🦋 Welcome to Raku! raku.org/ | evalbot usage: 'p6: say 3;' or /msg camelia p6: ... | irclog: colabti.org/irclogger/irclogger_log/raku
Set by ChanServ on 14 October 2019.
Geth doc: softmoth++ created pull request #3472:
Language/variables: Stronger warning about pre-compiled constants
04:04
Geth doc: 3cd9c3617d | (Tim Siegel)++ | doc/Language/variables.pod6
Language/variables: Stronger warning about pre-compiled constants
06:20
doc: 333a4f10c6 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | doc/Language/variables.pod6
Merge pull request #3472 from softmoth/patch-1

Language/variables: Stronger warning about pre-compiled constants
linkable6 Link: docs.raku.org/language/variables
Geth doc: b19f1f1ed1 | (Wenzel P. P. Peppmeyer)++ (committed using GitHub Web editor) | doc/Type/Block.pod6
$_ defaults to OUTER::<$_>

no reason to hide this fact
06:23
doc: 71c8d6e69d | (Wenzel P. P. Peppmeyer)++ (committed using GitHub Web editor) | doc/Type/Block.pod6
Update Block.pod6

Proper examples should go to /language/variables. I added the hint mostly to be able to link to the description of the Topic. I believe the Topic is quite an important topic and deserves more love. Just not here.
linkable6 Link: docs.raku.org/type/Block
doc: 3e922c4554 | (Juan Julián Merelo Guervós)++ (committed using GitHub Web editor) | doc/Type/Block.pod6
Merge pull request #3471 from gfldex/patch-15

  $_ defaults to OUTER::<$_>
finanalyst hi. Could anyone point me at a reasonable appveyor script? I want to install raku, a module and run prove. I'm having trouble working out what I need 08:47
tellable6 2020-06-09T14:45:26Z #raku <tbrowder> finanalyst: i’ve been very pleased with the rakudo-pkg for debian
finanalyst tbrowder: I like rakudo-pkg too. 08:48
sena_kun finanalyst, e.g. OpenSSL module has appveyor config, see github.com/sergot/openssl/blob/mas...pveyor.yml 09:11
Maybe can be a starting point.
finanalyst sena kun: thanx 09:12
finanalyst sena kun: looked at OpenSSL module. looks good. In prove invocation, can I use `raku` in place of `perl6` now? 09:18
patrickb .tell rba There are two new versions of rakubrew (v10 and v11) on rooster.uber.space/patcloud/index....EJLQDZcjbf Can you upload? 10:00
tellable6 patrickb, I'll pass your message to rba
moritz I'm getting errors installing Inline::Perl5: perlpunks.de/paste/show/5ee22ef3.23bb.e1 can anybody please give me a hint how to debug this? 13:18
[Coke] huh. realized my windows raku is old. I must have compiled a version late last year, didn't remember doing that. (updated to latest rakudo star, thank you!) 14:53
[Coke] ... the rakudo-star install on windows is broken. 14:55
sena_kun [Coke], which one? 15:05
I mean, star version. 15:06
melezhik we are trying to figure out one baffling bug with Alpine/SQLite/DBIish/Docker , I wonder if it rakudo related ? raku devs help would be appreciated - github.com/raku-community-modules/...issues/190 15:11
so basically it seems to I know what's happening, but I don't know _why_ 15:12
github.com/raku-community-modules/...-642384713 15:13
[Coke] github.com/rakudo/star/issues/156 15:18
sena_kun [Coke], can you try out www.nntp.perl.org/group/perl.perl6...g8497.html ? 15:23
sena_kun [Coke], oops, rc is not an option anymore. Left a link in the ticket. 15:28
melezhik .tell vrurg_ RakuDist fails for the latest release of Vikna, i don't know, may be just docker related - rakudist.raku.org/sparky/report/debian/298 15:48
tellable6 melezhik, I'll pass your message to vrurg
moritz for the record, I got some help in private message regarding my question earlier. Turns out I had to install some build dependencies (build-essential and libperl-dev) 16:08
melezhik moritz, yeah Inline::Perl requires system packages, for RakuDist I use that one - github.com/melezhik/RakuDist/blob/...parrowfile 16:09
and probably something else, like make, c compiler , but that is all covered by this list - github.com/melezhik/RakuDist/blob/...parrowfile 16:10
moritz melezhik: I was lazy enough to just use `build-essential`, which includes make, gcc, a linker, dunno what else 16:14
actually, could reduce to libc-dev make gcc 16:17
melezhik moritz, sure, RakuDist supports now alpine/centos/ubuntu/debian , so for various platforms there are various requirements 16:23
this is why I have patches systems for certain modules, for example - github.com/melezhik/RakuDist/tree/...es/patches 16:24
so either have some common requirements defined when one bootstrap docker instance - github.com/melezhik/RakuDist/tree/...er/dockers
or some patches that address specific needs if required - github.com/melezhik/RakuDist/tree/...es/patches 16:25
cpan-raku New module released to CPAN! Vikna (0.0.3) by 03VRURG 17:06
New module released to CPAN! UNIX::Privileges (0.1.3) by 03JSTOWE
New module released to CPAN! Prometheus::Client (0.2.0) by 03HANENKAMP
New module released to CPAN! Grammar::DiceRolls (0.3.1) by 03TYIL
New module released to CPAN! IRC::Client::Plugin::DiceRolls (0.2.1) by 03TYIL
New module released to CPAN! IRC::Client::Plugin::Reminders (0.1.0) by 03TYIL
melezhik upss ... cpan-raku is fixed again ))) 17:07
[Coke] where's the repo for the star website? 17:19
is it just rakudo/star ? 17:20
ok, anyone else able to succesfully use the R-* MSI on windows? 17:23
even 2020.05.01 explodes here, just trying to get the --version
timotimo does at least moar --version work?
