»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_log/perl6 | UTF-8 is our friend! 🦋
Set by Zoffix on 25 July 2018.
00:00 telex left 00:02 telex joined 00:03 p9s``` joined 00:08 p9s``` left 00:33 wildtrees left 00:42 p9s``` joined 00:48 p9s``` left
vrurg Who remembers in which section of docs.perl6.org FQN method call is described? /me is stuck... 01:04
01:13 reach_satori_ joined 01:14 molaf left 01:15 p9s``` joined, satori__ left
softmoth vrurg, maybe docs.perl6.org/language/modules#require ? 01:17
vrurg softmoth: No. I need the section about self.Parent::method() 01:18
timotimo i see two matches in the source 01:19
language/objects.pod6 has Steerable::steer and language/operators.pod6 has Foo.Bar::baz
01:20 p9s``` left
timotimo also, X::Method::InvalidQualifier has an example of it, and the Metamodel/MROBasedMethodDispatch doc also mentions that syntax 01:20
01:21 sftp left, sftp joined 01:26 snarkyboojum_ left, molaf joined, snarkyboojum_ joined, ufobat_ left 01:27 ufobat_ joined, llfourn_ left 01:28 llfourn joined 01:30 sftp left, cpan-p6 left
Xliff Hmmm... I don't like that write-up with role Bull-Like and role Steerable. 01:30
It's missing a step.
01:30 agentzh left, abraxxa left 01:31 jjatria left, agentzh joined, agentzh left, agentzh joined, TimToady joined, sftp joined, jjatria joined 01:32 cibs joined 01:39 snarkyboojum_ left
vrurg timotimo: thanks! 01:40
Ok, I wanted to make sure I miss nothing. 01:42
m: role R[::T] { method foo { } }; class C does R[Int] { method foo { self.R[Int]::foo } }; C.new.foo
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3 does R[Int] { method foo { self.R[Int]:7⏏5:foo } }; C.new.foo
expecting any of:
colon pair
Xliff m: role R[::T] { method foo { } }; class C does R[Int] { method foo { self.R::foo } }; C.new.foo 01:43
camelia No concretization found for R
in method foo at <tmp> line 1
in block <unit> at <tmp> line 1
Xliff m: role R[] { method foo { } }; class C does R[Int] { method foo { self.R::foo } }; C.new.foo
camelia 5===SORRY!5=== Error while compiling <tmp>
No appropriate parametric role variant available for 'R'
at <tmp>:1
Xliff m: role R { method foo { } }; class C does R[Int] { method foo { self.R::foo } }; C.new.foo 01:44
camelia 5===SORRY!5=== Error while compiling <tmp>
No appropriate parametric role variant available for 'R'
at <tmp>:1
Xliff m: role R { method foo { } }; class C does R { method foo { self.R::foo } }; C.new.foo
camelia ( no output )
Xliff m: role R[::T] { method foo { } }; class C does R[Int] { method foo { self.R::foo } }; C.new.foo
camelia No concretization found for R
in method foo at <tmp> line 1
in block <unit> at <tmp> line 1
Xliff vrurg: That looks like a parsing misstep.
m: role R[::T] { method foo { } }; class C does R[Int] { method foo { self.R[Int]::foo } }; C.new.foo
camelia 5===SORRY!5=== Error while compiling <tmp>
Confused
at <tmp>:1
------> 3 does R[Int] { method foo { self.R[Int]:7⏏5:foo } }; C.new.foo
expecting any of:
colon pair
vrurg Xliff: no way. Besides, self.R::foo is totally incorrect. Imagine class C does R[Int] does R[Str] { }. 01:45
Xliff: you're right. I was just checking if it worth a ticket and nothing is missed.
Xliff vrurg: In that situation you WILL have conflicting roles.
vrurg Xliff: no, I won't. 01:46
Xliff Yes. You will.
vrurg role R[::T] { method foo { } }; class C does R[Int] does R[Str] { method foo { } }; C.new.foo
m: role R[::T] { method foo { } }; class C does R[Int] does R[Str] { method foo { } }; C.new.foo
camelia ( no output )
vrurg ^ Xliff
The point is that R[Int] and R[Str] are different type objects. 01:47
Xliff O I C
Still, the parser will balk if you try to specify a role-with-parameter in the self.Parent::method format. 01:48
You should probably send that to problem-solving.
vrurg is not sure wether it's a bug or a new feature. ;) 01:50
But you're probably right. I was is doubt about this, but now I'm pushed toward problem-solving... 01:51
01:53 p9s``` joined
Xliff vrurg: I think what you are trying is something no one thought of. 01:54
Which is why the parser doesn't handle it.
