined, gfldex joined, rblackwe joined, apejens joined, r3m_ joined, ilbelkyr joined, DrForr joined, huf joined 10:21 saki joined 10:22 vcv_ joined
LeCamarade But back to the effusive pro-Perl 6 tribalism that got me in here today, in the first place: Py 3 people are still so, so far from even knowing that they could quote their strings properly, and they too broke backward compatibility and took 10+ years at it. Their problem is that they never had TimToady. 10:22
10:22 zostay left, zakharyas left 10:24 timotimo joined, timeless joined
LeCamarade Now, I am writing programs I expect my grandchildren to run. I can be that confident because I am coding in Perl 6 (and, where necessary or inevitable, Haskell). On the other hand, Py 3 will _have_ to quickly move to another botched attempt at ++. Py 4 will also fail, because it will not be Perl 6. :-p 10:24
10:25 g4 joined, g4 left, g4 joined, musca joined, Matthew[m] left, baest left, r3m left, effbiai left, nchambers left, r3m_ is now known as r3m 10:26 zostay joined, nchambers joined
pmurias LeCamarade: I don't think triabalism is a commonly held attitude on this channel 10:26
LeCamarade Even if TimToady now said he had Perl 7, I would be excited but deeply sceptical. We know how much work and time it was to improve on Perl 5, and, good as that was, it was still something that could desperately use some improvement. Now, Perl 6 is a completely different opus. It would take a dedicated monastery about 100 years of humble monks trying things, praying, then coding, before Perl 7 could happen. And yet Perl 7 could happen.
10:26 itaipu joined 10:27 Khisanth joined
AlexDaniel LeCamarade: I use a custom layout 10:28
LeCamarade For one, I am only now focussed on changing the runtime. Rakudo Star is fine for getting me off the ground. Ultimately, it must be a Forth that is running Perl 6. Imagine that: Perl 6 (where Haskell won’t do)and the lot running on a Forth machine. That is what I am working towards. Now, I genuinely was going to take the time to do a proper “hundred-year language”, but then Perl 6 happened. I have only the other parts to work on, now. 10:29
AlexDaniel it is based on dvorak (that is, all letters are in normal dvorak positions), but everything else is different
however, you must have a japanese keyboard to use it properly
LeCamarade pmurias I know this channel is not nearly as tribalistic and jingoistic as it has earned the right to be. Oh, well. I don’t come that often.
10:30 ilmari[m] joined, dakkar joined, setty1 joined 10:31 setty1 left, dakkar left
LeCamarade pmurias And people do not really know what Perl 6 is about. They will not know the truth of it unless someone says “It is the best scripting language out there right now.” Someone has to tell the truth at some point. A fierce tribalist like me, no doubt. It is the truth, after all. 10:31
AlexDaniel and that being said… I am happy that bash does not support ‘’ quotes :P
dalek c: 99fdca2 | gfldex++ | doc/Type/Routine.pod6:
clearly state responsibilities when using `is pure`
synopsebot6 Link: doc.perl6.org/type/Routine
10:31 grondilu joined 10:32 Matthew[m] joined, shadowpaste joined
AlexDaniel LeCamarade: it would be nice if this fierce tribalist contributed to the core development in some way… all this excitement has to be put into the right direction :P 10:32
10:32 ribasushi joined
AlexDaniel and given that most of rakudo is written in Perl 6, I suspect you will like it 10:33
LeCamarade AlexDaniel Yes. The right direction does not exclude loud declarations, even as it includes contributing to the core development.
Meanwhile, I am only going to be taking on development starting next year, and primarily on ARM64. On FreeBSD. A very, very difficult mix; I wonder if I am well-prepared (enough). 10:34
10:35 effbiai joined
AlexDaniel do we even support arm64 properly right now? 10:35
LeCamarade Before that, though, there is donations for Rakudo development, et cetera.
10:36 ShoaibJahejo joined, Spot__ joined, setty1 joined, dakkar joined, Spot__ left, Spot__ joined, ShoaibJahejo left, ShoaibJahejo joined
LeCamarade AlexDaniel No. In fact, on FreeBSD, I cannot even get a compiler that will work. At present (12-CURRENT). 10:36
10:36 ShoaibJahejo left, ShoaibJahejo joined
AlexDaniel well, that sucks 10:37
LeCamarade But consider: instead of all the stuff people are spouting about Py 3, which does not even try to get things correct, if someone shouted from the rooftops “Actually, ye trogs, ‘tis Perl 6 ye are lookin’ for!” you would have the massive surge in core developers that you want.
AlexDaniel ok, I have to run. See you around! 10:38
LeCamarade Glad for your quotes and Unicode resource! :-D
ZzZombo m: my @a=1;dd @a[0,3] 10:40
camelia rakudo-moar 843a6b: OUTPUT«(1, Any)␤»
ZzZombo why is there only one Any?
10:40 zacts joined
gfldex ZzZombo: you asked for 2 values and one of them is defined 10:41
ZzZombo oh 10:42
m: my @a=1;dd @a[0..3]
camelia rakudo-moar 843a6b: OUTPUT«(1, Any, Any, Any)␤»
10:42 AlexDaniel left
gfldex m: my @a=1;dd @a[0;3] 10:42
camelia rakudo-moar 843a6b: OUTPUT«Failure.new(exception => X::OutOfRange.new(what => "Index", got => 3, range => "0..0", comment => Any), backtrace => Backtrace.new)␤»
ZzZombo shitty comma, I swear it slipped w/o my consent!
gfldex m: my @a=1;dd @a[0|,3] 10:43
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Preceding context expects a term, but found infix , instead␤at <tmp>:1␤------> 3my @a=1;dd @a[0|,7⏏053]␤»
LeCamarade m: my @a = 1; dd @[0 … 3]
m: my @a = 1; dd @a[0 … 3]
camelia rakudo-moar 843a6b: OUTPUT«[0, 1, 2, 3]␤»
rakudo-moar 843a6b: OUTPUT«(1, Any, Any, Any)␤»
gfldex commas can Slip
*can't
seatek m: subset PosInt of Int where * > 0; CATCH { say $_.perl }; my PosInt $i = -2; 10:48
camelia rakudo-moar 843a6b: OUTPUT«X::TypeCheck::Assignment.new(symbol => "\$i", operation => Any, got => -2, expected => PosInt)␤Type check failed in assignment to $i; expected PosInt but got Int (-2)␤ in block <unit> at <tmp> line 1␤␤»
ZzZombo so is here a shorter version of
m: my @a=1;dd @a[0..(@a>1??1!!0)]
camelia rakudo-moar 843a6b: OUTPUT«(1,)␤»
seatek that's good
this confuses me
m: subset Password of Str where m/^<[0..z]> ** 10..100$/; CATCH { say $_.perl }; my Password $i = 'wefwef'; 10:49
camelia rakudo-moar 843a6b: OUTPUT«X::TypeCheck::Assignment.new(symbol => "\$i", operation => Any, got => "wefwef", expected => Password)␤Type check failed in assignment to $i; expected Password but got Str ("wefwef")␤ in block <unit> at <tmp> line 1␤␤»
seatek oh got it right on this camelia
must be my version
gfldex m: my @a = 1; say @a.head(2)
camelia rakudo-moar 843a6b: OUTPUT«(1)␤»
gfldex m: my @a = 1,2; say @a.head(2)
camelia rakudo-moar 843a6b: OUTPUT«(1 2)␤»
gfldex m: my @a = 1,2,3; say @a.head(2)
camelia rakudo-moar 843a6b: OUTPUT«(1 2)␤»
ZzZombo yay 10:50
gfldex ZzZombo: ^^^ depends a bit who you expect to read your code. If they are fluent in bash .head is the way to go.
pmurias nqp-m: my $foo := 12; my int $bar := $foo || 200;say($bar) 10:52
camelia nqp-moarvm: OUTPUT«0␤»
pmurias ^^ is this nqp-m bug known?
