00:19 tbrowder left 00:24 tbrowder joined 00:48 tbrowder left 00:54 tbrowder joined 00:56 AlexDaniel joined, AlexDaniel left, AlexDaniel joined
AlexDaniel yup, they didn't like it :) 01:37
Geth ¦ problem-solving: vrurg assigned to jnthn Issue Implementation of importing of classes with composite names perhaps needs better consideration github.com/perl6/problem-solving/issues/128 01:44
01:48 tbrowder left 01:54 tbrowder joined 02:48 tbrowder left 02:53 tbrowder joined 03:48 squashable6 joined 05:48 tbrowder left 05:49 tbrowder joined 06:04 jmerelo joined 06:07 titsuki left
Kaiepi i'm not sure how i can handle exposing nqp's type parameterization ops through Metamodel::Primitives methods in a way that doesn't make them dependent on which HOW the type given is using 06:13
Metamodel::ParametricRoleGroupHOW always makes the last parameter either a hash of named parameters or a NO_NAMEDS type, but none of the other HOWs in rakudo do anything like this 06:15
the best i can think of is giving the methods that return any of a type's parameters something like a :named parameter, but then people would need to know whether or not the type can actually take named parameters, which idt people will always know to document 06:16
07:10 go|dfish left 07:23 go|dfish joined 07:38 jmerelo left 08:56 sena_kun joined
Geth star: Tyil++ created pull request #144:
Get Rakudo Star working for 2019.07.1
09:24
09:28 Altai-man_ joined 09:31 sena_kun left
El_Che tyil: your parameters order for rm shocks me: "rm -fr work" :) 10:44
tyil alphabetical order? :p 10:46
11:08 tobs joined 11:29 sena_kun joined 11:31 Altai-man_ left
Geth rakudo: lizmat++ created pull request #3279:
Make Instant.perl a bit less noisy
11:54
rakudo/instant-perl: c0b5fb282c | (Elizabeth Mattijsen)++ | src/core.c/Instant.pm6
Make Instant.perl a bit less noisy

By removing the "Bool::False" when it is not needed, and use normal stringification for the Rat representing the posix value.
12:17
lizmat hmmm.. weird that the PR precedes the commit ? 12:18
sena_kun maybe the bot is just not sync enough? shouldn't be too harmful 12:36
12:39 AlexDaniel left 13:12 lucasb joined
lizmat hmmm... am I understanding the lexicality of $_ wrong? 13:16
m: $_ = 42; if True { say $_ }
camelia 42
jnthn I don't think so
Well, or maybe you're understanding it wrong if you find that surprising ;)
lizmat I guess I am 13:17
jnthn `if` doesn't topicalize
Unlike `with`
So maybe that's the confusion.
lizmat but we agree on that $_ is lexical, right ?
jnthn Yes, no disagreement there :) I'm not sure how what you wrote would disprove that :)
lizmat I guess I'm confused because I know that each scope has its own slot for $_ ? 13:18
jnthn I mean, the $_ assignment you make is lexically outside of the place you say $_, so if anything that illustrates its lexicality
lizmat *apper to know
jnthn It does, but it defaults to OUTER::<$_>
Run perl6 --target=ast -e '$_ = 42; if True { say $_ }' and look for getlexouter 13:19
lizmat ok, thanks for the clarification 13:20
trying to write up the difference in handling of $_ between Perl and Raku 13:21
m: sub a { say $_ }; $_ = 42; a # 42 in Perl 13:22
camelia (Any)
jnthn Right, 'cus a routine gets a fresh $_ with no defaulting to the outer one; that's a property of blocks rather than routines. 13:24
lizmat hmmm.. but if that's the property of blocks, why does this say 42? 13:26