Geth_ ¦ problem-solving: vrurg assigned to jnthn Issue FQN method call needs support for parameterized roles. github.com/perl6/problem-solving/issues/110 01:55
Xliff Still.... expect questions asking: "Why?"
vrurg Simple. Because parameterized roles can have methods of the same name we one would need to specify which one he is interested in. 01:56
Xliff Actually, the write-up you did does the job quite well! :)
Actually, I stand corrected. 01:57
In that example, don't you duplicate $!a?
vrurg Xliff: actually, yes, I do.
Xliff I don't think Perl6 will like that... ;)
vrurg is too tired after one of the craziest days of the last weeks... 01:58
Xliff: though I have fixed today `role R { has $.a; }; role R1 does R { }; role R2 does R2 { }; class C does R1 does R2 { }` – this one is valid.
I have simplified the example. Common role was redundant in first place. 02:00
02:00 p9s``` left
Xliff OK. I see. 02:02
My problem with such examples (and I just submitted a bug to docs about this). 02:03
Why do you have R1 and R2, but do NOTHING with R2?
It might help if you show how they are to be used together.
vrurg: Sorry you had a crazy day. Good crazy or bad crazy.... 02:05
... or just crazy?
vrurg Xliff: I think anyone can find a scenario where both are needed.
A bit of everything. So, just crazy maybe? ;)
Xliff vrurg: OK, if you say so. However, I suspect it's one of the first questions that will be asked. 02:06
Hopefully tomorrow will be more sane.
vrurg In either case I'm a step close to complete R#3199 and this is good for sure.
synopsebot R#3199 [open]: github.com/rakudo/rakudo/pull/3199 [WIP][roles] [WIP] Implement perl6/problem-solving#103
Xliff Ah! That's good. 02:07
Maybe you could check something for me.
vrurg: Do you think this is a Method::Also issue or a rakudo one?
github.com/lizmat/Method-Also/issues/1 02:08
02:08 p9s``` joined
vrurg Xliff: at the first glance I suspect Method::Also. Did you try it without the trait? 02:10
Xliff Yesh. The first two examples basically ignore the trait.
vrurg Ah, right.. Sure, 90% for Method::Also, 10% for rakudo. 02:11
Xliff Uh. Hold on thar, pardnah 02:12
vrurg Looks like the trait doesn't install proto for multi methods. Let me see if add_method is supposed to do it implicitly. But I'd rather say it should not.
02:14 softmoth left
Xliff vrurg: See update. 02:15
vrurg Xliff: Nah, wrong. You alias the proto. Try introspecting my-method and see if candidates are there.
Xliff Not for roles, but they are for class.
See updated bug
According to lizmat, aliasing the proto is the expected mechanism. 02:16
vrurg Right. Role makes things different. When it's applied to a class you get two new protos on the class. But then I don't remember the details of the order in which things happen. Looks like candidates are not installed on my-method. 02:18
Xliff Right, but if it works for class, then wouldn't it NOT working for roles make it internal?
vrurg As I said, try to introspect my-method on the class. 02:19
Xliff OK. One sec.
02:22 Actualeyes joined
Xliff .say for D.^find_method('my-method').candidates --> (my_method, my_method) 02:22
vrurg I don't currently see what could go wrong because same proto-method object is just getting installed under a different name. All candidates are installed on that object and thus should be available. This is how it works with class. But for the role it looks like protos are cloned early, before candidates are installed. But this must not happen as role is already composed by the time it is applied to the class.
Xliff .say for B.^find_method('my-method').candidates --> () 02:24
^^ Role
02:24 Cabanossi left
Xliff .say for B.^find_method('my_method').candidates --> (my_method, my_method) 02:26
^^ Role, non-alias
So maybe Method::Also isn't doing enough?
vrurg Looks like it's Method::Also, after all. It's using add_method on proto, but for multis things are a bit more complicated. 02:27
Xliff Aha!
vrurg Xliff: I think you're right.
Xliff Thanks!
Yeah. I'm doing that all over GTK, so I kinda want this to work.
What extra would need to be done for a role?
If I can figure that out, I can submit a PR.
I think the big difference is that it's ClassHOW vs ParametricRoleHOW. 02:28
Adding the alias to the proto will work for ClassHOW. 02:29
vrurg Maybe it would be enough to use add_multi_method but I'm not sure if this would work if only applied to the proto. Needs testing.
