»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.
SmokeMachine does this make any sense to someone? when I compile the rakudo, the perl6 bin works, but only on the first time... after that it doesn't work anymore... 01:04
www.irccloud.com/pastebin/nCaAmU7e/
I mean with this code... 01:05
I did change the usage-name
im trying to remove my changes and try again... 01:06
yes... it works... 01:07
SmokeMachine i can't get it 01:49
Xliff SmokeMachine: What can't you get? 02:17
SmokeMachine Xliff: the problem I was getting... but it seems its something with my environment... I cloned rakudo with the zscript, made the same change and that worked... 02:19
ZzZombo Hm... apparently it's easy enough to do what I want... 02:27
m: use nqp;class A {has $.a='asd'};say nqp::getattr(nqp::decont(my $a=A.new),A,'$!a')=1;say $a
camelia 1
A.new(a => 1)
ZzZombo m: use nqp;class A {has Str:D $.a where *.trim.chars='asd'};say nqp::getattr(nqp::decont(my $a=A.new),A,'$!a')='';say $a 02:28
camelia Type check failed in assignment to $!a; expected <anon> but got Str ("")
in block <unit> at <tmp> line 1
ZzZombo m: use nqp;class A {has Str:D $.a where *.trim.chars='asd'};say nqp::getattr(nqp::decont(my $a=A.new),A,'$!a')='qwe';say $a 02:29
camelia qwe
A.new(a => "qwe")
ZzZombo Good thing I didn't start on writing any workarounds before I go this brilliant idea.
got*
Zoffix ZzZombo: note that nqp is not meant for end-users. There's absolutely no support for it and its behaviour can change without warning. 02:32
Xliff has Str:D $.a where *.trim.chars='asd 02:32
I keep having problems reading this ^^ 02:33
Xliff I would expect that would be a subtype only allowing strings ending in 'asd'. 02:33
But that's not what appears to be the case.
Zoffix Xliff: `='asd'` portion is the default value
Juerd Why is undefine not available as a method? 02:34
ZzZombo Yea, but do I have a choice?
Zoffix ^ 02:35
Xliff ZzZombo: What are you really trying to accomplish?
Zoffix ZzZombo: don't know. What were you trying to do?
Xliff m: class A {has Str:D $.a = 'a'}; say A.new.a
camelia a
Xliff m: class A {has Str:D $.a = 'asd'}; say A.new.a
camelia asd
Xliff Why the need for nqp?
ZzZombo I need to generate a setter to an attribute (well, all of them), and this gets in the way: 02:36
[23:31:16] <ZzZombo> Why does this work?
[23:31:17] <ZzZombo> m: class C {has Str:D $.a='1'};my $a=C.new;$a.^attributes[0].set_value($a,False);say $a
Xliff Ooh.
Check the Proxy object.
Allows you to create setter methods.
And getter methods too. 02:37
Zoffix ZzZombo: I don't see the connection between your nqp code and the tweaking of attriute values via MOP
m: class A {has Str:D $.a is rw = 'a'}; say A.new.a = False
camelia Type check failed in assignment to $!a; expected Str:D but got Bool (Bool::False)
in block <unit> at <tmp> line 1
SmokeMachine m: class A {has Str:D $.a where *.trim.chars is rw ='asd'}; my A $a .= new; $a.a ='qwe';say $a
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3lass A {has Str:D $.a where *.trim.chars7⏏5 is rw ='asd'}; my A $a .= new; $a.a ='q
expecting any of:
constraint
infix
…
Zoffix `is rw` goes before post constraints 02:38
Juerd: IMO the better question is why we have &undefine
Awful name for what it actually does 02:39
Xliff Yeah, well "is rw" is a shortcut, Proxy lets you establish custom coding when setting the attribute.
Zoffix eco: Proxee
buggable Zoffix, Proxee 'A more usable Proxy with bells': github.com/zoffixznet/perl6-Proxee
Xliff Hooo
ZzZombo Zoffix, what do you mean? How can my method change the attribute's value other than 1) MOP, 2) NQP? If I do it via 1, it doesn't respect any constraints.
SmokeMachine m: class A {has Str:D $.a is rw where *.trim.chars ='asd'}; my A $a .= new; $a.a ='qwe';say $a
camelia A.new(a => "qwe")
Juerd Zoffix: Agreed! 02:40
Zoffix: In the 2 use cases I came up with thus far, I use "is default", and I find myself wanting to write $foo.reset
Zoffix ZzZombo: are you trying to change private attributes on the object that doesn't trust your tweaking code?
Juerd $foo = Nil to get $foo back to its default value feels utterly wrong, and so does the term "undefine", given that my default value is definitely defined. 02:41
ZzZombo heh, it's a class I'm composing.
So even if it doesn't trust me, what can it really do?
Xliff Zoffix++ # Proxee 02:42
Juerd Zoffix: In fact, I wanted to use this in some Perl 6 demo code before an audience, but now I'll just leave it out. It's too confusing and would need a lot of explaining.
ZzZombo I don't care about proxies, given they can't solve anything in this case. 02:43
Juerd Also, undefine doesn't seem to be documented, even!
(Does this mean we can get rid of it? :P)
Zoffix It's part of 6.c language 02:44
Xliff ZzZombo: I'm still not clear on what "this case" really is! :p
Juerd Zoffix: Are you sure? I'm struggling to find references to its specification
Xliff ZzZombo: Do you have a pastebin or gist so I can see things in context?
Juerd I only learned about undefine by trying undef
Zoffix 2018.01.121 zoffix@VirtualBox~/R/rakudo/t/spec (master)$ G 'undefine' | wc -l
407
Some are false positives, but tons of undefine usage 02:45
ZzZombo I'm composing a class, as in, I have created a custom metaclass out of ClassHOW, and I need custom setters for its attributes to do what derived classes supposed to do. 02:45
Juerd clones roast 02:46
Zoffix m: class C {has Str:D $.a ="1" };my $a = C.new; $a.^attributes.head.container = 42;
camelia Type check failed in assignment to $!a; expected Str:D but got Int (42)
in block <unit> at <tmp> line 1
Zoffix ZzZombo: looks like ^ that could give you a way to test whether a value can go in, before using .^set_value
Not ideal, but IMO better than nqp
Juerd 1;0 juerd@cxien:~/tmp/roast master$ ack undefine | wc -l 02:47
246
1;0 juerd@cxien:~/tmp/roast master$ ack 'undefine(?!d)' | wc -l
80
Heh.
ZzZombo And if it doesn't throw, will it have any effect?
Zoffix s: do { class C {has Str:D $.a ="1" };my $a = C.new; $a.^attributes.head }, 'container', \()
SourceBaby Zoffix, Sauce is at github.com/rakudo/rakudo/blob/300f...te.pm#L144
Zoffix No idea. 02:48
Not familiar with that part of codebase
m: class C {has Str:D $.a ="1" };my $a = C.new; $a.^attributes.head.container.clone = 42;
camelia Cannot modify an immutable Str:D ((Str:D))
in block <unit> at <tmp> line 1
Zoffix -_-
m: class C {has Str:D $.a ="1" };my $a = C.new; $a.^attributes.head.container ~~ 42; 02:49
camelia ( no output )
Zoffix m: class C {has Str:D $.a ="1" };my $a = C.new; say $a.^attributes.head.container ~~ 42;
camelia False
Zoffix m: class C {has Str:D $.a ="1" };my $a = C.new; say $a.^attributes.head.container ~~ "x";
camelia False
Zoffix *shrug*
Zoffix &
ZzZombo It changes the return value of subsequent class to `container`, whether this is/may be undesired I want to know. 02:52
SmokeMachine m: class C {has Str:D $.a ="1" };my $a = C.new; say “x” ~~ $a.^attributes.head.container # Zoffix
camelia True
SmokeMachine 00:51 <SmokeMachine> m: class C {has Str:D $.a ="1" };my $a = C.new; say 42 ~~ $a.^attributes.head.container 02:53
m: class C {has Str:D $.a ="1" };my $a = C.new; say 42 ~~ $a.^attributes.head.container
camelia False
ZzZombo O_o 02:58
m: my $a=Proxy.new(STORE=>sub (|c) {say c},FETCH=>sub (|c) {say c})
m: my $a=Proxy.new(STORE=>sub{},FETCH=>sub{})
camelia (timeout)
Too many positionals passed; expected 0 arguments but got 1
in sub at <tmp> line 1
in block <unit> at <tmp> line 1
ZzZombo Why the heck? 02:59
SmokeMachine m: class C {has Str:D $.a ="1" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “!$name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, “x”; say $a 03:02
camelia C.new(a => "x")
ZzZombo m: class C {has Str:D $.a where *.chars>2="asd" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “$!name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, “x”; say $a 03:03
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable $!name used where no 'self' is available
at <tmp>:1
------> 3tributes.first(*.name.ends-with: “$!name7⏏5”); $attr.set_value($obj, $val) if $val
ZzZombo m: class C {has Str:D $.a where *.chars>2="asd" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “\$!name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, “x”; say $a 03:04
camelia No such method 'container' for invocant of type 'Any'
in sub change-value at <tmp> line 1
in block <unit> at <tmp> line 1
ZzZombo m: class C {has Str:D $.a where *.chars>2="asd" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes[0]; $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, “x”; say $a 03:05
camelia C.new(a => "asd")
SmokeMachine m: my $a:=Proxy.new(STORE=>method (|c) {say c},FETCH=>method {42}); $a = 13; say $a 03:05
camelia \(13)
42
ZzZombo m: class C {has Str:D $.a where *.chars>2="asd" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “!$name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, “xxx”; say $a 03:06
camelia C.new(a => "xxx")
ZzZombo m: class C {has Str:D $.a where *.chars>2="asd" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “!$name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, “xx”; say $a
camelia C.new(a => "asd")
SmokeMachine 01:02 <SmokeMachine> m: class C {has Str:D $.a ="1" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “!$name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, 42; say $a 03:07
ZzZombo Seems like a better way to do it, thanks.
SmokeMachine m: class C {has Str:D $.a ="1" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “!$name”); $attr.set_value($obj, $val) if $val ~~ $attr.container}; change-value $a, “a”, 42; say $a
camelia C.new(a => "1")
ZzZombo This shouldn't be like that: 03:11
m: my $a=Proxy.new(STORE=>sub (|c) {say c},FETCH=>sub (|c) {42})
camelia ( no output )
ZzZombo m: my $a=Proxy.new(STORE=>sub (|c) {say c},FETCH=>sub (|c) {say |c})
camelia (timeout)
SmokeMachine m: class C {has Str:D $.a ="1" }; my $a = C.new; sub change-value($obj, $name, $val) {my $attr = $a.^attributes.first(*.name.ends-with: “!$name”); $attr.set_value($obj, $val) if $attr.container = $val}; change-value $a, “a”, 42; say $a 03:12
camelia Type check failed in assignment to $!a; expected Str:D but got Int (42)
in sub change-value at <tmp> line 1
in block <unit> at <tmp> line 1
SmokeMachine ZzZombo: maybe this one is better☝️
SmokeMachine m: my $a=Proxy.new(STORE=> method (|c) {say c},FETCH=> method (|c) {say |c}) 03:13
camelia
SmokeMachine 01:13 <SmokeMachine> m: my $a:=Proxy.new(STORE=> method (|c) {say c},FETCH=> method (|c) {say |c}); $a = 42; say $a 03:14
m: my $a:=Proxy.new(STORE=> method (|c) {say c},FETCH=> method (|c) {say |c}); $a = 42; say $a 03:15
camelia \(42)