jnthn pmurias: I don't recognize it. 10:55
10:58 jonas1 joined
LeCamarade And let me say one last thing, since seatek noticed I am crusader proper: after the revolution, Perl 6 will be taught in our schools. So ultimately it will be the BASIC for a certain generation, no matter how many or how few tribalistic jingoists you have on the ‘Net. Remember this. :-) Good thing it is being logged. Also, just because it is always fitting: jnthn++ 10:59
11:00 LeCamarade left
sjn hello, #perl6 11:00
viki \o
seatek m: subset Password of Str where m/^<[0..z]> ** 10..100$/; CATCH { say $_.perl }; class P { method pset(Password $pw) { say $pw }; }; my $p = P.new; $p.pset('wefwef'); 11:01
camelia rakudo-moar 843a6b: OUTPUT«X::AdHoc.new(payload => "Constraint type check failed for parameter '\$pw'")␤Constraint type check failed for parameter '$pw'␤ in method pset at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
seatek ok that weird me out a bit
it's changed to an AdHoc instead of Type error
viki reads a few of LeCamarade posts above.... 11:02
damn, I wish my drugs were that good...
seatek he teeters on the edge :) 11:03
viki -- can you solve my problems?
viki nadim: no, we can't change minute details of the theme because we run this on wordpress.com and they want a paid subscription to be able to do that.
seatek: which are? 11:04
seatek if exceptions have to come up through multiple classes for subset type constraints, do the loose their original exception type?
m: subset Password of Str where m/^<[0..z]> ** 10..100$/; CATCH { say $_.perl }; my Password $i = 'wefwef';
camelia rakudo-moar 843a6b: OUTPUT«X::TypeCheck::Assignment.new(symbol => "\$i", operation => Any, got => "wefwef", expected => Password)␤Type check failed in assignment to $i; expected Password but got Str ("wefwef")␤ in block <unit> at <tmp> line 1␤␤»
seatek This is right
m: subset Password of Str where m/^<[0..z]> ** 10..100$/; CATCH { say $_.perl }; class P { method pset(Password $pw) { say $pw }; }; my $p = P.new; $p.pset('wefwef');
11:04 geekosaur left
camelia rakudo-moar 843a6b: OUTPUT«X::AdHoc.new(payload => "Constraint type check failed for parameter '\$pw'")␤Constraint type check failed for parameter '$pw'␤ in method pset at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤» 11:04
seatek this is not
viki There's a ticket for that. 11:05
And no, I can't solve it.
seatek ah! :) ok i'm NOT insane :)
that's reassuring
ok :)
11:05 geekosaur joined
seatek i was just looking for reassurance or info. i'll resort to smartmatching the text of the error message... 11:05
hehehe
btw nobody can change the text of exception message any more from this point on... ;) 11:07
viki m: subset Password of Str where m/^<[0..z]> ** 10..100$/ or die "cry me a river"; CATCH { say $_.perl }; class P { method pset(Password $pw) { say $pw }; }; my $p = P.new; $p.pset('wefwef'); 11:08
camelia rakudo-moar 843a6b: OUTPUT«X::AdHoc.new(payload => "cry me a river")␤cry me a river␤ in block <unit> at <tmp> line 1␤␤»
seatek hehehe :) :) 11:09
viki seatek: you can throw proper exception yourself. Also there's modules.perl6.org/dist/Subset::Helper
seatek if i do down there, will it stick all the way up?
or will it be stripped to adhoc too?
oh this is nice 11:10
11:12 pierre_ left 11:19 TEttinger left 11:20 nadim left 11:25 CIAvash left
dalek : d5f0f32 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Remove pun from title (it's better elsewhere)
 11:27
andrzeju_ hey 11:28
viki \o
andrzeju_ :D
viki, could you look fast what I wrote? 11:29
dalek : d4025fc | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Book another day

With "Set In Your Ways: Perl 6's Setty and Baggy Types"
I'm just filling up holes. Feel free to remove me and claim those spots for yourself.		 11:30
viki andrzeju_: probably not, as I'm breakfasting and leaving to work shortly.
andrzeju_ :< 11:31
I am temporary do it at work, have no special tasks by these days 11:32
dalek : 7c0ac6f | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
mark Zofspots as claimable
viki andrzeju_: well, there are 318-$bot-count people in this channel. If you post a link, they could read 11:33
andrzeju_ sure, www.gitbook.com/book/damaxi/weird-...le/details
dalek : 2f59851 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
s/spot/post/
 11:34
11:34 _28_ria left
andrzeju_ viki, I am planning to read a bit information how the Perl6 design process look and architecture and then put it into introduction :P 11:34
viki Cool. 11:35
We use the whirlpool metaphor.... 11:36
.oO( swirling toilet? )
The design and real-world usage goes back and forth like a water spinning in the toilet (or are we doing the whirlpool?), and with each spin, we get closer to the center, which represents the perfect design :) 11:37
jnthn That's the first time I've heard it described as like being in a toilet :P 11:38
seatek with snakes in the plumming 11:39
viki >:)
seatek and a nice plant to the right
dalek osystem: f273743 | spebern++ | META.list:
add module
 11:41
osystem: 22ea7cb | (Zoffix Znet)++ | META.list:
Merge pull request #270 from spebern/master

add module
viki github.com/spebern/Parser-FreeXL-Native/
11:42 spebern joined
dalek : ec7ed54 | (brian d foy)++ | misc/perl6advent-2016/schedule:
I changed my topic

Just so other people know.		 11:43
: b3d1338 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Merge pull request #18 from briandfoy/patch-2

I changed my topic
viki andrzeju_: hehe, you should make ingrish your gimmick in the book 11:46
don't try to fix it :) 11:47
"I amn't Perl developer either." that's just gold
andrzeju_ huh, does it look not so bad? 11:48
11:48 Wiertek joined
viki andrzeju_: well, I know what you meant. It looks funny :) 11:48
And this way you don't need to worry about the language. If anyone tries to diss it, just say it was made that way on purpose :) 11:49
Like this dude: www.youtube.com/watch?v=As10jr1Vvj0
andrzeju_ hah, okay :D indeed I am quite funny guy
viki His heavy Russian accent is a gimmick
11:50 wamba joined, iH2O joined
andrzeju_ viki, okay thnks for hot words :D then I wil read some material and continue with it 11:50
11:51 iH2O left 11:53 itaipu left 11:55 effbiai left, effbiai joined 11:56 setty1 left, bjz joined, bjz_ left 11:59 Actualeyes left 12:06 CIAvash joined 12:14 canopus left 12:15 zakharyas joined 12:20 kalkin- left 12:21 pierre_ joined, canopus joined
ZzZombo Somebody got a minute to read stuff you probably won't agree to? 12:22
gist.github.com/ZzZombo/34803d3cc6...41b6cdbfc3
seatek if you use a role instead of a class, you can "subclass" in a private attribute -- is that what would help? 12:33
12:33 pat_js joined, wamba left
seatek role XML::GraphItemGroup does XML::GraphItem... or somesuch 12:35
i'd split it down into ever further granularity with roles though if it was me
seatek goes afk for some much needs breaktime 12:37
12:39 wamba joined 12:40 baest_ is now known as baest 12:43 Actualeyes joined
sjn looks at perl6.org/compilers/, seeing that Parrot isn't mentioned... Should there be a "Historical Runtimes" section there? 12:43
ZzZombo seatek: each of those attributes would have distinct values, you would have to go out of your way and somehow initialize them all. 12:49
I assume you were talking about xmlnode.