Xliff OK. Well this is stuff that I can share with lizmat. 02:30
vrurg Xliff: if you want to find out more, look in Perl6/Metamodel/MultiMethodContainer.nqp and Perl6/Metamodel/RoleToRoleApplier.nqp
Xliff Ooh. I missed MultiMethodContainer
vrurg++: Thanks!
vrurg my pleasure! 02:31
It was good timing because this is exactly the area where I currently wandering. ;) 02:32
Xliff Weirdly enough... I was too.
02:32 Doc_Holliwood joined 02:37 Cabanossi joined 02:46 p9s``` left, p9s``` joined 03:08 p9s``` left 03:21 p9s``` joined 04:02 Doc_Holliwood left 04:07 p9s``` left, p9s``` joined 04:19 p9s``` left 04:20 p9s``` joined, wamba joined 04:34 jmerelo joined 04:36 wamba left 04:49 ponbiki joined 04:52 wamba joined
Xliff m: use NativeCall; class MockOperation is repr<CStruct> { has uint32 $.requested; has uint32 $.done }; my $m = MockOperation.new; say nativecast(Pointer, $m) 04:58
camelia NativeCall::Types::Pointer<0x4281660>
05:15 robertle left 05:34 sauvin joined 05:35 sauvin left 05:39 discoD left 05:41 domidumont joined 05:53 domidumont left 06:04 kini is now known as fs, fs left, fs joined 06:07 fs is now known as kini, kini left, kini joined 06:25 ravenousmoose joined 06:30 xinming left, kensanata joined, xinming joined 06:39 jmerelo left 06:44 wamba left 06:45 wamba joined 06:54 domidumont joined 06:55 ravenousmoose left 06:56 ravenousmoose joined 07:09 sena_kun joined, HarmtH joined 07:14 Kaffe left 07:17 wamba left 07:22 kensanata left 07:26 p9s``` left 07:36 abraxxa joined 07:39 ravenous_ joined 07:41 ravenousmoose left 07:45 agentzh left 07:47 agentzh joined 07:58 p9s``` joined 08:02 wamba joined 08:04 p9s``` left 08:17 Cabanossi left 08:21 Cabanossi joined 08:28 Doc_Holliwood joined 08:36 p9s``` joined, Sgeo_ joined
Xliff m: say $?*A 08:38
camelia 5===SORRY!5=== Error while compiling <tmp>
Unsupported use of $? variable; in Perl 6 please use $! for handling child errors also
at <tmp>:1
------> 3say $?7⏏5*A
Xliff m: say ?$*A
camelia False
Xliff m: sub bb { say $*A; }; sub aa { my $*A = 4; bb; }; say ?$*A; aa; 08:39
camelia False
4
Xliff m: sub bb { say ?$*A; }; sub aa { my $*A = 4; bb; }; say ?$*A; aa;
camelia False
True
Xliff m: sub bb { say $*A if ?$*A; }; sub aa { my $*A = 4; bb; }; say ?$*A; aa; 08:40
camelia False
4
08:40 Sgeo left
Xliff m: say "aaa" ~~ / (.) {} ($0)+ /; 08:50
camelia Nil
Xliff m: say "aaa" ~~ / (.) {} :my $c = $0; ($c)+ /; 08:51
camelia 「aaa」
0 => 「a」
1 => 「a」
1 => 「a」
Xliff m: say "aba" ~~ / (.) {} :my $c = $0; ($c)+ /;
camelia Nil
Xliff m: say "bbbbbb" ~~ / (.) {} :my $c = $0; ($c)+ /;
camelia 「bbbbbb」
0 => 「b」
1 => 「b」
1 => 「b」
1 => 「b」
1 => 「b」
1 => 「b」
Xliff m: say "aaa" ~~ / (.) {} ($0)+ /; 08:54
camelia Nil
08:54 ufobat__ joined
Xliff m: say "aaa" ~~ / (.) {} $0+ /; 08:54
camelia 「aaa」
0 => 「a」
Xliff m: say "aaa" ~~ / (.) {} $0 /;
camelia 「aa」
0 => 「a」
Xliff m: say "aaa" ~~ / (.) {} ($0)+ /;
camelia Nil
Xliff m: say "bbbbbb" ~~ / (.) {} :my $c = $0; ($c)+ /;
camelia 「bbbbbb」
0 => 「b」
1 => 「b」
1 => 「b」
1 => 「b」
1 => 「b」
1 => 「b」
Xliff m: "bbbbbb" ~~ / (.) {} :my $c = $0; ($c)+ /; $c.say 08:55
camelia 「b」
Xliff m: "bbbbbb" ~~ / (.) {} :my $c = $0; ($c)+ /; $c.perl.say
camelia Match.new(hash => Map.new(()), from => 0, list => (), orig => "bbbbbb", pos => 1, made => Any) 08:56
08:57 ufobat_ left 09:01 ravenous_ left 09:04 ravenousmoose joined, xinming_ joined 09:06 Black_Ribbon left 09:07 xinming left 09:08 pecastro joined 09:11 p9s``` left, Doc_Holliwood left
Xliff m: "_transfer_full_" ~~ / '_' <[\w _]>+ '_'/; say $0 09:19
camelia Nil
Xliff m: "_transfer_full_" ~~ / '_' <[\w _]>+? '_'/; say $0
camelia Nil
Xliff m: "_transfer_full_" ~~ / '_' <[a..z _]>+? '_'/; say $0
camelia Nil