True
SmokeMachine m: my $a:=Proxy.new(STORE=> method (|c) {say c},FETCH=> method (|c) {say c}); $a = 42; say $a
camelia \(42)
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
True
SmokeMachine I don’t know why the fetch was called so many times... 03:16
m: my $a:=Proxy.new(STORE=> method (|c) {say c},FETCH=> method (\a: |c) {say c}); $a = 42; say $a 03:17
camelia \(42)
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
True
SmokeMachine m: my $a:=Proxy.new(STORE=> method (\a: |c) {say c},FETCH=> method (\a: |c) {say c}); $a = 42; say $a
camelia \(42)
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
True
SmokeMachine m: my $a:=Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; say $a 03:18
camelia \(42)
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
True
perlawhirl bisectable6: say [0 ^... 10] 07:38
bisectable6 perlawhirl, On both starting points (old=2015.12 new=311ef07) the exit code is 1 and the output is identical as well
perlawhirl, Output on both points: «04===SORRY!04=== Error while compiling /tmp/NF5wNgS5hJ␤Unsupported use of . to concatenate strings; in Perl 6 please use ~␤at /tmp/NF5wNgS5hJ:1␤------> 03say [0 ^...08⏏04 10]␤»
Xliff SmokeMachine: You're in an endless loop. In FETCH you call say which then needs to FETCH, which calls say again, which needs to FETCH.... 07:50
Xliff m: my $a := Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; 07:53
camelia \(42)
Xliff ^^ No extra calls to fetch.
m: my $a := Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; $a;
camelia WARNINGS for <tmp>:
\(42)
Useless use of $a in sink context (line 1)
Xliff Hrm. 07:54
If something is doine, then it's not a "Useless" use, is it?
s/doine/done/
geekosaur just mentioning $a like that is not considered "something is done" 07:55
Xliff m: $a = -> { say "Something"; }; $a;
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$a' is not declared
at <tmp>:1
------> 3<BOL>7⏏5$a = -> { say "Something"; }; $a;
Xliff m: my $a = -> { say "Something"; }; $a; 07:56
camelia WARNINGS for <tmp>:
Useless use of $a in sink context (line 1)
Xliff So it's a syntax thing.
m: my $a = -> { say "Something"; }; $a();
camelia Something
Xliff geekosaur: In a meta language like Perl6, I'd almost consider that LTA, because something is being done, even if the lexer can't see it. 07:57
But that's a harder thing to catch.
Because this causes problems: 07:58
m: my $a := Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; say $a;
camelia \(42)
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
\()
True
Xliff ^^
m: my $a := Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; $a.say; 07:59
camelia \(42)
\()
\()
\()
True
Xliff Odd that there are less repetitions in that codepath.
But not surprising.
m: my $a := Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; (my $b = $a).say; 08:00
camelia \(42)
\()
True
Xliff m: my $a := Proxy.new(STORE=> method (\a is raw: |c) {say c},FETCH=> method (\a is raw: |c) {say c}); $a = 42; (my $b = $a); 08:01
camelia \(42)
\()
Xliff geekosaur: ^^ That is the output I'd expect from the sink version.
But that works around the lexical issues.
geekosaur Xliff, if retrieving a value has a side effect other than jsut retrieving the value, Id imagine more than just a sink warning is due. Optimizers are free to assume you didn;t launch missiles in the process.... 08:02
Xliff Hehehe.
geekosaur not that Proxy is currently optimized, but I think that is intended at some point
Xliff No one expects a FISH in the middle of the night.
(That's a webcomic joke for anyone familiar with Otter's work) 08:03
ZzZombo Why oh why 08:03
m: role R[] {};class C is R[] {};
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot resolve caller trait_mod:<is>(C, R, Array); none of these signatures match:
(Mu:U $child, Mu:U $parent)
(Mu:U $child, :$DEPRECATED!)
(Mu:U $type, :$rw!)
(Mu:U $type, :$nat…
ZzZombo m: role R[] {};class C does R[] {};
camelia ( no output )
Xliff ZzZombo: Classes DO roles. 08:04
That took me a long time to figure out.
m: role R[] {}; role RB[] is R {}; class A does RB;
camelia 5===SORRY!5=== Error while compiling <tmp>
Too late for unit-scoped class definition;
Please use the block form.
at <tmp>:1
------> 3 {}; role RB[] is R {}; class A does RB;7⏏5<EOL>
Xliff m: role R[] {}; role RB[] is R {}; class A does RB {}; 08:05
camelia ( no output )
geekosaur one could argue that it should pun the role into a class there, but one could counter-argue that the magic is getting a bit too thick at that point
ZzZombo Yes, I expected auto punning here.
Xliff geekosaur++ # Too much magic is not a good thing.
geekosaur specificaly I'd be wary of something assuming that you could subsequently treat it as a role, when you punned it into a class and it's no longer role-like 08:06
too many chances for confusion
Xliff However I've never heard of "pun" being used in a Perl context. When (and how) did that enter the lexicon?
"pin" I could understand.... 08:07
TEttinger I've also heard pun as a verb in other languages 08:08
geekosaur docs.perl6.org/language/objects#Au...le_Punning 08:10
geekosaur I had thought it went back to the speculations, but I didn't notice it there in a quick look 08:12
ZzZombo p6: class A {};constant B=class :: is A[] {}; 08:13
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot resolve caller trait_mod:<is>(<anon|69121440>, A, Array); none of these signatures match:
(Mu:U $child, Mu:U $parent)
(Mu:U $child, :$DEPRECATED!)
(Mu:U $type, :$rw!)
(Mu:…
ZzZombo e: class A {};constant B=class :: is A[] {};
evalable6 (exit code 1) 04===SORRY!04=== Error while compiling /tmp/DL04gpvPQ2
Cannot …
ZzZombo, Full output: gist.github.com/4c6504dba51720ac2c...4a3fd35caa
geekosaur oh, it's in S12 instead of S14. figures 08:14
design.perl6.org/S12.html#Autovivifying_objects
(plus a reference to it earlier in the same file. it's about as organized as I am >.> ) 08:15
Xliff LOL
geekosaur++ # Docs reference. 08:16
But why punning? 08:17
geekosaur (in another window I'm editing a text file that I kinda stream-of-consciousness-ed together. It's got all sorts of weird forward and backward references, that I'm trying to clean up)
Xliff I mean, a role as an autovivifying class reminds someone of wordplay that tickes the funny bone? >.> 08:18
geekosaur of wordplay, period. doesn't have to be funny, necessarily
teatime geekosaur: just get Tarantino on-board with it, and make a feature out of it.
Xliff geekosaur: As in ... haha my role is now a temporary class. Got-ya! 08:19
And puns must be punny! It's a rule.
geekosaur (for example, the Bible is full of wordplay --- intended to make it easier to remember, not to be jokes.)
Xliff But those aren't called puns, are they?
geekosaur sometimes 08:20
Xliff Geez. Someone just had to go and become punitancal. :P
geekosaur there's a common usage, and a technical one
Xliff Please don't force me to keep punning.
geekosaur (linguistics, to be specific)
Xliff I will pun out.
geekosaur (but I'll leave the obvious rejoinder there to someone else...) 08:21
Xliff And it will get messy before the pun ends.
geekosaur: Hoo! I see your punning trap.
And I will not .... oh... w8. :P
geekosaur anyway, in a computer language context, a pun is a bit more controlled than C-style automatic casting 08:22
but its still using something in an "unexpected" way 08:23
Xliff So is referring to a NULL punter.
teatime groans audibly.
Xliff I said it was going to get messy. 08:24
You thought I was puntending?
geekosaur punts...
Xliff I still think that "pin" would be a better term in this case. 08:27
Because "pinning" a role into a class via .new makes more sense, and is less likely to be interpreted as something punintended.
Xliff flees.
Geth doc: 7cd9320474 | (JJ Merelo)++ | 3 files
Eliminates file that has been moved to the Rakudo wiki

