»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.
kjk so when perl6 wants a list or a collection of things from an object, what class or role of the object does it assume or expect? Is it just @ (the Positional role?) or Seq or Iterable or ... ? 01:33
kjk like when I have a slurpy sub (*@args) and I pass it @some-args, it knows the sub wants a bunch of objects and it is given a bunch of objects so the objects are all collected in @args 01:37
but if I have an object of my own class, what interface does it expect my class to support? 01:38
kjk hmm Positional and PositionalBindFailover ? 01:44
Geth doc: 1f70c83ebd | Coke++ | doc/Language/101-basics.pod6
Pass example compilation test
 02:50
synopsebot Link: doc.perl6.org/language/101-basics
b2gills .tell lizmat in Weekly 2018.27: my account is b2gills not bdgills 03:21
masak has anyone here tried Marpa for parsing? 05:54
jmerelo masak: not that I now. If you're in Twitter, @jluis10 has given several talks on the subject. 06:08
yoleaux 9 Jul 2018 21:04Z <tbrowder_> jmerelo: what is your final destination in Japan?
jmerelo .tell tbrowder_ I'm going to Osaka and then to Kyoto.
yoleaux jmerelo: I'll pass your message to tbrowder_.
masak I'm fascinated by Jeffrey Kegler's blogging output, but I've yet to understand exactly what the Earley parser _is_ 06:09
jmerelo masak: link? 06:10
masak jmerelo: I just finished reading jeffreykegler.github.io/Ocean-of-Aw...965_2.html 06:12
but I've recently read jeffreykegler.github.io/personal/timeline_v3 and found it very interesting
Kegler's main thesis can maybe be summarized as "people who care about parsers have been doing mostly the wrong thing for decades, and I can show you how to do it right" 06:13
masak (I hope I'm doing it justice by phrasing it that way. it sounds a bit hubristic when phrased like that, but I'm pretty sure that's the message.) 06:14
masak here he weighs in on Perl 6 parsing: jeffreykegler.github.io/Ocean-of-Aw...pdown.html 06:17
jmerelo masak: sounds totally hubristic, but there could be a bit of true there... 06:23
bit of truth, I mean. 06:24
masak I'll say this: this has the *structure* of the kind of hubristic assertions that happen to be true
(which is still not a total guarantee, but at least puts it in my "interesting, to read" inbox) 06:25
jmerelo masak: I've heard many good things about Marpa, most of which are over the top of my head. Heard them mostly from that person, a Barcelona Perl Monger.
Geth doc/master: 4 commits pushed by (JJ Merelo)++ 06:37
Geth doc: ef5badcc1e | (JJ Merelo)++ | doc/Language/operators.pod6
Uniforms capitalization refs #1945
 07:04
synopsebot Link: doc.perl6.org/language/operators
lizmat clickbaits p6weekly.wordpress.com/2018/07/09/...to-perl-6/ 07:28
Qwerasd How could I take an array, say ['foo', 'bar', 'baz'] and turn it in to a multi-level hash access (i.e. %hash{'foo'}{'bar'}{'baz'})? 08:07
moritz m: my %h = foo => { bar => {baz => 'x'}}; my $current = %h; $current = $current{$_} for <foo bar baz>; say $current 08:10
camelia x
moritz Qwerasd: ^^
Qwerasd :o
Qwerasd What if I wanted to set the value at that location though? is $current at the end a reference to the actual location in the hash? 08:11
moritz m: my %h = foo => { bar => {}}; my $current = %h; $current = $current{$_} for <foo bar baz>; $current = 'flurb'; say %h
camelia {foo => {bar => {}}}
moritz m: my %h = foo => { bar => {}}; my $current := %h; $current := $current{$_} for <foo bar baz>; $current = 'flurb'; say %h 08:12
camelia {foo => {bar => {baz => flurb}}}
moritz works with binding
Qwerasd That's really cool.
How could I filter out empty strings from an array? 08:15
moritz m: say ('abc', '', 'def').grep(&chars).perl
camelia ("abc", "def").Seq
Qwerasd coool 08:16
moritz m: say ('abc', '', 'def').grep(&so).perl
camelia ("abc", "def").Seq
Kaiepi love when my irc client refuses to reconnect to channels 08:22
Qwerasd How can I do a breadth-first search on a hash? 08:30
masak Qwerasd: manually
Qwerasd I essentially want to get a layer at a depth n of a hash
masak Qwerasd: I think I'd do it as a helper function that takes a list of hashes, and gives back all the values of all the hashes 08:31
that'd be "one layer" of the BFS
Qwerasd if I have {a => { b => 'c' }, 1 => { 2 => 3 }}, and I want to get [[a, 1], [b, 2], ['c', 3]] how would I do that?
masak I don't think I can generalize exactly what you want from that example. 08:32
Qwerasd ** ignore that ['c', 3], I only am interested in keys.
masak are there always two keys at the top level? what if there are more?
Qwerasd There are an arbitrary number of keys at any level 08:33
Qwerasd I'm generating a sitemap and I want to do a breadth-first search through it. 08:33
I guess I could do a recursive function. 08:35
masak a recursive function won't give you BFS 08:36
if it's DFS you want, a recursive function is fine
Qwerasd No, I mean a recursive function in order to get all keys at a certain level. Kinda a DFS for keys at a given level 08:37
masak m: my %h = a => { b => { c => 0, d => 1 } }; sub level(%h, $l) { my @hashes = [%h]; for ^$l { @hashes = @hashes.map(*.values).flat }; return @hashes.map(*.keys).flat }; for 0..2 -> $L { say "Keys on level $L: ", level(%h, $L) } 08:39
camelia Keys on level 0: (a)
Keys on level 1: (b)
Keys on level 2: (c d)
Qwerasd woah 08:40
Are you a wizard?
quester Yes, indeed Qwerasd. Masak is the author of strangelyconsistent.org/, designed most of what exists of the Perl6 implementation of macros, and... well, you can make up your own mind about wizardhood. 08:47
masak according to strangelyconsistent.org/blog/perl-6-is-my-mmorpg I'm a warrior ;) 08:49
masak there are plenty of mages around here, though 08:49
moritz masak: I was about to bring that one up :-) 08:50
masak moritz: I get to be quicker occasionally :P
masak .oO( meet my brother, Tweedledee )
moritz being a book author, am I an Archer now? 08:54
masak I think books count as crossbows 09:00
Qwerasd OK, I'm stumped. How could I modify that level function to ignore anything named with an empty string? 09:02
{'' => 'ignored', foo => {not => 'ignored'}} 09:03
(not => 'ignored' is not ignored) 09:04
Qwerasd Nvm solved it. 09:07
Maybe 0 sleep isn't the best for writing code :p
masak m: my %h = a => { b => { c => 0, d => 1 }, "" => { e => 1 } }; sub level(%h, $l) { my @hashes = [%h]; for ^$l { @hashes = @hashes.map(*.values).flat }; return @hashes.map(*.keys).flat.grep(* ne "") }; for 0..2 -> $L { say "Keys on level $L: ", level(%h, $L).perl } 09:08
camelia Keys on level 0: ("a",).Seq
Keys on level 1: ("b",).Seq
Keys on level 2: ("e", "d", "c").Seq
masak the `.grep(* ne "")` is the addition
Qwerasd Yep
Am I mistaken or would `.grep(*)` remove everything falsey? 09:09
masak you are mistaken :/ 09:10
m: my @a = "", 1, 0, "hi"; say @a.grep(*).perl
camelia ("", 1, 0, "hi").Seq
masak I mean, your instinct is sound, but the generalization is false 09:10
Qwerasd I guess .map(-> $e {$e if $e}) will have to do
moritz Qwerasd: .grep(&so) removes everthing falsey
masak you can also do .grep({ $_ }) or .grep(*.so)
or what moritz said 09:11
Qwerasd ah
Thank you so much.
masak it's important to know that `* ne ""` is a WhateverCode but `*` is a Whatever, and they work differently in (e.g.) `.grep`
masak think of it as a "wart" in the language 09:12
lizmat has been tempted many times to add a Whatever candidate for .grep that would do ?$_ 09:13
masak that temptation is there, for sure
moritz lizmat: you mean one without a matcher argument?