Xliff m: "_transfer_full_" ~~ / '_' ~ '_' (\w+) /; say $0 09:20
camelia 「transfer_full」
Xliff m: "_transfer_full_" ~~ / '_' ~ '_' (\w+) /; say $/.perl
camelia Match.new(from => 0, orig => "_transfer_full_", pos => 15, list => (Match.new(list => (), hash => Map.new(()), made => Any, from => 1, pos => 14, orig => "_transfer_full_"),), made => Any, hash => Map.new(()))
Xliff m: "_transfer_full_" ~~ / '_' ~ '_' (\w+) /; say $0.perl
camelia Match.new(list => (), orig => "_transfer_full_", pos => 14, from => 1, hash => Map.new(()), made => Any)
Xliff m: "_transfer_full_" ~~ / '_' ~ '_' (\w+) /; say $0.gist
camelia 「transfer_full」
Xliff m: "_transfer_full_" ~~ / '_' (\w+) '_' /; say $0 09:21
camelia 「transfer_full」
Xliff m: "_transfer_full_" ~~ / '_' (\w+) /; say $0
camelia 「transfer_full_」
09:32 ravenousmoose left 09:34 ravenousmoose joined 09:35 Doc_Holliwood joined 09:37 pat_js joined 09:38 p9s``` joined 09:44 p9s``` left 09:59 reach_satori_ left 10:06 aborazmeh joined, aborazmeh left, aborazmeh joined 10:15 p9s``` joined 10:18 softmoth joined 10:21 p9s``` left 10:27 domidumont left 10:38 Voldenet left 10:41 Voldenet joined, Voldenet left, Voldenet joined 10:49 reach_satori_ joined 10:50 p9s``` joined 10:51 isBEKaml joined 10:56 p9s``` left 10:58 Sgeo__ joined 11:01 Sgeo_ left 11:04 dogbert17 joined 11:11 cpan-p6 joined, cpan-p6 left, cpan-p6 joined
Xliff m: 0x4488.Str.say 11:16
camelia 17544
Xliff m: 17544.fmt('%x').say
camelia 4488
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new($x.comb(2).map( *.Int )); $b.say 11:18
camelia Blob:0x<2C 58>
11:18 aborazmeh left
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new($x.comb(2)); $b.say 11:18
camelia Type check failed in initializing element #0 to Blob; expected uint8 but got Str ("44")
in block <unit> at <tmp> line 1
Xliff m: use experimental; my $x = 17544.fmt('%x'); $x.comb(2).say
camelia (44 88)
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).map( "0x{$_}".Int ) ); $b.say 11:19
camelia Use of uninitialized value $_ of type Any in string context.
Methods .^name, .perl, .gist, or .say can be used to stringify it to something meaningful.
in block <unit> at <tmp> line 1
Cannot map a Seq using 'Failure.new(exception => X::Str::Nu
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).map({ "0x{$_}".Int }) ); $b.say
camelia Blob:0x<44 88>
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).reverse.map({ "0x{$_}".Int }) ); $b.say 11:20
camelia Blob:0x<88 44>
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).reverse.map({ "0x{$_}".Int }) ); $b.say; $b.read-int32.say
camelia Blob:0x<88 44>
Too few positionals passed; expected 2 or 3 arguments but got 1
in block <unit> at <tmp> line 1
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).reverse.map({ "0x{$_}".Int }) ); $b.say; $b.read-int32(0).say 11:21
camelia Blob:0x<88 44>
MVMArray: read_buf out of bounds offset 0 start 0 elems 2 count 4
in block <unit> at <tmp> line 1
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).reverse.map({ "0x{$_}".Int }) ); $b.say; $b.read-int16(0).say
camelia Blob:0x<88 44>
17544
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).reverse.map({ "0x{$_}".Int }) ); $b.say; $b.read-int16(0, BigEndian).say 11:22
camelia Blob:0x<88 44>
-30652
Xliff m: use experimental; my $x = 17544.fmt('%x'); my $b = Blob.new( $x.comb(2).reverse.map({ "0x{$_}".Int }) ); $b.say; $b.read-uint16(0, BigEndian).say
camelia Blob:0x<88 44>
34884
11:22 MilkmanDan left 11:23 p9s``` joined 11:24 MilkmanDan joined
Xliff Does perl6 provide an easier way to swap endian-ness of an Integer? 11:24
timotimo why would you use read-int but not write-int? 11:26
Xliff Is there a write-int?!
