»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 Zoffix on 25 May 2018.
jnthn El_Che: Yes, I think the disconnect was that I designed precomp as a cache that you should never care about the insides of, and had nowhere near enough idea of how packaging/distribution works to know this might not really fly. Thankfully, folks more clueful on those things are taking care of that area now. :-) 00:01
El_Che \o/ 00:03
jnthn: I think of it as a necessary evil for now, that will disapear once perl6 is faster
jnthn Oh, we won't ever get rid of precomp, I don't think.
El_Che or hide it better :) 00:04
jnthn Yes, "hide it" is precisely what I was trying to do. The whole point of the SHA-1s was "you should totally not be poking in here, take a hint" :)
El_Che as it, show the original module names in errors instead of sha1
jnthn I thought that already was happening?
El_Che not on the last stacktrace I saw 00:05
jnthn Hmmm. I thought, sans --ll-exception, they were translated
El_Che lemme check
jnthn I may well be wrong on that, but at least my expectation is that they are :)
geekosaur the original name should be displayed along with the hashed one 00:06
MasterDuke jnthn: thanks, i'll take a look there
jnthn geekosaur: By default, or by using --ll-exception to ask for it?
geekosaur by default
jnthn OK, reasoning?
geekosaur that I don't know 00:07
possibly in case it was something internal/precomp-related going wrong
since it was done while precomp was still a bit troublesome
jnthn Oh, hang on, are you saying how things are, or arguing how things should be? 00:08
kkrev considering linenoise is apparently problematic and stock readline is broken on many distros I might humbly suggest you get rid of the line perl6 spits out on the console suggesting both of them. rlwrap works pretty well.
El_Che nope
jnthn I'd assumed the latter, but perhaps shouldn't have
El_Che jnthn: i see sha1 paths gist.github.com/nxadm/e84b72cc23d5...6e08ad3dd9 00:09
geekosaur I'm saying what it has been doing ever snce people complained about the hashes back then
jnthn Ah, OK
geekosaur El_Che, note the names after the hashes? 00:10
that is what was added
El_Che that's the good part
the sha1 is the one that make me grind my teeth
geekosaur ok, you;re saying the hash should be made not to be at all. suspect you will get argyument there
no I anot going to try to dig up the old discussion that led to this 00:11
jnthn Yeah, I'd somehow thought we only showed the "good part" and you got the hash under --ll-exception
El_Che geekosaur: that's what I meant by well hidden (an implementation detail)
jnthn e.g. concise by default, all detail upon request
El_Che +1
00:11 TreyHarris left
El_Che I was told the line nr matched, so for debuggin purposes it should be transparent 00:12
geekosaur but note that if you follow that path and look at that line number, you should see thecode in question
so shall everything provide full filesystem mappings so that editors that parse these things work, or is that an unnecessary detail? 00:13
it mis not present in a file called Zef::Client
not even in that directory
El_Che geekosaur: yes, it's usable and useful. But ugly
geekosaur bnecause you can have multiple versions installed
or potentially a module with the same name but a different authority 00:14
El_Che true
about: kkrev's sugestion, if someone is in touch with Linenoise or Readline authors, please ping them 00:15
as perl6 indeed recommends the modules and they both are broken at least on the latest LTS ubuntu (probably the biggest install base) 00:16
(I was thinking of adding of the 2 to rakudo-pkg) 00:17
00:25 TreyHarris joined 00:32 comborico1611 left 00:35 Sgeo_ joined 00:37 Sgeo__ left 00:40 Kaiepi joined 00:55 Kaiepi left, Kaiepi joined 01:01 markoong left 01:05 Schepeers left 01:07 Schepeers joined 01:08 kkrev left 01:35 BenGoldberg joined 01:36 molaf left 01:45 ilbot3 left 01:49 molaf joined 01:53 Kaiepi left 01:54 Kaiepi joined 01:57 ilbot3 joined, ChanServ sets mode: +v ilbot3 02:08 lizmat left 02:18 aborazmeh joined, aborazmeh left, aborazmeh joined 02:33 araraloren joined 02:35 chakli left, chakli_ joined
araraloren morning :) 02:36
02:45 chakli_ left 02:46 MasterDuke left 02:48 MasterDuke joined 02:50 araraloren left 02:51 araraloren joined 02:53 ZofBot joined, ChanServ sets mode: +v ZofBot, huggable joined, ChanServ sets mode: +v huggable, buggable joined, ChanServ sets mode: +v buggable 03:09 chakli joined 03:11 eliasr left 03:31 aborazmeh left 03:34 ZeonPeon joined 03:36 Jonta left 03:37 ZeonPeon left 04:25 TimToady left 04:28 kaare_ joined 04:32 TimToady joined 04:47 zpc left 04:50 ZeonPeon joined
ZeonPeon o/ 04:50
Is there any perl6 language construct like lisp's "quote"? 04:51
04:54 BenGoldberg left
MasterDuke ZeonPeon: what does that do? 05:07
ZeonPeon Are you familiar with lisp? 05:13
if so, it transforms lisp code into a list which can be manipulated (and turned back into lisp code with unquote) 05:16
geekosaur I'm familiar, I just don;t see how it fits. A name like $foo is *already* quoted, in that sense; you don't need syntax to prevent a name from being dereferenced.