masak I think a discontinuity would show up _somewhere_ in the language, though, just further out
lizmat no. @a.grep(*) meaning @a.grep: { so $_ }
masak because we mean a whole lot of things by Whatever, not just that thing 09:14
lizmat yeah, so the shortest way would really be:
@a.grep: ?*
masak lizmat: I mean, an equally valid argument (I think) could be made for `@a.grep(*)` meaning "retain all the elements of @a"
lizmat yes, indeed, that's why I haven't done it :-) 09:15
masak as in "whatever, I don't care to grep for things today"
lookatme work out a normal solution
m: multi sub f(Hash $f, @a) { @a[@a.end].append: $f.keys; $f.values; } ; multi sub f($f, @a) { @a[@a.end].push($f); $*f = True;}; my @h = $%{a => { b => "c" }, 1 => { 2 => 3 }}; my @a = [[], ]; my $*f; while not $*f { @h = flat (f $_, @a for @h); @a.push: [];}; dd @a;
camelia Array @a = [["1", "a"], ["2", "b"], [3, "c"], []]
lookatme off work
moritz reminds me of the Date.new situation
where it used to return Christmas, but people were surprised by that
moritz and I thought that if you want a specific date, they should say which one 09:16
lizmat m: my @a; @a[9] = 1; dd @a.grep(*) # masak: seems it already does what you mean :-)
camelia (Any, Any, Any, Any, Any, Any, Any, Any, Any, 1).Seq
masak my _absolute favorite_ use of the whatever star is still `.indent(*)` where it means "the most aggressive deindentation possible" 09:17
lizmat: only about half of me meant that :P
jnthn grep smartmatches against the argument, and anything ~~ * is True 09:18
masak actually, my main point is that an irreducible discontinuity exists
between "smartmatch anything" (because of Whatever's smartmatch semantics) and "boolify" (because of consistency with WhateverCode in grep), in this case 09:20
jnthn Not sure I see the discontinuity. A WhateverCode happens to be written with a syntax using *, but it's a distinct type from Whatever and so can have its own smartmatch semantics. 09:24
masak jnthn:
<Qwerasd> Am I mistaken or would `.grep(*)` remove everything falsey?
masak jnthn: the discontinuity is not so much in the language as in people's expectations 09:25
jnthn Well, "some people's", but yes, fair enough. :)
moritz we also have [] or [*] for "give me everything
jnthn And I guess the difference between Whatever and WhateverCode is not the first thing one would learn. :) 09:26
Altreus I just learned it 09:29
I just learned *of* it
I have yet to learn it :P
Qwerasd So I'm using Data::Dumper::Tree and trying to set the option display_address to False, however: Type check failed in assignment to $!display_address; expected DDT_Address_Display but got Bool (Bool::False) 09:37
lizmat Qwerasd: that feels like it should be reported to the DDT author 09:43
Qwerasd Is it a bug then? I wasn't sure if I was just being dumb, cause the documentation says it's True by default implying it's boolean 09:44
lizmat Qwerasd: I have no idea, it's not even in the ecosystem ?? 09:45
jmerelo lizmat: DDT? Isn't it part of Rakudo? 09:46
lizmat is it?
Qwerasd I found it via modules.perl6.org and installed it via zef
jmerelo lizmat: ah, no, it's not github.com/nkh/P6-Data-Dump-Tree
Hum, but that's data _Dump_, not dumper... 09:47
Qwerasd Oh oops
my bad, tired brain made a typo
I indeed am talking about Data::Dump::Tree
lizmat Qwerasd: I suggest filing an issue at github.com/nkh/P6-Data-Dump-Tree/issues 09:48
jmerelo lizmat: the tiny data dumper is part of Rakudo... And you wrote it :-) My mistake
lizmat jmerelo: yeah, so tiny it lost the T
pmurias Qwerasd: the docs to that module are outdated 09:53
Qwerasd Oh?
pmurias Qwerasd: the modules used to expect a boolean for that parameter but it has been changed to an enum
Qwerasd How would I do that? 09:54
pmurias github.com/nkh/P6-Data-Dump-Tree/b...nums.pm#L5
Qwerasd Idk how 2 use enum 09:55
lizmat try using DDT_DISPLAY_NONE instead of False 09:56
pmurias github.com/nkh/P6-Data-Dump-Tree/b...se_class.t
Qwerasd: ^ use can use that test as an example
Qwerasd err 09:59
Undeclared name: DDT_DISPLAY_NONE
lizmat try 0 then :-)
Qwerasd Oh I need to use the enums
Thanks for the help! 10:01
AlexDaniel If anyone here is using Matrix and is not in perl 6 community, please let me know and I'll add you
pmurias Qwerasd: if you like the module, consider making a PR to fix the docs ;)
Qwerasd Can I get stdout as an IO::Handle? 10:10
AlexDaniel Qwerasd: a handle to what? 10:12
Qwerasd I'm not sure what you're asking me
jmerelo Qwerasd: you probably want $*OUT
Qwerasd: docs.perl6.org/language/variables#...-%24%2AOUT
Qwerasd Ah yep, thanks
AlexDaniel ah, I thought this was about Procs :) 10:13
tyil AlexDaniel: is the Matrix perl 6 not linked to freenode's #perl6? 10:24
pmurias what is this Matrix thing? 10:30
tyil pmurias: check matrix.org 10:31
it's a federated chat protocol
some would describe it as "the next best thing after irc"
I was unconvinced mostly due to the lack of good clients about a year ago
Qwerasd How would I make a flag option for the command line? 10:33
lizmat sub MAIN(:$foo) { }
would give you --foo
Qwerasd Cool thanks 10:34
lizmat docs.perl6.org/routine/MAIN#declar...%28MAIN%29 10:36
tyil using riot.im I get the message from firefox that the webpage is slowing down the entire browser :'D 10:39
AlexDaniel pmurias: demonstration: riot.im/app/#/group/+perl6:matrix.org 10:40
there's also a twitter feed there, but looks like it's not visible to guests
tyil: there's no native matrix perl 6 room 10:41
tyil hmm
it's been over a year since I last logged in there
AlexDaniel I mean, #freenode_#perl6:matrix.org works just fine for that purpose
tyil what's the speciality of a community
I dont think I've seen that before
AlexDaniel yeah, it's relatively new
tyil I thought you ment just a Perl 6 channel
but on Matrix 10:42
AlexDaniel basically it's a group of rooms + a basic html page
well, a list of rooms
tyil[m] ah
does it automatically join all the Perl 6 related channels/room when joining a community?
pmurias so it's sort of slack like thing 10:42
?
AlexDaniel`` I don't think it joins all of them 10:43
AlexDaniel`` but at least you can easily see the list :) 10:43
AlexDaniel pmurias: yeah 10:43
pmurias: but bettah!
because open-source all the way through and federated
tyil[m] open source and federation are very nice
slack will hold your chat history hostage 10:44
which doesn't sound open software/project friendly
tyil the riot.im app is incredibly slow and using almost a full core :( 10:45
the web app I mean
tbrowder_ .tell jmerelo i was in japan for a couple of years in my youth. i will share stories of that time if we cross paths at a perl mtg someday
yoleaux 06:08Z <jmerelo> tbrowder_: I'm going to Osaka and then to Kyoto.
tbrowder_: I'll pass your message to jmerelo.
