MasterDuke jnthn, timotimo, lizmat: any thoughts on github.com/rakudo/rakudo/pull/1317 ? 00:15
jnthn MasterDuke: I suspect more effective would be a set of candidates like (Slip:D), (Any) for one-item, and (Slip:D, Slip:D), (Slip:D, Any), (Any, Slip:D) and (Any, Any) for the two argument case 00:25
Those would deal with a lot more cases
And specialize decently into the Int/Int and Str/Str ones 00:26
m: say Slip ~~ Cool 00:28
camelia True 00:29
jnthn lizmat: ^^
(wrt the comment you just put on the ticket)
lizmat hehe... there's no escaping, I'm being watched :-)
would have been nice if the commit message had mentioned the reason :-)
lizmat is back from travelling for 3 days and goes to get some rest in a good bed for a change 00:30
timotimo have a good one!
jnthn Rest well 00:31
lizmat thanks all!
jnthn As I noted above, though, adding candidates for handling Slip would help
timotimo mostly for handling anything but slips? 00:37
jnthn timotimo: If there's a candidate for Slip, it's tighter than the one for Any, so it means the Any ones know there's no Slip 00:40
timotimo yeah 00:41
jnthn 'night o/ 00:45
AlexDaniel releasable6: next 02:05
releasable6 AlexDaniel, Next release in 6 days and ≈16 hours. Blockers: github.com/rakudo/rakudo/issues?q=...%9A%A0%22. Changelog for this release was not started yet
AlexDaniel, Details: gist.github.com/0bb2234029a2a03e69...3a5a9e1f14
jstuder timotimo: I updated the statevar init PR (github.com/rakudo/rakudo/pull/1542) with the spectest info we spoke about earlier. It appears that the test did not actually fail but I read the results wrong. 02:55
samcv nice! i got open() working with a replacement character when an unencodable character is encountered for windows-1252/1 07:13
pmurias geekosaur: It's still does a dynamic lookup, just that we have extra fallback to a global semantics when it fails 08:50
nine m: multi foo() { say "none" }; multi foo(|args) { say "many" }; multi foo(*@args, *%args) { say "named" }; foo(a => 1) 10:35
camelia Unexpected named argument 'a' passed
in sub foo at <tmp> line 1
in block <unit> at <tmp> line 1
nine jnthn: what's going on here ^^^? 10:36
MasterDuke .tell pmurias i added that @*comp_line_directives stuff when i was adding support for `#line 2 SETTING::src/core/core_prologue.pm6` (so error message gave the real file instead of gen/moar/CORE.setting). but i barely knew what i was doing, so feel free to do something better 11:33
yoleaux MasterDuke: I'll pass your message to pmurias.
AlexDaniel Zoffix: fwiw just as a thought, maybe eventually we should start recommending to install rakudo from the official distro repos. Ubuntu Artful (17.10) and Bionic (upcoming in a month) have 2017.06, but their LTS version has 2015.11 (oh no!), Debian stable has 2016.12 which is not bad at all, Debian testing has 2017.06. Fedora should have at least 2017.06 starting from f26. But all versions of Linux Mint have 2015.11 I think… 12:16
so I guess eventually = not now, but perhaps we should keep in mind that this will be possible soon
or, I don't know, am I the only one who likes when the download page says “just do `apt install thiscoolthing`”? 12:17
and, um… I guess there's no good way to detect a distro using JS 12:18
nine AlexDaniel: openSUSE Tumbleweed currently ships 2018.01 and so does the upcoming openSUSE Leap 15.0 12:25
Geth nqp: a6b9e845f4 | usev6++ | src/vm/jvm/runtime/org/perl6/nqp/io/FileHandle.java
[JVM] Support opening files with :read and :append

Makes some more test in S32-io/open.t pass. Compare also RT #131145.
Thanks to cygx++ for giving a pointer on how this could be done.		 12:32
synopsebot RT#131145 [open]: rt.perl.org/Ticket/Display.html?id=131145 [JVM] Modes for IO::Handle.open need clarification
Geth features: JJ++ created pull request #7:
Eliminates niecza refs closes #6 when accepted		 13:00
MasterDuke jnthn and anyone else: i would have thought this patch would work, but there are a lot of test failures. gist.github.com/MasterDuke17/86c7c...29cec3d9e8 13:18
timotimo MasterDuke: you're not slipping the slips at all there 14:31
m: say (|(1, 2, 3), |(9, 9, 9)).perl
camelia (1, 2, 3, 9, 9, 9)
timotimo oh, d'oh
you are binding a reifier there
i see a problem with the second candidate, though 14:32
you're binding b to the second slot, but what if the slip in a has more than one element? 14:33
MasterDuke: could that be the problem? 14:34
MasterDuke i didn't correctly follow what was happening (in the original code) when there was a Slip, so yeah, i definitely need to change something 14:36
but i don't think the number of elements in the slip matter, right?
m: use nqp; sub a(Any \b, Slip:D \c) { say nqp::elems(nqp::p6argvmarray()) }; my @d = <a b c>; a(4, @d.Slip) 14:37
camelia 2
MasterDuke m: use nqp; sub a(Any \b, Slip:D \c) { say nqp::elems(nqp::p6argvmarray()) }; my @d = <a b c>; a(4, |@d)
camelia Too many positionals passed; expected 2 arguments but got 4
in sub a at <tmp> line 1
in block <unit> at <tmp> line 1
timotimo well, you're binding into slot 2 of the reification target 14:39
when the slip gets slipped, its contents will spill into the reification target, too
MasterDuke hm, i guess the `Slip:D \a, Any \b` candidate shouldn't actually bind b at all 14:42
timotimo i agree 14:43
Geth nqp/master: 7 commits pushed by pmurias++ 15:37
AlexDaniel hmm, how can I make synopsebot bot #whateverable? 17:34
join*
MasterDuke synopsebot: help 17:35
synopsebot MasterDuke, See: github.com/perl6/synopsebot
MasterDuke just add #whateverable here? github.com/perl6/synopsebot/blob/m...bot.p6#L15 17:36
Geth synopsebot: d02bfb7932 | (Aleks-Daniel Jakimenko-Aleksejev)++ | synopsebot.p6
Add #whateverable to the list
 17:37
timotimo i'm not sure synopsebot has any "check out the source regularly" mechanisms in place; didn't when i last touched it, which may have been sufficiently long ago for this message to not matter at all 17:38
AlexDaniel `ssh hack.p6c.org` – hmm, can I have access to that? 17:39
Geth features: 5ad23566e3 | (JJ Merelo)++ | 3 files
Eliminates niecza refs closes #6 when accepted
 17:48
features: ee7fbb17da | (Moritz Lenz)++ (committed using GitHub Web editor) | 3 files
Merge pull request #7 from JJ/master