And works with links to it pointing them to a reasonable place. Refs #949		 08:29
doc: ac8bfd9611 | (JJ Merelo)++ | doc/Programs/01-debugging.pod6
Revising referring files

And checking that everything is OK after searching for the file in the repo, so this closes #494
doc: 0c678b4b23 | (JJ Merelo)++ | doc/Programs/02-reading-docs.pod6
Minor changes

Basically to make a commit that closes #949 since I typed the wrong issue number before.		 08:32
geekosaur to me, "pin" implies that from that point on, it'd always be a class. But the role is still there and usable as such; the punned class is a one-off. 08:33
Xliff So pins are commitment and puns are just a one night stand? 08:35
I can almost see that...
geekosaur it's the old joke about ham and eggs
geekosaur the chicken/pun was involved; the pig/pin was committed 08:36
Xliff Now I can groan. 08:36
teatime gentlemen, it's far too early for this.
Xliff teatime: Or waaaay too late? 08:37
teatime indeed, you know me well.
Xliff :D
Of course, it is the same for me, you know.
teatime Of course.
Xliff This is what happens when you slip into insomniac territory and cannot leave. 08:38
teatime I spent about 20 hours trying to meditate on the nature of Supply's
Xliff Ahh... 08:38
teatime can't shake this nagging feeling that there's a level of grok'ing these concurrency concepts that I am brushing up against but not finding the right materials to point me toward the light 08:39
Xliff teatime: A Supply to react, is but another Promise to be kept....
(someone please stop me...)
teatime haha
Xliff Actually... 08:40
"A Supply that react, is but another Promise to be kept...." # More accurate.
Xliff s/react/reacts/ 08:40
teatime Lua is an odd bird of a language, to me.
Xliff And on that note.... it's time to sleep.
teatime I wouldn't have expected to like it, given its common uses etc., and I def. wouldn't have expected the code people share in it to be as readable and enjoyable as it often is. 08:41
I can't really explain why, but it's feels like a kind of anti-javascript.
it could easily have the same kind of *ugh* ecosystem as nodejs, it seems to me 08:42
yet does not.
(many of those 20 hours were spent reading Lua-centric docs and examples of concurrency stuff.)
Xliff teatime++: You are a better man than I. 08:42
teatime alas, it didn't help much if at all. 08:43
Xliff I tried Lua while not in the proper mindset to get it. Have not gone back, yet.
I don't understand why it is the go-to gaming script language.
teatime I suppose only because it is small and light and extremely easy to embed
i.e. that was a design goal
Xliff Easy to embed, perhaps?
LOL! Jinx!
Those are good reasons, though. 08:44
Xliff &
teatime it's got some superficial weirdness, but the syntax is not very complex
if you find the practical need, I think you should revisit it w/o fear
Geth doc: b3f8233fa8 | (JJ Merelo)++ | doc/Language/pod.pod6
Including a definition for `defn`

