»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.
Kaiepi lizmat: try making it my sub gethostname(CArray[uint8] is rw, size_t --> int32) is native('c') { * } 00:00
my $host = CArray[uint8].new(("\x00" x 255).ords); my $errno = gethostname($host, 255); my $hostname = Str.new($host.list).trim; say $errno, $hostname 00:02
Zoffix shinobi_cl: there's an open Issue R#1414 for differences between variable and parameter subset constraints. Looks like there's additional issues with `Subset @foo` parameters. I'm hoping it's just a bug and doesn't interfere with how the behaviour in the Issue is gonna be implemented. If you omit the `of` in your subset, then it seems to work (which makes me think there's a bug): m: subset Pos where * >= 1; my 00:03
synopsebot R#1414 [open]: github.com/rakudo/rakudo/issues/1414 Class attribute constraints applied on elements of array attribute, not the attribute itself
Zoffix Pos @arr = 1,2,3; say @arr.^name; sub testPosArray(Pos @arr) { say @arr.^name }(@arr)
shinobi_cl: note that looking at your code, I'd say you're over-doing constraints. I hear "type is king" in Haskell. It's not in Perl 6. Trying to constraint every little thing will likely cause tears; though I'm happy your attempts are discovering bugs in less-used corners of the compiler. 00:05
00:05 pecastro left 00:09 kent\n left
Zoffix timotimo++ # explaining getlexref 00:10
m: use nqp; my int $x; nqp::if($x,$x) = 42; say $x 00:11
camelia 42
00:12 comborico1611 left
Zoffix I guess that answers R#1416 and it can be closed 00:12
synopsebot R#1416 [open]: github.com/rakudo/rakudo/issues/1416 [perf] my int $n; nqp::if($n,1) is 4.5x slower than non-native version
00:14 comborico1611 joined 00:16 markong left
shinobi_cl Zoffix: Understood. I'm was just hoping to rely in constraints to avoid checking the data myself :) 00:20
I was* 00:21
geekosaur tbh as consumer of such an API I'd much rather actual error checking and useful messages therefrom that know what the data is, instead of requiring me to understand the program that consumes it 00:25
...even if it's a one-off for my own use (consider that one a lesson learned the hard way) 00:26
00:28 TEttinger left 00:30 yoleaux joined, ChanServ sets mode: +v yoleaux 00:39 trnh left 00:41 wamba left
Xliff_ comborico1611: No, it's that PLUS the use of a different inline language that has nothing to do with the language you were initially writing in. Kinda like mixing C and ASM, but WORSE! 00:44
At least C and ASM have relations. 00:45
comborico1611 Xliff_: that sounds bazaar.
timotimo Xliff_: don't forget that the best practice is to differentiate between presentational components and functional components 00:50
or i guess presentational and container components 00:51
and only the presentational components are supposed to have noticable amounts of html in them 00:52
Xliff_ Well, I am OK with business and storage logic in code, and an intelligent presentation layer (read templates) to handle everything else. 00:53
Quite frankly if the presentation layer is dumb, I am ok with another layer to handle display logic only.
However React seems to be a step backwards in code readability. 00:54
comborico1611: I see C and ASM in the same file like kissing cousins. Code and HTML is more like.... Trump and Ivanka. 00:55
And now, after reading what I just typed, I will consign myself to timotimo's dungeon.
(we know you have one (timo x 2)) 00:56
comborico1611: In other words, there is livable discomfort, and then there's just plain yuck. 00:57
OK...definitely moving in to timotimo's dungeon, now.
timotimo huh? you mean like "timotimo enjoys react, so they must be a sadist and/or masochist"? 00:58
i'm rather confused 01:01
01:09 aborazmeh joined, aborazmeh left, aborazmeh joined 01:14 astj joined 01:20 kent\n joined
dmaestro b2gills: thanks for the suggestion - I'll be looking further at it :-) 01:22
01:25 TEttinger joined
Kaiepi t/04-nativecall/08-callbacks.t .............. Failed 8/8 subtests 01:33
yoleaux 12 Jan 2018 22:56Z <Zoffix> Kaiepi: RE: C book. I'm currently learning using this one knking.com/books/c2/index.html It was recommended by regulars in ##c channel and it's an almost perfect book (the imperfection is exercises often re-use code from exercises of previous chapters and if you don't feel like doing a particular program, you can't just skip it, because you'll come across it later again)
Kaiepi hm
ah thanks zoffix 01:34
01:34 astj_ joined, astj left 01:36 Kaiepi left, TEttinger left 01:37 Kaiepi joined 01:38 Kaiepi left 01:39 Kaiepi joined 01:52 TEttinger joined 01:55 comborico1611 left 02:30 Nanjizal joined
Xliff_ timotimo: I was just kidding. Ignore me. 02:35
timotimo: Are you denying you have a dungeon, tho? ¬_¬ 02:37
02:44 lookatme joined, ilbot3 left, khw left 02:56 ilbot3 joined, ChanServ sets mode: +v ilbot3 02:57 mtj_ left 03:02 shinobi_cl left 03:06 mtj_ joined 03:08 shinobi_cl joined 03:12 ponbiki left, ponbiki joined 03:13 AlexDaniel joined
AlexDaniel it takes a little bit of time to get a feel for react 03:15
at first I didn't like it too
03:26 mtj_ left 03:39 traxex left 03:42 traxex joined
AlexDaniel www.reddit.com/r/perl6/comments/5g...t/dapgo5w/ 03:44
hah, that seems to be accurate?
03:54 TEttinger left 03:58 TEttinger joined, khw joined 04:02 shinobi_cl left 04:07 TEttinger left 04:09 TEttinger joined 04:14 mtj_ joined 04:21 shinobi-cl joined 04:25 faraco joined
faraco Hi. Is there any similar module or a clone/port of Perl 5's WebService::Gitter? 04:26
I want to put this into most wanted list since I found this may be very useful for rapid API wrapping development. 04:27
If there isn't any available of course.
Oh, this is the WebService::Gitter link to save some of your time searching - metacpan.org/pod/WebService::Client 04:29
04:31 Zoffix left
faraco Well, I'll just leave the repo name in the list I guess and remove it in case anyone found it is not necessary. 04:31
04:32 Cabanossi left
shinobi-cl this looks a bit too complicated... any prettier way to do it? 04:32
m: my @a = ( "1,2,3", "4,5,6"); my @b = @a.map(*.split(',').list).flat; say @b.perl;
camelia ["1", "2", "3", "4", "5", "6"]
04:33 Cabanossi joined
AlexDaniel shinobi-cl: .list seems to be unnecessary 04:38
m: my @a = ( "1,2,3", "4,5,6"); my @b = @a.map(*.split(',')).flat; say @b.perl; 04:39
camelia ["1", "2", "3", "4", "5", "6"]
04:41 khw left
AlexDaniel faraco: does not look like there's a perl 6 alternative, but… 04:41
faraco AlexDaniel: oh...but what?