12:49 spebern left 12:58 grondilu left, cibs left 12:59 chris2 left 13:00 cibs joined
viki ZzZombo: read stuff you probably won't agree to.... to what end? 13:05
13:06 chris2 joined
pmurias why does perl6.org/compilers mention that Rakudo is copyrighted by the TPF? 13:09
moritz because it's true 13:12
viki pmurias: whose is the copyright then? :)
moritz the CLA assigns them a non-exclusive copyright
pmurias moritz: why is that relevant enough to mention on the download page? 13:14
13:14 viki left, viki joined, viki left
moritz pmurias: that's not the download page, it's the info page about compiler(s) 13:15
pmurias: I don't find it misplaced
when we had multiple compilers, we had the information who was responsible and/or the driving force, characterstics of the compiler, dowload links etc. 13:16
sjn: please no historical runtimes; we try very hard to keep outdated information off perl6.org 13:17
13:17 avalokite left, notviki joined
notviki heh 13:17
The real viki ghosted me :)
moritz what's the relevance of teaching a newbie that rakudo was based on parrot, years ago?
pmurias moritz: why is that relevant?
sjn moritz: fair enough :) 13:18
pmurias moritz: and is the CLA granting the TPF a license to use Rakudo the same thing as the "TPF copyrighting Rakudo"?
moritz pmurias: my legalese fu is not strong enough to answer that 13:19
ZzZombo <viki> to what end?
eh, what?
notviki ZzZombo: for what purpose would we read it?
ZzZombo ah, it's to help me with me having issues with P6 class system, I guess. 13:20
13:22 notviki is now known as cassini 13:23 cassini is now known as Guest80605
pmurias what are the arguments for keeping the "Copywrite by the TPF" mention 13:23
?
I suspect that's actually not the case, and it might lead people to believe that Rakudo is closed source 13:24
13:26 Guest80605 left
dalek href="https://perl6.org:">perl6.org: b3935c5 | (Pawel Murias)++ | source/compilers/index.html:
Remove confusing mention of Rakudo being "copyrighted by the TPF".
 13:30
13:32 babydrop joined
babydrop pmurias: section 4.1 gives TFP the non-exclusive copyright, 13:35
(of CLA)
perlpilot um ... Rakudo is still copyright TPF as far as I know. IF that's no longer the case, there are a *bunch* of files in the Rakudo repo that need to change. 13:37
13:38 CIAvash left
perlpilot okay ... maybe not a bunch, but several :) 13:38
babydrop sjn: see what you started! :P 13:39
sjn oh noes!
perlpilot Also, if TPF doesn't hold the copyright, who does?
13:40 dataf3l_ joined, dataf4l joined, CIAvash joined
babydrop perlpilot: TPF does hold the non-exclusive copyright 13:40
We've just changed the compiler info page.
perlpilot To what end? 13:41
dalek /advent_day_4_title_change: 7af7314 | (Brian Duggan)++ | misc/perl6advent-2016/schedule:
title change
pmurias babydrop: 4.1 mentions granting a license
babydrop pmurias: license to what? 13:43
pmurias: "all intellectual property rights (excluding patent and trademark, but including copyright)"
pmurias: the contributor still retains their rights and can do whatever they want with their contribution, but they can't sue TPF under copyright law 13:44
pmurias perlpilot: I removed the mention that Rakudo is coprighted by the TPF from the compilers page 13:47
babydrop Yeah, it's not overly relevant on that page.
13:50 zakharyas left
dalek : 7af7314 | (Brian Duggan)++ | misc/perl6advent-2016/schedule:
title change
 13:51
: 22c650d | (Brian Duggan)++ | misc/perl6advent-2016/schedule:
Merge pull request #19 from perl6/advent_day_4_title_change

title change
13:53 dataf3l_ left, dataf4l left 14:02 itaipu joined, user9 left 14:03 user9 joined
ZzZombo So... anybody? Please, I need your insight. 14:03
babydrop heh, Google just called me... asking what I used their compute engine for. Told them Perl 6 compiler... 14:04
like a badass
ZzZombo: but you said we wouldn't like reading it! :) 14:05
14:06 cdg joined, cdg left 14:07 cdg joined
ZzZombo Yes, because the previous time I tried to bring this topic, I got turned down, because apparently everything was thought of, and that P6 has is vastly superior to other languages. 14:07
babydrop hehe 14:08
14:08 CIAvash left, vendethiel left
babydrop glances at "this class hierarchy" 14:08
Well, I disagree, the previous time you brought it up, you got a nice, long blog post saying that OO are about passing messages around and that you should think about messages for and the class hierarchy will pop out out of it. 14:09
But what you have in that gist is two classes with methods like set_name and set_count 14:10
14:10 avalokite joined
pmurias ZzZombo: yes, that was the case ;) 14:10
ZzZombo babydrop, I take it as you missed I stripped it down? 14:11
so then tell me what should I do?
babydrop ZzZombo: and someone already mentioned roles
m: role Meow { has $!private-business = 42; }; class Foos does Meow { method meow { $!private-business } }.meow.say 14:12
camelia rakudo-moar 843a6b: OUTPUT«Cannot look up attributes in a Foos type object␤ in method meow at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
babydrop m: role Meow { has $!private-business = 42; }; class Foos does Meow { method meow { $!private-business } }.new.meow.say
camelia rakudo-moar 843a6b: OUTPUT«42␤»
ZzZombo and the bit about them that I don't really understand is 14:13
m: role R { has $!a='foo' };role R2 { has $!a='bar' };class A does R {method m{put $!a}};class B is A does R2 {method n{put $!a}};B.new.m;B.new.n; 14:14
camelia rakudo-moar 843a6b: OUTPUT«foo␤bar␤»
ZzZombo how it's supposed to work in cases like that.
babydrop doesn't see what the case is about
ZzZombo I just don't really understand them.
babydrop roles get inlined into your class 14:15
ZzZombo I tried to experiment wit them and got confusing results like ^
with*
babydrop So above, your class A still has role R in it, but class has its own $!a coming from R2
*but class B 14:16
m: class A { has $!a = 'foo'; method m { put $!a } }; class B { has $!a = 'bar'; method m { put $!a }; B.new.m; B.new.A::m 14:18
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3ethod m { put $!a }; B.new.m; B.new.A::m7⏏5<EOL>␤ expecting any of:␤ statement end␤ statement modifier␤ statement modifier loop␤»
babydrop m: class A { has $!a = 'foo'; method m { put $!a } }; class B { has $!a = 'bar'; method m { put $!a }; B.new.m; B.new.A::m() 14:19
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3hod m { put $!a }; B.new.m; B.new.A::m()7⏏5<EOL>␤ expecting any of:␤ statement end␤ statement modifier␤ statement modifier loop␤»
babydrop :/
m: class A { has $!a = 'foo'; method m { put $!a } }; class B { has $!a = 'bar'; method m { put $!a }; dd B.new.*m
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3'bar'; method m { put $!a }; dd B.new.*m7⏏5<EOL>␤ expecting any of:␤ postfix␤ statement end␤ statement modifier␤ statement…»
jnthn Missing clsoing } on the class
babydrop m: class A { has $!a = 'foo'; method m { put $!a } }; class B { has $!a = 'bar'; method m { put $!a } }; dd B.new.*m
camelia rakudo-moar 843a6b: OUTPUT«bar␤(Bool::True,)␤»
babydrop Thanks
ZzZombo *m? 14:20
perlpilot ZzZombo: no, .*m :)
babydrop m: class A { has $!a = "foo"; method m { put $!a } }; class B is A { has $!a = "bar"; method m { put $!a } }; B.new.*m
camelia rakudo-moar 843a6b: OUTPUT«bar␤foo␤»
babydrop ZzZombo: ^ there. does that make more sence?
ZzZombo yea 14:21
babydrop ZzZombo: both classes have their own $!a with different value and you get different value depending on whose class's method you call
ZzZombo but what's this *m thing
babydrop same with your role example.