AlexDaniel hmmm did you enable stickers? 10:45
I think there's a litle bug with loading the stickers, but could be anything
I'm not seeing that here
tyil idk, I just tried logging in and accepting the Perl 6 community invite 10:46
AlexDaniel tyil[m]: I'd say the biggest problem currently is that most people use matrix.org which is really overloaded 10:47
tyil got about 50 invites waiting on me ;~;
that could be true, I don't know about the stats or load of any matrix servers
AlexDaniel well, on matrix.org you'd normally get like… uh… um… 3-4 seconds delay when sending a message… 10:48
also presence events are disabled because that would make it even worse…
so everyone appears offline! :)
but yeah, all that goes away once you start running your own homeserver :) 10:49
tyil do we have a Perl homeserver planned yet? :p 10:49
AlexDaniel but I wonder, don't we want matrix homeserver on perl6.org? That'd be fun 10:50
yeah
tyil it doesn't have to be Perl 6 specific, just Perl in general to cater to both 5 and 6 would be cool 10:51
AlexDaniel I think you're right
tyil oh, pmurias: Matrix can bridge with an existing Slack community 10:52
so you can use Matrix to be part of all your Slack communities you're already in
(but that will require some admins from your slack communities to approve it and configure it) 10:53
Zoffix New blog post: "Cancellation of Perl 6 Constants and Rationals Grant": blogs.perl.org/users/zoffix_znet/20...grant.html 11:05
jmerelo tbrowder_: I hope that will happen in the near future. Going to Glasgow, maybe? 11:19
yoleaux 10:45Z <tbrowder_> jmerelo: i was in japan for a couple of years in my youth. i will share stories of that time if we cross paths at a perl mtg someday
tbrowder_ a long shot, but trying to convince wife... 11:20
jmerelo tbrowder_: Best of luck!
tbrowder_ thnx! are you there on academic business 11:21
jmerelo tbrowder_: yep. 11:22
tbrowder_: first to Osaka for this summer school sigevo-summer-school-2018.github.io/
tbrowder_ presenting? if so, title?
jmerelo then to Kyoto for GECCO: gecco-2018.sigevo.org/index.html/tiki-index.php
pmurias 6~/exit
jmerelo I'm actually presenting a couple of papers of stuff I did with Perl 6 :-)
tbrowder_ very nice, good PR! 11:23
jmerelo tbrowder_: I'll share the slides as soon as they are ready, but the papers are available from my GitHub account. 11:23
tbrowder_: first one is about stateless concurrent evolutionary algorithms, and another about my EA library in Perl 6. My Mexican colleague, Mario, will be presenting that one. 11:24
tbrowder_: thanks for asking :-) We people in academia can't seem to stop talking about our shit
Here's the EA library: modules.perl6.org/dist/Algorithm::...an:JMERELO
Qwerasd How do I find the index of a certain elem in an array?
tbrowder_ have you used tha p6 genetic stuff written in pdf on the p6 site? 11:25
jmerelo tbrowder_: He, no. I just wrote a few examples with a dummy DNA class. It could be used to optimize layout, for instance. I don't know. If someone can think about something, would gladly give it a try. 11:26
tbrowder_ ref academia: Prof. Donald Knuth is one of my favorite people. i wish i could attend one of his regular seminars. 11:28
tbrowder_ can’t belive the mountains of work he has accomplished 11:29
jmerelo tbrowder_: I never had the chance. Read some interviews 11:44
lizmat Qwerasd: you can't, but you can loop like this: 11:45
Qwerasd Thanks - but it turns out there was a better way to go about what I was doing.
lizmat m: my @a =<a b c>; for @a.kv -> $i, $letter { say "$i: $letter }
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in double quotes; couldn't find final '"' (corresponding starter was at line 1)
at <tmp>:1
------> 3a.kv -> $i, $letter { say "$i: $letter }7⏏5<EOL>
expecting …
lizmat m: my @a =<a b c>; for @a.kv -> $i, $letter { say "$i: $letter" }
camelia 0: a
1: b
2: c
Zoffix m: say <a b c>.first: :k, "b" 12:18
camelia 1
timo m: say <a b c>.index("b") 12:20
camelia 2
timo oh?
ah, that's the string one
Zoffix Yes
timo so the space in between counts
masak .oO( significant whitespace ) 12:28
Altreus m: say <a b c>.index("b") 12:31
camelia 2
Altreus not *that* significant :P 12:31
standard significant
tyil[m] AlexDaniel: heh, I get to see camelia 's output before I see what timo and zoffix were trying to run with `m:`
on matrix, that is
timo fabulous 12:32
tadzik yeah, the irc bridge can be kind of wacky, especially with the ordering sometimes :)
Zoffix m: say <a b c>.Str.index("b") 12:33
camelia 2
Zoffix m: say <a b c>.Str.perl.say # this whitespace is what was meant
camelia "a b c"
True
masak significant-ish. significant-oid. 12:34
Altreus it's the usual significance of whitespace, i.e. "any" vs "none" 12:39
is what I meant :)
Zoffix But I meant that you seem to mean that how you write `<a b c>` is where the significance is at, where that doesn't matter at all, it's how a List stringifies. 12:40
masak it's not the same whitespace at all 12:41
Zoffix m: my @l := <a b c> but "abc"; say @l.index: "b"
camelia 1
masak the whitespace that you put into the source code is lovingly handcrafted. it's made by the sweat of your brow.
Zoffix m: my @l := "abc".comb.List; say @l.index: "b"
camelia 2
masak the whitespace that .perl spits out is totally synthetic. I wouldn't eat it if I got paid to. 12:42
Zoffix It's .Str tho
masak my point still stands. 12:42
I refuse to eat whitespace that's been made by a machine.
Altreus indeed, I don't think the whitespace that .Str gives us is worthy of attention 12:43
just set $" xD
masak m: $"
camelia 5===SORRY!5=== Error while compiling <tmp>
Name must begin with alphabetic character
at <tmp>:1
------> 3$7⏏5"
expecting any of:
infix
infix stopper
statement end
statement modifier
…
Altreus I presume there is some analogy to that
masak checks S28
Altreus perl 5-to-6 doesn't seem to mention the punctuation variables 12:44
masak no, there is no analogue 12:44
Altreus nope, just uses spaces 12:45
masak m: my $long-boi = <a b c> but role { method Str { self.join(" ") } }; say $long-boi.Str
camelia a b c
masak there ya go
Altreus nevertheless, "significant whitespace" to me is a term that means that the whitespace you put in the code has an effect on the outcome of the code
masak who needs special globals
Altreus And by this definition perl6 has significant whitespace only in that if you put none it is different from putting some 12:46
masak Altreus: you're correct in so terming it ;)
Altreus As a derogatory term, "significant whitespace" simply means that the amount of whitespace you use is significant
masak we don't have time to be derogatory here
Altreus as opposed to being a binary state of existance
Altreus indeed, we have a future to craft 12:47
masak I don't care how your significant your whitespace is -- hurry up and help! :P
s/your //
Ulti :cool: my tests are running about 20% faster than at xmas 12:55
++ all the peeps doing optimisation 12:56
Ulti two sudden 10% jumps in the last month or so too 12:56
masak m: say 100 * 1.1 * 1.1
camelia 121
daxim <blogs.perl.org/users/zoffix_znet/20...t.html> which direction does the word "remitted" mean, TPF→Zoffix or other way around? 12:57
lizmat since TPF has not approved any payment for the grant yet, it means that Zoffix will not get paid for any of the work he's done 13:01
lizmat at least that's the way I've interpreted it 13:01
Zoffix Right, the work that's been done on constants has been done on volunteer basis. 13:21
pmurias another naming question, what would be a good file extension for Perl 6/NQP precompiled to binary for use by the GraalVM/Truffle backend? 13:32
lizmat .truffle ? 13:34
timo that's a bit over-generic?
there's more truffle than nqp 13:35
El_Che is .truffel6 ?
timo .tru64
El_Che ooh
:)
lizmat timo: that same argument would apply to .moarvm ? 13:36
tbrowder_ “.nqp.truf”
timo but .moarvm files have their format dictated by moarvm itself, right?
lizmat is that so? no Perl 6 specific things in there ?
I thought there were ?
timo i was assuming the perl6/nqp precompiled binary stuff for the gvm/t backend was a home-baked format
tbrowder_ any rumors about site of next TPC::NA? A city in Canada would be nice, like Quebec City 13:38
lizmat Pittsburgh, afaik 13:39
pmurias lizmat: there are p6 specific ops in the .moarvm files Rakudo generate but the .moarvm format could be used by other languages
Qwerasd I'm getting this error w/ no stacktrace: Will not decode invalid ASCII (code point > 127 found) 13:44
lizmat Qwerasd: please supply a gist 13:46
Qwerasd Huh?
lizmat I mean, you haven't supplied us with any info that would allow us to reproduce the problem 13:47
masak a gist is just a text pasted into gist.github.com -- any similar pastebin is fine
jnthn Can also try --ll-exception option to perl6
Qwerasd It's an issue w/ LWP::Simple it would seem
Qwerasd LWP::Simple.get('eckva.net') something about the response from this server causes the error 13:48
lizmat well, maybe it claims to return ascii, but in fact returns utf-8 ? 13:50
Qwerasd Hmmm that would make sense
lizmat FWIW, ASCII is a valid subset of UTF-8 afaik 13:51
so maybe you can override the encoding ?