m: $_ = 42; { say $_ }
camelia 42
jnthn 'cus the {...} there is a block? :) 13:27
13:29 Altai-man_ joined
lizmat "that's a property of blocks rather than routines" ?? 13:30
sorry to be persistent, but I want to truly grok this for fear of spreading misinformation otherwise :-) 13:31
13:32 sena_kun left
lucasb iiuc, the property in question is: $_ defaulting to OUTER::<$_> in blocks 13:38
m: my &f = { .say }; $_ = 42; for ^3 { .say; f } 13:49
camelia 0
42
1
42
2
42
lucasb is this right? &f sees 42, not the values from the range 13:50
m: my &f = { .say }; $_ = 42; { .say; f } # for comparison :) 13:55
camelia 42
42
lizmat lucasb: yeah, I think that's correct
jnthn Yes, that illustrates the lexicality nicely, even. 13:56
And yes, the property I was referring to is defaulting to OUTER::<$_>
lizmat m: my &f = -> $_? { .say }; $_ = 42; for ^3 { .say; f } 13:57
camelia 0
(Mu)
1
(Mu)
2
(Mu)
lizmat shouldn't that be (Any) ? 13:58
jnthn: still, there is something funky going on: 13:59
m: my &f = -> ;; $_? is raw { .say }; $_ = 42; f
camelia (Mu)
lizmat m: my &f = { .say }; $_ = 42; f 14:00
camelia 42
lizmat introspection shows that signature of both cases is identical
m: my &f = { .say }; dd &f.signature
camelia :(;; $_? is raw)
lizmat m: my &f = -> ;; $_? is raw { .say }; dd &f.signature
camelia :(;; $_? is raw)
lizmat so there's something else that makes the tow cases different
*two 14:01
jnthn I think somehting is off with signature dumping 14:02
There's a special flag for "obtain the value from the outer"
And probably .perl on the signature doesn't account for it
lizmat ok, then I will look at that 14:03
jnthn: do we actually have syntax for that flag ? 14:04
jnthn = OUTER::<$the-variable-name> 14:07
But I think actually we only ever code-gen it for the $_ case
(And it's there so we can easily spit out better code) 14:08
lizmat ok, so the signature in that case should be:
-> ;; $_ = OUTER:<$_> is raw
am about to go on the road for Amsterdam PM meeting, will look at this later today / tomorrow 14:09
El_Che jnthn: new comma, new cro, new rakudo soon: wow 14:11
jnthn lizmat: Looks right to me 14:21
El_Che: Yes, nice to have things happening :) 14:26
14:30 AlexDaniel joined, AlexDaniel left, AlexDaniel joined
AlexDaniel tyil: can you submit a CLA? 14:31
tyil: so that in the future you can commit to star directly 14:32
tyil CLA?
[Coke] contributor license agreement. 14:33
AlexDaniel notable6: new-topic CLA
notable6 AlexDaniel, Topic “CLA” already exists
AlexDaniel notable6: CLA
notable6 AlexDaniel, 1 note: 2019-08-01T07:40:37Z <AlexDaniel>: www.perlfoundation.org/contributor...-agreement print it out, sign it, scan it in (resize your scan to 250KB or less). Email to trademark -at- perlfoundation.org and don't forget to CC will at coleda.com
tyil skimming over this, I retain all my rights to my contributions, but I cannot revoke access to them to TPF? 14:35
lizmat afk& 14:37
14:38 lizmat left
[Coke] tyil: That's part of section 4.1/4.2, yes. 14:38
tyil I'll have to read it in more detail before signing it, but it looks OK to me so far
[Coke] I imagine that helps the foundation not worry about someone removing access to key functionality. 14:39
Let us know if you find anything concerning.
tyil as I don't have a scanner at home, is a picture made by a phone from 8 years ago acceptable?