.* method call calls the method on all classes in the MRO
so it calls B's m, then A's m 14:22
perlpilot ZzZombo: There's also .?meth which I find useful and you may too. It's like "call meth if it exists"
14:22 wamba left
babydrop m: use MONKEY; augment class Any { method m { put "meows!" } }; class A { has $!a = "foo"; method m { put $!a } }; class B is A { has $!a = "bar"; method m { put $!a } }; B.new.*m 14:22
camelia rakudo-moar 843a6b: OUTPUT«bar␤foo␤meows!␤»
babydrop And here I added `m` to Any too, so it called that too 14:23
14:23 cygx joined
cygx o/ 14:23
babydrop \o\
cygx perhaps a shorter example in P6 would help
m: role Named { has $.name; method !rename($name) { $!n
ame = $name } }; class A does Named { submethod TWEAK { $!name = "default" } };
class B is A { method rename($name) { self!A::rename($name) } }; my $b = B.new
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Missing block␤at <tmp>:1␤------> 3 has $.name; method !rename($name) { $!n7⏏5<EOL>␤ expecting any of:␤ statement end␤ statement modifier␤ statement modifier loop␤»
cygx ; say $b.name; $b.rename("seuss"); say $b.name
babydrop hehehe
cygx c&p fail 14:24
ZzZombo nicely failed
cygx m: role Named { has $.name; method !rename($name) { $!name = $name } }; class A does Named { submethod TWEAK { $!name = "default" } }; class B is A { method rename($name) { self!A::rename($name) } }; my $b = B.new; say $b.name; $b.rename("seuss"); say $b.name
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Cannot call private method 'rename' on package A because it does not trust B␤at <tmp>:1␤------> 3od rename($name) { self!A::rename($name)7⏏5 } }; my $b = B.new; say $b.name; $b.ren␤»
cygx how would I implement something like that without having to trust all child classes beforehand
ZzZombo yeah, that trust thing
just what I had described there as well 14:25
wha solutions exist for that? 14:26
BRB, dinner.
babydrop m: gist.github.com/zoffixznet/fafbdd1...c3bb108edc 14:29
camelia rakudo-moar 843a6b: OUTPUT«default␤seuss␤»
babydrop IMO the mistake you're making is having the assumption that subclases must have privilidged access to ancestor's private guts
moritz cygx: right, make it a public attribute
anyone can subclass your class, so it must be a stable API and documented anyway 14:30
ZzZombo But what if I wanna have my own TWEAK, for example, and also the one from the role? 14:31
14:31 CIAvash joined 14:32 eisen74 joined, CIAvash left
cygx "even though you can set the name via .name = ..., use .rename(...) instead" 14:32
14:32 nightfrog joined
cygx that's smelly 14:32
babydrop That's where jnthn's article comes in about thinking of messages in OO :) 14:33
( 6guts.wordpress.com/2016/11/25/per...ood-thing/ ) 14:34
cygx rename is a perfectly fine message
the issue is a failure to express certain relationships without breaking encapsulation 14:35
14:35 cdg left
babydrop It may be, but why is it in a subclass? :) 14:35
dalek c: 4a5da6b | coke++ | xt/words.pws:
learn new "word"
 14:36
14:36 cdg joined 14:37 skids joined
cygx babydrop: having MutableVersion inherit from ImmutableVersion can be useful 14:38
pmurias cygx: if you want to break encapsulation you have to use trusts 14:39
babydrop cygx: well, we have List/Array, Map/Hash, Set/SetHash, Bag/BagHash, Mix/MixHash in core.... as mutable complements to the immutable ones. Do they have the issue you describe?
pmurias if class B was aware that class was using a Named role it would be break encapsulation 14:40
cygx think about it this way: there are internal methods, infrastructural methods and interface methods 14:45
subclasses may require access to infrastructural methods, so in p6 they have to be promoted to the interface
14:46 cdg left
cygx naming conventions can help with that, eg SHOUTING 14:47
14:48 cdg joined
babydrop "PHP Parse error: syntax error, unexpected '='" 14:48
That's what Perl6-ism of using kebob-case looks like in PHP :) 14:49
14:56 avalokite left 14:57 lukaramu joined 14:59 firstdayonthejob left, vendethiel joined
ZzZombo cygx: yes, I'm still not sold on that protected visibility is bad and therefore P6 is fine w/o it. 14:59
that exactly why it was invented. 15:00
15:03 seatek left
babydrop What's protected visibility? 15:04
pmurias only subclasses can access it
babydrop If your subclass is using my $!name, that means I can't rename it something else willy nilly.
Without breaking that subclass. That's why it's private 15:05
moritz aka "you have to wear a funny yellow hat to see my privates"
babydrop hheh
pmurias babydrop: protected ones are not privates just, "it's public but you shouldn't use it on a whim, so we add an arbitary restriction" 15:06
babydrop :/
cygx in Java, funny hat means subclass and/or same package
moritz forcing people who *really* need to write subclasses, instead of using a more appropriate relationship
forcing people who *really* need to access the attribute to write subclasses, instead of using a more appropriate relationship 15:07
that's what I meant
cygx that's why if I were to design an OO system, I'd invert the trust relationship
"on request, give me access to stuff that you deem infrastructural"
pmurias that breaks encapsulation 15:08
moritz cygx: isn't that exactly what 'trusts' does? The author of the subclass requests access, and you trust it in the parent class
cygx so does slapping an "is rw" on things to make my example code work
moritz: no, because it requires the truster to know about all trustees, instead of marking things accessible to anyone who requests access 15:09
babydrop cygx: I think you took my version of your example too literally. You don't "slap" `is rw`. You design proper interface. 15:10
15:10 eisen74 left
moritz cygx: I don't understand what benefits the "requesting access" step gives if it's granted automatically, always 15:10
15:11 g4 left
pmurias seperates the clean suggested API from the dirty infrastructual methods? 15:11
cygx moritz: it's granted on request and keeps the public API clean
pmurias: exactly
babydrop cygx: but then nothing is private? Or are you proposing an "it's their own dumb fault" as a response for anyone who uses ancestor's privates willy-nilly, when their code breaks all of a sudden? 15:13
cygx no, you have to be declare which privates to make accessible on request, instead of who can touch them (as is the current approach) 15:14
s/to be/to/ 15:15
pmurias cygx: the ones that are accessible on request are not privates 15:16
babydrop Then they no longer privates. I can't change them without breaking someone's code.
pmurias cygx: you could use a naming convention
cygx pmurias: indeed, cf irclog.perlgeek.de/perl6/2016-11-30#i_13655193 15:17
babydrop And I bet such a thing can be done with a trait on a method 15:18
So, in return to ZzZombo not buying things, I'm not buying that this paradigm is impossible in Perl 6 or that the system defeats itself or whatever that gist said. 15:19
ZzZombo but nobody has shown a way to do so, did they? The suggested workaround to hack into class internals and trust subclasses this way is dirty and dosn't allow for selective exposure of certain members only. 15:21
babydrop ZzZombo: so that somehow means it's impossible? 15:22
ZzZombo: I'd show, but I have X amount of other things to take care of.
ZzZombo: or rather, I'd try to make an `is protected` trait. I have no guarantee it works, since I don't know traits.
And it'd install a check on a normally-public method to croak unless the class asked for protected attributes 15:23
cygx another approach besides noming convention: gist.github.com/cygx/a3de36abd3984...f13bd619e4
babydrop or methods. And that would be obtained from caller() on a method that must be called to obtain access to prootected stuff
callframe I mean 15:24
ZzZombo docs.perl6.org/type/$AMPERSAND$QUE..._MARKBLOCK is dead link on docs.perl6.org/type/CallFrame 15:28
babydrop m: class A { method trustme ($who) { trusts $who }; method !meow { say "meow" } }; class B is A { BEGIN A.trustme(B); method meow { self!A::meow } }
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤An exception occurred while evaluating a BEGIN␤at <tmp>:1␤Exception details:␤ 5===SORRY!5=== Error while compiling ␤ Cannot invoke this object (REPR: Null; VMNull)␤ at :␤»
babydrop something like this I envisioned
dunoo how to make it work or whatever 15:29
And we can make our own metaclasses, I'm sure we can do something like that.
