05:14 sena_kun joined 05:22 Kaiepi left 05:29 Kaiepi joined 05:30 Kaiepi left 05:31 Kaiepi joined 05:37 jmerelo joined 05:42 Altai-man joined 05:44 sena_kun left 06:13 squashable6 left 06:15 squashable6 joined 06:48 patrickb joined 07:02 leont joined 07:03 domidumont joined 07:38 MasterDuke joined 08:46 MasterDuke left 09:08 MasterDuke joined
lizmat Files=1336, Tests=113630, 219 wallclock secs (29.46 usr 8.60 sys + 3053.82 cusr 291.63 csys = 3383.51 CPU) 09:19
which feels in the same ballpark as before: colabti.org/irclogger//irclogger_l...xt=checked 09:20
^^^ was on the nqpif2ternary branch 09:41
09:43 sena_kun joined 09:45 Altai-man left 09:46 lizmat_ joined 09:48 lizmat left, lizmat_ is now known as lizmat 10:18 domidumont left
Geth_ rakudo: 94d8968669 | (Elizabeth Mattijsen)++ | src/core.e/hash_slice.pm6
Allow for current Hash.DELETE-KEY behaviour

In v6.e I'm assuming for now that %h{a;b;c} will return Nil when trying to delete a non-existing key. Currently that returns the default for the hash. See also github.com/rakudo/rakudo/pull/3954
roast: d93c105a6e | (Elizabeth Mattijsen)++ | S32-hash/multislice-6e.t
Referring to a non-existing key returns a container
11:49 Kaiepi left, Kaiepi joined 12:12 AlexDaniel left 12:40 jmerelo left 12:42 jmerelo joined 12:49 domidumont joined 13:06 jmerelo left 13:42 Altai-man joined 13:44 sena_kun left 14:27 lucasb joined 14:55 patrickb left 14:57 Kaiepi left 14:58 Kaiepi joined 15:05 finsternis left 15:30 Kaeipi joined 15:31 Kaiepi left, domidumont left 16:47 Kaeipi left, Kaiepi joined 16:50 Kaiepi left, Kaiepi joined 17:22 Altai-man left, sena_kun joined 18:13 Ven`` joined
Geth_ roast: 6852df0781 | (Elizabeth Mattijsen)++ | S32-hash/multislice-6e.t
Add %h{a;b;c} tests with any of them being * or a list
roast: bca1a03162 | (Elizabeth Mattijsen)++ | S32-hash/multislice-6e.t
Use <a b c> notation instead of ("a","b","c")

For readability, really.
lizmat m: () = 42 # I wonder if this should be a runtime error 18:54
camelia ( no output )
lizmat technical explanation: because the left-had side of the assignment is an empty list, it will not try to assign anything from the right 18:55
m: dd (my $a, my $b) = 42,666,314; dd $a, $b # just as the 314 is silently dropped on the floor here 18:56
camelia (42, 666)
Int $a = 42
Int $b = 666
lizmat or maybe it should warn 18:57
19:00 patrickb joined 19:05 Ven`` left 19:06 Ven`` joined 19:08 Ven`` left
Geth_ rakudo: b2gills++ created pull request #3960:
[RAK(:!foo).IMPL-INTERPRET should have a False value
19:15 Ven`` joined, vrurg joined 19:17 vrurg_ left 19:18 vrurg left 19:20 vrurg joined 19:23 patrickb left 19:30 patrickb joined
[Coke] "Useless assignment of discarded value" 19:49
lizmat thing is, a check for that will slow down things such as ($a,$b) = 42,666 significantly :-( 20:01
it would involve an extra pull-one on the iterator, and a check for IterationEnd 20:02
I bet that's the reason it's not being tested
Geth_ roast: a715c44706 | (Elizabeth Mattijsen)++ | S32-hash/multislice-6e.t
Add tests for success of %h{a;b;c}:delete