[Coke] yes 17:24
github.com/rakudo/star/issues/156
timotimo is it picking up existing precomp files?! you don't happen to have anything set in $PATH? could it have something to do with relocatable rakudo binary stuff? 17:35
JJMerelo releasable6 status 17:36
Why o why
[Coke] I uninstalled the old one first (didn't manually remove C:\rakudo though). I don't have another rakudo in my path. (I did before I started, the one I built from scratch)
JJMerelo releasable6 status
[Coke] timotimo: I'm seeing if I can recreate the magical "build just works" step I did back in 2019 17:37
(so far, so good)
Will be hard to sell anyone else at work on using Raku, though.
timotimo yeah we'll probably have to investigate this thoroughly 17:38
[Coke] as time permits, happy to test stuff.
chloekek . 17:39
timotimo gist.github.com/timo/1024223e0f6bb...c3092e2df3 17:40
[Coke] timotimo: that python works fine here. 17:41
timotimo then you didn't input it the same way i did, LOL
raku-bridge <Rustem B.> for me python fine too
timotimo m: say uninames("  ")
camelia (SPACE NO-BREAK SPACE)
timotimo m: say uninames("-c 'a = 44'") 17:42
camelia (HYPHEN-MINUS LATIN SMALL LETTER C SPACE APOSTROPHE LATIN SMALL LETTER A SPACE EQUALS SIGN SPACE DIGIT FOUR DIGIT FOUR APOSTROPHE)
[Coke] I cut and pasted from your gist
timotimo damn, it didn't copy over huh?
[Coke] ... I *thought* I did, one sec. 17:43
timotimo somewhere along the way the space got made harmless
[Coke] yup
raku-bridge <Rustem B.> UwU lol cdn.discordapp.com/attachments/695...nknown.png 17:52
<Rustem B.> i curl this from gist and hexize it 17:53
<Rustem B.> thereis inviz syms instead of space 😄 17:54
timotimo interesting! so copy-pasting into the browser worked, and sending it up to gist also worked (would be terrible if gist would "clean" your data for you tbh) but copying it *out* of the browser somehow missed the special values 18:00
ShimmerFairy is that utf-c8 striking again?
timotimo using $foo<no-break-space here> gives me an entry in the lex pad that is just $foo, so it isn't considered part of the identifier 18:01
which would have been a fantastic surprise
but you're right, what about b0rked input into -e 18:02
i don't have a simple, reliable way to generate some silly stuff that utf8-c8 would turn into "special stuff" 18:03
ShimmerFairy I encountered the same problem trying to type something like ä (that is, in decomposed form) as an arg to -e and it fell apart. Note that the precomposed form ä works, and things that require NFG to be one character (like s̈) work too. 18:04
*same kind of problem
timotimo unicodable6: COMBINING 18:05
unicodable6 timotimo, U+0300 COMBINING GRAVE ACCENT [Mn] ( ̀)
timotimo, U+0301 COMBINING ACUTE ACCENT [Mn] ( ́)
timotimo, 458 characters in total: gist.github.com/1741b7f48f6737c687...2f3fee7f90
timotimo there we go, those are what i wanted
m: say ("a".encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8")).decode("utf8-c8") 18:06
camelia a􏿽xCC􏿽x80
timotimo m: say ("a".encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8")).decode("utf8-c8").encode("utf8-c8").decode("utf8").&uninames
camelia (LATIN SMALL LETTER A WITH GRAVE)
ShimmerFairy utf8-c8 screws up at the NQP level, btw, as I recall.