Oh, I guess this is different from what I originally said. 15:30
ZzZombo That would be useful if you could arbitrarily attach a different one to a type. I've found no such way though. 15:31
15:31 firstdayonthejob joined 15:43 cibs left, khw joined
pmurias cygx: that breaks polymorphisms 15:44
15:44 cibs joined
cygx pmurias: which 'that'? 15:45
15:45 user9 left
pmurias cygx: using a global sub 15:45
15:45 wamba joined
[Coke] has a brief concern over bdf's new advent post title. 15:46
(sounds very perl fivey. :)
babydrop huggable: advent
huggable babydrop, github.com/perl6/mu/blob/master/mi...6/schedule
jnthn Object hashes? 15:47
[Coke] 9 slots still open. 9 days til first open slot (zoffix probably overburdened at 3 posts)
jnthn doesn't even know if Perl 5 has those...
babydrop [Coke]: yeah, reminds me of Perl 5 too... but I don't really know what he'll write about
jnthn I mean, I figure it's about `my %hash{Mu}`
That is, hashes with object keys
babydrop jnthn: the most common way to make an object in Perl 5 is to `bless` a hash :)
jnthn That was my first guess anyway :)
babydrop: Sure, but I don't think I've ever called that an object hash ;) 15:48
babydrop true :)
jnthn Maybe a hash object but... :)
cygx pmurias: how is that a problem for my use case? the inheritance chain is known statically 15:49
ZzZombo babydrop: that still lacks selectiveness. You trust all or nothing, so it doesn't emulate protected access very well. 15:50
15:51 user9 joined
pmurias cygx: if you inherit from B you can't override A::rename 15:51
babydrop ZzZombo: selectiveness of what?
ZzZombo what members allows access to 15:52
allow*
cygx pmurias: sure - you'd have to override make-me-a-foo
babydrop ZzZombo: in the example above yeah
ZzZombo: but I think it should be possible to make a trait that you mark methods with. Methods that in your code you'll write as public. And if they're accessed like normal public methods, the trait will croak, but if a class requested access to them, then they'll function normally, like public methods 15:54
s: &trait_mod:<is>, \(:export)
SourceBaby babydrop, Something's wrong: ␤ERR: Could not find candidate that can do \(:export)␤ in sub sourcery at /home/zoffix/services/lib/CoreHackers-Sourcery/lib/CoreHackers/Sourcery.pm6 (CoreHackers::Sourcery) line 37␤ in block <unit> at -e line 6␤␤
babydrop s: &trait_mod:<is>, \(:export, |c) 15:55
SourceBaby babydrop, Something's wrong: ␤ERR: ===SORRY!=== Error while compiling -e␤Undeclared routine:␤ c used at line 6␤␤
babydrop blahg
15:55 eisen74 joined 16:00 wamba left
nicq20 ufobat: Are you online? 16:01
16:01 wamba joined
babydrop how come we don't sign R* releases.. 16:02
cygx pmurias: having thought on it a bit longer, I agree that there are probably cases where not being virtual could be an issue
so for now, naming convention its is...
16:03 jonas1 left
nicq20 Anyone know if Bailador supports being a HTTPS server? 16:09
babydrop You can always drop it behind reverse proxy
16:11 domidumont left 16:16 Wiertek93 joined, Wiertek left
ufobat nicq20, yes i am there 16:16
nicq20 ufobat: Sorry, for bothering you. I was trying to figure out if Bailador supports being a HTTPS server. From what I can tell it does not, but I can get around that using a reverse proxy like babydrop suggested. 16:18
ufobat no worries
github.com/ufobat/Bailador/blob/ma...dor.pm#L97
it depends on the http:: whatever module you want to use 16:19
nicq20 ufobat: Oh, ok. Makes sense. :) 16:20
ufobat you could use different http backends, some work
some wont :-/
nicq20, github.com/ufobat/Bailador#get-psgi-app there is a method you need if you want to use a different http backend 16:22
nicq20 ufobat: Do you happen to know of one that would support being a HTTPS server? I looked at a few of them and it seems that none of them support it. :(
ufobat sorry, i dont know it :-( 16:23
nicq20 ufobat: That's ok, I'll just use the reverse proxy method. Thatnk you! :D
*thank
ufobat i would .. yeah exactly :)
16:24 andrzejku joined
ufobat gl with it :) 16:24
16:25 andrzejku left
ufobat does anyone know if the london workshop talks will be on youtube? 16:27
16:28 Wiertek joined 16:30 Wiertek93 left 16:32 lizmat joined 16:33 Wiertek left
ilmari ufobat: #london.pm on irc.perl.org is probably a better place to ask 16:35
16:39 andrzejku joined
babydrop "The Perl 6 organization has their default repository permission set to write. This means that every member of this organization has write access to this repository, regardless of the team and collaborator access specified below." 16:39
So what's `write` access? Does that mean anyone in P6 org can push to it?
babydrop assumes the default's good and moves on 16:41
ufobat ilmari, aye, thanks :) 16:42
16:47 Actualeyes left 16:51 Actualeyes joined
dalek b-rakudo: e8a69fe | (Zoffix Znet)++ | / (4 files):
First working version of download page viewer
 16:51
16:54 andrzejku left
dalek b-rakudo: cc124ce | (Zoffix Znet)++ | download-page.php:
Comment out commandline script usage hack
 17:00
b-rakudo: fdc68cd | (Zoffix Znet)++ | download-page.css:
Make PGP sig links less in-your-face
babydrop I'm moving star/archive/ files to star/ 'cause the download page script now clearly shows the latest release of each kind anyway and adding support for folders like that is extra work. 17:02
ummm... apparently rakudo.org is running PHP version, and I quote, "<Woet> yea.. thats been end of life for so long its not even mentioned on the end of life page" 17:09
.seen pmichaud 17:10
yoleaux I saw pmichaud 15 Sep 2016 01:34Z in #perl6: <pmichaud> m: my $input = '(\d\d\d)'; my $m = 'a 123' ~~ /$0=<$input>/; say $m
17:12 dogbert2 left
[Coke] babydrop: should perl6/web-rakudo be under rakudo/ ? probably just a stylistic thing. 17:12
babydrop .ask pmichaud any way we could touch base on rakudo.org maintenance? Could a person be given a sudo or something along those lines? We need HTTPs, upgraded Wordpress, and upgraded `php` + upgraded something else too, I'm sure 17:13
yoleaux babydrop: I'll pass your message to pmichaud.
babydrop [Coke]: probably, I mentioned the need to create it in #perl6-dev but nothing happened, so I created it where I had the perms to create it in
17:13 dogbert17 joined
[Coke] the distinction was mainly to enforce the copyright issue on the code, so it's fine. no worries. 17:14
17:17 Vynce joined
dalek c: 800e445 | (Jan-Olof Hendig)++ | doc/Type/Supply.pod6:
Clarified what happens if the interval is small. See RT #130168
 17:17
synopsebot6 Link: doc.perl6.org/type/Supply
synopsebot6 Link: rt.perl.org/rt3//Public/Bug/Displa...?id=130168
17:20 wamba left 17:21 andreoss joined
andreoss can i define "bottom" objects in perl6? (as Nil/Nothing in Scala or undefined in Haskell) 17:22
dalek b-rakudo: 4b7416f | (Zoffix Znet)++ | download-page.php:
Add array workaround to support ancient PHP
 17:23
timotimo well, every type has its type object to stand in for an undefined value
moritz is on his way to Berlin, to give a Perl 6 training
and amazingly, the airport has free wifi. I thought German province uncapable of such advances :-) 17:24
andreoss m: class Foo {}; my Foo $x = Nil; say $x.^mro 17:27
camelia rakudo-moar 843a6b: OUTPUT«((Foo) (Any) (Mu))␤»
andreoss how does this work?
babydrop andreoss: Nil is absence of value, so $x gets its default() whose default is Any 17:29
Oh
andreoss: this being what?