[Coke] Probably
tyil: gist.github.com/coke/a0cc69f1558d7...a28167b807 14:51
tyil [Coke]: thanks 15:17
15:21 lizmat joined
lizmat jnthn: where does that special flag live for = OUTER:<$_> ? I don't see it in the Parameter class, is it something in the Block? 15:23
ah, found it in bootstrap :-) 15:25
15:29 sena_kun joined 15:31 Altai-man_ left
moritz killall TabNine 15:33
[Coke] O_o; 15:34
moritz sorry, wrong window 15:38
and for the curious: tabnine is a machine learning autocompletion engine that can be plugged into several editors and IDEs
it's pretty great, except that occasionally, it gobbles up all RAM and CPU :( 15:39
lizmat m: $_ = 66; my &f := -> $_ = OUTER::<$_> { .say }; f 15:46
camelia (Mu)
lizmat m: $_ = 66; my &f := { .say }; f 15:47
camelia 66
lizmat looks like the syntax -> $_ = OUTER:<$_> won't cut it, roundtrippable wise jnthn
but I guess .perl on Parameters is mostly FYI anyway, rather than to roundtrip 15:48
jnthn Yeah...though I think we have an issue somewhere noting that default thunks are not marked as thunks and so aren't skipped properly 15:49
|Tux| Rakudo version 2019.07.1-460-g044b33902 - MoarVM version 2019.07.1-313-g836396c74
csv-ip5xs0.694 - 0.710
csv-ip5xs-206.373 - 6.538
csv-parser20.981 - 21.346
csv-test-xs-200.423 - 0.425
test6.881 - 7.082
test-t1.745 - 1.760
test-t --race0.826 - 0.830
test-t-2028.675 - 29.298
test-t-20 --race9.089 - 9.116
15:57
lizmat yeah, Geth is definitely lagging... 16:17
Geth rakudo: lizmat++ created pull request #3280:
Add support for implicit signature of { ... }
16:23
AlexDaniel lizmat: it could also be that github itself is delivering hooks in a weird way 16:26
lizmat yeah, it's probably that
AlexDaniel actually, yeah: www.githubstatus.com/ 16:27
Notifications – “Degraded”
lizmat m: $_ = 666; sub a($a = OUTER::<$_>) { say $a }; a # sigh 16:28
camelia (Any)
lizmat m: INIT $_ = 666; sub a($a = OUTER::<$_>) { say $a }; a # sigh
camelia (Any)
AlexDaniel lizmat: I don't think that's the OUTER you have in mind
16:28 lizmat left 16:30 lizmat joined
lizmat m: BEGIN $_ = 666; sub a($a = OUTER::<$_>) { say $a }; a # huh ? 16:30
camelia 666
AlexDaniel lizmat: well, question is what's OUTER in this case
lizmat so it looks like it is taking the OUTER::<$_> as a literal value at compile time ?
16:33 jmerelo joined 16:43 lizmat left 17:01 lizmat joined 17:31 sena_kun left
Geth rakudo/outer-perl: e7c20386cf | (Elizabeth Mattijsen)++ | src/core.c/Parameter.pm6
Add support for implicit signature of { ... }

The .perl of parameters did not account for the case where a parameter was taking the $_ of the outer scope. This is now simulated by adding
  " = OUTER<$_>" in that case. Please note however that this is a bit of
a kludge, as this will *not* roundtrip correctly. But it will at least convey the difference with a block that had a specific signature that was otherwise identical.
17:38
17:42 lucasb left 18:18 patrickb joined 18:55 robertle joined 19:16 patrickb left 19:23 patrickb joined 19:33 jmerelo left 20:07 lizmat left 20:08 lizmat joined 20:13 ufobat__ joined 20:17 ufobat_ left 21:01 lizmat left
Geth rakudo: 88db15f73a | (Aleks-Daniel Jakimenko-Aleksejev)++ | .github/FUNDING.yml
Append myself to the list
21:16
21:33 TreyHarris joined 21:51 go|dfish left 22:05 go|dfish joined
Geth rakudo: e7c20386cf | (Elizabeth Mattijsen)++ | src/core.c/Parameter.pm6
Add support for implicit signature of { ... }

The .perl of parameters did not account for the case where a parameter was taking the $_ of the outer scope. This is now simulated by adding
  " = OUTER<$_>" in that case. Please note however that this is a bit of
a kludge, as this will *not* roundtrip correctly. But it will at least convey the difference with a block that had a specific signature that was otherwise identical.
23:34
rakudo: 1d84c64d63 | (Jonathan Worthington)++ (committed using GitHub Web editor) | src/core.c/Parameter.pm6
Merge pull request #3280 from rakudo/outer-perl

Add support for implicit signature of { ... }