Qwerasd where does it report the encoding? There's nothing in the response headers... 13:53
jnthn a curl -v of it shows it doesn't declare a content type in the headers 13:53
Qwerasd Yes it does. text/html 13:54
jnthn gah!
doesn't declare *an encoding in the content type" in the headers
Sorry, doing 3 things at once
Qwerasd yeah :p
jnthn Tried it with another Perl 6 HTTP client and it works there 13:55
I'm guessing LWP::Simple is defaulting to ASCII
Qwerasd Hmmm 13:56
Odd, as LWP::Simple has Str $.default_encoding = 'utf-8';
jnthn If you can get it to give you the body as a blob, then you can .decode it yourself, I guess 13:57
Qwerasd Oh, in the github for perl6 LWP::Simple it says "At this point it is recommended that you use HTTP:UserAgent instead. I think I'll try that. I may not have this bug 14:01
Zoffix eco: WWW 14:02
buggable Zoffix, WWW 'No-nonsense, simple HTTPS client with JSON decoder': github.com/zoffixznet/perl6-WWW 5 other matching results: modules.perl6.org/s/WWW
Zoffix Qwerasd: you can try that ^ module. It's like LWP::Simple, but simpler... and uses HTTP::UserAgent under the hood 14:03
Zoffix $ perl6 -MWWW -e 'say get "eckva.net"' 14:03
Will not decode invalid ASCII (code point > 127 found)
Qwerasd: no love
Qwerasd Oh RIP
jnthn used Cro::HTTP::Client :) 14:05
Zoffix eco: Cro::HTTP::Client 14:05
buggable Zoffix, Nothing found 14:06
jnthn eco: Cro::HTTP
buggable jnthn, Cro::HTTP 'Asynchronous HTTP, both client and server side. Includes HTTP/2.0 support.': github.com/croservices/cro-http.git 2 other matching results: modules.perl6.org/s/Cro%3A%3AHTTP
jnthn Part of that distribution
Qwerasd I recall using Cro::Websocket for a previous project, and discovering multiple bugs w/ it ;p
I guess I can try Cro::HTTP
Zoffix m: with IO::Socket::INET.new: :host<eckva.net>, :80port { .print: "GET / HTTP/1.1\r\nHost: eckva.net\r\n\r\n"; say .recv; .close } 14:11
camelia IO::Socket::INET is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in method new at src/RESTRICTED.setting line 32
in block <unit> at <tmp> line 1
Zoffix e: with IO::Socket::INET.new: :host<eckva.net>, :80port { .print: "GET / HTTP/1.1\r\nHost: eckva.net\r\n\r\n"; say .recv; .close }
evalable6 HTTP/1.1 200 OK␦Date: Tue, 10 Jul 2018 14:11:59 GMT␦Server: Apache␦Last-Modified: Mon, …
Zoffix, Full output: gist.github.com/38f5c8cb500c250338...88ce77543e 14:12
Zoffix e: with IO::Socket::INET.new: :host<eckva.net>, :80port { .print: "GET / HTTP/1.1\r\nHost: eckva.net\r\n\r\n"; say .recv.split("\r\n\r\n", 2).tail; .close }
evalable6 <!DOCTYPE html>
<html lang="en">
<head><meta http-equiv="Content-Type" content="text/html;…
Zoffix, Full output: gist.github.com/d571328f7259fcdd6b...186b587879
Zoffix ^_^
Qwerasd Cro::HTTP solved it for me 14:13
Zoffix Filed github.com/perl6/perl6-lwp-simple/issues/29 and github.com/sergot/http-useragent/issues/207 14:22
tyil[m] Qwerasd: have you reported the bugs you had with `Cro::Websocket` on the repository? 14:37
Qwerasd I actually ended up getting them fixed by complaining about them here. 14:38
tyil Qwerasd: ah, that's fine too :> 14:39
tyil[m] .tell AlexDaniel if the perl community were to get it's own Matrix homeserver, would it be interesting to link both perl.org and freenode's #perl6 channels to them?
yoleaux tyil[m]: I'll pass your message to AlexDaniel.
tbrowder_ .tell jmerelo see updated doc wiki for phase 2 plan 14:45
yoleaux tbrowder_: I'll pass your message to jmerelo.
timo Qwerasd: did you report the bugs you found in Cro::WebSocket? 14:55
Qwerasd I was already asked and responded. "I actually ended up getting them fixed by complaining about them here."
timo oh, OK 14:56
very good :)
buggable New CPAN upload: Dist-Helper-0.21.0.tar.gz by TYIL modules.perl6.org/dist/Dist::Helper:cpan:TYIL 15:05
Geth p6-sake: wbiker++ created pull request #23:
Look for Sakefile in parent folders as well		 15:23
b2gills .tell Qwerasd to get the index of an element add :k to .first `[<a b c d>].first('b'):k` or use .grep if you want more than one. Can also be inside `.first(:k,'b')` `.first('b',:k)` 15:25
yoleaux b2gills: I'll pass your message to Qwerasd.
pukku Hi! A while back, some of you were willing to review some code I had written to suggest more idiomatic Perl6-isms. I've written another program, and I was wondering if I could ask for a review again? I think I've used most of the suggestions from last time, but I'm sure there are still places I could improve. 16:01
moritz pukku: just paste a link here, and somebody will likely do it 16:11
pukku: I have to work a bit more, but will review in ~2 hours if nobody else has done by then
pukku Thanks! It's at github.com/pukku/ringing_halfsheets ; the particular file is gentroff.pl6.
I don't want to take too much of people's time though -- if you (or others) are busy, don't make any special effort on my behalf. 16:12
Thanks!
tbrowder_ pukku: looking now, but see typo on bcr logo 16:22
pukku Thanks -- the misspelling of "Boston Change Ringers" is intended -- change ringing (the hobby this is all related to) is all about permutations, so it's a (very poor) joke. 16:25
jmerelo o/ 16:26
yoleaux 14:45Z <tbrowder_> jmerelo: see updated doc wiki for phase 2 plan
jmerelo .tell tbrowder OK
yoleaux jmerelo: I'll pass your message to tbrowder.
jmerelo .tell tbrowder Right now sources are picked up by htmlify.p6 to generate URLs. I guess that wouldn't change, right? 16:28
yoleaux jmerelo: I'll pass your message to tbrowder.
tbrowder_ correct. i see you’re up very early! ohaiyogozamaisu (sp?) 16:30
Geth doc: 1775e07b82 | (JJ Merelo)++ | type-groups.json
Moved type-groups.json to new branch and deleted closes #2164
 16:32
[Coke] jmerelo: can you note which branch in the ticket? 16:33
tbrowder_ the only diff is htmlify.p6 would pick up the generated sources in their Language subdirectory (or, maybe better yet, a more obscure subdir somewhere else). separating the source in another repo would be best solution imo. 16:33
[Coke] ah, the obv. one.
[Coke] having another repo doesn't change how crappy the current htmlify is. 16:34
[Coke] reviews briefly
the main problem I've heard here is that star is including the website build tools: another fix could be to make star smarter about how it's dealing with the doc directory. 16:35
We need to at least include p6doc, so there's already something slightly more than the raw pod6 files. 16:36
but: is there another problem aside from star that's driving this request?
jmerelo [Coke]: hey, htmlify.p6 also has feelings! 16:38
[Coke]: or maybe not.
tbrowder_ just general untidiness and keeping source from the build area... 16:39
jmerelo [Coke]: I would say there is. Mainly the modules, which are not tested and might be useful outside the perl6/doc repo. 16:39
[Coke] if we move them out, it's another maintenance issue. I would assume they are utility modules only unless someone has a particular case to make about individual ones. 16:40
if they are inside the repo, they can be elminated, method signatures changed, no impact to downstream users.
jmerelo [Coke]: Well, all Pod6 modules could be moved elsewhere. We can keep in the perl6 org in case we want to keep a close eye on them. 16:41
jmerelo [Coke]: plus I agree with the general untidiness that tbrowder_ has mentioned. 16:42
[Coke] What general untidiness? 16:44
Can you point to specific untidiness?
(that we can fix.)
[Coke] sorry, emphasis on that; meaning that if it's specific, we can then fix it. 16:44
tbrowder_ that was a careless comment i made...sorry. i just like to clearly separate often-edited source files from tooling 16:47
i remember first time i looked at doc repo seeing doc subdir was confusing...just the way i view things 16:49
jmerelo [Coke]: basically having so many unrelated things in the same dir, with many directories hanging from the main one, I guess... 16:54
tbrowder_: right. There's a doc repo with a doc directory. Hum.