AlexDaniel faraco: here's a crazy idea, it's very likely that it won't work for you: there's github.com/matiaslina/perl6-matrix-client 04:42
faraco: given that gitter seems to be bridged with matrix, maybe you can get something useful out of it?
faraco AlexDaniel: Maybe I can use this to port WebService::Gitter myself. :) 04:43
AlexDaniel: Thanks. Will see how this one will do.
Anyway, I'll put the name into the list regardless. Maybe someone else is already working on it.. 04:45
AlexDaniel faraco: yes, sure! 04:46
faraco :)
Geth perl6-most-wanted: 47e0b78812 | faraco++ (committed using GitHub Web editor) | most-wanted/modules.md
Add WebService::Client

  - A module to make wrapping internet services API, easier.
04:48
shinobi-cl what functions i can use to measure time spent in an execution of a method? 04:49
i want to add some performance tests in one .t file
AlexDaniel` Hello from the Matrix network!!! (and Matrix::Client module) 04:50
AlexDaniel wow, it works
faraco: ↑ :)
faraco shinobi-cl: docs.perl6.org/language/performanc...-_INIT_now 04:51
may be helpful
s/may/might/ 04:52
04:53 nativecallable6 left, nativecallable6 joined
faraco oooh. Decentralized communication. 04:56
05:05 faraco left 05:09 cpup left 05:13 Kaiepi left 05:16 Kaiepi joined 05:19 aborazmeh left 05:23 shinobi-cl left 05:49 cpup joined 05:50 statisfiable6 left, statisfiable6 joined 06:06 darutoko joined 06:27 Nanjizal left 07:04 espadrine left 07:26 rindolf joined 07:32 ufobat joined 07:33 wamba joined 07:46 jast left
Geth doc: c77565e445 | (Luca Ferrari)++ | doc/Language/regexes.pod6
Change output line to match exactly.
07:46
doc: 550ce558fa | (Luca Ferrari)++ | doc/Language/regexes.pod6
Itemize the benefits of non capturing groups in regex.

Also format another couple of OUTPUTs to match the exact statement.
synopsebot Link: doc.perl6.org/language/regexes
doc: dcaebf0a19 | (Luca Ferrari)++ | doc/Language/regexes.pod6
Extend explaination of nesting capture markers.
07:47 jast joined 07:52 domidumont joined
Geth doc: c9b7a70f57 | (Luca Ferrari)++ | doc/Language/variables.pod6
Minor changes to the sigilles introduction.

Remove the quotes around variabiles, move Perl 6 at the beginning of the sentence and make 'sigil' singular.
07:55
synopsebot Link: doc.perl6.org/language/variables
doc: 2e937b77f6 | (Luca Ferrari)++ | doc/Language/variables.pod6
Minor rewording in the sigilles introduction.
08:03 lowbro joined, lowbro left, lowbro joined 08:12 domidumont left 08:13 domidumont joined 08:15 stmuk joined 08:16 stmuk_ left 08:24 abraxxa joined, S007 left, S007 joined 08:30 mempko left 08:33 mempko joined 08:34 zakharyas joined 08:35 melezhik joined, S007 left, S007 joined 08:55 zakharyas left 08:56 zakharyas joined 08:59 lookatme left 09:00 zakharyas left 09:01 zakharyas joined 09:02 johnnymacs is now known as dorothyw 09:07 lowbro left, jsimonet left 09:09 abraxxa left 09:10 Krishna joined, Krishna left
Geth doc: 9cfac0f3ec | (Luca Ferrari)++ | doc/Language/variables.pod6
Place OUTPUT in context variabiles and compile time ones.
09:13
doc: 27d5f3f334 | (Luca Ferrari)++ | doc/Language/variables.pod6
Make '!' and '.' twigils consecutive sections.

Move '?' after dynamic variables, and make class-wide twigil
  ('.' and '!') together: it is simple for the reader to read about
these two sections also because they refer each other.
synopsebot Link: doc.perl6.org/language/variables
09:14 dakkar joined 09:20 jsimonet joined 09:21 lowbro joined 09:27 zakharyas left, zakharyas joined 09:30 lowbro left, lowbro joined 10:19 tyil left 10:25 markong joined
Kaiepi does LD_LIBRARY_PATH have any effect on where perl'll look for libraries for NativeCall? 10:25
melezhik Hi! Any suggestions on Perl6 module/helpers to create cli apps, exect "sub MAIN" ? 10:27
exect -> exept 10:28
markong exept -> except :) 10:31
AlexDaniel melezhik: perhaps look through Term::* modules modules.perl6.org/search/?q=TERM 10:33
melezhik markong, AlexDaniel . Thanks, will do )) 10:34
moritz also modules.perl6.org/search/?q=getopt 10:37
AlexDaniel oh yea
10:41 raschipi joined
melezhik Thanks, moritz. This one is probably is what I looked for ... 10:41
10:42 tyil joined 10:43 sena_kun joined 10:48 setty1 joined 10:56 piojo joined 10:59 natrys joined
piojo Hi 11:01
is it a known issue that IO::Spec::*.abs2rel isn't threadsafe?
(or was it fixed already?)
11:02 wamba left
piojo I'm about 80% sure that IO::Path.Str showed the same error, but it's hard to reproduce that one 11:04
11:04 zakharyas left 11:05 Zoffix joined
Zoffix piojo: IO::Path is not a thread-safe object. It's mutable 11:05
jkramer Also $bar might be a typo in the documentation: $bar defaults to $*CWD
Should be $base I guess
11:07 TEttinger left
Geth doc/post-release: 3f25ea5589 | (Zoffix Znet)++ | 2 files
Fix typo; jkramer++
11:07
piojo Zoffix: so if I call this in a threaded context, I should expect inconsistency? It looks reasonable to me: say $path.IO.basename ~ " is " ~ IO::Spec::Unix($path.Str);
I know that the .IO or the .Str is redundant, but I haven't tested what happens if I remove it 11:08
Zoffix piojo: that code is a runtime error... 11:09
piojo Sorry, the current exact line is: 11:10
say "Basename of "~IO::Spec::Cygwin.abs2rel($file-path.Str)~" is "~$file-path.basename;
evalable6 (exit code 1) 04===SORRY!04=== Error while compiling /tmp/QdiLGMzppC
Variabl…
piojo, Full output: gist.github.com/4438f8cdb612e2fd50...cdce5f596f
piojo bot is too clever :)
Zoffix piojo: and $file-path is accessed from multiple threads? 11:11
jnthn Why is IO::Path mutable, ooc? 11:12
(I'd always assumed it was value-like, something like Str is)
Zoffix Dunno, it was that way before my IO work. %!parts is mutable and is filled only when one of the methods (e.g. dirname/.basename) are accessed. $!is-absolute is also mutable and is set on first request to is-absolute 11:13
piojo Zoffix: something about the object is being modified. I'll be able to tell after I refactor it
Zoffix piojo: IO::Spec::Unix.abs2rel looks thread-safe to me.