m: class Foo {}; my Foo $x; dd $x 17:30
camelia rakudo-moar 843a6b: OUTPUT«Foo $x = Foo␤»
andreoss m: role Abst { has $.x = die }; class Foo does Abst { has $.x = 1 }; class FooNil is Nil does Abst { has $.x = 2 }; my Abst $x = FooNil.new ; say $x, $x.^mro
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Attribute '$!x' already exists in the class 'Foo', but a role also wishes to compose it␤at <tmp>:1␤»
babydrop Oh, I guess when used that way the default is the type you specified
m: my $x is default('meow'); dd $x
camelia rakudo-moar 843a6b: OUTPUT«Str $x = "meow"␤»
andreoss by this I mean Nil being subtype of Foo
babydrop You lost me 17:31
Nil passes type checks
m: sub foo () returns Int { 42 }() 17:32
camelia ( no output )
babydrop m: sub foo () returns Int { 'meow' }()
camelia rakudo-moar 843a6b: OUTPUT«Type check failed for return value; expected Int but got Str ("meow")␤ in sub foo at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
babydrop m: sub foo () returns Int { Nil }()
camelia ( no output )
babydrop see
s/passes/bypasses/;
andreoss what if I want my own class to bypass this ? 17:33
m: role Abst { }; class Foo does Abst { has $.x = 1 }; class FooNil is Nil does Abst { has $.x = 2 }; my Abst $x = FooNil.new ; say $x.WHAT; say $x.x
camelia rakudo-moar 843a6b: OUTPUT«(Abst)␤No such method 'x' for invocant of type 'Abst'␤ in block <unit> at <tmp> line 1␤␤»
babydrop andreoss: any subclasses of Nill also bypass typecheck IIRC 17:34
17:34 wamba joined
timotimo that's right 17:36
Failure will bypass, as well
moritz m: class Foo is Nil { }; sub f(Int:D $x) { }; f Foo.new
camelia rakudo-moar 843a6b: OUTPUT«Type check failed in binding to $x; expected Int but got Nil (Nil)␤ in sub f at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
moritz m: class Foo is Nil { }; sub f(Int $x) { }; f Foo.new 17:37
camelia rakudo-moar 843a6b: OUTPUT«Type check failed in binding to $x; expected Int but got Nil (Nil)␤ in sub f at <tmp> line 1␤ in block <unit> at <tmp> line 1␤␤»
moritz only in assignment to variables, it seems
andreoss m: my Int:D $x = Nil.new
camelia rakudo-moar 843a6b: OUTPUT«Type check failed in assignment to $x; expected type Int:D cannot be itself (perhaps Nil was assigned to a :D which had no default?)␤ in block <unit> at <tmp> line 1␤␤»
babydrop Oh, only return type constraints
oh no, that's the :D failing 17:38
m: class Foo is Nil { }; sub f(Int $x) { }; f Nil
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Calling f(Nil) will never work with declared signature (Int $x)␤at <tmp>:1␤------> 3lass Foo is Nil { }; sub f(Int $x) { }; 7⏏5f Nil␤»
andreoss so Nil is not a subclass of every class, just a compiler trick?
babydrop m: class Foo is Nil { }; sub f(Int $x?) { }; f Nil
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Calling f(Nil) will never work with declared signature (Int $x?)␤at <tmp>:1␤------> 3ass Foo is Nil { }; sub f(Int $x?) { }; 7⏏5f Nil␤»
babydrop andreoss: Nil is absence of value
That's its use
ugh.... the logos aren't showing up on my download page when I host it off rakudo :S 17:40
andreoss what if I want an abstract class (A), classes which implement A (B, C, D...) and a "bottom" class (Z) which is subtype of them all
babydrop I've no idea what either "bottom" or "subtype" mean :/ 17:41
andreoss *subclass
[Coke] moritz: Perl 6 Training! Sehr güt!
moritz [Coke]: I hope it will be :-)
babydrop andreoss: ok, what do you mean by "bottom" tho? 17:42
cygx andreoss: you can't really do that in p6
andreoss en.wikipedia.org/wiki/Bottom_type
cygx andreoss: as far as I'm aware, you'd have to do something like `subset BOrBottom where B | Bottom` and use that everywhere instead of B 17:43
17:49 andreoss left 17:51 CIAvash joined, CIAvash left 17:58 labster joined
babydrop huh 17:59
TIL: I can't copy-paste files with uber-long lines into a ssh session
That's why the logos weren't showing up. The base64 background image URLs got abridged :/ 18:00
18:03 dakkar left 18:05 nicq20 left
dalek b-rakudo: 42ca829 | (Zoffix Znet)++ | download-page.htaccess-sliver:
Include domain in Redirect rule

Server on rakudo.org doesn't support domainless rule		 18:06
timotimo moritz: yeah, binding is what signatures do, and binding Nil will not do the magic "reset to default value" thing
18:06 andrzejku joined
babydrop ah 18:06
m: say 42 ~~ Nil 18:07
camelia rakudo-moar 843a6b: OUTPUT«False␤»
18:07 mohae joined
babydrop Aanndd... my stuff's live rakudo.org/downloads/star/ 18:13
huzzah 18:14
18:15 zacts left
babydrop .tell stmuk_ FYI, I've moved star/archive into just star/ since the new download page shows clearly what the latest releases are and not having to code for archive/ is less code. Also, there's no longer a need to update the `-latest-` URLs after each release, so I sent a PR to remove that from release guide 18:18
yoleaux babydrop: I'll pass your message to stmuk_.
babydrop tbrowder: rakudo.org/downloads/rakudo/rakudo-latest.tar.gz 18:19
18:23 andrzejku left 18:26 domidumont joined
babydrop oh shit 18:30
I'm blind as a duck lol :)
For some reason the page sorts different extensions together 18:31
I mean same extensions together
18:34 wamba left
babydrop s/some reason/I see why/; 18:34
18:37 pierre_ left 18:39 zacts joined
pmurias what would be a good tool to chart/track memory use of process? (node.js running rakudo.js) 18:39
dalek b-rakudo: f52241f | (Zoffix Znet)++ | download-page.php:
Use different way to filter pgp sig files

The brace glob() returns result sorted by extension groups and we don't want that.
[Coke] pmurias: stackoverflow.com/questions/200185...of-node-js ? 18:40
18:51 ggoebel left 18:54 andrzejku joined 19:03 xylixy joined 19:04 xylixy left 19:06 setty1 joined 19:08 AlexDaniel joined 19:09 pecastro joined 19:10 darutoko left
timotimo babydrop++ # fantastic download page 19:21
19:28 domidumont left
babydrop \o/ 19:30
samcv: what top bar? 19:32
19:32 Goros joined
babydrop samcv: you mean on rakudo.org/ ? 19:32
19:32 Goros left
babydrop shrugs 19:33
I'll leave it for those who like working with Wordpress :)
19:35 wamba joined
babydrop FWIW the accepted answer on this SO confuses even me: stackoverflow.com/questions/4081493...e-elements 19:36
19:38 ggoebel joined, pierre_ joined, cdg left 19:42 pierre_ left
dalek c: 7d2a711 | coke++ | doc/Language/variables.pod6:
Remove section calling $<> a $< twigil - closes #998
 19:43
synopsebot6 Link: doc.perl6.org/language/variables
19:45 RabidGravy joined
babydrop m: class Foo is Rat is Str {has $!numerator = 42; has $!denominator = 55}; my Rat $r = Foo.new 19:47
camelia ( no output )
babydrop didn't realize that would work :\
m: class Foo {}; class Bar is Foo {}; has Foo $x = Bar.new 19:48
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤You cannot declare attribute '$x' here; maybe you'd like a class or a role?␤at <tmp>:1␤------> 3 Foo {}; class Bar is Foo {}; has Foo $x7⏏5 = Bar.new␤ expecting any of:␤ constraint␤»
19:48 itaipu left
babydrop wat? 19:48
oh
:)
m: class Foo {}; class Bar is Foo {}; my Foo $x = Bar.new 19:49
camelia ( no output )
babydrop Oh ok. I see my confusion now.