[Coke]: there are also many files in the main directory. Some unused files like the one just deleted go undetected... 16:55
Geth doc: 535144cc61 | (JJ Merelo)++ | doc/Language/101-basics.pod6
Re-indexes adding 'Basics' many terms. Closes #2165
synopsebot Link: doc.perl6.org/language/101-basics
[Coke] ... Yes, there is clean that can be done, sure. here's what'll happen to fix that if you move doc's doc/ into a new repository: nothing. 16:56
if there is cleanup that needs to happen, sure, let's clean it up. 16:57
These docs aren't intended to be read raw from the repo, so again, not sure why the sub dir is a problem. 16:58
but "cleanup crufty things", absolutely, that's been a low priority on going thing.
tbrowder_: how do you feel about t/ and xt/, for example? 16:59
(in my mind, they are critical to be in the same repo as the docs to insure doc quality)
tbrowder_ pukku: casual glance looks pretty good to me. do you ring bells at same site mostly? old church i assume. how old are bells? 17:00
Geth doc: 0264be340b | (JJ Merelo)++ | doc/Type/IO/Notification.pod6
Index FileChangedEvent closes #1558
synopsebot Link: doc.perl6.org/type/IO::Notification
jmerelo [Coke]: most of the tests are actually doc tests, but some are general document tests. For instance, testing the tests is now a, well, test 17:01
[Coke] Sure. imagine my comment is just about the doc tests. 17:02
tbrowder_ in my maybe anal view i would change doc dir to something like doc-src or src-docs and never put build products in it. 17:04
more than once i’ve blown away the html dir forgetting that some input stuff was ther as well as the generated buid output 17:06
p6noob Is there a more idiomatic way to destructure an object than: my ($x,$y,$h) = do given $object { .x, .y, .h } 17:08
TimToady m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3 }; my (:$x, :$y, :$z) := Foo.new; say "$x $y $z" 17:13
camelia 1 2 3
TimToady named binding will call methods on an object that is not Associative 17:14
[Coke] tbrowder_: what build products are going in doc/ ? 17:16
I tought all the build stuff was going in html/
p6noob TimToady: of course, that's great.
[Coke] ah. yes, html/ having gen'd stuff and source stuff is problematic, yes. 17:16
p6noob TimToady: the only thing is that it complains if there are any extra attributes that you don't provide a variable for..
[Coke] tbrowder_: but 'make clean' should do that for you. 17:17
p6noob class Foo { has $.x = 1; has $.y = 2; has $.z = 3; has $.a = 4 }; my (:$x, :$y, :$z) := Foo.new; say "$x $y $z" # Unexpected named argument 'a' passed
Zoffix m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3 }; my ($x, $y, $z) = Foo.new.Capture<x y z>; 17:18
camelia ( no output )
Zoffix m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3 }; my ($x, $y, $z) = Foo.new.Capture<x y z>; dd [$x, $y, $z]
camelia [1, 2, 3]
[Coke] that part of the build process could be cleaned up for sure
p6noob Zoffix, great. Thanks to you both
Zoffix m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3; has $.meows = 42; method Capture { \(:$!x, :$!y, :$!z) } }; my (:$x, :$y, :$z) := Foo.new; dd [$x, $y, $z] 17:19
camelia [1, 2, 3]
Zoffix (^ it doesn't call individual methods, it extracts all the things from the object's Capture)
Also, you can stick a `|` into params
p6noob Zoffix, interesting. In my current case, that's even better though.
Zoffix m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3; has $.meows = 42; }; my (:$x, :$y, :$z, |) := Foo.new; dd [$x, $y, $z]
camelia [1, 2, 3]
TimToady m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3; has $.t = 4 }; my (:$x, :$y, :$z, *%) := Foo.new; say "$x $y $z"
camelia 1 2 3
Zoffix or that :) 17:20
hobbs I was just writing that. Sniped by TimToady :)
p6noob lol.. okay i'll take a look at all the above, thanks again TimToady & Zoffix
hobbs in that version the *% in the capture just means "there might be some more keys/values... accept them, but don't put them anywhere" 17:21
p6noob Makes sense, thx hobbs
Zoffix p6noob: the `(:$x, :$y, :$z, *%)` thing in that constract is basically a signature, so all the features supported in `sub (...) {}`'s signature are supported there on language level (on Rakudo level, some things aren't implemented yet) 17:23
m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3; has $.t = 4 }; my (:$x, :$y, :$z, *% (:$t)) := Foo.new; say "$x $y $z $t"
camelia Cannot call method 'Stringy' on a null object
in block <unit> at <tmp> line 1
Zoffix like this would normally stick 4 into $t
m: class Foo { has $.x = 1; has $.y = 2; has $.z = 3; has $.t = 4 }; my (:$x, :$y, :$z, :$meows = "default meows", |) := Foo.new; dd $meows 17:24
camelia "default meows"
Zoffix m: my @a := <a b c>; my ($first, "b", $third) := @a; dd [$first, $third] 17:25
camelia ["a", "c"]
Zoffix m: my @a := <a Z c>; my ($first, "b", $third) := @a; dd [$first, $third]
camelia Constraint type check failed in binding to parameter '<anon>'; expected "b" but got "Z"
in block <unit> at <tmp> line 1
Zoffix :)
Zoffix hm 17:33
m: :(&s :())
camelia ( no output )
Zoffix well weird. 17:34
That parses as invocant marker actually 17:35
But, the weird part is this doesn't match:
m: say :(&s :(), *%_) ~~ my method (&s: $ ()) {}.signature
camelia False
Zoffix m: .say for :(&s :(), *%_), my method (&s: $ ()) {}.signature
camelia (&s: $ (), *%_)
(&s: $ (), *%_)
Zoffix how come it's False?
Zoffix m: say :() ~~ my method () {} 17:36
camelia Nil
Zoffix Nil? :S
oh, forgot .signature
rindolf hi all
Zoffix \o
rindolf Zoffix: o/ 17:37
Zoffix m: say :(&s:():) 17:40
camelia No such method 'multi-invocant' for invocant of type 'Any'
in block <unit> at <tmp> line 1
Zoffix Filed all the things as R#2044 17:44
synopsebot R#2044 [open]: github.com/rakudo/rakudo/issues/2044 LTAness with putting space before sig unpack of Callable
TimToady note that &foo:() is not supposed to mean sig unpacking, but to accept a typed function that takes 0 arguments 17:47
pukku tbrowder_: Thanks! Yes, I ring at the Old North Church and Church of the Advent (since they are only about a mile apart). The bells at Old North are from 1745, and although their fittings have been redone a few times, the bells themselves are original and haven't been "fixed" in any way (ie, their overtones are all still intact).
timo does that mean "a function that has a proto of '0 args only'" or would a multi that has a candidate that accepts 0 args be fine, too? 17:48
TimToady shrugs a covariant/contravariant shrug 17:49
lizmat pictures NLPW last weekend: du-chains-and-opts 17:53
eh...
www.flickr.com/photos/wendyga/sets...1034878058
:-)
Altreus Presumably, dying in a then block breaks the promise returned by then? 18:11
timo that's right 18:12
Altreus so I can do $promise1.then(-> $r { if ($r) { ... } else { die $r.cause } })
is there a shortcut to that? 18:13
Zoffix timo: I would've thought the multi is fine, since my understand is the goal here is to take a callable that you can call with particular arguments, but looks like multies aren't actually accepted :(
timo yes, using $r.result will throw the exception if the promise was broken
Zoffix And I think maybe that's a bug because smartmatching doesn't work right either
m: multi foo ($) {}; multi foo () {}; say &foo.candidates.tail ~~ :(&:())
camelia False
Zoffix m: multi foo ($) {}; multi foo () {}; -> &:() { }( &foo.candidates.tail )
camelia ( no output )
Zoffix m: multi foo ($) {}; multi foo () {}; -> &:() { }( &foo )
camelia Constraint type check failed in binding to parameter '<anon>'; expected anonymous constraint to be met but got Sub (proto sub foo (;; Mu ...)
in block <unit> at <tmp> line 1
moritz pukku: github.com/pukku/ringing_halfsheet...ff.pl6#L97 I'd do that with a hash, but that's mostly a question of style 18:14
Altreus oh! that makes sense, neat
Zoffix TimToady: what's the word for it? "signature specifier"? parametarizer? for the :(...) part in &foo:(...)