raschipi Zoffix: Can't put a lock on that? 11:14
Zoffix raschipi: locks aren't free
jnthn It sounds like those are just cache
In which case we can 1) compute them and 2) bind the result of that computation into the attribute 11:15
Since the cached result would be immutable once computed
piojo oh, I should also mention that I'm occasionally getting weird output like "Basename of . is <the name of a file>"
jnthn Two threads may race to compute that cached data, but they'd never see incomplete data
piojo so it doesn't sound like anything I could possibly be doing wrong
there's no way I would be assigning "." to a file path string anywhere in the script. 11:16
Zoffix jnthn: ah right. Then never mind me.
jnthn Well, you're right in that a lock is heavyweight, I'm just pointing out we can achieve thread safety without one here, by the sounds of it :)
Zoffix piojo: IO::Path has .absolute method. The ::Spec:: objects isn't something you typically need to use.
jnthn I can take a look at some point 11:17
11:17 zakharyas joined
jnthn (Not now, alas...) 11:17
piojo Zoffix: now I'm just trying to verify that this is a bug and figure out how to report. I don't need this script to always have correct output
Zoffix jnthn: but I think it might already be. 'cause IIRC caching is all that's done
piojo: "this" being what? I don't see you mentioning what specific issue you were having in the log. 11:18
jnthn Zoffix: Seems it is indeed: github.com/rakudo/rakudo/blob/mast...ath.pm#L73
And it's then a Map which is immutable after creation 11:19
piojo Zoffix: I'm getting this output sometimes:
Basename of Core/UI/Screens/CTooltip_HeroIconLayout.cs is CTooltip_PlainTextLayout.cs
or this, which isn't wrong but doesn't make sense:
Basename of ../CTooltip_HeroIconLayout.cs is CTooltip_HeroIconLayout.cs
In the first case, the basename is wrong. In the second case, the path is being unpredictable (or sometimes showing as ".") 11:20
Zoffix piojo: no idea. Too tired. You can file an Issue here: github.com/rakudo/rakudo/issues/new 11:21
Zoffix drops to bed
11:21 Zoffix left
piojo Zoffix: thanks. I'll try to make it happen with a set of dummy files 11:22
11:30 jast left 11:31 jast joined
buggable New CPAN upload: P5times-0.0.1.tar.gz by ELIZABETH cpan.metacpan.org/authors/id/E/EL/...0.1.tar.gz 11:32
11:32 travis-ci joined
travis-ci Doc build failed. Zoffix Znet 'Fix typo; jkramer++' 11:32
travis-ci.org/perl6/doc/builds/331757499 github.com/perl6/doc/compare/9d07d...25ea558935
11:32 travis-ci left
buggable [travis build above] ☠ Did not recognize some failures. Check results manually. 11:33
lizmat I'm thinking of starting a #perl6-cpan5 channel in support of the CPAN Butterfly Plan
opinions?
11:33 yoleaux left 11:34 yoleaux joined, ChanServ sets mode: +v yoleaux 11:35 lowbro left 11:44 psch joined 11:46 zakharyas left 11:48 releasable6 left, releasable6 joined 12:02 Nanjizal joined 12:06 AlexDani` joined 12:07 AlexDaniel left 12:09 AlexDani` is now known as AlexDaniel 12:14 yoleaux left, yoleaux joined, ChanServ sets mode: +v yoleaux 12:25 wamba joined 12:27 abraxxa joined
buggable New CPAN upload: Sparrowform-0.0.1.tar.gz by MELEZHIK cpan.metacpan.org/authors/id/M/ME/...0.1.tar.gz 12:32
12:36 benchable6 left, coverable6 left, reportable6 left, benchable6 joined, ChanServ sets mode: +v benchable6 12:37 coverable6 joined, reportable6 joined, ChanServ sets mode: +v reportable6 12:46 wamba left 13:10 cgfbee left
buggable New CPAN upload: Sparrowform-0.0.2.tar.gz by MELEZHIK cpan.metacpan.org/authors/id/M/ME/...0.2.tar.gz 13:12
13:20 cgfbee joined 13:27 zakharyas joined 13:28 MasterDuke left 13:32 AlexDaniel is now known as AlexDanie, AlexDanie is now known as AlexDaniel 13:40 astj_ left, astj joined
buggable New CPAN upload: Sparrowform-0.0.3.tar.gz by MELEZHIK cpan.metacpan.org/authors/id/M/ME/...0.3.tar.gz 13:42
New CPAN upload: SQL-Lexer-0.2.1.tar.gz by DMAESTRO cpan.metacpan.org/authors/id/D/DM/...2.1.tar.gz
13:42 jeromelanteri left 13:43 AlexDaniel left
El_Che lizmat: something for your faq? What is someone want to port a perl5 module; shoud (s)he use a perl6 api or a straight port of the perl 5 api (even if weird in perl6)? In short, what is more useful, native or p5->6modules? 13:44
lizmat for the CPAN Butterfly Plan, the API should be as close to P5 as possible
I've used the tag "CPAN5" for these 13:45
El_Che I get that, but in general. It looks like competing efforts
geekosaur sounds like it as well 13:46
not to mention that it does neither perl 5 nor perl 6 justice
El_Che do you want perl 5 code to port to perl 6 or new perl 6 code?
ideally, both, of course
just pointing it out, because it may be something that pops up
lizmat 1. I want to lower the threshold for people coming from Perl 5
2. I want to create an ecosystem on which a Butterfly Perl 5 can be built 13:47
3. I don't care how the modules with the P5 API look inside, they could well be a shim around a true Perl 6 port
El_Che I was thinking of 3 for a faq 13:48
or rather, the other way round
(sound easier to design something native, and put a layer on top of that than the other way around)
13:49 angelds joined
lizmat El_Che: that may very well depend on the situation 13:49
and the complexity of the module
El_Che true, but if you prefer one way or the other it's better to define that in de guidelines, while not making them obligatory 13:50
lizmat ok, will do :-) 13:51
El_Che maybe it does not matter at all, but I think it may be one of the questions people may have
lizmat: you covered XS on reddit already :)
lizmat oh, which one specifically do you mean? 13:52
El_Che in the discussion with chromatic you kind of classified what was out of scope 13:53
and what was built-in in the perl6 ecosystem
introspection, exceptions, dual vars, etc
lizmat ah, that list 13:54
ok
BTW, maybe you'd care to comment on daxim's comment? 13:55
El_Che url?