timotimo of course
Xliff docs.perl6.org/type/Blob#method_read-uint32 ?!??
timotimo i'm still wishing for write-* on Buf:U
well, yeah, blobs are immutable 11:27
lizmat on Buf:U ??
timotimo yes
lizmat that should auto-vivify ?
timotimo it'd allocate a new buf, write to it, then return the object
docs.perl6.org/type/Buf#method_write-uint8 - xliff
Xliff m: my $b = Blob.new.write-int(0x88884444).read-int(0, LittleEndian).say
camelia No such method 'write-int' for invocant of type 'Blob'
in block <unit> at <tmp> line 1
Xliff m: my $b = Blob.new.write-int(0x88884444).read-uint32(0, LittleEndian).say 11:28
camelia No such method 'write-int' for invocant of type 'Blob'
in block <unit> at <tmp> line 1
timotimo Buf!
Xliff m: my $b = Bug.new.write-uint32(0x88884444).read-uint32(0, LittleEndian).say
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared name:
Bug used at line 1. Did you mean 'Bag', 'Buf'?
lizmat it could work for Blob:U as well
Xliff m: my $b = Buf.new.write-uint32(0x88884444).read-uint32(0, LittleEndian).say
11:28 p9s``` left
camelia Too few positionals passed; expected 3 or 4 arguments but got 2
in block <unit> at <tmp> line 1
11:28
timotimo that's true!
lizmat as long as its :U
Xliff m: my $b = Buf.new.write-uint32(0, 0x88884444).read-uint32(0, LittleEndian).say
camelia Nil
timotimo Xliff: write-uint32 doesn't return self
Xliff m: my $b = Buf.new.write-uint32(0, 0x88884444); $b.read-uint32(0, LittleEndian).say
camelia No such method 'read-uint32' for invocant of type 'Any'
in block <unit> at <tmp> line 1
timotimo m: (my $b = Buf.new).write-uint32(0, 0x88884444); $b.read-uint32(0, LittleEndian).say
camelia 2290631748
Xliff m: my $b = Buf.new; $b.write-uint32(0, 0x88884444); $b.read-uint32(0, LittleEndian).say
camelia 2290631748
Xliff !sleep... wheee! 11:29
m: 0x44448888.say
camelia 1145342088
Xliff m: my $b = Buf.new; $b.write-uint32(0, 0x88884444); $b.read-uint32(0, BigEndian).say
camelia 1145342088
timotimo m: my $b = Buf.new; $b.write-uint32(0, 0x88884444); $b.read-uint32(0, BigEndian).base(16).say
camelia 44448888
Xliff OK. So now I have my answer. timotimo++ 11:30
Trying to simplify this: github.com/GNOME/glib/blob/master/...t-stream.c
11:31 wamba left
lizmat so I wonder, why don't the write-* methods return self ? 11:31
Xliff lizmat++
lizmat looking at them now, it seems that having both -> Nil *and* "is raw" feel weiird
.oO( what was I thinking )
timotimo :) 11:32
Xliff lizmat: I'm looking at github.com/lizmat/Method-Also/issues/1. If you have some time, could you comment on it? Thanks.
lizmat commented, but not with anything useful 11:33
Xliff heh 11:34
lizmat: Weekend is good! Thanks!