Zoffix
.oO( crombabulator! )
 18:15
moritz pukku: line 141, would be more robust with s:g/\n**2..*/\n/
Zoffix pukku: also, instead of repeating `%rdata<urpic><img>` in each branch of `when` you can just write it with `%rdata<urpic><img> = do given ... { when ... { 'nagcr' } ... }` 18:16
moritz and I'd write s:g/\&quot\;/"/ as s:g/'&quot'/"/ instead
Zoffix I'd just use HTML::Entities module :)
eco: HTML::Entities
buggable Zoffix, Nothing found
Zoffix bah
eco: HTML::Escape 18:17
buggable Zoffix, HTML::Escape 'Utility of HTML escaping': www.cpan.org/authors/id/M/MO/MOZNIO...0.1.tar.gz 1 other matching results: modules.perl6.org/s/HTML%3A%3AEscape
timo eco: slashes
buggable timo, Acme::Addslashes 'PHP security. Now in Perl 6.': modules.perl6.org/dist/Acme::Addsla...github:N'A
moritz pukku: finally, I like your style of having a pretty tight MAIN routine, and most of the log in subroutines
</review>
pukku Moritz: thanks! I'm not sure why I didn't use a hash there. And I see what you mean about using the specifiers. I think I may have meant to type `\n\n+`, but `**2..*` is the new idiom, I guess? 18:18
Zoffix yeah 18:19
pukku Zoffix: I had tried to do that without the "do" keyword, but it didn't work, and the documentation of "do given" made it sound like it wouldn't work in this case. I'll see why I thought that and maybe suggest a change to the wording.
Zoffix pukku: weird, maybe documentation needs to be fixed? FWIW—since it's a common assumption—`given` and `when` are entirely separate constructs. `given` simply aliases a thing to `$_` topic variable and `when` smartmatches against `$_` (which is why you often see `given` and `when` used together), but if you already have the stuff in $_, you can just use `when` without `given` 18:20
m: for <a b c> { when 'a' { say "tis a" }; when "b" { say "tis b" }; say "something else, bruh" } 18:21
camelia tis a
tis b
something else, bruh
wbn so the `when` block also advances to the next iteration of the loop when it ends? 18:23
timo there's `succeed` and `proceed` that control that behaviour
Zoffix wbn: the block form of it returns from current block, yes (doesn't have to be a loop) and you can control that behaviour with the keywords timo mentioned 18:25
m: { when * { say "meow" }; say "foo" }; say "bar"
camelia meow
bar
Zoffix m: { say "meow" when *; say "foo" }; say "bar"
camelia meow
foo
bar
Zoffix m: { when * { say "meow"; proceed }; say "foo" }; say "bar"
camelia meow
foo
bar
timo `succeed` and `proceed` don't require a `given`, they can be used in other contexts, too
Zoffix m: { proceed; say 42 }; say 10 18:26
camelia proceed without when clause
in block <unit> at <tmp> line 1
timo oh, seems i was wrong
Zoffix Would be handy in some whenless contexts, like inside `.tap: { ... }` 18:26
wbn docs.perl6.org/language/control#proceed 18:27
i see
Zoffix Filed R#2045 for the earlier discussion on sig matching 18:28
synopsebot R#2045 [open]: github.com/rakudo/rakudo/issues/2045 [@LARRY] Signature specifiers on Callables do not consider multi candidates
Zoffix ah, damn, and I realize why it doesn't function this way right after I file >_< 18:30
pukku Thanks! 18:31
pukku Thanks for looking at the code. I've updated it with the suggested changes (I think). 18:44
lindylex If I have a class such as :: my $testMe = Euclidean_algorithm.new( a=>+@*ARGS[0], b=>+@*ARGS[1] ); and I would like to pass new values later for variables a and b. How can I do this? 18:50
timo you can set the attributes to "is rw" in the class declaration 18:51
geekosaur if you declared them 'is rw' then you can assign to $testMe.a and $testMe.b. otherwise, you probably want a method to set them
timo then it'll let you $testMe.a = 1; $testMe.b = 9; 18:52
lindylex Ok ok one sec let me try this.
SmokeMachine m: sub run(&foo:(--> Int)) { say foo + 1 }; run(-> --> Int { 41 }) 18:56
camelia Too many positionals passed; expected 0 arguments but got 1
in block <unit> at <tmp> line 1
lindylex This is what I am trying. pastebin.com/bkdFv2wr I tried $testMe.$!a=120; and $testMe.$a=120; neither worked. 18:57
[Coke] lindylex: you're not in the class there. 18:59
timo it has to be $testMe.a = 120 instead
DrForr o/
[Coke] +1
timo it works by giving you an accessor method named the same as the attribute, but without the sigil (because method names don't have sigils in them)
lindylex Ok got it now.
timo when i started out with perl6, this stumped me for like three days 19:00
lindylex Thanks all for the help.
timo YW
lindylex timo : it is kicking my butt. I am really enjoying this damn language so much!
timo glad to hear 19:01
DrForr lindylex: Yay! Where'd you hear about it?
(Not taking an official poll, just curious.) 19:02
lindylex DrForr : you mean the language? 19:05
DrForr Yep.
lindylex Oh I have been programming since I was 14 years of age. I have perl on my resume. When the P stack was a thing on resumes. I need to get back into programming for financial reasons. I started to study many languages again. I decided to increase my understanding of Perl and realized the language was at version 6. I decided it would a be a good time to dedicate to learning this new version. It has been such a joy to do so. I do 19:08
not know of many places searching for Perl 6 programmers but the knowledge has been worth the continued investment in gaining proficiency.
timo well, perl 5 is still being developed as well, perl 6 isn't meant to replace it 19:10
lindylex timo : what is the purpose of perl 6? 19:11
timo oh, many things
there were things in perl 5 that were impossible to change due to backwards compatibility, so perl 6 was meant to break backwards compatibility and Get Everything Right™ 19:12
[Coke] Zoffix++: added a comment to perl6advent.wordpress.com/2015/12/...ment-21093
DrForr If you're looking for actual paying jobs, I'd probably look at Perl 5, at least for the next few years. But your investment in Perl 6 isn't wasted, there are many modules that bring Perl 5 almost to where Perl 6 is, albeit on the other syntax layer.
timo among other things, perl 6 is meant to be modifiable to adjust to any major changes in the future; the "100 year language"
lindylex That is my understanding of newer versions of programming languages.
lindylex DrForr : so I am correct. Perl 6 is not going to land me a job as a perl 6 developer. 19:14
DrForr Well, it will, there just aren't that many jobs out there for it right now.
lindylex What sector do you project the growth for perl 6 will occur? 19:15
DrForr Lots of us are using it alongside Perl 5, and have found that while Perl 5 is where the jobs are, the tasks that Perl 5 can do, Perl 6 can do just as well but it hasn't quite reached that critical mass yet. 19:15
timo perl 6 currently tends to be slower than perl 5 for many things 19:16
but perl 6 has more features in the parallelism and asynchronous space
DrForr Beats me, I'm no market analyst.
timo interfacing with C libraries is simpler in perl 6 as well 19:17
DrForr Though I'd probably lean towards higher-end parsing tasks because regexen are vastly more powerful in Perl 6, and as timo suggests, concurrency and asynch processing in the near term.
lindylex I can not find any data suggesting any trends with Perl 6. So this is why I have not though of it as a great way to generate income.