If what you were looking for was the opposite, that does exist
oh, you mean backquotes.
there's a macro extension, which is not well defined as yet
ZeonPeon I was looking to capture a function call to be manipulated by an operator
I was hoping to make a dirty hack version of elixers pipe ( |> ) operator 05:17
AlexDaniel squashable6: status 05:20
squashable6 AlexDaniel, 🍕🍕 SQUASHathon is in progress! The end of the event in ≈6 hours. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
AlexDaniel, Log and stats: gist.github.com/cdcc39cf4c63a50c9c...ea07baa2cb
geekosaur anyway I don't think you can do that as such currently 05:22
ZeonPeon bummer, I guess it's still "early days"
geekosaur perl 6 is not lisp 05:23
ZeonPeon sadly
araraloren thinking in Perl 6 :) 05:24
ZeonPeon I guess that's a week ban?
araraloren what ?
ZeonPeon off colour jokes? 05:25
araraloren haha
geekosaur lisp is somewhat unusual in letting you capture any expression as an AST. I don't think a form as general as lisp's is ever intended. not sure it's even possible as rakudo is implemented; already things like thunking need to be special cases in the grammar, not a general mechanism 05:26
ZeonPeon I like it as a possiblility (not that I would use it effectively often) 05:28
geekosaur I don't much. mostly because I know what it costs to support it ghenerally
araraloren people always want some they may don't need it :) 05:29
ZeonPeon I know b'y 05:31
05:44 quotable6 left, quotable6 joined 05:46 kurahaupo left 05:47 kurahaupo joined 06:00 robertle joined 06:02 psychoslave joined 06:26 chakli left 06:36 TimToady left, TimToady joined 06:43 wamba joined 06:48 Tison joined 06:49 Tison left, Tison joined
Tison \o 06:49
hi guys
06:50 jmerelo joined
Tison is `knowhow` a perl6 INTERNAL keyword? I've seen it at jnthn's OO::Monitor but it is not documented. 06:51
anyway, how `knowhow` works
interesting sentence :-)
btw, our ecosystem can have modules having same name, right? IIRC we use :auth to distinguish it, but I'm not that sure 06:52
araraloren they are in different repo, such as github/cpan
jmerelo Tison: I'll have to check. Anything with HOW is usually part of the metamodel, and this one tends to be implementation-dependent. So it might be Rakudo internal 06:54
araraloren where the knowhow, I am not see it
jmerelo araraloren: github.com/rakudo/rakudo/search?q=..._q=knowHOW 06:55
06:55 damaxi joined
araraloren I mean in OO::Monitor.. 06:55
damaxi Hey :)
araraloren okay, that's fine
Tison docs.perl6.org/language/mop#Bootst...g_concerns 06:56
some sentences about `knowhow`
araraloren so, it said NQP has a .. 06:57
it's maybe not stable in mean
jmerelo Tison: yep, but it's NQP.