JJMerelo releasable6 status 18:10
hum
[Coke] timotimo: build is going fine, forgot how slow it was. :) 18:12
AlexDaniel` JJMerelo: you forget the colon 18:13
JJMerelo releasable6: status 18:14
releasable6 JJMerelo, Next release in ≈9 days and ≈0 hours. There are no known blockers. 143 out of 275 commits logged (⚠ 4 warnings)
JJMerelo, Details: gist.github.com/ed637d2b6c8a7ba9b9...0ee1883baa
JJMerelo hey! thanks AlexDaniel
timotimo m: say ("my $a".encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") ~ " = 5; say $a\c[COMBINING GRAVE ACCENT]".encode("utf8")).decode("utf8-c8").encode("utf8-c8") 18:17
camelia 5===SORRY!5=== Error while compiling <tmp>
Variable '$a' is not declared
at <tmp>:1
------> 3say ("my 7⏏5$a".encode("utf8") ~ uniparse("COMBINING
timotimo of course
m: say ('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") ~ ' = 5; say $a\c[COMBINING GRAVE ACCENT]'.encode("utf8")).decode("utf8-c8").encode("utf8-c8") 18:18
camelia Blob[uint8]:0x<6D 79 20 24 61 CC 80 20 3D 20 35 3B 20 73 61 79 20 24 61 5C 63 5B 43 4F 4D 42 49 4E 49 4E 47 20 47 52 41 56 45 20 41 43 43 45 4E 54 5D>
timotimo m: use MONKEY; say ('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") ~ ' = 5; say $a\c[COMBINING GRAVE ACCENT]'.encode("utf8")).decode("utf8-c8").&EVAL
camelia 5===SORRY!5=== Error while compiling /home/camelia/EVAL_0
Bogus postfix
at /home/camelia/EVAL_0:1
------> 3my $a7⏏5􏿽xCC􏿽x80 = 5; say $a\c[COMBINING GRAVE ACCENT]
expecting any of:
constraint
infix
timotimo m: use MONKEY; say ('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") ~ ' = 5; say $a\c[COMBINING GRAVE ACCENT]'.encode("utf8")).decode("utf8").&EVAL
camelia 5===SORRY!5=== Error while compiling /home/camelia/EVAL_0
Variable '$a' is not declared
at /home/camelia/EVAL_0:1
------> 3my $à = 5; say 7⏏5$a\c[COMBINING GRAVE ACCENT]
timotimo m: use MONKEY; say ('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") ~ " = 5; say \$a\c[COMBINING GRAVE ACCENT]".encode("utf8")).decode("utf8").&EVAL
camelia 5
True
timotimo m: use MONKEY; say ('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") ~ " = 5; say \$a\c[COMBINING GRAVE ACCENT]".encode("utf8-c8")).decode("utf8-c8").&EVAL
camelia 5===SORRY!5=== Error while compiling /home/camelia/EVAL_0
Bogus postfix
at /home/camelia/EVAL_0:1
------> 3my $a7⏏5􏿽xCC􏿽x80 = 5; say $à
expecting any of:
constraint
infix
infix stopper
timotimo not sure what exactly i was meaning to test before i left my keyboard %) 18:19
timotimo ok, if we take @*ARGS in as utf8-c8 encoded, we'd be getting the string that has the private-use characters there 18:20
and those aren't allowed inside of identifiers
dominix hi everyone. I would like to have a list of system modules provided with rakudo-star (the link is broken on web site) 18:21
where do I get it from ?
timotimo m: $*OUT.write('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8") 18:22
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
------> 3"COMBINING GRAVE ACCENT").encode("utf8")7⏏5<EOL>
timotimo m: $*OUT.write('my $a'.encode("utf8") ~ uniparse("COMBINING GRAVE ACCENT").encode("utf8"))
camelia my $à
timotimo OK, when i put the a with the combining grave accent separately in the commandline, it'll give the error about the private use characters 18:23
i feel like we really ought to put something into our parse error messages that handles those, because it's absolutely ungoogleable
moritz dominix: perlpunks.de/paste/show/5ee276a4.753e.33d from the rakudo-star:2020.1 docker image
timotimo we may want to have a little thread in the problem-solving repository 18:24
dominix thanks moritz
timotimo when i put the same string into a file and invoke it from the commandline, it apparently gets decoded as utf8 without c8, so combined and uncombined $à both refer to the same thing there 18:25
at best it's a surprising inconsistency between -c and running from file
jdv79 any easy way to see what is using so much mem? 19:01
jdv79 i have what is basically a concurrent http crawler and /usr/bin/time -v says: Maximum resident set size (kbytes): 2066912 19:03
MasterDuke best bet is a heap snapshot 19:07
timotimo annoyingly, "max resident set size" is literally the max over the lifetime of the process
so you can't easily tell if it's just growing without bounds, or if it has reached a peak and went down again 19:08
but yes, heap snapshot can help you there
either run the process with heap snapshot profiling turned on, i.e. --profile=/tmp/blah.mvmheap, or use the function from the Telemetry module that i think it has?