11:35 p9s``` joined
Kaiepi i thought there weren't write-* methods as well, did they get added after the read-* ones or st? 11:35
timotimo no, at the same time
i think? 11:36
not more than a week or two apart in any case
Kaiepi " Introducing buf8/blob8.read-(u)bits/write-(u)bits"
ig not 11:37
11:43 domidumont joined
Xliff m: my $b; say ?$b 11:44
camelia False
Xliff m: my $b; say ?$b; $b = 2; say ?$b
camelia False
True
AlexDaniel squashable6: status 11:46
squashable6 AlexDaniel, 🍕🍕 SQUASHathon is in progress! The end of the event in 2 days and ≈0 hours. See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
AlexDaniel hold on
squashable6: status 11:47
squashable6 AlexDaniel, ⚠🍕 Next SQUASHathon in 6 days and ≈16 hours (2019-10-12 UTC-12⌁UTC+20). See github.com/rakudo/rakudo/wiki/Mont...Squash-Day
AlexDaniel ok that's better
still looking for ideas for this squashathon!
and also for someone to take care of it :)
11:51 isBEKaml left 11:59 abraxxa left 12:01 pat_js left 12:04 pat_js joined 12:17 Cabanossi left 12:18 Cabanossi joined 12:22 gabiruh_ joined 12:25 gabiruh left, Xliff left 12:26 scimon joined 12:31 lucasb joined
lizmat timotimo Kaiepi looks like massive test breakage if write-int* methods return self 12:37
timotimo huh, that's interesting, i wouldn't expect that 12:38
lizmat so either we let only the :U versions return the Buf/Blob
or we change the spectest
12:39 Guest93 left, Guest93 joined, Guest93 left 12:41 softmoth left
Kaiepi which tests break lizmat? 12:45
timotimo i hope no tests will ever break lizmat
12:45 softmoth joined
lizmat Kaiepi: all the tests that check for Nil :-) 12:49
12:50 Guest93 joined 12:53 camelCaser left 12:55 camelCaser joined 12:56 pat_js left, pat_js joined
timotimo whyyy :) 12:56
12:59 camelCaser left 13:04 Sgeo_ joined 13:07 Sgeo__ left 13:08 gabiruh_ left, gabiruh joined 13:09 ravenousmoose left 13:10 ravenousmoose joined 13:17 ufobat__ left 13:21 ravenousmoose left 13:28 scimon_ joined 13:31 scimon left 13:32 ravenousmoose joined 13:34 AlexDaniel left, AlexDaniel joined, AlexDaniel left, AlexDaniel joined 13:35 Kaiepi left
lizmat weekly: news.perlfoundation.org/post/grants...2019_votes 13:37
notable6 lizmat, Noted! (weekly)
lizmat timotimo: because I wrote thorough tests :-)
timotimo i wonder if there's something to be said for having a bit less rigid tests every now and then 13:38
13:43 camelCaser joined 13:44 Sgeo joined 13:47 Sgeo_ left
lizmat timotimo: I don't agree: it just means more potential fallout when code is depending on untested behaviour, should we change that behaviour 13:48
timotimo mhh 13:49
lizmat s/code/code out there/
13:52 ravenous_ joined 13:53 ravenousmoose left 14:01 stux|RC left 14:02 agentzh left 14:04 agentzh joined, wamba joined 14:05 ravenous_ left 14:08 ravenousmoose joined 14:11 p9s``` left 14:15 stux|RC joined 14:44 ravenous_ joined, ravenousmoose left 14:48 p9s``` joined 14:53 p9s``` left 14:56 ravenous_ left 14:57 ravenousmoose joined 14:58 softmoth left 15:01 softmoth joined 15:04 silug joined, domidumont left
lizmat timotimo: otoh, write-int* is post 6.d, so I guess it's ok to change the tests 15:21
15:23 p9s``` joined 15:29 p9s``` left 15:35 Kaiepi joined 15:41 Doc_Holliwood left 15:44 Doc_Holliwood joined 15:45 reach_satori_ left 15:51 sno left, pat_js left 15:52 sno joined 15:54 aindilis left 15:55 aindilis` joined 15:56 p9s``` joined 15:58 aindilis` left 15:59 aindilis joined
timotimo <3 15:59
15:59 scimon_ left 16:00 ravenousmoose left
timotimo there's this thing going around on twitter with "first language, had difficulties, most used, totally hate, most loved, for beginners", each one with a number, 1 through 6 16:01
since "most loved" comes before "6. for beginners", i get all the tweets with "most loved: perl" in my "perl 6" search column
16:02 p9s``` left, ravenousmoose joined 16:12 aindilis left 16:15 ravenousmoose left 16:16 ravenousmoose joined 16:23 Xliff joined, MilkmanDan left 16:25 MilkmanDan joined 16:27 p9s``` joined
lizmat timotimo: yeah :-) 16:28
16:33 p9s``` left 16:44 jmerelo joined 17:01 Sgeo_ joined 17:02 p9s``` joined 17:04 Sgeo left, robertle joined 17:09 p9s``` left
uzl[m] AlexDaniel: I think providing a P6 version for opendatastructures.org/ would be a great idea for 17:19
a squashathon (or possible multiple of them). It could help putting P6 out there, especially given
that this textbook might be used in college courses.
