00:36 travis-ci joined
travis-ci Rakudo build passed. Vadim Belman 'Merge pull request #3590 from vrurg/problem-solving-170 00:36
travis-ci.org/rakudo/rakudo/builds/670360172 github.com/rakudo/rakudo/compare/a...bf58372fa7
00:36 travis-ci left 01:01 Kaiepi left, Kaiepi joined 01:08 Kaiepi left 01:13 Kaiepi joined 01:21 Altai-man_ joined 01:23 sena_kun left 01:48 upupbb-user2 left
Geth nqp: 1517cf77db | (Vadim Belman)++ | tools/templates/jvm/nqp-j.in
Don't quote `%*` in the command line

It causes java to treat everything `nqp-j.bat` gets on the command line as a single parameter.
02:11 sivoais left 02:23 sivoais joined
Geth ¦ rakudo: vrurg assigned to patrickbkr Issue Rakudo JVM backend build broken on Windows github.com/rakudo/rakudo/issues/3591 02:35
vrurg .seen patrickb 02:37
tellable6 vrurg, I saw patrickb 2020-03-31T22:22:32Z in #raku: <patrickb> rba: Thanks for the previous upload!
vrurg .ask patrickb Would you pls have a look at R#3591? Not all nqp runners are exe files yet, as it turns out. :) 02:38
linkable6 R#3591 [open]: github.com/rakudo/rakudo/issues/3591 [JVM][Windows][build][severe] Rakudo JVM backend build broken on Windows
tellable6 vrurg, I'll pass your message to patrickb
Geth nqp/master: 4 commits pushed by Coke++ 02:52
03:22 sena_kun joined 03:23 Altai-man_ left 04:22 upupbb-user3 joined 05:02 Xliff joined 05:21 Altai-man_ joined 05:23 sena_kun left
Xliff \o 05:41
Off again! :)
05:41 Xliff left 06:15 upupbb-user3 left 07:22 sena_kun joined 07:23 Altai-man_ left
AlexDaniel sourceable6: 42.return 07:39
sourceable6 AlexDaniel, github.com/rakudo/rakudo/blob/2cbf...Mu.pm6#L75
AlexDaniel greppable6: \.return 07:41
greppable6 AlexDaniel, 197 lines, 37 modules: gist.github.com/a27f670061904a8fe4...6030db285c
AlexDaniel TIL .take .return 07:42
jjatria tyil: I've built a couple of more images based on different distros, and put them here for now in case you want to check them out: quay.io/repository/jjatria/rakudo?tab=tags 08:14
I've ironed out a bunch of small issues with the Dockerfiles, which is nice 08:15
tyil oh, security scans on those images
very nice
jjatria I think I should be able to send you the patch tonight, maybe, or during the weekend at some point 08:16
What do we want to do about labels? You want to keep the ones in that other Dockerfile you sent me? 08:17
This is how the patch currently looks, by the way: gitlab.com/jjatria/rstar/-/merge_requests/1 08:18
tyil I'm honestly not too invested in how Docker will be arranged, I used the tags I'm using because they were clear to me 08:24
if you think you have a more clear way to distinguish them, that's fine too
09:21 Altai-man_ joined 09:24 sena_kun left 10:19 Kaiepi left 10:20 Kaiepi joined 10:40 samcv left 10:49 MasterDuke left
lizmat Files=1306, Tests=111228, 213 wallclock secs (29.09 usr 8.04 sys + 2998.59 cusr 264.43 csys = 3300.15 CPU) 10:51
11:07 hungryd56 joined 11:21 sena_kun joined 11:23 Altai-man_ left 11:48 Xliff joined
Xliff \o 11:48
What does this mean: "Cannot invoke this object (REPR: P6int; uint32)" 11:49
sena_kun Xliff, "something is wrong" and "a gist is helpful". 11:50
Xliff Would help if I knew what to gist! :P
sena_kun I know... :/
Xliff No backtrace information in this.
sena_kun Xliff, is it reliable?