19:51 cibs left
AlexDaniel m: my \selfie = π; say selfie 19:52
camelia rakudo-moar 843a6b: OUTPUT«3.14159265358979␤»
19:52 cibs joined
AlexDaniel m: my \self = π; say self 19:52
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤'self' used where no object is available␤at <tmp>:1␤------> 3my \self = π; say 7⏏5self␤ expecting any of:␤ argument list␤ term␤»
mspo m: say self.WHAT 19:54
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤'self' used where no object is available␤at <tmp>:1␤------> 3say 7⏏5self.WHAT␤ expecting any of:␤ argument list␤ term␤»
19:54 andrzejku left
dalek c: 4ba2aff | (Zoffix Znet)++ | doc/Type/ (4 files):
Use clearer language in allomorphs

  - Avoid magicalness. They're just subclasses.
  - Explicitly show they don't share object identity, in, e.g.
   set operatoions
  - Fix broken ComplexStr example that used a literal Complex
   instead of the allomorph.		 19:57
babydrop m: my \sеlf = π; say sеlf 19:58
camelia rakudo-moar 843a6b: OUTPUT«3.14159265358979␤»
babydrop doesn't see a problem :)
19:59 AlexDaniel left
dalek c: 6f9f3be | coke++ | doc/Type/Instant.pod6:
Reword Leap Second Blurb

Closes #768		 20:00
synopsebot6 Link: doc.perl6.org/type/Instant
20:01 itaipu joined 20:05 kurahaupo__ joined
stmuk_ babydrop: I would have rather the pre-releases would have remained under 'archive' since I think having them obviousily visible is just confusing 20:05
yoleaux 18:18Z <babydrop> stmuk_: FYI, I've moved star/archive into just star/ since the new download page shows clearly what the latest releases are and not having to code for archive/ is less code. Also, there's no longer a need to update the `-latest-` URLs after each release, so I sent a PR to remove that from release guide
babydrop stmuk_: you mean the pre-Christmass stuff?
20:05 kurahaupo__ is now known as kurahaupo 20:07 nchambers left, nchambers joined, girafe joined 20:08 obfusk left 20:09 obfusk joined
babydrop I'll add back the archive.. on the weekend or Friday 20:10
20:10 domidumont joined 20:12 andrzejku joined
dalek b-rakudo: 14683f3 | (Zoffix Znet)++ | download-page.htaccess-sliver:
Fix typo in rakudo `latest` URL asset name

  tbrowder++		 20:13
stmuk_ babydrop: thanks 20:14
its just I see no advantage and much risk in enabling people to easily download 2010.07 versions etc
babydrop Sure.
stmuk_ BTW I was wondering if anyone had ran a WP security scanner recently :/ 20:15
babydrop Well, I'm sure we have holes :) 20:16
We need an update of stuff
Folks in #php didn't want to help me with my script because of how ancient our php installation is 20:18
stmuk_ :) 20:19
20:23 domidumont left 20:25 eisen74_ joined, girafe2 joined 20:27 eisen74 left 20:28 girafe left
moritz hates wifi that blocks port 22 20:28
mst I remember when my father was in a hospice 20:29
I borrowed every spare 443 on shadowcat's systems to get enough tunnels out of there I could do ssh and he could use email ;)
stmuk_ tunnel over dns? :) 20:30
jonadab stmuk_: tunnel IP over IP, more like.
moritz I have a host with ssh on port 443 that I can use for tunnels 20:33
20:35 firstdayonthejob left
moritz is it coincidence that the wifi disconnected me as soon as I googled "ssh jumphost configuration"? :-) 20:40
now using tethering over the mobile phone. At least there's decent LTE connection. 20:41
20:43 wisti joined 20:44 raiph joined 20:45 firstdayonthejob joined
stmuk_ I just had to tender since a chromecast (powered off for months) went all windows update on us 20:46
20:49 skids left 20:51 Tonik left 20:58 eisen74_ left 21:04 cdg joined
yoleaux AlexDaniel: harass mst 21:11
timotimo what. 21:15
babydrop hehe :) 21:17
timotimo, irclog.perlgeek.de/perl6/2016-11-28#i_13643736
[Coke] pod6 - any idea what I did wrong here? gist.github.com/coke/1946617c44717...e5cc518098
babydrop is =pod ok instead of =end pod? 21:18
[Coke] heh. 21:19
no, just saw that.
timotimo ah
21:19 Matthew[m] left, tadzik left, ilmari[m] left, Matias[m] left, dp[m] left, M-Illandan left
MilkmanDan Nice to see that the new Llama/Butterfly KS book is getting close to funded. 21:23
21:23 andrzejku left, ChoHag left
babydrop ayeah 21:24
MilkmanDan $32,012 pledged of $37,000 goal 384 backers
21:24 TEttinger joined
MilkmanDan Any word on the new Camel Book? 21:24
[Coke] nope 21:26
21:29 bjz left
timotimo is without credit card 21:29
21:38 ChoHag joined
dudz oh nice, almost there MilkmanDan 21:39
though not enough backers 21:40
but with the monetary we can get started i reckon, keep our community small
MilkmanDan dudz: There is time to get more. :)
dudz :)
once something gets big it must be abandoned anyway 21:41
amazon, microsoft, google.
[Coke] wonders why building the docs generates so many duplicated path warnings.
Do we want to avoid duplication (fix the problem avoid the warning) or do we not care (don't emit the warning) 21:42
babydrop dudz, but those are evil corporations, not ideas :) 21:43
dudz :)
21:47 cygx left, kyclark joined
kyclark I would expect the answer to this to be 1: 21:47
m: put ((1,2)).elems
camelia rakudo-moar 843a6b: OUTPUT«2␤»
kyclark While this would be 2:
timotimo extra parenthesis doesn't change anything
kyclark m: put ((1,2), (3,4)).elems
camelia rakudo-moar 843a6b: OUTPUT«2␤»
timotimo m: put ((1,2),).elems 21:48
camelia rakudo-moar 843a6b: OUTPUT«1␤»
timotimo ^- that's what you want
kyclark Ah, hmm, OK.
I’m struggling with turing a list of lists (tuples, really) into a hash
Backing up, I have a two-column file that I want to ‘file’.IO.lines.map(*.split(/\t/) -> hash 21:49
But I have to .flat.pairup to make it work. Seems overkill.
Any suggestions?
21:50 bjz joined
timotimo hm, you really need to do the pairup? i thought a flat list of values coerced to a hash gives you key, value, key, value, ... 21:50
but it could be .flat doesn't do enough and you have to .list or something
kyclark No, you’re right. That works.
21:51 itaipu left
kyclark So that’s the best way, then? 21:51
jnthn m: my %h = flat "a\t1\nb\t2\n".lines.map(*.split(/\t/)); say %h.perl
camelia rakudo-moar 843a6b: OUTPUT«{:a("1"), :b("2")}␤»
timotimo you can split on two different separators
with a single call, i mean
kyclark Can you show? 21:52
timotimo oh, also: it'll be a thousand times faster to split on "\t" instead of /\t/
kyclark Noted. 21:53
timotimo m: my %h = flat "a\t1\nb\t2\n".split(("\t", "\n")); say %h.perl
camelia rakudo-moar 843a6b: OUTPUT«X::Hash::Store::OddNumber exception produced no message␤ in block <unit> at <tmp> line 1␤␤»
timotimo m: my %h = flat "a\t1\nb\t2\n".chomp.split(("\t", "\n")); say %h.perl
camelia rakudo-moar 843a6b: OUTPUT«{:a("1"), :b("2")}␤»
timotimo m: my %h = flat "a\t1\nb\t2\n".split(("\t", "\n"), :skip-empty); say %h.perl
camelia rakudo-moar 843a6b: OUTPUT«{:a("1"), :b("2")}␤»
timotimo take your pick
[Coke] GAH. doc build is soooooooooooo slow 21:54
kyclark Is it better than “lines”?
timotimo no
it might be better than lines + split, though
kyclark “lines” seems to self-document
timotimo that's right
kyclark Since it’s a small file, I’ll stick with lines + split
Thanks.