(I would need to create an account for that though)
lizmat www.reddit.com/r/perl/comments/7r1...y/dsyy8hw/ 13:56
13:58 domidumont left
El_Che mention the pkgs or something deeper? 13:59
13:59 domidumont joined
stmuk github.com/stmuk/rakudo-packages 14:00
lizmat El_Che: starting with the packages would be a good start ?
stmuk I didn't reply because the question didn't make any sense
what's a "default language"?
modern computers aren't like turning on a C64 and seeing BASIC :) 14:01
El_Che stmuk: default installed
installed by default in English
I think it's more appropiate to answer the parent 14:02
stmuk El_Che: many of the linux installs ask questions about packages
I'm sure there are debian installs which don't install ruby or python
raschipi Mine for example. 14:03
stmuk there will be no redhat installs without python because their installers use it
El_Che stmuk: yes, I am wrting similar 14:05
"Living in a container world, I don't care much about a default install. You'll end up installing modules, compilers and whatnot anyway, no matter if its Perl5, Ruby or Python."
"That said..."
14:05 yoleaux left
El_Che I need to type the rest :) 14:05
14:05 cdg joined, yoleaux joined, ChanServ sets mode: +v yoleaux
lizmat El_Che++ 14:05
stmuk its a nonsense question 14:08
what does he mean "Python and Go are doing now"? python has been available for over 20 years
what self respecting perl 5 programmer would use a system supplied perl 5 anyway :) 14:09
geekosaur that is one of the things programmers learn with experience; it doesn't come prewired 14:11
14:11 astj left
geekosaur I mean "don't use system supplied X" is a faq in all of perl(5)/python/ruby/etc. 14:11
and starting to be for haskell (thanks, Arch...)
14:12 astj joined 14:14 bisectable6 left, bisectable6 joined, ChanServ sets mode: +v bisectable6
teatime Any Debian install will have both perl and python because Apt uses both 14:16
14:16 astj left
Geth ecosystem: 10c2cd9b14 | (Elizabeth Mattijsen)++ | META.list
Remove PackUnpack from ecosystem

It is superseded by P5pack on CPAN.
14:16
14:17 Kaiepi left, cgfbee left
El_Che lizmat: www.reddit.com/r/perl/comments/7r1...y/dt274qj/ 14:17
14:18 Kaiepi joined
lizmat El_Che: thanks! 14:18
geekosaur has already had to help someone try to recover from installing the wrong upgraded module into system perl5 on debian and blowing apt's brains out 14:19
raschipi teatime: I can't find a depedency from apt/dpkg on python on my desktop, where is it used? 14:20
El_Che lizmat: idea is not to be confrontational, as the question was slightly weird but not aggressive 14:21
stmuk raschipi: my guess is its probably an apt backend
raschipi Cumulus Linux has a project called ifupdown2 that can't susbstitute ifupdown in Debian because it's written in Python basic system can't depend on Python. It could if it was in Perl 5. 14:22
buggable New CPAN upload: P5pack-0.0.1.tar.gz by ELIZABETH cpan.metacpan.org/authors/id/E/EL/...0.1.tar.gz
raschipi Missed a comma.
stmuk anyway I still think a "default language" is pretty meaningless and not very useful for the average programmer on a modern linux system 14:23
teatime I mean, it's something I def. think about when writing something like an installer script
but I missed the original discussion
stmuk it was used basically just a way of bashing perl 6 on reddit 14:24
14:24 cgfbee joined
raschipi I need to use system perl to do system administration, because documenting a non-default perl isn't worth it. Well, I wouyld have to use it at least to keep the other installation going, so it would be wasted effort anyway. 14:24
mspo default/system perl ruined perl just like it ruined python
linux distros hate software, afaict 14:25
El_Che it sounds more like "mspo hate linux distros, afaict" :)
mspo also true 14:26
moritz hates software, hardware and meatware
there, I said it :-)
lizmat hugs moritz
mspo El_Che: they should have called it rh-perl and rh-python, just like rpm forked its own version of bdb
El_Che: then there would have been much fewer problems
but the whole linux distro philosophy is .. 14:27
El_Che mspo: that's actually a good idea
mspo nevermind
El_Che: no shit
El_Che: then yum update python doesn't destroy your system because yum depends on python 14:28
raschipi Debian has a project to put upstream maintainers in control of releasing software in Debian itself, automating the distro gates, but there has to be a test suit hooked into autopackagetest. The package would be built and if the tests pass, the new version would be submitted to the archive immediately.
Of course all that depend on it would have to be tested too.
teatime in my experience developers don't understand the benefits of package management enough and sysadmins don't understand developers' needs enough
mspo or packages, like apt/yum, that depend on perl/python should be statically linked 14:29
raschipi teatime: sysadmins understand their needs perfectly well, it just say no.
mspo perl5 even has methods to do that!
teatime mspo: why?
timotimo not being included by default hasn't harmed nodejs's appeal, has it?
teatime raschipi: hmm?
mspo timotimo: it enhances it because you can actually update it 14:30
timotimo well, yeah, in your world, sure. but this is arguing against the "perl 6 sucks because it's not preinstalled on the 100 megabyte debian netinst image" viewpoint
14:30 st_elmo joined
teatime your best bet w/ e.g. python is leave the system python alone, completely, and use venvs / private python installation for your developed code. 14:31
timotimo ^ obvious reductio ad absurdum, of course
teatime but again, sysadmins and developers both think the other group is stupid. and are both usually wrong.
mspo teatime: yes I know and it's terrible
teatime mspo: I don't see how to improve it. 14:32
mspo teatime: although, ironically, redhat thinks pyenv is such a good idea that they have duplicated it in the form of software collections
stmuk I do see advantages to using system languages sometimes .. at least they are likely to get security patched .. unlike that old 5.8.7 install noone dare touch
mspo teatime: so now to use newer *git* you have to run a special bash :)
speaking of pyenv, is there a way to get out of it? 14:33
14:33 yoleaux left
raschipi Since CVE-2002-0059, static linking is completely forbidden in Debian. 14:34
mspo stmuk: as someone who had to learn shell scripting on solaris 7 I'm not sure about patches :)
raschipi: no one tell golang :)
anyway I am highly biasted to the BSD method of base and packages/ports being different things 14:35
raschipi My system tells me "Packages that depend on golang (0)". 14:36
mspo I consider docker a reaction to linux distro complexity and language packaging problems
raschipi Having system perl being in a different package has the precedent of it being done with GCC. 14:37
El_Che mspo: you'd be surprised how many bad practices are migrated fron hw/vm to a container
raschipi Yep, containers are just a way to enable bad practices.