Xliff Yes. 11:54
sena_kun Xliff, --ll-exception shows more output? 11:55
Xliff No. 11:56
sena_kun Xliff, sigh... well, all this as a ticket to rakudo repo, I think.
Xliff sena_kun: --ll-exception will not work for Nativecall 11:57
Especially if it's from Moar code thrown inside C 11:58
Hah! Found it. 12:05
m: constant G := guint32; G(1) 12:06
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routine:
guint32 used at line 1
Xliff m: use NativeCall; constant G := uint32; G(1)
camelia 5===SORRY!5=== Error while compiling <tmp>
Failed to create directory '/home/camelia/.raku/precomp' with mode '0o777': Failed to mkdir: Permission denied
at <tmp>:1
Actually thrown at:
in any statement_control at /home/camelia/rakudo-m-i…
Xliff *sigh*
p6: use NativeCall; constant G := uint32; G(1)
camelia 5===SORRY!5=== Error while compiling <tmp>
Failed to create directory '/home/camelia/.raku/precomp' with mode '0o777': Failed to mkdir: Permission denied
at <tmp>:1
Actually thrown at:
in any statement_control at /home/camelia/rakudo-m-i…
Xliff Is camelia out of disk space? 12:07
lizmat nine ^^ 12:08
nine No, there's space left and when I try manually with camelia's user it works 12:09
m: use NativeCall; constant G := uint32; G(1) 12:10
camelia Cannot invoke this object (REPR: P6int; uint32)
in block <unit> at <tmp> line 1
Xliff sena_kun: Hey! A golf! 12:21
sena_kun Xliff, nice, now the ticket can be updated. :) 12:27
Xliff sena_kun: That error message is more LTA, I think. 12:32
12:58 hungryd56 left 12:59 hungrydonkey joined 13:17 MasterDuke joined 13:21 Altai-man_ joined 13:23 sena_kun left 13:32 hungryd87 joined 13:35 hungrydonkey left
lizmat yuck: TIL you can end a ... sequence with "last" 13:42
AlexDaniel lizmat: you can but I'm not sure if you should… it's weird 13:58
[Coke] m: say 1...last
camelia ===SORRY!===
last without loop construct
AlexDaniel lizmat: as I complained a few days ago, it's the is-lazy that's making it different
[Coke] assumes not literally. :)
AlexDaniel [Coke]: I think it's about 1, { last if $_ > 5; $_ + 1 } … ∞ 13:59
m: my $a = 1, { last if $_ > 5; $_ + 1 } … ∞; say $a
camelia Potential difficulties:
Useless use of … in sink context
at <tmp>:1
------> 3my $a = 1, { last if $_ > 5; $_ + 1 } …7⏏5 ∞; say $a
AlexDaniel m: my $a = (1, { last if $_ > 5; $_ + 1 } … ∞); say $a
camelia (...)
AlexDaniel m: my $a = (1, { last if $_ > 5; $_ + 1 } … ∞); say $a[5]
camelia 6
AlexDaniel m: my $a = (1, { last if $_ > 5; $_ + 1 } … ∞); say $a[*]
camelia Cannot .elems a lazy list
in block <unit> at <tmp> line 1
AlexDaniel m: my $a = (1, { last if $_ > 5; $_ + 1 } … ∞); .say far $a
camelia 5===SORRY!5=== Error while compiling <tmp>
Two terms in a row
at <tmp>:1
------> 031, { last if $_ > 5; $_ + 1 } … ∞); .say7⏏5 far $a
expecting any of:
infix stopper
statement end
AlexDaniel m: my $a = (1, { last if $_ > 5; $_ + 1 } … ∞); .say for $a
camelia (...)