[Coke] it seems like every time I try to do some doc stuff I end up yak shaving.
timotimo nothing about that split call says "the \t and \n will always take turns"
babydrop Is flat even needed? 21:57
timotimo oh
not with the split invocation, i don't think
babydrop m: my %h = "a\t1\nb\t2\n".words; say %h.perl
camelia rakudo-moar 843a6b: OUTPUT«{:a("1"), :b("2")}␤»
babydrop shrugs
timotimo well, with .words your code will b0rk if there's multiple words in a single field
21:58 bjz left
babydrop #WorksOnMyComputer :) 21:58
[Coke] DrForr: how close is your module to letting us do syntax highlighting of perl 6 from perl 6? 21:59
babydrop m: my %h = "a\t1\nb\t2\n".comb: not("\n", "\t"); say %h.perl
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Calling not(Str, Str) will never work with proto signature (Mu)␤at <tmp>:1␤------> 3my %h = "a\t1\nb\t2\n".comb: 7⏏5not("\n", "\t"); say %h.perl␤»
babydrop heh
oh, right, it's none, not not :) 22:01
But it just autothreads, not uses it as a matcher
timotimo you mean none, not not 22:02
oh
you noticed
the result ought to be funny, though. one string split by one sep, the other by the other
babydrop nah, it just finds the stuff 22:03
m: dd "a\t1\nb\t2\n".comb: none("\n", "\t");
camelia rakudo-moar 843a6b: OUTPUT«none(("\n", "\n").Seq, ("\t", "\t").Seq)␤»
timotimo oh
babydrop m: dd "a\t1\nb\t2\n".split: none("\n", "\t");
camelia rakudo-moar 843a6b: OUTPUT«none(("a\t1", "b\t2", ""), ("a", "1\nb", "2\n"))␤»
timotimo i thought you had split
babydrop ehehe
that's cool
timotimo well, check this out:
m: say "cool" if "hello how are you".comb(any("a", "e", "i", "o", "u")) > 3 22:04
camelia ( no output )
timotimo m: say "cool" if "hello how are you".comb(any("a", "e", "i", "o", "u")) > 2
camelia rakudo-moar 843a6b: OUTPUT«cool␤»
timotimo m: say perl "hello how are you".comb(any("a", "e", "i", "o", "u")) > 2
camelia rakudo-moar 843a6b: OUTPUT«5===SORRY!5=== Error while compiling <tmp>␤Undeclared routine:␤ perl used at line 1␤␤»
timotimo m: say ("hello how are you".comb(any("a", "e", "i", "o", "u")) > 2).perl
camelia rakudo-moar 843a6b: OUTPUT«any(Bool::False, Bool::False, Bool::False, Bool::True, Bool::False)␤»
babydrop I don't get it 22:05
Why does it think a Bool is > 2?
oh
OK :)
Neat :)
22:06 RabidGravy left
babydrop m: say "No repeating stuff!" if .comb(.comb.any) > 1 given "There is repeating stuff here" 22:07
camelia rakudo-moar 843a6b: OUTPUT«No repeating stuff!␤»
timotimo that's not correct :) 22:08
you need unless there
babydrop well, it is if you read "No repeating stuff!" as a reprimand :)
timotimo oh
well, right
babydrop m: say "there is no repeating stuff" unless .comb(.comb.any) > 1 given "There is repeating stuff here"
camelia ( no output )
perlpilot That is some non-obvious behavior 22:09
babydrop m: say "there is no repeating stuff" unless .comb(.comb.any) > 1 given "There is repeating stuff here".comb.Bag.join
camelia ( no output )
babydrop :(
m: say "there is no repeating stuff" unless .comb(.comb.any) > 1 given "There is repeating stuff here".comb.Bag.keys.join
camelia rakudo-moar 843a6b: OUTPUT«there is no repeating stuff␤»
babydrop \o/
perlpilot: it's advanced quantum stuff :D
timotimo %)
stmuk_ [+] WordPress version 4.6.1 (Released on 2016-09-07) identified from meta generator 22:15
looks like someone did update the WP!
someone++
22:16 ilmari[m] joined
babydrop \o/ 22:16
22:16 kyclark left
timotimo hopefully with a fully functional backup 22:17
22:19 tadzik joined, Matthew[m] joined, M-Illandan joined, Matias[m] joined, dp[m] joined 22:23 kurahaupo left
babydrop huggable: advent 22:25
huggable babydrop, github.com/perl6/mu/blob/master/mi...6/schedule
timotimo yo babydrop, will you try to come up with a topic for me to write about? :S
dalek : b6f9046 | (Zoffix Znet)++ | misc/perl6advent-2016/schedule:
Add another zofspot

  "Perl 6 Is My Drummer: Playing Music With Grammars"		 22:26
babydrop Wanted to write that for perl6.party for ages, but always put it off :) Should be neat 22:27
timotimo sounds neat, yeah
will you be using PortMIDI for that?
babydrop Audio::MIDI::Note (which uses it under the hood)
Gonna try to make it play through my keyboard and then record me on guitars with Perl 6 doing the drums :D 22:28
(hence the title :P)
*keyboard as in musical instrument
timotimo :D
"through" your keyboard, as in: the keyboard receives midi and synthesizes sound?
babydrop yeah 22:29
timotimo understood
my keyboard is midi-only, so my laptop synthesizes for me
not that i have anything interesting i can input into the keyboard yet
babydrop As for the topic... dunno. I'd personally would read it if you wrote something about MoarVM guts :) 22:30
stmuk_ haha the 23rd isn't the most positive topic 22:35
22:35 AlexDaniel joined
babydrop stmuk_: AlexDaniel said that it'll be positive :) 22:35
22:46 girafe2 left
AlexDaniel moritz: by the way, I need the access to the advent thing 22:48
22:53 ufobat left
AlexDaniel do I have to create a wordpress account for that? 22:54
babydrop Yes.
AlexDaniel “To get started, tell us what your blog or website is about.”… 22:55
Fashion / Beauty, yes 22:56
timotimo haha
you could do that
nothing wrong with that, really
AlexDaniel “What would you like your homepage to look like?” – wait, I don't need a blog!
babydrop Just click "Other" and mash on the keyboard :) 22:57
hm 22:58
AlexDaniel well, have to come up with a username
babydrop AlexDaniel: /msg me your email I think I can just add you
AlexDaniel alexdaniel is taken
timotimo alexDOOMiel
22:59 cpage left, cpage joined
babydrop "Invitation sent successfully" 22:59
timotimo wait, what? 23:00
oh, email, not user name
AlexDaniel “Password must be at least 6 characters.” – okay, it didn't like my 64-char long password 23:01
To accept this invitation you will need to: 1. Signup for a WordPress.com account
ok I need an account anyway
timotimo yeah
geekosaur *golfclap* 23:02
(password)
babydrop Yeah, but it prolly won't be asking you all those questions about creating a blog.
I was hoping
23:02 Vynce left
AlexDaniel okay, here is my new blog: alexdanielprincess.fashion.blog/ 23:03
babydrop hehe 23:04
AlexDaniel babydrop: thank you 23:10
23:16 firstdayonthejob left 23:20 lukaramu left 23:21 benchable6 left, unicodable6 left, committable6 left, bisectable6 left, evalable6 left 23:27 bisectable6 joined, ChanServ sets mode: +v bisectable6, bisectable6_ joined, bisectable6__ joined 23:28 bisectable6 left, bisectable6_ left, bisectable6__ left 23:29 bisectable6 joined, ChanServ sets mode: +v bisectable6 23:30 bisectable6 left, unicodable6 joined, committable6 joined, ChanServ sets mode: +v committable6, bisectable6 joined, ChanServ sets mode: +v bisectable6, evalable6 joined, benchable6 joined, pierre_ joined 23:32 unicodable6 left, pierre_ left 23:35 Actualeyes left 23:37 unicodable6 joined 23:38 committable6 left 23:39 unicodable6 left, committable6 joined, ChanServ sets mode: +v committable6 23:42 seatek joined
[Coke] ... still building local docs 23:55
some of that time is from my commute home, but not most of it 23:57
babydrop :o
timotimo could be because of inline python vs invoking python for our syntax highlighting?
23:58 rindolf 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!