In fact, its only use in this documentation is commented out, and a real use was in the file that was removed in #949. Anyway, it's used throughout the documentation and I think it's fine to include just this little definition. Refs #1763.		 08:52
synopsebot Link: doc.perl6.org/language/pod
travis-ci Doc build passed. JJ Merelo 'Revising referring files 08:57
travis-ci.org/perl6/doc/builds/339785535 github.com/perl6/doc/compare/02ed6...8bfd961141
travis-ci Doc build passed. JJ Merelo 'Minor changes 09:01
travis-ci.org/perl6/doc/builds/339786050 github.com/perl6/doc/compare/ac8bf...678b4b2341
Geth doc: 58ade6326b | (JJ Merelo)++ | 2 files
defn included in POD6 documentation

I see now why it wasn't included, since it's not actually rendered using definition lists. That's maybe a NYI feature somewhere upstream, might fix it too. Anyway, closes #1763.		 09:17
travis-ci Doc build passed. JJ Merelo 'Including a definition for `defn` 09:22
travis-ci.org/perl6/doc/builds/339788766 github.com/perl6/doc/compare/0c678...f8233fa8b1
ZzZombo s: trait_mod:<is>, (Mu:U $child, Mu:U $parent) 09:48
SourceBaby ZzZombo, Something's wrong: ␤ERR: ===SORRY!=== Error while compiling -e␤Two terms in a row␤at -e:6␤------> put sourcery( trait_mod:<is>, (Mu:U<HERE> $child, Mu:U $parent) )[1];␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ statement modifier loop␤
ZzZombo s: trait_mod:<is>, \(Mu:U $child, Mu:U $parent) 09:50
SourceBaby ZzZombo, Something's wrong: ␤ERR: ===SORRY!=== Error while compiling -e␤Two terms in a row␤at -e:6␤------> put sourcery( trait_mod:<is>, \(Mu:U<HERE> $child, Mu:U $parent) )[1];␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ statement modifier loop␤
ZzZombo s: &trait_mod:<is>, \(Mu:U $child, Mu:U $parent) 09:51
SourceBaby ZzZombo, Something's wrong: ␤ERR: ===SORRY!=== Error while compiling -e␤Two terms in a row␤at -e:6␤------> put sourcery( &trait_mod:<is>, \(Mu:U<HERE> $child, Mu:U $parent) )[1];␤ expecting any of:␤ infix␤ infix stopper␤ statement end␤ statement modifier␤ statement modifier loop␤
ZzZombo m: &trait_mod:<is>.cando(\(Mu:U $child, Mu:U $parent))
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3&trait_mod:<is>.cando(\(Mu:U7⏏5 $child, Mu:U $parent))
expecting any of:
infix
infix stopper
statement end
…
ZzZombo m: &trait_mod<is>.cando(:(Mu:U $child, Mu:U $parent)) 09:52
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
trait_mod used at line 1
ZzZombo m: &trait_mod:<is>.cando(:(Mu:U $child, Mu:U $parent)) 09:52
camelia Type check failed in binding to parameter '$c'; expected Capture but got Signature (:(Mu:U $child, Mu:U $...)
in block <unit> at <tmp> line 1
ZzZombo m: &trait_mod:<is>.cando(|\(Mu:U $child, Mu:U $parent)) 09:53
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3&trait_mod:<is>.cando(|\(Mu:U7⏏5 $child, Mu:U $parent))
expecting any of:
infix
infix stopper
statement end
…
ZzZombo m: my \c=\(Mu:U $child, Mu:U $parent);&trait_mod:<is>.cando(|c)
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 3my \c=\(Mu:U7⏏5 $child, Mu:U $parent);&trait_mod:<is>.c
expecting any of:
infix
infix stopper
statement end
…
ZzZombo wtf
ZzZombo m: &trait_mod:<is>.cando(\(Mu:U,Mu:U)) 09:57
camelia ( no output )
ZzZombo s: trait_mod:<is>, \(Mu:U,Mu:U) 09:58
SourceBaby ZzZombo, Something's wrong: ␤ERR: ===SORRY!=== Error while compiling -e␤Calling trait_mod:<is>() will never work with any of these multi signatures:␤ (Mu:U $child, Mu:U $parent) ␤ (Mu:U $child, :$DEPRECATED!) ␤ (Mu:U $type, :$rw!) ␤ (Mu:U $type, :$nativesize!) ␤ (Mu:U $type, :$ctype!) ␤ (Mu:U $type, :$unsigned!) ␤ (Mu:U $type, :$hidden!) ␤ (Mu:U $type, Mu :$array_type!) ␤ (Mu:U $type, *%fail) ␤
ZzZombo s: &trait_mod:<is>, \(Mu:U,Mu:U)
SourceBaby ZzZombo, Sauce is at github.com/rakudo/rakudo/blob/311e...its.pm#L14
Geth doc: 0213908498 | (JJ Merelo)++ | doc/Type/Str.pod6
Test changing to routine

To see if anything breaks. Refs #683		 10:41
synopsebot Link: doc.perl6.org/type/Str
ZzZombo When MOP `new_type` is called? 11:23
pmurias m: say &OUR::access_lexical_a().perl;{my $a = 42; our sub access_lexical_a() {$a}} 11:25
camelia Any
pmurias do we have a use for that behavior ^^
ZzZombo What behavior exactly? 11:26
Geth doc: ddd6b24a65 | (JJ Merelo)++ | doc/Type/Str.pod6
Changes `sub` to `routine`

Actually, methods and subroutines are usually listed under
  "methods", however 'routine' and not 'sub' is normally used
there. This is not peculiar to this file, there are `sub`s elsewhere in the documentation; however, they will need their very own issue, because this closes #683		 11:29
synopsebot Link: doc.perl6.org/type/Str
pmurias ZzZombo: being able to access variables that don't exist yet 11:43
ZzZombo That's wrong perception. 11:58
Geth doc: a90adc268c | (JJ Merelo)++ | doc/Language/about.pod6
Changed real infix to imaginary one