AlexDaniel m: my $a = (1, { last if $_ > 5; $_ + 1 } … ∞); .say for @$a
camelia 1
AlexDaniel I'm clearly not good at this
anyway, you're much better of having this condition on the right side of …
because then .is-lazy is False
off* 14:01
but can you skip elements with next? :)
m: my $a = (1, { next if $_ == 5; $_ + 1 } … 10); .say for @$a
camelia (signal XCPU)1
AlexDaniel huh! 14:02
6c: my $a = (1, { next if $_ == 5; $_ + 1 } … 10); .say for @$a
committable6 AlexDaniel, gist.github.com/abebaf898806e43f2e...befdc9132c 14:09
Geth nqp: 789c41e540 | Coke++ | 2 files
standardize QAST Op formatting

  ... and update the test to be able to find these ops in the docs.
[Coke] down to 351 failures in the nqp docs. We are down to errors around "this op isn't documented" 14:26
... which is actually more because there are a lot of 'stub' entries that quite the test but don't actually have docs. 14:28
gist.github.com/coke/325a98074f317...7266df7da4 is the list the test complains about 14:33
14:38 hungryd87 left, hungrydonkey joined 14:48 hungryd85 joined 14:49 hungrydonkey left 15:22 sena_kun joined 15:23 Altai-man_ left 15:24 ufobat__ joined 15:26 patrickb joined 15:28 ufobat_ left
Geth nqp: patrickbkr++ created pull request #610:
Always handle proc exec arguments verbatim on JS
16:19 Xliff left
lizmat m: dd ({ |(1,9) } ... 9).head(10) # I'd argue this is wrong, should be (1,9).Seq 16:39
camelia (1, 9, 1, 9, 1, 9, 1, 9, 1, 9).Seq
lizmat m: dd ({ 9 } ... 9).head(10) 16:41
camelia (9,).Seq
17:21 Altai-man_ joined, hungryd85 left 17:22 hungrydonkey joined 17:23 sena_kun left
lizmat notices that the :nohandler named argument to nqp::until / nqp::while is not documented #jnthn 17:29
m: dd ("A" ... "zz").tail # surprised to see that end at "ZZ" (aka uppercase) 17:35
camelia "ZZ"
lizmat m: dd ("A" ... "aa").tail # aah, it's halting as soon as there are more chars in the produced value than are in the endpoint 17:36
camelia "ZZ"
17:40 finsternis joined 18:02 hungrydonkey left
[Coke] bartolin_: hio 18:25
18:28 samcv joined
[Coke] there are 22 js-only opcodes not documented; 30 jvm only. (none for moar) 18:28
Is there a way to get nqp to programmatically tell me its opcodes rather than looking at source and guessing? 18:30
There are a few cases where the ops are added programmatically, so the doc test reports JS has "isle" when it really has "isle_I", "isle_.." 18:31
AlexDaniel e: gist.github.com/AlexDaniel/e25346a...81bcb27626 18:37
evalable6 (exit code 1) 04===SORRY!04=== Er…
AlexDaniel, Full output: gist.github.com/e3f13bab58a5e5a989...d99f79cdc5
AlexDaniel e: gist.github.com/AlexDaniel/e25346a...81bcb27626 18:38
AlexDaniel c: all gist.github.com/AlexDaniel/e25346a...81bcb27626
committable6 AlexDaniel, gist.github.com/86dd35c4e4524042bb...f2b2062e4d
AlexDaniel c: all gist.github.com/AlexDaniel/e25346a...81bcb27626 18:39
committable6 AlexDaniel, gist.github.com/be404f776790049c75...940ed7b3e2 18:40
AlexDaniel bisect: gist.github.com/AlexDaniel/e25346a...81bcb27626
bisectable6 AlexDaniel, Bisecting by output (old=2015.12 new=2cbf583) because on both starting points the exit code is 0
AlexDaniel, bisect log: gist.github.com/93040317b21224f693...e10f2ea71e 18:41
AlexDaniel, (2016-02-13) github.com/rakudo/rakudo/commit/95...be0ac6ee60
AlexDaniel c: 9576596690^,9576596690 gist.github.com/AlexDaniel/e25346a...81bcb27626
committable6 AlexDaniel, ¦9576596690^: «42␤» ¦9576596: «»
AlexDaniel R#3593 18:50
linkable6 R#3593 [open]: github.com/rakudo/rakudo/issues/3593 File ending with backslash silently exits with 0 status
AlexDaniel btw I hate the issue template :)
“Steps to Reproduce” is fine, Environment is something I never look at, Expected Behavior is obvious in almost all cases 18:51
it's extremely easy to ask for more info if it's needed 18:52
vrurg AlexDaniel++ 19:05
On top of it, 'steps to reproduce' are usually explained as a part of the problem description. 19:06
jnthn Given how to reproduce the problem is probably the most important thing for having a chance of solving it, I'd rather keep that and leave the problem description 19:17
In many cases a decent issue name captures that
Uh, *leave out* the problem description
Expected behavior is desirable IMO; how to repro + what it should do = ingredients for a test case. 19:19
As for "easy to ask for more info" - yeah, maybe, but at least for me my time to fix things is limited, and if a ticket doesn't give me enough to go on, I tend to just pass over it for one that did. 19:20
nine The expected behavior also shows if the reporter has the right expectations in the first place. Also it's just not always clear what the expected outcome should actually be 19:21
jnthn Well, yes, sometimes there's no bug, just wrong expectations.