17:21 p9s``` joined 17:26 p9s``` left 17:27 ravenousmoose left, ravenousmoose joined 17:31 wamba left
AlexDaniel uzl[m]: interesting! 17:32
uzl[m]: someone needs to organize that though
jmerelo uzl[m]: Um, not another book, please. We need a lot of help in the documentation. Let's try and get a good documentation first. 17:37
uzl[m] jmerelo: Totally. All efforts should go to improve the documentation first. It's less of "do this now" and more of a "possible squashaton in the no-so distant future as long as it doesn't conflict with more important issues". ;-) 17:41
So was there a book-related squashaton before?
AlexDaniel no, but there were grants 17:45
jmerelo uzl[m]: Some time ago. I would really wish some to happen again. 17:48
uzl[m]: ah, OK, not book-related. More documentation related.
17:49 domidumont joined 17:56 aborazmeh joined, aborazmeh left, aborazmeh joined
jmerelo BTW, does anyone know how Andrew Shitov's book is doing? 18:00
Sorry, course.
18:00 aindilis joined 18:01 p9s``` joined 18:07 p9s``` left
AlexDaniel jmerelo: no clue, but I guess we'll hear about it in grant progress reports 18:10
jmerelo I just saw finanalyst's request was rejected. 18:20
AlexDaniel: I don't think the PR in the documentation is the way to go. I will open an issue to discuss that, although my plans have been exposed in several places already. 18:33
AlexDaniel jmerelo: it's a fine PR, I think it makes sense to do that first and then do more dramatic changes later 18:39
18:41 p9s``` joined
jmerelo AlexDaniel: it's not in the big picture, it's not in the details. "Rakudo Raku" does not make sense, for instance. Using raku fencing in markdown (like ```raku) does not either. Also just changing stuff in master will not cut it. 18:46
AlexDaniel jmerelo: yeah, just Rakudo is fine 18:47
jmerelo AlexDaniel: preparing an issue right now to discuss that. Anyway, the general idea can be used... But once we get the details hammered out in an issue.
AlexDaniel Rakudo Raku, hahahaha
18:48 p9s``` left
jmerelo AlexDaniel: OK, here goes the issue github.com/perl6/doc/issues/3046 18:51
18:53 aborazmeh left 18:54 wokurka joined 18:56 domidumont left
tbrowder raku raku sounds like a phrase out of a limbo song 19:00
jmerelo: progress bogged down on Documentable, but at least my Build.pm6 is doing what i want it to do. 19:01
19:02 aborazmeh joined, aborazmeh left, aborazmeh joined, aborazmeh left
jmerelo tbrowder: we're working down there in the basement, trying to iron out problems with Pod::to::Cached and what might be eventually called Document::Cache 19:03
tbrowder: I don't know if you have seen this github.com/perl6/Pod-To-Cached/issues/11
tbrowder: also, Antonio is minus his computer now. And I have started classes, so I get bogged now preparing them, tutoring students... 19:04
19:04 Black_Ribbon joined
lizmat twitter.com/melezhik2/status/11801...8736169987 #rakulang 19:09
AlexDaniel it's not early btw 19:10
we haven't renamed to raku but we have raku alias
I mean, yet
lizmat from the PR: "When using social media that use hash-tags, users are suggested to use the `#rakulang` hash-tag" 19:11
lizmat is not on Twitter
19:13 aborazmeh joined, aborazmeh left, aborazmeh joined 19:16 p9s``` joined 19:19 sena_kun left 19:22 aborazmeh left, p9s``` left 19:26 aborazmeh joined, aborazmeh left, aborazmeh joined, aborazmeh left 19:29 aborazmeh joined, aborazmeh left, aborazmeh joined 19:30 aborazmeh left 19:33 aborazmeh joined, aborazmeh left, aborazmeh joined 19:36 cpan-p6 left, jmerelo left, cpan-p6 joined, cpan-p6 left, cpan-p6 joined 19:43 Sgeo_ left 19:47 Actualeyes left
guifa Okay, here comes another “What’s the best practice for something no one has really done yet” question from guifa :-) 19:48
I’m going to make an Intl::Date class, subclassing regular Date
19:49 p9s``` joined
guifa Idea being I’d like it to be compatible with anything else that takes a Date. 19:49
19:50 aborazmeh left
guifa If someone passes a Intl::Date::Hebrew, for instance, though, should .day .month .year, etc, pass the Gregorian or the Hebrew number? 19:51
19:51 cinch__ is now known as cinch
lizmat guifa: I think you've answered your own question with: "I’d like it to be compatible with anything else that takes a Date" 19:52