For an example, I gave a CVE above. 14:39
stmuk I did experiment a bit with a linux appimage of rakudo star but appimage didn't seem to fit programming language use very well 14:41
seems to be intended just to run a single app .. so it worked well with the REPL but little else
El_Che stmuk: it looks to me like a base that appimage requires
stmuk I couldn't see an easy way of running a program outside the appimage 14:42
El_Che coffee 14:43
rakudo is compiling on on opensuse, let's see how far we get on this run 14:44
so far so good
geekosaur there are some fairly significant misunderstandings here about the point of distributions and the differences between end users and developers (and between different styles of development, too!)
mspo the point of distributions? 14:45
teatime raschipi: I am bored of looking, but package 'apt' suggests python-apt (which depends python), and package 'python' is priority standard.
mspo computers, in general, are for solving business problems :)
stmuk "apt" is a new front end to "apt-get"
raschipi Right, a suggestion, not a depedency.
geekosaur yes. and many businesses want or even need stable platforms.
teatime raschipi: I am fairly sure you can't not have python on debian, but perhaps I am wrong it is only a few niche apt-* utils and scripts that would break.
geekosaur distributions are stable platforms 14:46
teatime stmuk: they're both front ends to the apt system, really.
mspo geekosaur: they are scoped too broadly in that stability
raschipi I know you can have Debian without Python, it's standard procedure for me.
mspo geekosaur: and violate it constantly in the name of "security"
teatime it's frustrating because now you can't say 'apt' to mean the whole system w/o people thinking you mean the new command.
geekosaur afaik debian's need for python is not for apt, it is that a lot of "standard" programs/utilities use it. if you avoid those, you can install debian without python
14:47 domidumont left
raschipi Yes, it has to be a stripped down system. 14:48
stmuk apt(1) is new 14:49
mspo geekosaur: I believe in busybox I guess :)
14:50 jeromelanteri joined, wamba joined
teatime stmuk: when I said it's needed for apt, I was not referring to apt(1). 14:50
stmuk I think there are probably some backends which use python
14:51 domidumont joined
ilmari stmuk: for values of "new" including oldstable 14:51
teatime oh, maybe I'd have more luck looking at rdepends for python3; they're working to get python2 out of the default install
14:52 eliasr joined
mspo bringing it back around; perl (5) as a "better awk" is great to have around and I use it all the time 14:52
but I wouldn't want to run my business on the "default install" since I know it would eventually not be a good choice
raschipi It is installed by default but programs on the minbase install are forbiden from depending on Python.
mspo since I would need a newer $whatever or an older $somethingelse because $business and $techdebt 14:53
14:57 astj joined
stmuk arguing about whether debian needs python or not at least covers some tech issues whereas arguing about Perl 6 naming doesn't :) 15:00
15:02 astj left
[Coke] one of the docs has a variable called "þing" in a snippet; should probably avoid the unicode there unless that word means something. (also it fails the spelling test, so if we keep it, needs to be added to the code word listing) 15:10
15:11 bwisti joined
raschipi Here is a list of all the required packages in Debian currently: gist.github.com/anonymous/daa39961...4b476ad0c4 15:11
It has Perl 5 but no Python. 15:12
15:16 zakharyas left 15:17 zakharyas joined
teatime raschipi: I take it that's the essential + priority=required packages? 15:19
raschipi I just installed minbase and that's what I got. 15:20
teatime oh. well I stand corrected, sorry for spreading misinfo
raschipi Policy changed and now essential and required packages can depend on not-essential/not-required packages, so just listing essential and required doesn't actually show what's in the minimal set anymore, you need to solve their depedencies. The easier way of doing that is performing the minbase bootstrap. 15:22
teatime yeah, I know actually; I was just now struggling to do exactly that (resolve their deps) 15:23
El_Che lizmat: I wonder if it's worth answering in a 300 msg reddit thread. Do people keep reading it?
teatime I expected to be able to do it w/ just aptitude search but I think I'd have to aptitude | apt-rdepends instead 15:24
El_Che Reddit makes it impossible to follow a discussion without re-reading most of it
lizmat El_Che: well, there's that
if you think it's an important enough point to make :-)
El_Che lizmat: like "someone is wrong on the Internet"?
:)
raschipi It's perfectly possible to follow older discussions in reddit as long as you pay them. 15:25
El_Che raschipi: I mean fresh ones
raschipi Yes, paid subscribers have a feature to highlight new comments in a thread. 15:26
El_Che those people must love to argue on the net :)
15:27 angelds left 15:28 yoleaux joined, ChanServ sets mode: +v yoleaux 15:40 wictory[m] left, nebuchadnezzar left, timotimo left, ab5tract left, ab5tract_ joined, nebuchadnezzar joined, timotimo joined 15:41 peteretep joined, skaji joined, mithaldu_ joined 15:42 releasable6 left, nativecallable6 left 15:43 releasable6 joined, coverable6 left, unicodable6 left, greppable6 left
lizmat afk for a few hoursd& 15:44
15:44 tadzik joined 15:45 wictory[m] joined 15:46 unclechu joined 15:50 itaipu joined 15:52 zakharyas left 15:53 yoleaux left 15:54 yoleaux joined, ChanServ sets mode: +v yoleaux 15:58 ggoebel joined 16:06 Jjh joined
Jjh I have a scalar that may be a string or Nil. I attempt to test for Nil, but I always get the ‘Use of Nil in string context.’ What am I doing wrong? 16:10
raschipi Jjh: Can you give us a gist?