19:22 sena_kun joined
jnthn If anything is to be dropped from the template, I'd vote for problem description. 19:22
Environment is a tricky one, often it doesn't matter, occasionally it really matters. 19:23
19:23 Altai-man_ left
AlexDaniel maybe “Compiler version” is enough 19:24
I wish we motivated users to provide executable code blocks or gists 19:30
then I'd be able to automatically run committable and bisectable
but I don't know how to do that, everybody is using their own format 19:31
MasterDuke social distancing con: you don't get to throw a birthday party for your three-year-old daughter. social distancing pro: you don't have to throw a birthday party for your three-year-old daughter 19:32
19:36 lucasb joined
AlexDaniel MasterDuke: yeah, ignoring for a moment people dying and losing their jobs and stuff, it's an interesting way of life 19:40
perhaps a bit less fun but definitely more efficient :) 19:41
patrickb vrurg: Is there a way to access NQP configure variables containing '-' in rakudo Makefiles? I have the impression the '-' makes it impossible to access them. 19:49
MasterDuke definitely different. less daily spending, but there's a bit of a tendency to "i've already been in lockdown three whole days! time to treat myself to something crazy i order online!" 19:50
vrurg patrickb: I think it's been translated into _, doesn't it? If not we must do it then. 19:54
patrickb Does that happen implicitly or do we have to do that explicitly for every parameter? 19:55
Could we just replace all - with _ when reading a backend configuration? 19:56
vrurg patrickb: 1min 19:57
patrickb: actually, macros may support - too. If they not already. Let me check it... 19:58
patrickb: No, - isn't supported, but it's easy to add. I would have to go in 10mins, won't have time to test. You could do it by modifying regexp in Macros.pm lines 267 and 273. 20:00
patrickb: just replace [:\w] with [:\w\-] 20:01
patrickb I'll have a look
vrurg is afk& 20:03
21:21 Altai-man_ joined 21:23 sena_kun left
jjatria tyil: I've been playing around with these rakudo star images I've built, and I'm wondering what we want them to be able to do. If rakudo star is supposed to be "a user-oriented distribution", then do we want them to be able to eg. install modules? 21:41
Because currently, although zef is there, it's not very usable without eg. git, make, etc
Geth nqp-configure: patrickbkr++ created pull request #18:
Allow '-' in config variables
nqp-configure: d9e0ab97d0 | (Patrick Böker)++ | lib/NQP/Macros.pm
Allow '-' in config variables
nqp-configure: f1a2d81971 | (Patrick Böker)++ (committed using GitHub Web editor) | lib/NQP/Macros.pm
Merge pull request #18 from patrickbkr/minus-in-confvars

Allow '-' in config variables
22:40 Altai-man_ left 23:15 patrickb left 23:45 lucasb left 23:58 hungrydonkey joined