Tison araraloren: seems like I misremembered it :-( 06:58
jmerelo You can apparently use it github.com/perl6/mu/blob/bafe57579...d_method.t
araraloren okay
jmerelo But that code is old, so who knows. It's probably called something else now. 06:59
07:00 Sgeo joined
Tison ok, back to modules have same name. I know that they are in different repos, but do we distinguish them by `zef install <module>:auth<...>`? 07:02
..if we use zef to install module
araraloren yeah, I am not well know it
07:02 Sgeo_ left
jmerelo Tison: what else could you use? Yep, that's the way to distinguish them. 07:03
Tison thanks :P 07:04
squashable6 🍕 JJ++ wrote a comment on “Running bots without zstd leaves lock dirs behind”: github.com/perl6/whateverable/issu...-394142075 07:08
07:17 Schepeers left
squashable6 🍕 JJ++ wrote a comment on “Running bots without zstd leaves lock dirs behind”: github.com/perl6/whateverable/issu...-394142710 07:20
07:23 darutoko joined
squashable6 🍕 JJ++ opened issue “Tests hang with "... is locked, waiting"”: github.com/perl6/whateverable/issues/327 07:25
🍕 JJ++ edited issue “Tests hang with "... is locked, waiting"”: github.com/perl6/whateverable/issues/327 07:29
07:30 psychoslave left 07:33 HaraldJoerg joined 07:39 rindolf joined 07:47 domidumont joined 07:50 Schepeers joined 07:54 domidumont left 07:55 domidumont joined 07:56 Schepeers_ joined 07:58 Schepeers left, Schepeers_ left 07:59 Schepeers joined 08:07 sena_kun joined
Geth whateverable: JJ++ created pull request #328:
Adds test for correct MIME type		 08:09
squashable6 🍕 JJ++ opened pull request “Adds test for correct MIME type”: github.com/perl6/whateverable/pull/328
jmerelo squashable6: status
squashable6 jmerelo, 🍕🍕 SQUASHathon is in progress! The end of the event in ≈3 hours. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
jmerelo, Log and stats: gist.github.com/bdb24ff95e54c6362c...5afdac1a20
08:10 domidumont left 08:11 domidumont joined 08:21 domidumont left 08:22 domidumont joined 08:30 Schepeers left 08:42 lizmat joined 08:47 Schepeers joined 08:51 espadrine left 08:58 Schepeers left 08:59 Schepeers joined 09:04 Schepeers_ joined 09:05 Schepeers left 09:17 Schepeers_ left 09:18 Schepeers joined 09:25 Schepeers_ joined 09:26 Schepeers left 09:28 Schepeers_ left 09:30 Schepeers joined
squashable6 🍕 JJ++ wrote a comment on “Dockerize it?”: github.com/perl6/whateverable/issu...-394149235 09:31
09:31 Schepeers left 09:33 Schepeers joined 09:36 Schepeers left 09:37 Schepeers joined
squashable6 🍕 JJ++ wrote a comment on “Set ME env variable”: github.com/perl6/whateverable/issu...-394150769 09:59
🍕 JJ++ closed issue “Refuses to take a text/p…”: github.com/perl6/whateverable/issues/17 10:08
Geth whateverable: a828f95d28 | (JJ Merelo)++ | t/unicodable.t
Adds test for correct MIME type

This is basically part of the repo, so it's supposed to work. This would close #17
But there might be other issues here. I had to add it last, because it broke the next test, whatever that was.
squashable6 🍕 JJ++ merged pull request “Adds test for correct MIME type”: github.com/perl6/whateverable/pull/328
whateverable: b148c2a585 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | t/unicodable.t
Merge pull request #328 from JJ/master

Adds test for correct MIME type I'm going to accept this, since I have another one in the oven. It's just a commit, revert if not adequate.