This kind of solves #1394, creating in its turn a broken link, but at least it should show the actual documentation for the Z infix.
Closes #1394 if everything goes well.		 12:07
synopsebot Link: doc.perl6.org/language/about
DrForr_ notes jjmerelo++ is on fire today. 12:28
El_Che hi DrForr_ 12:31
I like the new _ thing :)
DrForr_ o/
ZzZombo Can I make Perl accept `const` as constant declarator? 12:32
DrForr On occasion I get the mental image of a piece of stray toilet paper.
yoleaux 31 Jan 2018 23:55Z <comborico1611> DrForr: Have you heard of this book: Coders at Work: Reflections on the Craft of Programming ?
El_Che DrForr: lol 12:32
El_Che DrForr: I see it more as a unicode problem: DrForrö or something :) 12:33
DrForr .tell comborico1611 I have a copy, but it's in storage, I believe. 12:34
yoleaux DrForr: I'll pass your message to comborico1611.
DrForr gives up on X11. 12:37
El_Che wayland? 12:38
DrForr No, I'm using a VM and running full-screen X11 eats up too much RAM.
teatime breathes a sigh of relief. 12:39
El_Che did you installed the tools/addons?
I run a win7 VM on linux and without the guest edition and enabling 2d/3d support + 256mb memory is sluggish 12:40
aka, it runs the graphical stuff on the gpu
El_Che in the past, before I formatted my work machine and instelled linux, I had a company win 7 with a ubuntu VM. Same, needed the additions + max video memoru for 3D settings and it was fine 12:41
DrForr I had guest additions installed, had to in order to get the seamless bit to run. 12:42
El_Che I formatted the thing because with a desktop vm you can use around half of your ram, and I was spending all my time in the vm anywaus
DrForr: did you change the 2d/3D setting and adapted video memory?
once you also enable 3D Virtualbox lets you go up to 256MB (from 128)
DrForr Hrm, I'll have to look. 12:43
El_Che vmware allows you to even give more video ram to the VM
I had a license, but I got fed up stuggeling with every OS update and went back to virtualbox
DrForr Good grief.
That... could be an issue. 12:44
El_Che (the times I patched their vmware tools, I lost count)
(it a perl script :) )
DrForr I know, I also know one of the people that wrote them :)
El_Che :)
DrForr gooses RAM up from (eep) 16Mb to 128 and enables 3D... 12:45
El_Che then go to 256 12:46
DrForr 128 was max. It's an early Win10... I gave up a bit of RAM for the touchscreen, though that didn't last nearly as long as I'd hoped.
DrForr Switching desktops feels smoother. 12:53
caa51h Hi. I'm learning perl6. 13:11
DrForr Well, you came to the right place :) 13:12
caa51h Can I use the hyper operator with lambda expression?
DrForr: cool thank you
El_Che ahaha, you're not loosing time 13:13
DrForr m: my @x=1,2,4; say[{$^a*$^b}] @x;
camelia 5===SORRY!5=== Error while compiling <tmp>
Use of non-subscript brackets after "say" where postfix is expected; please use whitespace before any arguments
at <tmp>:1
------> 3my @x=1,2,4; say7⏏5[{$^a*$^b}] @x;
expecting any of:…
SmokeMachine m: sub MAIN($*v) {}
camelia Usage:
<tmp> <v>
SmokeMachine :)
DrForr That was me just playing with an idea; I'd think there should be a way. 13:15
SmokeMachine m: say class :: {has $!a; method bla($!a) {}}.^find_method(“bla”).signature.params.[1].usage-name
camelia a
SmokeMachine :)
caa51h oh I get it.
m: say <0 1 2 3 4>>>.&(-> $x { $x %% 2 })
camelia (True False True False True)
jnthn m: say (1, 2, 3) >>[&(-> $a, $b { "$a x $b" })]<< (4, 5, 6) 13:16
camelia (1 x 4 2 x 5 3 x 6)
yoleaux 04:26Z <Zoffix> jnthn: Any tips for what could be causing `if 42 -> *@a { say @a.perl }` to never set `@a`? It works fine with `with 42 -> *@a { say @a.perl }` and the QAST of the two are basically identical (the `if` one is inside a Want, but other than that and the `if`vs`with` ops, the same)
jnthn Like that for infix
.tell Zoffix Perhaps an issue of the compiler (I think down in the QAST -> MAST stage) using .arity to make the decision, and that is 0 on such a block 13:17
yoleaux jnthn: I'll pass your message to Zoffix.
El_Che are there any perl5/6 projects participating in the GSoC? 13:24
DrForr I think TPF gave up on submitting to GSoC a few years ago.. 13:25
sjn why? 13:26
DrForr Getting denied 3 or 4 years in a row, I imagine. 13:28
El_Che DrForr: perl6 is newish :)
wasn't out 3y ago :)
I thought about it because e.g. JJ is in academia 13:34
Geth perl6.org: 1f63e45b32 | (Steve Mynott)++ | source/downloads/index.html
recommend updating zef shipped in Rakudo Star 2018.01
stmuk is ZzZombo zoffix I lose track? 13:36
El_Che no 13:37
DrForr I think you'd have to talk to TPF people directly. My memory is that last time there weren't enough GSoC students interested in Perl to cross the submission threshol.
caa51h m: say (1, 2, 3) >>[[[&(-> $a, $b { "$a x $b" })]]]>> 'a'
camelia (1 x a 2 x a 3 x a)
caa51h Why can I have any number of [ and ]?
stmuk can anyone but zoffix drive p6lert? 13:40
I wanted a message about updating zef similar to the one I just added to the website
DrForr I would *hope* that TPF would be willing to submit GSoC projects for either Perl; I would suspect it comes down to nobody submtting.. 13:41
Er, "nobody presenting projects..."
El_Che a new language with an old name will not create hype, so I get the TPF people 13:43
ZzZombo m: class A {method ^parameterize(\SELF, |c){SELF.new(|c)}};class B is A[:a(1)] {} 13:45
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot resolve caller trait_mod:<is>(B, A, Array); none of these signatures match:
(Mu:U $child, Mu:U $parent)
(Mu:U $child, :$DEPRECATED!)
(Mu:U $type, :$rw!)
(Mu:U $type, :$nat…
DrForr But Python2 -> Python3 is massive hype?...
ZzZombo What the horse?
MasterDuke m: my @a = [[[1]]]; dd @a
camelia Array @a = [1]
MasterDuke m: my @a = [[[1,],],]; dd @a 13:46
camelia Array @a = [[[1],],]
MasterDuke caa51h: without the `,`, single things get collapsed 13:46
(i feel that's a terrible explanation, but am not thinking quickly this morning) 13:47
El_Che DrForr: that the other famous foot shooting incident in FOSS :) 13:48
caa51h MasterDuke: cool. why make this design decision?
El_Che DrForr: python has been picked up in a lot universities to learn programming, hence the situation is different in this case
caa51h oh, so there are some deep meaning. # a terrible explanation 13:49
teatime announces (expecting little interest/relevence to y'all, but just in case I'm wrong) his intention to volunteer w/ the Debian Rakudo packaging team.
El_Che teatime++
DrForr Mmhmm. And it may be something to do with Google doing a lot of work with Python, and now Go... 13:50
caa51h so maybe someday I will understand the true meaning of [[[1]]] by reading more docs
El_Che I got a mail to maintain rakudo in fedora, but I declined to lack of time :(
MasterDuke caa51h: pretty sure i've known why in the past, but it's not coming to me
DrForr teatime++
MasterDuke caa51h: i recommend asking jnthn, Zoffix, lizmat, timotimo, TimToady 13:51
MasterDuke or does anyone else around have an answer? 13:51
caa51h MasterDuke: thank you. you're very kind. 13:52
MasterDuke it's related to docs.perl6.org/type/Signature#inde..._Slurpy%29 and the subsequently linked docs.perl6.org/language/functions#...onventions 13:54