guifa lizmat: I was secretly hoping there was a “If in a Date container, use X method, if in an Intl::Date container, use Y method” :-)
lizmat well, if you're subclassing Date, you *can* supply your own .day, .month, .year etc methods 19:53
guifa meant more like, my Date $d = Intl::Date.new(…); $d.day = 23; BUT my Intl::Date $d = Intl::Date.new(…); $d.day = 4 19:54
But I’m guessing variablse have no way of knowing anything about their container (plus I can see this causing no end of confusion) 19:55
lizmat then you're misunderstanding what "my Foo $a" means
19:55 p9s``` left
lizmat that just means that a type check will be done when assigning 19:55
and if Intl::Date is a subclass of Date, then my Date $d = Intl::Date.new(...) will just work without any additional semantics 19:56
guifa So suppose the next question is … is compatibility with Date something folks would see as important? (Beyond simple coercion, which is a given) 19:58
MasterDuke well, i guess you could try using greppable6 and seeing if a lot of current modules use Dates. if so, transparent compatibility with them might be desired 19:59
lizmat guifa: also: modules.perl6.org/dist/Date::Calen...an:JFORGET 20:01
and: modules.perl6.org/dist/Date::Calen...an:JFORGET
perhaps coordination with Jean Forget would be helpful :-) 20:02
El_Che One of the nicest thing about the (mini) web frameworks in Go is that their interfaces are compatible with the http interface in core. Everything you do is familiar and the structs/datatypes are compatible, so you can pass request, responses around and so on. I think that people would appreciate a similar mindset in p6 as well
lizmat El_Che: have you looked at Cro? 20:03
El_Che lizmat: I don't mean the web frameworks as such (Cro is nice), it's about the Date transparency
20:03 mensvaga left
El_Che an "add-on" instead of different thing altogther to be learnt 20:04
lizmat: agreeing with you, actually
guifa lizmat: I saw those. The Hebrew one has been helpful for doing some of the coding. I’m just really want the Intl:: stuff to work as intuitively as possible so people actually use them. But what’s intuitive for me may not be for others so I wanted to check (in particular in this case, where I see very valid reasons to subclass Date, or to keep Intl::Date its own thing). I might toss the question on the FB page too and see what people 20:11
there before deciding
lizmat sounds like a plan :-) 20:12
20:18 agentzh left 20:20 ravenousmoose left, agentzh joined 20:21 p9s``` joined 20:24 leah2 left 20:27 p9s``` left 20:37 wamba joined
Kaiepi none of the equality operators that exist currently allow you to type check stuff more strictly than ~~ by using the generic checks HOWs use while ignoring the ACCEPTS method, right? 20:38
if it doesn't exist i wanna make a library for it, but i'd rather not add another equality operator unless there isn't a way to do it already 20:39
20:39 softmoth left
lizmat Kaiepi: not sure what you're getting at ? 20:39
20:39 leah2 joined 20:40 softmoth joined
Kaiepi m: 1.^type_check: 1 | 2 | 3 20:44
camelia ( no output )
Kaiepi m: say 1.^type_check: 1 | 2 | 3
camelia 0
Kaiepi m: say 1.ACCEPTS: 1 | 2 | 3
camelia any(True, False, False)
Kaiepi basically i want something that type checks using the literal type
but works with parents and roles
Any.ACCEPTS was the closest i could find to what i want in rakudo but that's not an operator 20:46
lizmat so "say 1.ACCEPTS: 1 | 2 | 3" should be any(True,True,True) ?
Kaiepi it shouldn't say 1 and 1 | 2 | 3 are equal because Int is not Junction 20:47
lizmat I'm not following, sorry 20:50
Kaiepi Int and Junction are unrelated at the type level, so in this case they shouldn't be considered equal 20:54
but Int and Int should, and Int and UInt should since UInt is a subset of Int 20:55
lizmat but they aren't considered equal? 20:57
Int and Junction?
1.ACCEPTS: 1 | 2 | 3 returns a Junction
20:58 p9s``` joined
Kaiepi ik, but Int does not inherit or do Junction 20:59
lizmat indeed... 21:00
Junctions aren't even Any
which is what makes them work, btw
21:04 p9s``` left
Kaiepi lemme think of a better example, junction isn't a very good type to use to show what i mean 21:09
21:26 krychu joined
AlexDaniel yes, it's a raku bus stop :P github.com/AlexDaniel 21:29