Jjh Gist? Do you want a code example? 16:12
raschipi Yeah, code example
Jjh ($s ~~ Nil) ?? ‘No value’ || $s; 16:14
16:16 psch left
raschipi try (defined $s) ?? $s !! 'No value' 16:17
timotimo Jjh: the problem is most likely that assigning Nil to a scalar container will reset the scalar container to its default value, which is "Any" unless you defined a type or default for it
raschipi $s //= 'No value'
timotimo m: my $foo = "hello"; $foo = Nil; say $foo.perl
camelia Any
Jjh Wow! You da man! That worked! Thanks! 16:19
I used the defined test, just like Perl 5. 16:20
stmuk reddit like usenet only not as good 16:21
El_Che a mailing list archive would as least mark the visited pages as such 16:23
16:25 Jjh left
El_Che releasable6: status 16:28
releasable6 El_Che, Next release will happen when it's ready. Blockers: github.com/rakudo/rakudo/issues?q=...%9A%A0%22. Unknown changelog format
El_Che good 16:29
16:31 cdg_ joined 16:34 cdg left, setty1 left 16:37 chakli left
El_Che if you google "reddit perl", below the first link (reddit) you get the first indented sub result: "Is Perl a Dead Language?" 16:37
pretty awful marketing
16:37 chakli joined
timotimo great. 16:37
El_Che the second on the right? "Is learning Perl still worth it?" 16:38
Followed by "Learning Perl", "Perl 5 v 6", "Why is Perl considered awful?" and "Perl 6" 16:39
the Perl 6 thread is nice, but the rest is pretty awful
tadzik is there a precedent for a language that used to be considered awful but now people like it? 16:41
timotimo javascript?
raschipi C 16:42
El_Che php7 did a log of good for the lang 16:43
16:43 troys joined
El_Che I know that I will annoy stmuk, but the google result for reddit+perl is a pretty good argument for renaming this hipster language :) 16:44
raschipi Gotta convince TimToady
El_Che raschipi: I am pretty hopeful he's lurking :) 16:45
16:45 khw joined
stmuk so all the reddit posts are "perl a language so rubbish they had to rename it"? 16:45
16:46 chakli_ joined
moritz the world doesn't just consist of reddit 16:46
El_Che stmuk: that would be perl 5 problem then, but I am pretty sure many perl 5 people would take that gladly if it means freeing the major version number
16:47 chakli left
mspo moritz: the world is reddit + HN 16:47
obviously :)
El_Che moritz: I just created my account a few hours ago :)
moritz: because lizmat strong muscled me to answer :) 16:48
(jk)
mspo I'd argue that perl was a dead language, but now it isn't
or doesn't have to be since perl6 is moving
stmuk this Perl 7 is myth 1. Larry has said he wants Perl 7 2. Perl 5 people have no idea what to with a new major version 3. Perl 5 is unmaintable anyway and they can't even add a simple "given/when" 16:49
mspo so it took 15 years to stabilize; shit happens :)
stmuk ^ do with
raschipi Yep, it will be just, new perl, just the same as old perl
El_Che stmuk: it may be. But it's up to Perl5 to succeed or fuck it up
so, in the worst case you'll be proven right, in the best plesantly surprised (or the other way around, depending on your stance) 16:51
raschipi Don't try to help, though, you'll be accused of trying to kill Perl5 just for mentioning there's a problem.
stmuk its up to them now, messing around with names will do nothing to help
mspo I watched that youtube clip and it seemed more like a "I'd rather not deal with it" or "I'm too old for this shit" than anything :) 16:52
I guess being BDFL isn't always easy
16:52 zakharyas joined
El_Che raschipi: I th ink many perl people think there is a problem. But the reason of the problem varies. If the naming thing is out of the way, then there is no scapegoating/blocking factor any longer 16:53
stmuk El_Che: no they will just say you didn't change the name fast enough
El_Che stmuk: and at that point, you don't care while you're busy promoting this new cool hipster language 16:54
mst I'd note that given/when went wrong because we actually tried to be compatible with perl6 - but then the perl6 team spotted problems with the original smartmatch design, substantially redesigned it, and somehow that didn't get noticed until larry tried 5.10.0 the day after it was released
so that was 100% an organisational screw up rather than saying anything much about the maintainability of the perl5 VM 16:55
(there are plenty of valid criticisms of the maintainability of the perl5 VM, mind, just stmuk's attempt at an example isn't one at all)
El_Che mst: thx for the input. I didn't know that
stmuk I think the real reason they dislike perl 6 is they are scared of perl 6 and becoming beginners again in something they don't currently understand
El_Che stmuk: I think you're overestimating "them" 16:56
a lot of people use both langs
and want to keep using them
as both perl5 and 6 have weaknesses and strengths
mst that's definitely the line I follow 16:57
El_Che me too
16:57 abraxxa left
mst for a start, perl5 is very much a creature of *n?x in a way perl6 isn't, and that absolutely has advantages and disadvantages in both directions 16:58
timotimo maybe we should be promoting the F out of perl 6 on windows 16:59
stmuk timotimo: after the tests pass 100%! :) 17:00
timotimo sure
and some more bindings for winapi stuff
El_Che timotimo: windows is pretty tied to .net
if you're running windows, there is a very big chance you're running c#
timotimo you mean we'll have a hard time against c#? 17:01
El_Che being an order of magnitude faster than perl 6, I won't bet too much on it :)
c# is well integrated
timotimo c# also runs on linux, so we're already lost anyway :P
El_Che IDE's, courses, books, ...
train arrives, bbl 17:02
stmuk I once looked in the mono room at FOSDEM (when they had one) and there was noone in it :)
timotimo CoreCLR should have done a lot of good in this space
17:04 wamba left
stmuk note above I said "and they..." not "because they .." 17:04
Geth doc: 439ae20352 | (JJ Merelo)++ | doc/Language/operators.pod6
Changed example refs #1739

It might be convenient also to use different examples that would cover different use cases, not only grep and map, but other "chaining" operators too.
synopsebot Link: doc.perl6.org/language/operators
stmuk the Perl 5 community remind me a little of the Amiga community in the early 90s .. if only things were "marketed better" we would have more users 17:05
timotimo i wasn't around for it then, but wasn't amiga literally the best computer platform you could get? 17:06
stmuk in a lot of ways yes ... and they had dedicated users ... but events moved on
17:08 bmusin joined
raschipi It was just impossible to beat the contracts Microsoft signed with OEMs, except for making own hardware. No amount of marketing can overcome that. 17:10
17:10 zakharyas left, ChoHag left 17:11 zakharyas joined
stmuk hmmm I wonder how to sign R*? I tending towards just signing a file containing a SHA256 of the tgz rather than signing the tgz itself 17:17
moritz why?
that makes it harder to verify with a tool
if you sign the .tar.gz, you can just throw gpg --verify on it
stmuk but more useful to people without a tool
which is probably a larger number of people 17:18
moritz you mean, anybody will verify the signature with manual math?
stmuk uh?
moritz people without a tool?
stmuk without your tool gpg and an imported copy of the public key
moritz I don't see how you can verify signatures without a tool 17:19
and those who care about signature verification tend to have gpg installed
stmuk most people won't bother to get the public key and can just verify with sha256sum 17:20
moritz so, provide both?