10:12 telex left
Geth whateverable: JJ++ created pull request #329:
Adds Dockerfile		 10:12
squashable6 🍕 JJ++ opened pull request “Adds Dockerfile”: github.com/perl6/whateverable/pull/329
10:14 telex joined
jmerelo squashable6: status 10:26
squashable6 jmerelo, 🍕🍕 SQUASHathon is in progress! The end of the event in ≈1 hour. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
jmerelo, Log and stats: gist.github.com/5734baae11c8b3403d...0a342600f9
10:31 psychoslave joined 10:33 kurahaupo_ joined 10:36 kurahaupo_ left, kurahaupo_ joined 10:37 kurahaupo left 10:46 Jonta joined 11:05 markoong joined 11:11 jmerelo left 11:25 mscha joined
mscha m: my $r = 7/2 - 3/2; say $r; say $r.numerator, "/", $r.denominator; # wasn't this supposed to be fixed? 11:25
camelia 2
4/2
11:26 Jonta is now known as Jonta2 11:30 dakkar joined
Summertime is there a way to get perl6 to not output unicode versions of quotes around things or whatever? 11:36
makes debugging a grammar a bit of a pain
moritz Summertime: you can just call .Str on the match object to get only the matched string, and no information about captures 11:40
Summertime: there's also Grammar::Tracer
Summertime: I've also written more about debugging grammars in smile.amazon.com/dp/1484232275/
Summertime I don't mind the extra information! just some part of my environment is junked up in some way 11:42
moritz maybe fix that? :) 11:47
perlgeek.de/en/article/set-up-a-cl...nvironment 11:48
11:48 dct joined 11:51 sno joined 11:54 damaxi left
Summertime hmm 11:58
dunno how practical that'd be for me
12:01 Aaronepower left
El_Che Summertime: you may have the wrong priorities if you start adapting programs to be non utf8 :) 12:01
Summertime oh I don't doubt that, but... something's junked, and I'm not exactly sure what, and I'm not smart enough to work it out 12:04
El_Che most os'es come nowadays out of the box with urf8 support 12:09
anyway, moritz .Str is a good workaround 12:10
12:10 Kaiepi left 12:13 kurahaupo joined
lizmat or shorter: prefix ~ 12:16
12:16 Kaiepi joined 12:17 kurahaupo_ left
El_Che run a utf8 capable VM :) 12:17
jnthn Tison: knowhow is a mechanism for bootstrapping the object system. It declares a type using the most primitive existing meta-object, and in fact the only one that a Perl 6 runtime needs to provide. It knows how to have attributes and methods and type-match precisely against itself. No inheritance, no role composition, etc. All other OO features are implemented then in NQP or Perl 6 code. 12:31
Tison: There's very little use for it, unless you decide you'd like to bootstrap an object system from scratch. :-) 12:32
The main interesting observation is that we don't need to implement things like inheritance, role composition, subset types, enum types, etc. at the level of the VM. Just one very minimal object system that enables writing more interesting meta-objects is enough to base the rest on. 12:34
As it happens, roles and, to a limited degree, inheritance, are quite nice for factoring a complex meta-model, so NQP's relatively simple object system has meta-objects built on knowhow, and Perl 6 has meta-objects implemented atop of the NQP object system. 12:35
moritz and if your mind isn't bent enough yet, read "The Art of the Metaobject Protocol" by Kiczales et al 12:40
:)
jnthn My mind's so bent I don't find this stuff mind-bending any more. :P 12:41
El_Che just waiting to snap :) 12:42
P6 ook looks very nice to most people, I think. Haven't heard bad things about it 12:43
s/ool/OO/
buggable New CPAN upload: vars-0.0.1.tar.gz by ELIZABETH cpan.metacpan.org/authors/id/E/EL/...0.1.tar.gz 12:44
El_Che great job!