Zoffix stmuk: ATM I'm the only one with keys to p6lert, but if you got a stable IRC host mask, I can add you to list of admins
stmuk Zoffix: I'd rather not have the responsibility :) 13:55
Zoffix stmuk: OK, well give me the message and I send it
stmuk "Users of Rakudo Star 2018 are recommended to update zef via "zef install zef". Newer versions of zef include a work around for an apparent Rakudo bug which shows itself in zef being unable to update mirrors sometimes." 13:57
caa51h > All candidates are currently installed␤No reason to proceed. Use --force-install to continue anyway 13:58
Zoffix Yeah, I get the same. Even with `zef update; zef --/cached install zef` 14:01
stmuk I didn't get that message with a fresh Rakudo Star 2018.01 install
Zoffix zef info zef shows I got `0.2.0`, but I see there's 0.2.2 in repo github.com/ugexe/zef/blob/master/META6.json#L4
stmuk Also see github.com/ugexe/zef/issues/229 14:02
maybe "rm -rf ~/.zef ~/.perl6 && zef install --force-install zef" would be more defensive 14:03
Zoffix `zef --force install zef` (without any rm's) did the trick for me. 14:04
stmuk you could add that flag .. although zef itself will warn if its needed
Zoffix p6lert: add severity:low Users of Rakudo Star 2018 are recommended to update zef via "zef --force install zef". Newer versions of zef include a work around for an apparent Rakudo bug which shows itself in zef being unable to update mirrors sometimes.
p6lert Zoffix, Added alert ID 3: alerts.perl6.org/alert/3
Zoffix p6lert: update 3 severity:low Users of Rakudo Star 2018.01 are recommended to update zef via "zef --force install zef". Newer versions of zef include a work around for an apparent Rakudo bug which shows itself in zef being unable to update mirrors sometimes. 14:05
p6lert Zoffix, Updated alert ID 3: alerts.perl6.org/alert/3
stmuk thanks
Zoffix (changed `2018` to `2018.01`)
It'll take ~10m for it to get live on alerts home page and to get tweeted (the delay is so we could edit the alert before it gets tweeted if we need to)
stmuk you missed a dash from the zef flag
Zoffix stmuk: which dash? I see two `--force` 14:06
stmuk oh --force works
Zoffix Ah, `--force` is a general force option, `--force-install` is an install-only force option. And above it's not dash missing, it's `--force` flag + `install` command 14:07
stmuk I wish zef had full docs (inline pod6?) rather than me having to go to the github page each time
does github render pod6 yet? I forget 14:08
Zoffix No
Still open as D#167 14:09
synopsebot D#167 [open]: github.com/perl6/doc/issues/167 [build][external] GitHub Pod Parsing
Zoffix "You're certainly more than welcome to perform a PR with tests, and update the .travis.yml file so that our CI test suite verifies that Perl6 documentation is being generated."
Zoffix But I don't think GitHub realized that to parse POD6, they'd need to have a fully-functional Perl 6 compiler installed and that parsing pod can execute arbitrary code 14:09
Zoffix or that it's quite expensive to parse POD6 with current compiler (compared to say Perl 5's pod, which looks like it's generated on the fly when you visit a page and gets updated as soon as you make changes to source) 14:11
k, alert got shipped twitter.com/p6lert 14:16
Zoffix &
caa51h Is szabgab here? Part 1, 1.5, 2 and 6 are dead. # szabgab.com/perl6.html#screencast 14:19
stmuk caa51h: I think he has a cron starting things as a workaround .. maybe wait a bit 14:19
caa51h stmuk: ok 14:20
MasterDuke anybody here know anything about OpenSSL::SSL::SSL_read() ? 14:25
El_Che stmuk: how did the flatpack experiment go, btw? Was it accepted upstread as a base layer for other fatpacked soft? 14:42
stmuk: I see that snap is a lot less desktop oriented (and Ubuntu is pushing it aggressively the last few months) 14:45
caa51h it's cool to have a subset of an existing type. perl6++ 14:52
tbrowder AlexDaniel: is there anything wrong with my nqp PR #401? 14:59
AlexDaniel NQP#401
synopsebot NQP#401 [open]: github.com/perl6/nqp/pull/401 improve pod panic msg
AlexDaniel tbrowder: I don't think so 15:01
I just woke up to a pagefull of notifications :) JJ++ 15:02
AlexDaniel tbrowder: how do I reproduce it? 15:05
tbrowder hm, i think i have a test that goes with it....hold on, it’s been so long 15:06
AlexDaniel m: #=foo␤=cut 15:07
camelia 5===SORRY!5=== Error while compiling <tmp>
Preceding context expects a term, but found infix = instead.
Did you make a mistake in Pod syntax?
at <tmp>:2
------> 3=cut7⏏5<EOL>
AlexDaniel that's not it… 15:07
ZzZombo m: class A {method ^parameterize(\SELF, |c){SELF.new(|c)}};class B is A[:a(1)] {} 15:11
camelia 5===SORRY!5=== Error while compiling <tmp>
Cannot resolve caller trait_mod:<is>(B, A, Array); none of these signatures match:
(Mu:U $child, Mu:U $parent)
(Mu:U $child, :$DEPRECATED!)
(Mu:U $type, :$rw!)
(Mu:U $type, :$nat…
ZzZombo Why?
ZzZombo m: class A {method ^parameterize(\SELF, |c){SELF.new(|c)}};class B is A [:a(1)] {} 15:12
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse class definition
at <tmp>:1
------> 3e(\SELF, |c){SELF.new(|c)}};class B is A7⏏5 [:a(1)] {}
AlexDaniel s: &emit 15:19
SourceBaby AlexDaniel, Sauce is at github.com/rakudo/rakudo/blob/f559...ol.pm#L159
tbrowder m: =pod \n=cut 15:21
camelia ( no output )
tbrowder AlexDaniel: now i remember, i had a test that used perl 5 since i needed to write a file and eval it with shell but Zoffix shot it down. i haven’t been able to use the asynch replacement for nqp::shell successfully, thus no test yet exists for the panic. 15:24
tbrowder i also got hung up trying to write a file but that’s no longer a show-stopper. forget about the pr for now. thanks. 15:26
AlexDaniel :) 15:27
SmokeMachine m: say ^10 .unique: * %% 2 15:34
camelia Cannot resolve caller unique(Range: WhateverCode); none of these signatures match:
($: *%_)
($: :&as!, :&with!, *%_)
($: :&as!, *%_)
($: :&with!, *%_)
in block <unit> at <tmp> line 1
SmokeMachine m: say <1 2 3 4 5 6 7 8 9>.unique: * %% 2
camelia Cannot resolve caller unique(List: WhateverCode); none of these signatures match:
($: *%_)
($: :&as!, :&with!, *%_)
($: :&as!, *%_)
($: :&with!, *%_)
in block <unit> at <tmp> line 1
SmokeMachine m: say ^10 .unique: :as{ * %% 2 }
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed double closure; WhateverCode is already a closure without curlies, so either remove the curlies or use valid parameter syntax instead of *
at <tmp>:1
------> 3say ^10 .unique: :as{ * %% 2…
SmokeMachine m: say ^10 .unique: :as{$_ %% 2} 15:35
camelia (0 1)
MasterDuke m: say ^10 .unique: :as(* %% 2) 15:36
camelia (0 1)
SmokeMachine is it always true that unique will get the first values that &as answer the same value?
m: say ^10 .reverse.unique: :as{$_ %% 2} 15:37
camelia (9 8)
benjikun Anyone have any idea how to fix this? gist.github.com/anonymous/dfe93372...f4a198a990 16:38
"Too many open files" 16:39
I thought doing *.IO would close the file after retrieving the data, does it not? 16:41
Zoffix benjikun: it does. Are you running many versions of &scan at the same time perhaps? 16:44