teatime stmuk: aren't you just setting people up for fail by expecting them to verify the sha256sum but ignore the fact that it's gpg signed?
although it's still useful for corrupt downloads etc., I guess
stmuk I can provide both I suppose
moritz the sha256 checksum (for transmission errors) and the gpg signature of the tarball?
buggable New CPAN upload: Terminal-Getpass-0.0.4.tar.gz by TITSUKI cpan.metacpan.org/authors/id/T/TI/...0.4.tar.gz 17:22
stmuk actually perl5 only supplies checksums and nothing is signed AFAIK
moritz: yes 17:23
raschipi Debian signs the sums file instead of the actual content for performance, but I don't think that kind of performance is necessary in this case.
moritz raschipi: Debian also automatically verifies the signatures, which means the indirection doesn't hurt usability 17:24
raschipi Yes, that's why speed is important there but not for Perl6. 17:25
Also allows to check integrity of specific files instead of the whole. 17:29
For when it's already installed.
stmuk anyway I'll be using pgp.mit.edu/pks/lookup?op=get&s...6197190DD5 to sign R* 17:31
timotimo should we do something to R* to use tarballs from cpan instead of git submodules for our modules? 17:32
stmuk timotimo: yeah that did occur to me today 17:33
17:34 geospeck joined
timotimo i see you're already one step ahead of me :) 17:34
stmuk I swept it under my mental carpet :) 17:35
17:35 jeromelanteri left
jkramer Is there a nice way of mass-assignment to rw attributes of objects in an array? Something like @foo>>.color = $new-color; maybe? :) 17:35
Oops sorry 17:36
timotimo yeah, you can either hyper or X the assignment, too
m: class Col { has $.color = <red green blue>.pick }; my @cols = Col.new xx 10; say @cols; @cols>>.color >>=>> "rainbow"; say @cols
camelia 5===SORRY!5=== Error while compiling <tmp>
Missing << or >>
at <tmp>:1
------> 3new xx 10; say @cols; @cols>>.color >>=>7⏏5> "rainbow"; say @cols
expecting any of:
infix
infix stopper
timotimo m: class Col { has $.color = <red green blue>.pick }; my @cols = Col.new xx 10; say @cols; @cols>>.color X= "rainbow"; say @cols 17:37
camelia [Col.new(color => "green") Col.new(color => "blue") Col.new(color => "red") Col.new(color => "green") Col.new(color => "red") Col.new(color => "blue") Col.new(color => "blue") Col.new(color => "green") Col.new(color => "red") Col.new(color => "red")]…
timotimo too much text, hold on
m: class Col { has $.color = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols>>.color X= "rainbow"; say @cols
camelia [Col.new(color => "blue") Col.new(color => "green") Col.new(color => "green") Col.new(color => "red")]
Cannot modify an immutable Str (blue)
in block <unit> at <tmp> line 1
timotimo m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols>>.color X= "rainbow"; say @cols
camelia [Col.new(color => "blue") Col.new(color => "green") Col.new(color => "blue") Col.new(color => "red")]
[Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow")]
timotimo there we go
17:38 troys is now known as troys_
timotimo if you want to hyper the = operator you'll have to put brackets or use »=» instead of >>=>> because => wins over = in the middle there 17:38
m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols>>.color »=» "rainbow"; say @cols
camelia [Col.new(color => "blue") Col.new(color => "blue") Col.new(color => "green") Col.new(color => "green")]
[Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow")]
timotimo this is with »=» ^
m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols>>.color >>[=]>> "rainbow"; say @cols 17:39
camelia [Col.new(color => "red") Col.new(color => "red") Col.new(color => "blue") Col.new(color => "red")]
[Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow")]
timotimo this is with >>[=]>> ^
jkramer: looks good? :) 17:40
m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols>>.color>>.& "rainbow"; say @cols 17:41
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3l.new xx 4; say @cols; @cols>>.color>>.&7⏏5 "rainbow"; say @cols
expecting any of:
infix
infix stopper
statement end…
timotimo oops
jkramer That's a rediculously long operator :D
17:41 setty1 joined
timotimo m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols>>.&{ .color = "rainbow" }; say @cols 17:41
camelia [Col.new(color => "red") Col.new(color => "blue") Col.new(color => "green") Col.new(color => "red")]
[Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow")]
jkramer I'll probably just do >>.&{.color = $new-color} :) 17:42
timotimo m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols.map(*.color = "rainbow"); say @cols
camelia [Col.new(color => "green") Col.new(color => "red") Col.new(color => "blue") Col.new(color => "red")]
Cannot modify an immutable WhateverCode ({ ... })
in block <unit> at <tmp> line 1
timotimo whoops
m: class Col { has $.color is rw = <red green blue>.pick }; my @cols = Col.new xx 4; say @cols; @cols.map({.color = "rainbow"}); say @cols
camelia [Col.new(color => "red") Col.new(color => "green") Col.new(color => "blue") Col.new(color => "blue")]
[Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow") Col.new(color => "rainbow")]
timotimo but yeah, timtowtdi
17:43 evalable6 left, evalable6 joined, ChanServ sets mode: +v evalable6 17:44 cdg_ left 17:45 dakkar left
jkramer Gnaaah 17:45
m: my $foo = Array.new; my @bar = $foo; @bar.perl.say # this gets me every time
camelia [[],]
17:49 chakli_ left, chakli joined, troys_ is now known as troys 17:52 cdg joined 17:53 chakli left 17:54 mempko left 17:55 ChoHag joined, Kaiepi left 17:56 Kaiepi joined, chakli joined, AlexDani` joined
timotimo i'm tellin' ya, those scalar containers, man ... 17:58
raschipi Better than than yucky references 17:59
18:08 AlexDani` is now known as AlexDaniel
stmuk in the absence of PSIXDIST I assume only people registered on PAUSE are currently uploading perl6 modules to CPAN? 18:13