DrForr It certainly doesn't have the 30-year pedigree that Perl 5 has, especially for a language that was just released in 2015. 19:18
wbn perl 6 is just fun! 19:19
masak I stopped attaching my hopes and my happiness to Perl 6 massively gaining market share long ago. doesn't mean I don't wish Perl 6 well, of course. it happens to be one of my favorite languages.
but I can still feel good and happy without needing it to succeed in any wider sense.
warriors will perl 6 ever be fast enough
masak quite probably
timo we're making progress all the time 19:20
masak I mean, there's precedent in other languages
timo we just landed a major piece in the performance puzzle
masak they start out slow, then slowly get optimized. eventually the curves meet
timo we used to not be able to optimize much around assignment of scalars and such
warriors which was?
masak this has happened to Java, Python, Ruby...
timo we will be able to do that in the near future
pmurias lindylex: Perl 6 is not a good bet if you are looking for jobs in it 19:21
masak timo: ooh, that sounds like something I need to read more about
warriors timo are you a core perl 6 developer
timo i am
timotimo ha-ha! 19:21
masak now he's *two* core developers! :D
timotimo i was timotimo all along!
warriors :)
timotimo in disguise
masak *gasp*
I certainly couldn't tell
DrForr It's fast enough for me. Asking if something is "fast enough" kind of requires that you set performance goals *beforehand* otherwise the goalposts keep moving. 19:22
warriors the current trend in programming languages, people are moving code, from python to go, and from go to rust ... and from scala to scala native , so Perl 6 have it though
masak I know programming languages look like a zero-sum game, and maybe they partly are. 19:23
but I think there's also more to it than that. languages influence each other. the same developer gains insights in one languages and applies them in another.
I want Perl 6 to be able to *enrich* the world in that sense, regardless of market share. 19:24
to me, that's what the wider Perl culture is all about
warriors dynamic languages are loosing the race to static language, dart moved from optional typing, to statically typed ... clojure is kinda loosing
masak there's a trend, yes 19:25
warriors :S
DrForr Shrug. We started out with static typing, moved to dynamic when people learned what freedoms could be gotten, then the crowd that grew up with dynamic languages learned how useful static typing is, I see it as a cycle. 19:26
Zoffix SmokeMachine: you're missing parentheses on foo call. 19:26
m: sub run(&foo:(--> Int)) { say foo() + 1 }; run(-> --> Int { 41 })
camelia 42
warriors is there a nice article that describe what gradual typin in perl 6 really means
is it like dart's optional typing 19:27
pmurias the gradual typing Perl 6 sucks
* in Perl 6
warriors ohh
Zoffix heh
masak yes, it leaves something to be desired
SmokeMachine Zoffix: thanks!
Zoffix pmurias: why not improve it?
You got the commit bit!
masak I've come to believe that TypeScript's structural typing is, to use a technical term, "pretty rad"
masak m: sub foo(Int @a) { say @a }; foo([1, 2, 3]) 19:28
camelia Type check failed in binding to parameter '@a'; expected Positional[Int] but got Array ($[1, 2, 3])
in sub foo at <tmp> line 1
in block <unit> at <tmp> line 1
masak structural typing would help with... that
SmokeMachine Zoffix: why are those parentesis required?
DrForr I usually teach it as "ignore static typing until you run into a bug it could have prevented, then your program's probably in a consistent enough state to benefit."
Zoffix SmokeMachine: because listop precedence is lower than that of infix +, so you get `foo (+1)` operation 19:29
pmurias Zoffix: I have too many Perl 6 project to work on it already
SmokeMachine Zoffix: thanks!
masak DrForr: I agree, but I'm currently more radical because I see it doing good already during refactoring, long before you run the thing
pmurias but it's borderline abusing the name to call what we have gradual typing
masak pmurias: oh! why do you think so? 19:30
pmurias because that means if you add all the types you can your program becomes statically typed
masak ah, right
yes, that's... overselling it a bit :)
Zoffix pmurias: yeah, but "gradual typing in perl 6 sucks" in response to a person asking for an article is kinda out of the blue statement slagging the language.
warriors: is there actually a trend of "people moving code" or is that just a tiny subset of bloggers posting about the next shiny thing they found? 19:31
pmurias we shouldn't use it to sell the language
Zoffix warriors: moving existing code to another language is a very expensive and time consuming. I doubt many undertake that process for anything serious.
masak well, you *can* reliably start untyped and then gradually add type information. just don't expect a pot of gold at the end. 19:31
Zoffix Just because $foo is trending 19:32
DrForr masak: I've done it increasingly as I get more comfortable with hacking, but it's not something I'd recommend to first-timers. I guess if you've got experience in C/C++/Java, which is still pretty common
Zoffix And also, the "trending" I often see reported is based on google searchers of "$foo tutorial"... I personally don't recall every googling that for any languages I was interested in. I just went to the main website and looked for the docs/tutorials there. 19:33
warriors well, we can say its a trend .. in the blog-o-sphere :) .. so we can say, its a trend among the trendy
Zoffix I imagine a hypothetical language with super amazing docs would have very little google searches for it :)
warriors :) 19:33
masak languages as such aren't even really the point 19:35
Zoffix pmurias, there's a difference between not using something to sell the language and outright dissing the language without elaborating. 19:36
masak beyond some fairly shallow surface and semantics, they're just meeting points, nexuses, around which communities form
Zoffix Yeah, I agree, languages themselves matter little. 19:36
Zoffix AutoIt is godawful to write in, but it gets the job done, and because of that it's an amazing language :) 19:37
warriors everything matters
masak or rather, there are some *really* interesting effects around how a small core group of people "seed" a language community, and then that community goes on to realize and implement those values, whatever they are 19:38
Zoffix PHP is just weird, but you can write a working program even if you never programmed before by just googling for stuff, and because of that it's an amazing language :)
warriors the language feature matters, performance matters, the ecosystem matters ... everything is important
masak PHP has massively succeeded *despite* PHP-the-language is how I would put it ;)
but it's good to think of it that way because it means there *has* to be other merits to PHP than just the language 19:39
ease of deployment comes to mind
DrForr I'd say that it succeeded by virtue of being easy to install, which led to it being installed everywhere. 19:40
pmurias warriors: re blog-o-sphere trend it's interesting how much of it is caused by gigantic companies where CPU costs translate to tons of money
masak also the "ease" with which you can just wire the database directly into the HTML, which might not bite you for *months* if you're lucky
Zoffix warriors: depends on your purpose. Perl 6's tiny ecosystem is a huge plus for people who itch to develop some interesting projects from scratch instead of simply re-inventing the wheel that'll likely never get used because a language already got a go-to web framework, HTML parser, IRC client, etc. It's an unplowed field.
warriors Perl in general have a solid community, and i think Perl 6 is unique enough to attract many programmers ... I think once it becomes speedy, it will take a nice bounce in popularity
what is the go-to web framework for Perl6 19:41
Zoffix We don't got one.
:)
There's Cro, which is aiming for it
masak was gonna say
Zoffix Bailador is behind it
DrForr I'd say Cro right now, though there's an install issue (at least for me.)
Zoffix But there's no "go-to" polished web framework. If you start a web framework now, you have a good chance of coming up with a go-to framework :)
masak warriors: I've been using Perl 6 for 13 years. in a sense, it was fast enough for much of what I needed back then. it depends on the use. 19:42
Zoffix Yeah, here's a chart with speed improvements for the past ~3yr tux.nl/Talks/CSV6/speed4.html 19:44
s/Yeah, //; 19:45
warriors so .. is there a nice gradual typing article, or is it a concensus that its no good :) 19:45
Zoffix hahaha :D
pmurias warriors: a lot of people like it 19:46
warriors (Y)
cool
Zoffix warriors: I dunno, maybe this covers it: rakudo.party/post/Perl-6-Extra-Typical-Perl-6
warriors i like the idea
DrForr warriors: Not specifically that I'm aware of. I teach it in classes, come on down to TPC in Glasgow and we can talk about it :)
warriors :) 19:47
masak that article is nice, but it reminds me of one thing 19:48
masak then Perl 6 says "typing", it's definitely coming from the dynamic end of the spectrum 19:48
which I think is connected to pmurias' critique
timotimo yeah, we're no haskell or coq
masak right, and it's not realistic to expect Perl 6 to ever get there
pukku Unrelated, is there a function in Perl 6 to get the index in a list corresponding to a particular value? "index" in only for strings (ie, turns the list into a string and gives you that offset). 19:49
timotimo pukku: "first" with the :k named argument would get you there
Zoffix warriors: tho skimming it, looks like it doesn't. But I can explain it in two sentences: you can omit types, or you can stick them in and by sticking them in you get to catch some errors, such us stuff of wrong type being given to other stuff. And you can narrow down how specific you wanna go, from general stuff like `Cool` which takes any stringy/numeric things, down to `Numeric`, which takes only numeric
types, down to `Int`, which takes only Int types, down to `UInt`, which is a subset that takes only non-negative Ints, down to a subset of that subset that you can write as your own subset `subset UIntPrime of UInt where .is-prime` and that will restrict to non-negative integers that are also prime numbers. You can also use native machine types that can offer performance boosts in certain cases or the
overflow/underflow behaviour if your program needs it
Well, there was more than 2 sentences but still :)
masak I've seen some people ask now and then "Does Perl 6 have dependent types?" -- and I think the question is *ill-posed*, at least that's my conclusion, because Perl 6 doesn't have static types to begin with
DrForr looks at the "POSIX Standard" on his web browser and reaches for the 4-point harness.
pukku timotimo: thanks! I completely missed that. 19:50
timotimo if you expect multiple entries, you can grep instead
grep also has :k, right?