12:48 Schepeers left, Schepeers joined 12:51 Aaronepower joined
buggable New CPAN upload: vars-0.0.2.tar.gz by ELIZABETH modules.perl6.org/dist/vars:cpan:ELIZABETH 13:04
13:08 Aaronepower left 13:13 Aaronepower joined 13:18 AlexDani` joined 13:20 AlexDaniel left 13:22 dakkar left 13:27 Aaronepower left 13:44 daxim_ left 13:47 psychoslave left 13:48 psychoslave joined 13:52 Aaronepower joined 13:58 Aaronepower left, Kaiepi left, Kaiepi joined 13:59 Aaronepower joined 14:02 Aaronepower left, Kaiepi left, Kaiepi joined 14:04 BenGoldberg joined 14:12 john_parr_ joined, john_parr left, john_parr_ is now known as john_parr 14:25 andreoss joined
andreoss perl6.org/compilers/features this page seems to be empty 14:26
moritz I think we can stop maintaining that page; not much going on there anyway these days 14:28
14:28 zakharyas joined, zpc joined
El_Che it looks like a left over before the split of actual and historical compilers 14:30
andreoss is JVM backend historical now? 14:31
araraloren why we need so many backend
Geth perl6.org: 2c7cc21a18 | (Moritz Lenz)++ | source/compilers/index.html
Stop linking to the feature matrix

not of much interest going on there these days
moritz if anybody disagrees strongly, feel free to revert this commit 14:32
14:33 kaare_ left
andreoss for me it was just an empty page 14:33
Geth perl6.org: be2a3d8abb | (Claudio Ramirez)++ | source/compilers/index.html
remote empty referencen to feature matrix (pre split historical/present compilers
 14:34
perl6.org: c3ed280ee5 | (Claudio Ramirez)++ | 5 files
Merge branch 'master' of github.com:perl6/perl6.org
14:34 Kaiepi left
El_Che oops 14:35
we did the same commit
andreoss araraloren: I used to want JVM backend to run perl6 on Solaris
El_Che andreoss: jvm is part of the rakudo compiler
rakudo has a moarvm and jvm backend (where the moarvm is the more advanced one for now) 14:36
andreoss: rakudo already runs on Solaris. samcv did the work and tested it recently
(it used to fail on libuv in the past)
araraloren will rakudo have js backend in the future, NQP have a js backend under construction 14:37
andreoss El_Che: I've already changed the job 14:38
moritz araraloren: rakudo already has a "js" branch
araraloren oh, so I don't understand, why we need 3 backend ?
14:39 psychoslave left
moritz because you cannot rakudo Perl 6 in the browser right now (js) 14:39
moarvm because it's the most versatile one
araraloren Perl 6 will run on the browser ?
moritz and jvm because some shops are very experienced with JVM, and are suspicious of everything that's not JVM
El_Che araraloren: we don't need it. People work on what they find useful. And tha depends on what ecosystem they need to work 14:40
andreoss araraloren: it'd be much easier to sneak perl6 into prduction with JVM backend
moritz and with open source projects, the question typically isn't "do we need it?" but "does somebody implement/maintain it?"
araraloren okay 14:41
El_Che araraloren: if it makes it easier to digest, most efforts seem to be focused on moarvm 14:42
I certainly can see the value of a js and jvm backend, though
araraloren I found we build it with node.js, Is node.js is server side language ? or is my mistake 14:43
El_Che it can be used on the client
14:45 psychoslave joined, TimToady left
araraloren oh, great 14:45
thanks
14:47 TimToady joined
Summertime for / (<['"]>) <-[ ??? ]> * $0 /; how do I not match $0 in the center there? 14:48
andreoss pmd is not very happy about java code rakudo got
may be i should try to fix some of it? 14:49
jnthn What's pmd?
andreoss java's linter 14:50
araraloren m: say "aaa" ~~ /(\w)$0/;
camelia ｢aa｣
0 => ｢a｣
andreoss not actually java's 14:51
i guess it has rulesets even for perl
no, i'm wrong it has only copy-paste code detection for Perl 14:53
Summertime araraloren, looking to not match though, so, the opposite of $0 I guess? 14:55
andreoss >Stage parse : 339.055 14:56
jnthn m: say "'foo'" ~~ / (<['"]>) {} :my $stop = ~$0; [<!before $stop> .]* $stop /
camelia ｢'foo'｣
0 => ｢'｣
Summertime I'll see if I can work a lookahead in 14:59
araraloren JVM is slower than moarvm
andreoss araraloren: for Perl 6 perphaps 15:00
15:00 HaraldJoerg1 joined
araraloren hmm, yeah, only for Perl 6 15:00
15:02 HaraldJoerg left 15:08 rindolf left
andreoss 6 seconds just to warm up 15:10
15:11 rindolf joined 15:18 psychoslave left 15:19 araraloren left 15:21 andreoss left 15:24 aborazmeh joined, aborazmeh left, aborazmeh joined
El_Che Compilers 15:26
Rakudo
...