benjikun I'm using this to run the sub every 1200 seconds: `Supply.interval(1200).tap: { scan(); }
Zoffix benjikun: what's your compiler's version? perl6 -vv 16:46
I mean perl6 -v
benjikun `This is Rakudo version 2017.12 built on MoarVM version 2017.12 implementing Perl 6.c.`
Zoffix oh lol 16:47
I see now :)
benjikun Is that version too old?
Zoffix benjikun: it does close the files, but only if you consume the entire .lines Seq
Which you never do, so it sits and waits for you to eat all the lines
benjikun Oh, interesting 16:48
Should I just `my $lines = "posts/$i.html".IO.lines;`, or is there a more efficient way of doing things
AlexDaniel I guess you'll have to .open and .close it explicitly? 16:51
but also, that's a rakudo bug I'd say…
Zoffix benjikun: that won't consume the Seq either. with "posts/$i.html".IO.lines { @titles[$latest - $i] = .head; .sink } 16:52
benjikun: or with "posts/$i.html".IO.open { @titles[$latest - $i] = .get; .close } 16:53
AlexDaniel: it isn't
or "posts/$i.html".IO.lines andthen @titles[$latest - $i] = .head andthen .sink 16:54
oh wait
AlexDaniel Zoffix: why not? It leaves the file open when I won't be able to read from it anymore.
Zoffix benjikun: there's limit arg to lines. Just change .lines[0] to .lines: 1 16:55
benjikun What do you mean by "limit arg to lines" 16:56
Zoffix AlexDaniel: because it doesn't know what you've done with that Seq.
AlexDaniel Zoffix: sure, but then it will be garbage collected at some point, no? 16:56
Zoffix AlexDaniel: yes, and then it'll be closed.
Zoffix benjikun: the .lines method takes an argument. Second paragraph here: docs.perl6.org/routine/lines#(IO::...tine_lines 16:56
benjikun: change `@titles[$latest - $i] = "posts/$i.html".IO.lines[0]; ` to `@titles[$latest - $i] = "posts/$i.html".IO.lines: 1; 16:57
benjikun ah, I think I understand
thanks for the help
AlexDaniel oh, so if you stick a force_gc into that thing it starts “wörking”?
Zoffix or @titles[$latest - $i] = "posts/$i.html".IO.lines(1) in more familiar spelling
AlexDaniel: yeah. That's why .lines got $limit implemented. The original solution to this problem was making .lines slurp the entire file at once. I think that was around March 2017 and then in IO Grant it was reverted and $limit was added 16:58
Zoffix This was it I think github.com/rakudo/rakudo/blob/mast...discussion 17:00
Except for `:$leave-open` stuff, which was turned down
AlexDaniel Zoffix: I see. Thanks 17:02
maybe that's some language/traps material 17:03
SmokeMachine do you guys think this is the right usage in this case? www.irccloud.com/pastebin/2qDpbah0/ 17:04
Zoffix *shrug* The docs for .lines do say "NOTE: the lines are read lazily, so ensure the returned Seq is either fully reified or is no longer needed when you close the handle or attempt to use any other methods that change the file position." 17:05
That's for IO::Handle.lines and for IO::Path.lines it says "NOTE: the lines are ready lazily and the handle used under the hood won't get closed until the returned Seq is fully reified, so ensure it is, or you'll be leaking open file handles. (TIP: use the $limit argument)" 17:06
AlexDaniel that's pretty good I guess
SmokeMachine www.irccloud.com/pastebin/2EncEfke/
Zoffix It's also mentioned in the IO Guide: docs.perl6.org/language/io-guide#R...from_files 17:07
AlexDaniel Zoffix++
Zoffix SmokeMachine: does it even work? The call that sets :$*v in the proto can't ever get to the candidates that use $*v but don't take a :$v arg 17:08
m: proto z (:$*v) {*}; multi z { dd $*v }; z :42v
camelia Unexpected named argument 'v' passed
in sub z at <tmp> line 1
in block <unit> at <tmp> line 1
Zoffix m: proto z (:$*v) {*}; multi z (|) { dd $*v }; z :42v; # need this 17:09
camelia Int $*v = 42
Zoffix m: proto z (:$*v) {*}; multi z (*%) { dd $*v }; z :42v; # or this
camelia Int $*v = 42
Zoffix m: proto z (:$*v) {*}; multi z (:$v) { dd $*v }; z :42v; # or this, but then messing around with dynamics is kinda pointless 17:09
camelia Int $*v = 42
SmokeMachine Zoffix: www.irccloud.com/pastebin/m6F0VBsR/
Zoffix SmokeMachine: what I meant that in this candidate `multi MAIN($a, $b) {say $*v, $a, $b}`, $*v would always be an Any type object, because it's impossible for it to get set from the proto. If it were set in the proto, that candidate would not be able to take it 17:10
Zoffix Try `-v=42 foo bar` 17:11
SmokeMachine Zoffix: www.irccloud.com/pastebin/flKm9nZK/ 17:12
Zoffix Exactly.
SmokeMachine Zoffix: yes, but is the usage right?
Zoffix Ah 17:13
Zoffix SmokeMachine: no, it's wrong. It lists ` -e '...' [-v=<Any>] <a> <b> `. If you ignore the `'...'` bit, that's exactly the args you gave it. Yet it can't accept it 17:14
SmokeMachine: basically someone (maybe me) told you yesterday that it was a bug that stuff from proto wasn't listed in the usage... and I think it's possible it shouldn't be listed. 17:15
SmokeMachine :( 17:15
Zoffix It kinda gets complicated. `proto z (:$*v) {*}; multi z (|) { dd $*v };` means usage `[-v=<Any>]` only. `proto z (:$*v, |) {*}; multi z () { dd $*v };` means no valid canidates 17:16
Actually, no, last one can work with no args. But the `-v` from the proto never should be shown in usage, because no candidate can take it 17:17
Zoffix So it's possible it's best to leave this complexity out and let people who want to have proto MAIN write their own $*USAGE 17:18
SmokeMachine Zoffix: and if the usage says this, or something like that... www.irccloud.com/pastebin/tC05K7pI/
Zoffix SmokeMachine: what does "Global parameters" mean? 17:19
They're not quite global, are they, since only one candidate can take them. And in my last example above, no candidate can take it
Zoffix SmokeMachine: also, the usage still lists `'...'` from the proto, which doesn't apply to any candidate 17:20
Personally, I'd just leave it alone and ignore the args from proto. The usecase is too small for the maintenance burden.
Zoffix &
SmokeMachine Zoffix: I was meaning something like: proto params 17:21
SmokeMachine any other opinion about that? 17:28
jnthn: ?
travis-ci Doc build errored. JJ Merelo 'Testing with whateverable' 17:53
travis-ci.org/JJ/doc/builds/339903131 github.com/JJ/doc/compare/d9485954...daef5751d2
travis-ci Doc build errored. JJ Merelo 'Testing with sudo' 17:58
travis-ci.org/JJ/doc/builds/339903928 github.com/JJ/doc/compare/97daef57...2d2d3f5c68
travis-ci Doc build failed. JJ Merelo 'Test with travis' 18:16
travis-ci.org/JJ/doc/builds/339908047 github.com/JJ/doc/compare/552d2d3f...69f0a1c2c4
travis-ci Doc build failed. JJ Merelo 'Tests with SSL' 18:47
travis-ci.org/JJ/doc/builds/339914458 github.com/JJ/doc/compare/5769f0a1...8867f0adf0
travis-ci Doc build errored. JJ Merelo 'YAML syntax gets in the way...' 18:49
travis-ci.org/JJ/doc/builds/339915428 github.com/JJ/doc/compare/078867f0...f36978b46b
AlexDaniel weekly: 8000th commit github.com/perl6/doc 18:51
notable6 AlexDaniel, Noted!
travis-ci Doc build errored. JJ Merelo 'Putting stuff properly' 18:54
travis-ci.org/JJ/doc/builds/339916014 github.com/JJ/doc/compare/b3f36978...9751908240
Geth doc: 1b62109e13 | (Zoffix Znet)++ (committed using GitHub Web editor) | .travis.yml
Tell travis to fetch libssl-dev