18:19 konsolebox left, xinming left, cdg_ joined 18:21 cdg left 18:22 chakli left 18:23 domidumont left 18:24 konsolebox joined 18:25 Kaiepi left 18:34 Kaiepi joined 18:36 zakharyas left 18:38 domidumont joined, setty1 left 18:39 domidumont left 18:43 MasterDuke joined 18:44 kaare_ left, domidumont joined 18:46 domidumont1 joined 18:48 Kaiepi left 18:49 Kaiepi joined, domidumont left 18:51 domidumont joined
buggable New CPAN upload: P5each-0.0.1.tar.gz by ELIZABETH cpan.metacpan.org/authors/id/E/EL/...0.1.tar.gz 18:52
El_Che AlexDaniel: When 2018.01 is out, we'll have also openSUSE packages
18:53 domidumont1 left 18:54 vichib joined 18:55 vichib left, st_elmo left 18:58 domidumont left, Nanjizal left, darutoko left, domidumont joined
AlexDaniel El_Che: oh no, openSUSE packages! Stop the release process! This is unnacce… ah wait, that's a positive :) Cool! 19:04
19:04 wamba joined
El_Che AlexDaniel: stop me when I start to creat Lindows packages -> betanews.com/2018/01/02/lindows-fr...ire-lives/ 19:05
stmuk I've a hat here which would support you el_che! 19:07
"Windows MSI Builder" :)
suits you sir! 19:08
El_Che :)
19:08 domidumont left, domidumont joined, domidumont left
El_Che the AD a powershell instructor coming over for a week. He had a laptop full with Microsoft stickers. Someone, as hackers go, he a pretty-fly-for-a-white-guy vibe 19:09
19:09 domidumont joined, domidumont left, domidumont joined, cdg_ left
stmuk actually there seems to be something of a Windows renaissance with ubuntu subsystem etc. And I saw quite a nice powershell Space Invaders 19:10
19:10 domidumont left
El_Che Modern Windows 19:11
stmuk :D
mspo well osx is dropping the ball 19:13
stmuk I gave up on os x when it didn't let me dtrace a sh script 19:14
El_Che the osx machine that I have is mostly a typewriter for the train. But friends that are apple fans are really worried that they'll end with iOS in their macbooks pros
19:14 domidumont joined
El_Che stmuk: dtrace was a lot of fun on solaris 19:16
stmuk: well not fun, but useful :)
stmuk I did wonder if it could give insight into MoarVM 19:20
El_Che does moarvm run on systems with dtrace at all? 19:22
19:24 ojkl joined
stmuk FreeBSD 19:24
19:25 ojkl left
El_Che nice, I thought the *BSDs were a work in progress still 19:26
stmuk os x for scripts that don't use /bin and whatever in the shebang line
for fear of revealing Apple secrets
El_Che omg
I don't think that MS would go that far 19:27
19:29 ChoHag left 19:36 quotable6 left, quotable6 joined, ChanServ sets mode: +v quotable6 19:40 domidumont left, domidumont joined 19:43 kaare_ joined
bmusin Hi. Where zef installs packages? Thanks. 19:48
MasterDuke bmusin: i believe in ~/.perl6 19:50
19:52 ChoHag joined 19:54 natrys left 20:00 domidumont1 joined 20:01 raschipi left 20:03 domidumont left 20:04 troys is now known as troys_ 20:18 wamba left 20:20 stux|RC-only joined 20:22 mempko joined 20:23 domidumont1 left 20:29 softmoth joined 20:31 bmusin left 20:32 vike left, vike1 joined 20:34 masak joined, masak is now known as Guest10951, wamba joined 20:36 Kaiepi left, Kaiepi joined 20:37 domidumont joined 20:45 stmuk_ joined 20:47 stmuk left 20:50 Kaiepi left 20:51 andrzejku joined 20:56 tyil left 20:59 andrzejku left 21:00 Nanjizal joined, andrzejku joined 21:02 tyil joined 21:06 domidumont left 21:11 TEttinger joined 21:13 astj joined 21:14 pecastro joined 21:16 cdg joined 21:17 Nanjizal left, astj left 21:30 softmoth left, softmoth joined 21:31 itaipu left 21:33 jast left 21:46 cdg left 21:47 geospeck left 21:57 rudo joined 22:01 softmoth left
rudo Maybe someone can explain this to me. I played around with the Advent post: perl6advent.wordpress.com/2017/12/...of-perl-6/ paragraph 5. why does this work: 22:06
p6: sub d(*%data) {1}; d(1=> 1)
camelia Too many positionals passed; expected 0 arguments but got 1
in sub d at <tmp> line 1
in block <unit> at <tmp> line 1
rudo p6: sub d(*%data) {1}; d(x=> 1)
camelia ( no output )
rudo so why does the second work and not the first? 22:07
gfldex m: dd 1 => 1; dd x => 1; 22:09
camelia 1 => 1
block <unit>()
gfldex m: (1 => 1).^name.say; (x => 1).^name.say; 22:10
camelia Pair
Pair
moritz the <thing> => <otherthing> form is a named argument only if <thing> is a valid identifier
so f x => 1 calls f with a named param x 22:11
gfldex m: sub d(*%data) {1}; d(|(1 => 1))
camelia ( no output )
moritz but f 1 => x calls f with a positional argument which happens to be a Pair 22:12
m: sub f(*@pos, *%named) { say "pos: @pos.perl(); named: %named.perl()" }; f x => 1, 1 => x;
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
x used at line 1
moritz m: sub f(*@pos, *%named) { say "pos: @pos.perl(); named: %named.perl()" }; f x => 1, 1 => 'x';
camelia pos: [1 => "x"]; named: {:x(1)}
22:14 pecastro left 22:15 trnh joined, andrzejku left 22:17 mempko left
rudo ok, the slurpy hash expects named arguments. I get it now. thanks for the explanation 22:17
Geth doc: gfldex++ created pull request #1740:
1=>1 vs. a=>1 as named arguments
22:20
gfldex rudo++ # for asking good questions 22:21
22:22 squashable6 left, committable6 left, bloatable6 left, squashable6 joined, unicodable6 joined, ChanServ sets mode: +v unicodable6, committable6 joined, ChanServ sets mode: +v committable6, bloatable6 joined 22:23 ChoHag left 22:24 ChoHag joined 22:29 pecastro joined 22:30 andrzejku joined, Nanjizal joined
Geth doc: 9e66f47c20 | (Wenzel P. P. Peppmeyer)++ (committed by Zoffix Znet) | doc/Type/Signature.pod6
1=>1 vs. a=>1 as named arguments (#1740)
22:36
synopsebot Link: doc.perl6.org/type/Signature
22:40 espadrine joined, yoleaux left 22:41 yoleaux joined, ChanServ sets mode: +v yoleaux 22:45 pecastro left, rudo left 22:46 troys_ is now known as troys, yoleaux left 22:57 wamba left 22:59 Nanjizal left, trnh left 23:04 andrzejku left 23:09 greppable6 joined
lizmat and another Perl 6 Weekly hits the Net: p6weekly.wordpress.com/2018/01/22/...-optimism/ 23:13
23:24 cdg joined 23:26 SmokeMachine left, eliasr left 23:27 SmokeMachine joined, eliasr joined 23:39 Nanjizal joined 23:44 sena_kun left
lizmat goed to bed& 23:50
timotimo gnite lizmat 23:53
23:53 n9l joined 23:56 konsolebox left