jdv79 timotimo: even so, max is relevant in terms of resource constraints 19:12
and that's why i'm interested actually aka "where can i or can't i run this thing" 19:13
timotimo true 19:15
OTOH, if you have swap, and the main use of memory is just a huge array of integers where you do sequential accesses or so, that's a different thing
jdv79 the --profile=/tmp/blah.mvmheap addition seems to kill its speed 19:16
might take forever to complete:) 19:17
how do i use htat file when it does
timotimo that flag will run a heap snapshot collection on every GC run, which may be Often™
there's a commandline tool for it, and there's a web app tool for it 19:18
jdv79 based on how fast the term is being written to i'd say i have about 30-45m til its done
timotimo the first is called App::MoarVM::HeapAnalyzer (iirc), the other is moarperf
jdv79 ah, moarperf - i have that around
cool
timotimo you should be able to kill it in the middle and still get A Result that may be interesting
if the program's run is kind of homogenuous all in all
BBIAB 19:19
jdv79 ok
timotimo wakelift.de/2019/09/12/progressing...-progress/ the screenshot under "Frontend Work" here is the first thing you'll see with your heap snapshot loaded 19:20
i.e. a graph showing how the counts of objects by type evolve
the same for frames
atroxaper I have a question about Raku compiler's Grammar. Is there a reason to mention the vws token only in under unsp token but not under _ws token too as it did for unv? Because of that, a VSC conflict will be detected only after the escape symbol. 19:34
github.com/rakudo/rakudo/blob/mast...r.nqp#L703
dominix On macOS, mojave I got an error installing REadline [Readline] # Cannot locate native library '(null)': dlopen(libreadline.7.dylib, 10): image not found 19:40
it look like on Mojave there is libreadline.8.dylib but not 7
[Coke] so, a local build of raku on windows against HEAD was just fine, so it's just star, not rakudo 19:41
dominix: pretty sure there's an open ticket about that. one sec.
ugh. we need to adopt Readline if we haven't already. :( :( :( :( 19:42
jdv79 isn't LineNoise an alternative?
atroxaper dominix: I use only Linenoise on my Mac.
jdv79 i have no idea, don't use the repl - feels like i've heard that 19:43
[Coke] dominix: github.com/drforr/perl6-readline/issues/20 is not quite the same, but similar issue on os x
dominix thank atroxaper Coke, I'll install Linenoise, but that's not convenient, cause I got receipe with ReadLine that I depend on ... 19:49
Linenoise install with no errors
jdv79 timotimo: i overshot - was 23m 19:53
jdv79 that heap file is 21G. wonder if moarperf will handle it. 20:04
MasterDuke heh, how much ram do you have? 20:06
timotimo wow
well, it only reads the stuff you ask for
timotimo depending on how many snaps it took, it could be just fine with not terribly much ram 20:07
and it doesn't transfer very much to the browser, either
jdv79 timotimo: nopaste.linux-dev.org/?1319622
timotimo oh, very interesting! 20:08
i don't think i've seen this kind of error yet
jdv79 sweet
timotimo can you give me the git commit or release or whatevery ou're on? alternatively just a sha1sum or whatever of lib/HeapAnalyzerWeb.pm6? 20:09
jdv79 its the one you made... hold on
github.com/timo/moarperf/releases/...0e.tar.bz2
timotimo oh ... i wonder if it comes with the moarvm heapanalyzer module with commits i made, or if it uses the version jonathan has on his repo
that could be trouble perhaps
timotimo OK, it'd do that if it throws after line 39, so perhaps in self.request-snapshot(0) 20:12
ah, no, it does output "done" 20:13
wonder where that's from
ah, it made it out of sub load-file, but after that it asplodes 20:14
jdv79 did it eat too much data? 20:15
timotimo shouldn't be a problem, i hope! 20:16
the CATCH inside of the inner .then would have .say'd the exception, so that's probably not what happened
but it's possible that model-overview threw
timotimo can you put a CATCH { .say } into line 122 (make sure it doesn't change the return value, use explicit return statement if needed) 20:17
jdv79 nopaste.linux-dev.org/?1319626 20:23
i just inserted the CATCH 20:24
timotimo ah, that could actually be a version mismatch because i didn't push my newest changes to the "official" repo 20:27
timotimo damn i made so many branches 20:29
the very very user friendly setup now requires you to check out the branch "nqp-ops-make-stuff-faster" of github.com/timo/p6-app-moarvm-heap...r/branches 20:32
jdv79 i'll pass. i don't have time at the moment. 20:33
maybe next week
thanks
timotimo sorry that it's annoying, i really need to smooth this out 20:35
raku-bridge <draco100> hi, i have an existencial question, why this two act different perlpunks.de/paste/show/5ee2ade1.23bb.347 22:21