Zoffix Yes
pukku no, I just need the one...
Zoffix m: say <a b c d>.first: :k, "c"
camelia 2
Zoffix m: say <a b c d>.grep: :k, "c"|"b"
camelia (1 2)
timotimo check what it returns when nothing matches, too
Zoffix m: with <a b c d>.first: :k, "c" { say "index is $_" } else { say "not there, bruh" } 19:51
camelia index is 2
Zoffix m: with <a b c d>.first: :k, "meows" { say "index is $_" } else { say "not there, bruh" }
camelia not there, bruh
masak those :k parameters are a stroke of brilliance, because they unify so much stuff. I've always had an itching feeling we could do even more of that kind of unification :)
timotimo masak: it's a bit of a schwartzian transform, isn't it? 19:52
masak mebbe 19:53
masak mostly it's a neat way to think about whole families of subs/methods as just being a "mode" to searching 19:53
like "oh yeah search for this, but give me back the index, not the thing"
or "give me back both" 19:54
m: my @a = <foo bar baz>; @a.first("bar", :p).value = "wow!"; say @a 19:55
camelia [foo wow! baz]
masak huh :)
lex_ how do I display variable type? 19:57
pmurias masak: re Perl 6 getting there I don't think adding support for plugging static type systems from CPAN is that crazy 19:57
pmurias masak: crazy meaning something that we can expect to never happen 19:58
timotimo lex_: do you actually want the variable's type, like in "my Int $foo", or the type of what's currently in it?
DrForr m: my $x=32; say $x.WHAT; # One way 19:59
camelia (Int)
timotimo better to have .^name actually 19:59
lex_ DrForr: this thanks $x.WHAT 20:00
DrForr Nod, I think WHAT returns a type object.
m: my $x=32; say $x.WHAT.^name; # One way
camelia Int
timotimo yep, you can do more with the type object, but things like "put it into a string" will emit warnings
DrForr That's a stringified version.
timotimo m: my $x = 32; say "the type is " ~ $x.WHAT
camelia Use of uninitialized value of type Int in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
the type is
in block <unit> at <tmp> line 1
timotimo it'll warn and give the empty string instead
and you don't need a .WHAT before the .^name
m: my $x = 32; say "the type is " ~ $x.^name
camelia the type is Int
timotimo also, it can go into string interpolation if you put parens: 20:01
m: my $x = 32; say "the type is $x.^name()"
camelia the type is Int
pmurias masak: if I ever get tempted to try to get a PhD I might look into it 20:02
masak :) 20:02
pmurias masak: although Perl 6 is like the opposite of what Computer Science encourages 20:03
pukku If I want to have a parameter to MAIN that is optional, but must be one of a few values if supplied, how do I specify that so that the auto-generated usage knows what to say? Or is that not possible and I have to define USAGE? 20:07
moritz m: sub MAIN($x where any(<a b c>)) { } 20:10
camelia Usage:
<tmp> <x>
moritz maybe with an enum?
DrForr You could create a subtype that limits a string to a certain number of values, dunno if usage would pick up the type.
moritz m: enum PerformanceClass <mass ultra ssd>; sub MAIN(PerformanceClass $x) { } 20:11
camelia Usage:
<tmp> <x>
moritz :(
pukku Thanks -- I guess I'll have write a usage description. :-( 20:12
pukku Ooo -- it looks like I can add a die statement to the spec if it isn't one of the legitimate results. That's not the same, but it'll probably be close enough... 20:14
buggable New CPAN upload: Version-Semantic-0.1.0.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...1.0.tar.gz 20:15
DrForr Hrm, apparently forcing install of Cro doesn't help either.
[Coke] getting better information on usage on enums seems like a reasonable feature request
DrForr Cannot locate native library 'libssl.so': libssl.so: cannot open shared object file: No such file or directory 20:16
Yeah, it is. You could introspect the enum type to get a list of contents easily enough.
Altreus Is there a comprehensive list of what characters are special in p6 regexes? 20:17
cos it's a bit tedious checking the entire regex doc for everything that might be mentioned :)
lex_ timotimo : I like this ^name() . 20:20
DrForr Altreus: Not seeing one specifically, you could pull the list of topics from the regex page but that's by no means a guarantee that something wasn't missed. 20:22
perlreref had a summary entry in perl5... Well volunteered? :) 20:23
pukku bye
Altreus ;_; 20:27
Altreus What's the right way of returning a copy from s/// ? 20:31
I'm assuming from the doc, although it is not explicit, that it replaces it in situ 20:32
moritz Altreus: S///
Altreus oh cool
thank :)
DrForr ?+*.<[{}]>|\%$ # should be a good start though.
b2gills Altreus: All non-alphanumeric characters are special in regexes 20:39
Altreus aha! 20:40
b2gills m: say 'abcd' ~~ /.»/ 20:42
camelia ｢d｣
DrForr Special in the sense of "need to be escaped", not necessarily in the sense of "are metacharacters" though I think the sets are equivalent.
b2gills m: 'abcd' ~~ /``/
camelia 5===SORRY!5===
Unrecognized regex metacharacter ` (must be quoted to match literally)
at <tmp>:1
------> 3'abcd' ~~ /7⏏5``/
Unrecognized regex metacharacter ` (must be quoted to match literally)
at <tmp>:1
------> 3'abcd' ~~ /`…
Xliff \o 20:43
I'm trying to initialize a Duration object, but no matter what I use, I keep getting a type-check error message.
DrForr Informal poll - Should I simply open dir('/lib') to find the current libreadline version? (I know there are more paths to look in.) 20:44
I started looking at `uname -a`, realized that was fragile in the sense that if a user upgrades readline it'll break.
Xliff nm. I was forgetting "Duration.new(\d+)" 20:45
DrForr Though I do have to look at the version API to figure out how to convert the number 6 to version 6. 20:46
Xliff m: v6.^name.say 20:53
camelia Version
Xliff my $a = 6; v($a).^name.say 20:54
m: my $a = 6; v($a).^name.say
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
v used at line 1
Xliff m: my $a = 6; v$a.^name.say
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3my $a = 6; v7⏏5$a.^name.say
expecting any of:
infix
infix stopper
statement end
statement modifier
…
Xliff m: my $a = 6; v{$a}.^name.say
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
v used at line 1
Xliff Yeah. That's a tricky one.
Xliff m: my $v = Version.new( (6, 0, 0).join('.') ); say $v.parts 20:56
camelia (6 0 0)
Xliff m: my $v = Version.new( (6, 0, 0).join('.') ); say $v.parts; say $v.^name
camelia (6 0 0)
Version
Xliff DrForr: ^^
Xliff m: my ($a, $maj, $min) = (6, 2, 4); my $v = Version.new( ($a, $maj, $min).join('.') ); say $v.parts; say $v.^name 20:57
camelia (6 2 4)
Version
Xliff m: my ($a, $maj, $min) = (6, 2, 4); my $v = Version.new( ($a, $maj, $min).join('.') ); say $v.parts; say $v.^name; $v.say
camelia (6 2 4)
Version
v6.2.4
DrForr Yeah, just tried that after reading the Version page. 20:59
Xliff :)
DrForr Thanks though, that reminds me of something else.
SmokeMachine Is there a way to a module create a custom phaser? 22:47
timotimo a slang allows you to parse a new phaser, but where code to support it has to go is entirely dependent on when they're supposed to run 22:48
buggable New CPAN upload: Readline-0.1.1.tar.gz by JGOFF modules.perl6.org/dist/Readline:cpan:JGOFF 23:25
New CPAN upload: Readline-0.1.2.tar.gz by JGOFF modules.perl6.org/dist/Readline:cpan:JGOFF
tyil for those who want to score some SO rep: stackoverflow.com/questions/512753...e-matching 23:44

Powered by Cro and the Raku® Programming Language.

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

Thank you!