NQP
NQP (Not Quite Perl) is not a Perl 6 compiler.
that feels wrong
kurahaupo It's more like an assembler 15:28
APic Probably depends on how one defines „Perl 6“
kurahaupo Perl 5+ε
15:35 mkmk joined, aborazmeh left 15:36 HaraldJoerg1 left
Geth perl6.org: dbf248a13f | (Claudio Ramirez)++ | source/compilers/index.html
Put Rakudo and NQP in the same paragraph for compilers (NQP is not a compiler) + s/nom/master/g in github links
 15:41
rakudo.org: 80bb88123d | (Claudio Ramirez)++ | 31 files
s/nom/master/g
 15:47
15:56 kaare_ joined 16:19 jmerelo joined 16:23 Kaiepi joined 16:28 mkmk left
Geth perl6.org: 75af1865bd | (Claudio Ramirez)++ | source/compilers/index.html
typo
 16:34
perl6.org: a281da8777 | (Claudio Ramirez)++ | source/compilers/index.html
rewording multiple compiler ("there are" -> "it encourages") because there is only 1 active listed
 16:46
16:53 raschipi joined 17:07 zakharyas left 17:08 ufobat joined 17:09 zakharyas joined 17:18 raschipi left 17:19 zakharyas left
mscha I think I've got myself a Heisenbug... 17:24
m: my @m = [1,2,3],[4,5,6],[7,8,9]; for ^3 -> $i { @m[$i]<> X*= 2; }; dd @m; # doesn't work 17:25
camelia Array @m = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for ^3 -> $i { @m[$i]<> X*= 2; say @m[$i]; }; dd @m; # works after adding a .say
camelia [2 4 6]
Array @m = [[2, 4, 6], [8, 10, 12], [14, 16, 18]]
[8 10 12]
[14 16 18]
jmerelo mscha: when it works after adding a "say" it's a problem with doing stuff in sink context.
mscha I don't see how the context of the "X*=" is different in both cases... 17:26
jmerelo mscha: probably not sink context. Item context in this case. In the second case, you're reifying @m[$i] when you do "say". In the first case, you're not
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for ^3 -> $i { @m[$i]<> X*= 2; 1; }; dd @m; 17:27
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Array @m = [[2, 4, 6], [8, 10, 12], [14, 16, 18]]
jmerelo mscha: kinda similar to this docs.perl6.org/language/traps#Cont..._a_Capture
mscha This works as well, so it is indeed a context thing, but I'm not reifying @m now, am I? 17:28
geekosaur mscha, I think what goes on here is that the sink context is that of the 'for' 17:29
jmerelo mscha: you're using @m[$i]<>, decontainerizing and thus not reifying @m[$i]. That is why if you "say", it's reified. If you just did the multiplication without decontainerizing, the multiplication would reify
my @m = [1,2,3],[4,5,6],[7,8,9]; for ^3 -> $i { @m[$i] X*= 2; }; dd @m;
evalable6 Array @m = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
jmerelo it's complicated 17:31
17:31 domidumont left
jmerelo You're decontainerizing to eliminate item context. You should probably use deepmap here 17:31
my @m = [1,2,3],[4,5,6],[7,8,9]; say @m.deepmap( * * 2 ) 17:32
evalable6 [[2 4 6] [8 10 12] [14 16 18]]
mscha Without the decont it doesn't work correctly (wrong context)
m: my @m = [1,2,3],[4,5,6],[7,8,9]; for ^3 -> $i { @m[$i] X*= 2; 1; }; dd @m;
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Array @m = [6, 6, 6]
17:32 domidumont joined
jmerelo my @m = [1,2,3],[4,5,6],[7,8,9]; say @m.deepmap( * X*= 2 ) 17:32
evalable6 [[(2) (4) (6)] [(8) (10) (12)] [(14) (16) (18)]]
mscha In my actual code I'm multiplying (actually dividing) each row with a different number, so deepmap is not an option. I guess I'll go back to my nested for. :-( 17:34
m: my @m = [1,2,3],[4,5,6],[7,8,9]; for ^3 -> $i { my @r := @m[$i]; @r X*= 2; }; dd @m; # even this doesn't work 17:35
camelia Array @m = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
jmerelo mscha: if you need the row number it's going to be complicated no matter what. But think that anythink you can do with for you can do with nested maps, using adverbs.