Eliminates niecza refs closes #6
roast: ronaldxs++ created pull request #399:
Fix attribute mutator tests and add comment discouraging use		 17:50
mr_ron From what I can figure out a fix for the attribute mutator roast test could have been done a long time ago with subs but there was a hangup over using methods. Why is it not OK to solve with subs until we understand exactly how methods fit this case? 18:25
AlexDaniel timotimo: maybe you can just follow the steps here for me? github.com/perl6/synopsebot#pullin...ng-the-bot 19:09
or anybody else who has enough privs
Zoffix Yeah, I can 19:10
AlexDaniel thanks!
Zoffix AlexDaniel: actually, I'll create an account and you can do it yourself. What username/key? 19:11
AlexDaniel Zoffix: these two keys: github.com/AlexDaniel.keys 19:12
Zoffix: username “alex” if available
alexdaniel otherwise
Zoffix AlexDaniel: I added you to sudoers and msg'ed you the pass. I didn't setup the key. I think you'd be able to set it up on your own. 19:14
Are you able to log in and pull-restart synopsebot? 19:15
AlexDaniel: note that perms on your home dir are drwxr-xr-x. That's the default adduser makes but I changed mine to be 0 for the world 19:16
AlexDaniel Zoffix: OK, that works! Thank you very much! 19:17
Zoffix ok 19:18
Geth geth: be7c31d75e | (Zoffix Znet)++ (committed using GitHub Web editor) | README.md
Add info on webhook/code locations
 19:28
MasterDuke m: use nqp; sub a(\b) { say nqp::elems(b) }; sub c(|) { a(nqp::list(nqp::p6argvmarray())) }; c(1,2) 20:02
camelia This type (List) does not support elems
in sub a at <tmp> line 1
in sub c at <tmp> line 1
in block <unit> at <tmp> line 1
MasterDuke i am thinking very slowly today. how do it make ^^^ work? changing \b to $b or @b doesn't help 20:03
Geth nqp: 46fbb5e9d8 | usev6++ | src/vm/jvm/runtime/org/perl6/nqp/runtime/Ops.java
[JVM] Let seekfh return null, not file handle

This avoids errors in S32-io/seek.t caused by missing method 'sink' for BOOTIO when nqp::seekfh is the last command in a block. This mirrors the behaviour of MoarVM.		 20:09
Zoffix MasterDuke: looks like it gets hllized when it goes from one sub to another. 20:13
m: use nqp; sub a(\b) { say nqp::elems(nqp::getattr(b, List, '$!reified')) }; sub c(|) { a nqp::p6argvmarray }; c 1,2
camelia 2
Zoffix Don't know how solid that is, to poke into List guts that way (and assuming it's reified)
timotimo you can get around hllization with some trickery; not sure what exactly, perhaps involving findmethod
Zoffix m: use nqp; sub a(\b) { say nqp::elems(b) }; sub c(|) { nqp::callstatic(&a, nqp::p6argvmarray) }; c 1,2 20:14
camelia This type (List) does not support elems
in sub a at <tmp> line 1
in sub c at <tmp> line 1
in block <unit> at <tmp> line 1
timotimo oh, right, not a method 20:19
i wonder if what i'm thinking of is even still relevant or if it's already years old
MasterDuke Zoffix++, timotimo++ 20:28
timotimo what if you type the argument Mu?
MasterDuke same
Geth rakudo: MasterDuke17++ created pull request #1608:
Make 2 elem infix:<,> special cases more generic		 20:50
AlexDaniel releasable6: status 22:05
releasable6 AlexDaniel, Next release in 5 days and ≈20 hours. Blockers: github.com/rakudo/rakudo/issues?q=...%9A%A0%22. Changelog for this release was not started yet
AlexDaniel, Details: gist.github.com/613eeadedea306d077...0b03b7dacd
Kaiepi where would be the appropriate place to put tests for Perl6::Metamodel::Naming? 22:35
yoleaux 10 Mar 2018 06:42Z <Zoffix> Kaiepi: not 100% sure, but this seems to do the trick, without regexes: gist.github.com/zoffixznet/da3b0ce...f8483b2f05
Kaiepi would writing some for rakudo's test suite be needed, or just ones for roast? 22:39
Zoffix Kaiepi: just for roast. Tests in rakudo's suite are just for stuff we don't want to be part of the language spec. 23:01
Maybe in S14-roles/mixin.t, unless you already found a good place. Seems there's already a test for a similar thing: S14-roles/mixin.t:229: "mixing in a role from a deeper namespace doesn't clobber the targets shortname";

Powered by Cro and the Raku® Programming Language.

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

Thank you!