Looks like OpenSSL is now in prereqs		 19:01
Geth doc: b1b107f867 | (Zoffix Znet)++ | .travis.yml
Revert "Tell travis to fetch libssl-dev"

This reverts commit 1b62109e13fd0bb6f8cf0667d4cfac3ed6dfc32b.
Never mind... Travis was reporting failures for a *fork* in our chat.		 19:04
Zoffix That's kinda lame that travis bot reports failures for forks. 19:05
AlexDaniel hehe
geekosaur yes, it's a known wart. with a workaround where you store the IRC information in an encrypted form based on the project, so it won't match in a fork and reporting fails 19:16
(really, how's it going to know? it trusts the .travis.yml) 19:17
moritz well, it could ask github through the API if a project is a fork 19:18
and don't report, unless explicitly overrideen 19:19
*overridden
Geth whateverable: 319716d8f0 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
New bot: Shareable

Basically, it makes every build publicly available. The bot itself simply dumps links to whateverable.6lang.org/.
The code can be improved a bit but that's for later. Right now we're just testing the concept.		 19:22
whateverable: c8d1db6ec8 | (Aleks-Daniel Jakimenko-Aleksejev)++ | 2 files
⚠ Make all bots runnable locally (this is so cool!)

Previously if you wanted to run bisectable or some other bot locally you had to build tens (if not thousands) of rakudo versions to get something useful out of it. Now whenever it stumbles upon a missing build it will attempt to pull it from the main server. If you're on linux x86_64, this means that the bots you start are no worse than the ... (11 more lines)
Zoffix :o 19:23
On a related note, JJMerelo wanted to do previous versions of rakudo as docker images: www.facebook.com/groups/perl6/perm...926523156/ 19:24
AlexDaniel on the same note, they also attempted to use whateverable for that: github.com/perl6/doc/issues/1765#i...-364680210 19:25
without much success, because… ubuntu 14.04?? Seriously? 19:26
Geth whateverable: 64cbc09b6a | (Aleks-Daniel Jakimenko-Aleksejev)++ | services/whateverable-all.service
Start Shareable along with other bots

Forgot to do that in 319716d8f006f0af474700a57319aa457724b2f1. See also: #122		 19:27
MasterDuke i thought there was a way to get a newer ubuntu in travis?
AlexDaniel how? 19:28
El_Che no, you can't
MasterDuke dunno, istr samcv doing it for something
El_Che however, I work around that by using docker
samcv istr?
moritz "i seem to recall"
El_Che github.com/nxadm/rakudo-pkg/blob/m...travis.yml
samcv well you can get a newer ubuntu. but i think we already had the latest ubuntu? at least maybe on some repos 19:29
AlexDaniel shareable: HEAD 19:29
shareable AlexDaniel, whateverable.6lang.org/HEAD
AlexDaniel wrong nickname…
Zoffix samcv: FWIW one of your commits broke Windows build: github.com/rakudo/rakudo/issues/1511 I tried looking at it, but couldn't figure out if the files with `:` in their names were auto generated or not
samcv ohh 19:30
ok i can see how that could happen
::)
Geth whateverable: e3df1090a4 | (Aleks-Daniel Jakimenko-Aleksejev)++ | bin/Shareable.p6
Oops, wrong nickname
AlexDaniel shareable6: 2015.12^ 19:30
shareable6 AlexDaniel, whateverable.6lang.org/2015.12^
AlexDaniel well… that's not a proper link 19:31
AlexDaniel obviously needs a bit more work, but I'm happy with it anyway :) 19:31
MasterDuke but it works
AlexDaniel yea, it does
MasterDuke btw, i'm trying to --profile downloading just a single build with the HTTP::UserAgent version of your leak-some-memory.p6 script. it creates a 14m profile that even the qt viewer can't open 19:33
AlexDaniel MasterDuke: what about Cro version? 19:34
because it does leak as well
but with a smaller memory footprint
MasterDuke there may be some inefficiencies in HTTP::UA or its dependencies, just guessing
AlexDaniel weekly: 500th commit in github.com/perl6/whateverable, two new bots: github.com/perl6/whateverable/wiki/Notable github.com/perl6/whateverable/wiki/Shareable 19:35
notable6 AlexDaniel, Noted!
MasterDuke it's such a drastic difference between the Cro and HTTP::UA versions that i'm not sure it would be the same cause
samcv i am going to commit a change to fix the windows build in a few minutes 19:36
AlexDaniel shareable6: help 19:37
shareable6 AlexDaniel, Like this: shareable6: f583f22 # See wiki for more examples: github.com/perl6/whateverable/wiki/Shareable
AlexDaniel shareable6: f583f22
shareable6 AlexDaniel, whateverable.6lang.org/f583f22
AlexDaniel shareable6: help 19:48
shareable6 AlexDaniel, Like this: shareable6: f583f22 # See wiki for more examples: github.com/perl6/whateverable/wiki/Shareable
AlexDaniel ↑ some help here
AlexDaniel but I know that perhaps this is not super useful for most people right now 19:49
AlexDaniel fwiw moarvm builds are also available: whateverable.6lang.org/HEAD?arch=l...ype=moarvm 19:52
though not sure why anybody would need that given that moarvm can be compiled really fast
benjikun What other languages do you guys primarily use? 23:22
I'm suspecting perl5 & C the most 23:23
AlexDaniel … SystemVerilog 23:28
.oO( vague question gets a nonsensical answer )
benjikun: OK, actually most of my programming needs are satisfied by perl6 23:29
so if I'm using something else then it usually means that this project was started some time ago and was using something else in the first place
aaaand… yeah 23:30
jnthn It musta been nearly 15 years since I wrote something in Verilog. Fun to program in something where the default is "everything happens at once". :) 23:31
AlexDaniel jnthn: oh no! Verilog and “program” almost in one sentence! Now someone will have to come to say that it's a hardware *description* language :) 23:39
.oO( – but testbenches? – Testbenches are irrelevant! )
benjikun AlexDaniel: How do you feel about C++17? 23:51
AlexDaniel benjikun: my use of c++ is accidental, so I don't have any opinion on that 23:53
benjikun I haven't used C/C++ in a long while, but I'm thinking about heading back and using C++17 23:55

Powered by Cro and the Raku® Programming Language.

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

Thank you!