mscha: for loops are not p6y, but they do the job.
mscha I know. But I want to write p6y code. ;-) 17:36
m: my @m = [1,2,3],[4,5,6],[7,8,9]; for @m -> @r { @r X*= 2; }; dd @m; # still no good 17:37
camelia Array @m = [[1, 2, 3], [4, 5, 6], [7, 8, 9]]
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for @m -> @r { @r X*= 2; 1 }; dd @m; # works, with a warning
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Array @m = [[2, 4, 6], [8, 10, 12], [14, 16, 18]]
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for @m -> @r { my $div = @r[1]; @r /*= $div; 1 }; dd @m; # works, with a warning 17:39
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Cannot modify an immutable WhateverCode ({ ... })
in block <unit> at <tmp> line 1
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for @m -> @r { my $div = @r[1]; @r /= $div; 1 }; dd @m; # works, with a warning
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Array @m = [[1.5], [0.6], [0.375]]
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for @m -> @r { my $div = @r[1]; @r X/= $div; 1 }; dd @m; # works, with a warning
camelia WARNINGS for <tmp>:
Useless use of constant integer 1 in sink context (line 1)
Array @m = [[0.5, 1.0, 1.5], [0.8, 1.0, 1.2], [0.875, 1.0, 1.125]]
mscha That's pretty close to what I'm actually trying to do. But I still don't really understand why it doesn't work without the dummy statement. 17:40
17:42 kwb joined
geekosaur I feel like the for is optimizing it away because it ends up in the for's result context where it is discarded, possibly because it's part of a list 17:43
we've had bugs like that before
17:43 kwb left
geekosaur and "features" (Failures don't fire in that case because why loop through a list thats's being discarded?) 17:43
mscha m: my @m = [1,2,3],[4,5,6],[7,8,9]; for @m -> @r { my $div = @r[1]; @r X/= $div; $div = 0 }; dd @m; # works, still a dummy statement, but no warning
camelia Array @m = [[0.5, 1.0, 1.5], [0.8, 1.0, 1.2], [0.875, 1.0, 1.125]]
mscha So this is a bug, right, and worth reporting? 17:44
17:44 zakharyas joined
geekosaur I would say so, although others might differ 17:45
the line between "bug" and "normal but unexpected in some cases" is sometimes blurry 17:46
jmerelo mscha: I would say it's a trap. Things get complicated when you involve different contexts (item, sink...) 17:50
dummy statement == statement in sink context. There's usually no warning, it just sinks. 17:51
17:51 pmurias joined
jmerelo mscha: it's worth the while to ask in StackOverflow; there will someone very knowledgeable who will explain what's going on. And they someone (or me) will reap the explanation and improve the documentation. 17:52
17:52 MasterDuke left
jmerelo Generated report of issues closed during the second month of the Perl Foundation grant: github.com/JJ/TPF-Grant/blob/maste...ay-2018.md 17:57
17:58 MasterDuke joined
mscha jmerelo: already reported at github.com/rakudo/rakudo/issues/1898 . If it's not a bug (I don't see how), just close it. ;-) 17:59
jmerelo mscha: don't have the privs there :-) But definitely, not a bug. 18:00
18:12 ZeonPeon left 18:13 ZeonPeon joined 18:24 AlexDani` is now known as AlexDaniel
Geth doc: 509428b100 | (JJ Merelo)++ | doc/Language/classtut.pod6
Adds doc on class attributes, closes #2089
 18:29
synopsebot Link: doc.perl6.org/language/classtut
18:38 Kaiepi left, ZeonPeon left, f3ew_ left, f3ew joined 18:39 nnunley left 18:40 espadrine joined, ZeonPeon joined 18:41 nnunley joined 18:46 skids joined
Geth doc: 9b7993bec5 | (JJ Merelo)++ | 2 files
Adds TWEAK examples, closes #1393
 18:49