Also drop the idea that the number of tests is easily algorithmically determinable. Just set the number of tests expected.
20:29 Ven`` left 20:30 sena_kun left
patrickb How would I go about re-attaching STDOUT/STDERR/STDIN of ones own process to a file or a sink? 20:46
I'm asking because I have implemented a console less raku.exe on Windows. Those don't have their STD* attached. Writing to them causes a program abort. (It's the same with python.) It would be nice to have a way to prevent the abort and instead just silently do nothing or alternatively write to a file or so. 20:49
[Coke] patrickb: if you want to replace STDOUT with a bucket that just throws out the output, that's doable. 20:50
(is that what you're asking?)
patrickb [Coke]: Yes. 20:51
[Coke] github.com/Raku/roast/blob/master/...list.t#L23 20:52
that saves all the output. You could have it just be an empty block.
but this only touches the Raku-level items. If something hits the underlying stdin/stdout at the OS level, that's not capturable, I don't think
(there's some stuff in Moar/NQP that I think uses the underlying OS level handles) 20:53
er, empty *class*, even.
m: temp $*OUT=class {...} ; say 3 20:54
camelia The following packages were stubbed but not defined:
[Coke] m: temp $*OUT=class {print(*@args){}} ; say 3
camelia 5===SORRY!5=== Error while compiling <tmp>
Unable to parse expression in argument list; couldn't find final ')' (corresponding starter was at line 1)
at <tmp>:1
------> 3temp $*OUT=class {print(*7⏏5@args){}} ; say 3
expecting any…
[Coke] m: temp $*OUT=class {method print(*@args){}} ; say 3
camelia ( no output )
[Coke] ^^ there. 20:55
m: temp $*ERR = temp $*OUT=class {method print(*@args){}} ; say 3; note 'wtf'
camelia ( no output )
patrickb m: $*OUT=class {method print(*@args){}} ; say 3 20:56
camelia ( no output )
patrickb [Coke]: That works. Thanks! :-) 20:57
[Coke] (ah, right, don't need the temp if you never want it back. :) 21:02
Geth_ roast: 761c1c39ec | (Elizabeth Mattijsen)++ | S32-hash/multislice-6e.t
Add assignability tests for %h{a;b;c}
21:11 Kaeipi joined 21:12 Kaiepi left
patrickb [Coke]: Could you have a quick look whether this makes sense? : github.com/rakudo/rakudo/pull/3955...10bR85-R88 21:14
21:16 scovit joined
scovit patrickb : this is the m$ docs to obtain what I think you want to achieve docs.microsoft.com/en-us/cpp/c-run...quirements 21:18
stdin is 0, stdout is 1 and stderr is 2. In the past there was a number also for the printer 21:20
patrickb scovit: Thanks for the info! 21:21
I think using `dup` is currently not possible from within Rakudo. (Is it with nativecall?) 21:22
Geth_ roast: 311b65736c | (Elizabeth Mattijsen)++ | S32-hash/multislice-6e.t
Add tests for non-assignability of %h{a;b;c}:adverbs
patrickb scovit: What do you think about the bit of documentation I added? Do you have other ideas to recommend? 21:24
scovit patrickb if the user wants he can use dup with nativecall. Or the trick you document, that would not work with certain outputs. The alternative is to standardize the behavior in moarvm 21:25
I’m a bit concerned about what would happen if the vm aborted with an error 21:28
patrickb scovit: What would be a good default? Silent swallowing? (The python folks have been discussing this issue a few years ago and I think decided to not do anything: bugs.python.org/issue706263) 21:29
lizmat and that concludes my hacking for today&
patrickb lizmat: Sleep well! 21:32
scovit patrickb open NUL and assign it to 0 1 and 2 in moarvm, maybe ? Die with MessageBoxes :p 21:34
21:35 scovit left
patrickb scovit: Thanks for your input! I think I'll keep the current solution for now. Then we can discuss how to best do it in the future. I don't think there is much of a backwards compatibility issue here. 21:39
tellable6 patrickb, I'll pass your message to scovit
21:43 Merfont joined, Kaeipi left
patrickb rebooting... 21:49
21:49 patrickb left 21:59 patrickb joined 22:35 Merfont left 22:36 Merfont joined
Geth_ rakudo: e878e23a73 | (Patrick Böker)++ | docs/running.pod
Correct a wrong env var RAKU_HOME -> RAKUDO_HOME in docs
rakudo: 0ea6902dda | (Patrick Böker)++ | 6 files
Generate a `rakuw.exe` on Windows

On Windows executables are either marked as "console" applications or
  "windows" applications. The first always spawn a console window to which
STDIN, STDOUT and STDERR are connected, the second spawns no such window and STDIN, STDOUT and STDERR are not connected to anything. This can't be decided at runtime, but is a flag set on the executable file itself. So the ... (8 more lines)
rakudo: 71a749b2e7 | (Patrick Böker)++ (committed using GitHub Web editor) | 6 files
Merge pull request #3955 from patrickbkr/rakuw-executable

Generate a `rakuw.exe` on Windows
22:54 patrickb left 22:58 ggoebel joined 23:32 Merfont left 23:33 Kaiepi joined
rba patrickb: I have not been able to update rakubrew tonight. Have not been able to access my keys I need to access the docker remotly. I will be back at the weekend and I made a note to myself to not forget the update. I'm very sorry! 😐 Update follows... 23:55
tellable6 rba, I'll pass your message to patrickb