doc: 5d32afafa4 | (JJ Merelo)++ | doc/Language/classtut.pod6
Links the other section on TWEAK
 18:51
synopsebot Link: doc.perl6.org/language/classtut
18:52 ZeonPeon left 19:00 domidumont left 19:08 [Sno] joined, darutoko left 19:10 jmerelo left 19:11 sno left
tbrowder_ i just saw the p6 comma IDE is now on the market--woo woo! 19:11
19:12 rjbs left
tbrowder_ commaide.com 19:12
19:32 BenGoldberg left 19:36 rjbs joined
buggable New CPAN upload: PDF-Class-0.2.2.tar.gz by WARRINGD modules.perl6.org/dist/PDF::Class:cpan:WARRINGD 19:44
19:48 lizmat left, aindilis left 19:55 tyil left 19:56 rindolf left 19:58 tyil joined 19:59 dct left 20:00 andreoss joined 20:01 rindolf joined 20:04 dct joined 20:07 zakharyas left 20:13 lizmat joined 20:29 andreoss left 20:37 wamba left, wamba joined 20:49 rindolf left, Schepeers left
Juerd Why does Promise.in take an untyped $seconds, rather than a Duration? 20:50
20:52 Schepeers joined
moritz so that you don't have to coerce to Duration, if all you've got are seconds as an integer 20:56
El_Che doesn't type trumps untyped? 20:57
Juerd moritz: But wouldn't Duration() $d do that for you? 21:03
Although I wouldn't know what if that'd be any better
MasterDuke m: sub a(Duration() $d) { dd $d }; a(123) 21:04
camelia No such method 'Duration' for invocant of type 'Int'
in sub a at <tmp> line 1
in block <unit> at <tmp> line 1
MasterDuke though i agree that a Duration multi for Promise.in seems logical 21:09
21:10 rindolf joined 21:14 dct left, aindilis joined, Kaiepi joined
moritz aye 21:22
21:25 Kaiepi left 21:27 Kaiepi joined
Geth doc: 449d18ca23 | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
Whitespace *is* allowed before the method period
 21:28
synopsebot Link: doc.perl6.org/language/5to6-nutshell
21:30 rindolf left 21:31 dct joined
Geth doc: 1e17bb83cd | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
I don't think Range still has wrinkles :-)
 21:38
doc: 9e0dfe8602 | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
Point out that "{$foo}" is not just for variables
 21:40
synopsebot Link: doc.perl6.org/language/5to6-nutshell
doc: d2b0952449 | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
for @a { } is the same as for @a <-> $_ { }
 21:46
21:49 sena_kun left
Geth doc: 09a6e6d01f | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
Elaborate a bit on blockless NEXT phasers
 21:51
synopsebot Link: doc.perl6.org/language/5to6-nutshell
21:52 wamba left 21:56 dct left 21:58 molaf left
Geth doc: aa13bf9c24 | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
open() returns a Failure and doesn't throw

Give a little example on how to to test for Failure and easily get a
  $handle assigned on the fly.		 22:00
doc: ff74285647 | (Elizabeth Mattijsen)++ | doc/Language/5to6-nutshell.pod6
Mention τ, π, 𝑒 as built-in constants
 22:05
synopsebot Link: doc.perl6.org/language/5to6-nutshell
lizmat enough for today& 22:06
22:15 pmurias left 22:18 BenGoldberg joined 22:20 ufobat left 22:27 a3f_ left 22:31 a3f joined, a3f left, a3f joined
Geth doc: b6dc8f5140 | MasterDuke17++ (committed using GitHub Web editor) | doc/Language/5to6-nutshell.pod6
Remove an extra space
 22:41
synopsebot Link: doc.perl6.org/language/5to6-nutshell
22:47 donpdonp left 22:52 ufobat joined 22:55 itaipu joined 23:02 molaf joined 23:18 ZeonPeon joined 23:19 ZeonPeon left 23:29 raschipi joined 23:33 lizmat left 23:39 Khisanth left 23:42 lizmat joined 23:49 Tison left 23:55 Khisanth joined 23:58 w_richard_w joined

Powered by Cro and the Raku® Programming Language.

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

Thank you!