»ö« Welcome to Perl 6! | perl6.org/ | evalbot usage: 'p6: say 3;' or rakudo:, or /msg camelia p6: ... | irclog: irc.perl6.org or colabti.org/irclogger/irclogger_logs/perl6 | UTF-8 is our friend!
Set by moritz on 22 December 2015.
El_Che New 2017.11 Linux packages: github.com/nxadm/rakudo-pkg/releas...g/v2017.11 00:17
AlexDaniel El_Che++ 00:49
lucs My mistake, or how do I find out what is wrong?: gist.github.com/lucs/43aa5fa13f686...4a7b45d811 01:52
comborico1611 Quiet night. 01:55
APic *shrug* 01:56
geekosaur weird. I think it dereferenced and flattened it into the p5 arglist 02:02
comborico1611 Goodnight! 02:33
Geth mu: c6673ce250 | (Zoffix Znet)++ (committed using GitHub Web editor) | misc/perl6advent-2017/schedule
Add/rearrange articles

  - Postpone drummer article 'til next year (ain't got time to record the needed videos)
  - Add "Perl 6: Sigils, Variables, and Containers" article
  - Add "Perl 6: Iterate All The Things!" (an article expounding Perl 6's Iterable types)
  - Add "Extending Perl 6's Exception Reporting"
  - Add "GitHub to Twitter Notifier in Perl 6"
  - Rearrange ZofArticles to start filling holes from the *end* of the calendar, to let potential
   latecomer authors jump in if they want.		 03:27
Geth doc: 90f7097b4c | (Jeremy Studer)++ | 18 files
Replace "dd" in examples

Replacing instances of Rakudo-specific "dd" operator with standard Perl 6.
SEE ALSO:
  [Issue #1680](github.com/perl6/doc/issues/1680)		 03:53
doc: c71c154d74 | (Jeremy Studer)++ | doc/Language/variables.pod6
Modify example to show explicit cloning vs binding

Example previously mentions the pitfalls of implicit binding when creating a Pair and suggests cloning. Now shows cloning explicitly.
synopsebot Link: doc.perl6.org/language/variables
doc: a0ee6719c1 | (Jeremy Studer)++ | xt/examples-compilation.t
Fix false positive on dd test due to "yyyy-mm-dd"

The "yyyy-mm-dd" method on Dateish role was causing false positives in
  "dd" test.
Geth doc: ad276ed7d5 | (Jeremy Studer)++ | xt/examples-compilation.t
Rearrange "dd" regex; Reads better.

Moved negative lookbehind before boundary. Reads better.		 04:20
tyil m: my $a = [email@hidden.address] my @a = ($a, $a); dd @a; 10:02
camelia Array @a = ["test\@domain.org", "test\@domain.org"]
tyil why does it insert these backslashes here?
tyil it seems to write the email correctly to the file (without the backslash), but the test on it fails and shows it has a backslash 10:13
parv perhaps a function of dd? 10:14
tyil it seems so, but I'm not using dd in my test
and my test shows it with the \, which of course fails the test 10:15
jnthn .perl would also do that, I guess 10:19
tyil can I stop it from doing that? 10:23
jnthn Are you actually calling .perl in your test? 10:25
If so, call .perl on the thing you're testing it against?
tyil not yet
but I can add a .perl
jnthn Ah 10:26
Then I'm suspecting it's only showing the \ as part of the error reporting
Rather than it being there in the data
Or at least, it's worth checking if that's what is going on :)
jast I found a moarvm issue the other day on armv7l and I'm going to report the bug the next time I actually manage to get home before night, but just in case I wanted to try and figure it out myself, any smart ideas on what docs I could read to get an overview of the 6model under-the-hood stuff without going through all of the code? 10:28
jnthn jast: What kind of issue was it?
jast a SIGBUS due to alignment issues
possibly a compiler problem, too, I haven't tried too many yet 10:29
just one so far. lack of time and all that...
jnthn Ah. I can probably give more hints with a stacktrace 10:30
Does it get so far as building Rakudo, or does it explode well before that?
jast the build all succeeds 10:31
anyway, I've already traced it, it fails in P6num's set_num
the default case of the switch block, as I recall, but I'd better check again
for getting more info on where the data is actually coming from I'll probably need to collect more than just the stacktrace 10:32
jnthn Do you observe the failure in a nativecall test?
jast doesn't help that I'm not exactly literate in ARM-style instruction sets
no, it's when I launch perl6.moarvm 10:33
jnthn Ah
Then maybe take a look in P6opaque.c which does the layout of objects
It may be doing insufficient alignment ahead of a P6num
But iirc that's a data-driven process, based on alignment configuration which may be in P6num.c 10:34
jast well, the SIGBUS happens in a 'bx' instruction, that would be code-driven, right?
jnthn I meant that the calculation of object layout and alignment may be 10:35
jast ah looky here, I kept a copy of the output... gist.github.com/jast/df9ba1a023222...7ad1dfce45
in another run I checked the register contents and lr was, indeed, not aligned
but I know nothing about the registers and all that, so that didn't tell me a lot :)
I suppose this is an opportunity to start learning more about ARM 10:37
jnthn p &(((MVMP6numBody *)data)->value) may be interesting 10:38
(To see what the lower bits of the address are, which may hint an the alignment issue)
jast $2 = (union {...} *) 0xb413266a 10:42
so I guess it wants 64 bit values aligned at 64 bits, which this isn't 10:45
jnthn jast: Yeah, looks like. Though oddly it already should be accounting for the platform's alignment desires: github.com/MoarVM/MoarVM/blob/mast...6num.c#L12 11:18
ZzZombo #perl6, I'm still struggling to make attribute binding to work, what I need is to make an attribute bound to value in the setter. Can somebody show a quick example? 12:05
timotimo maybe you need to set the incoming parameter "is raw" so that it doesn't recontainerize what you're passing into a read-only container? 12:08
what is your setup? are you using a proxy?
ZzZombo yes, it is result of `Str.substr-rw` 12:09
timotimo ok, try an "is raw" on the argument to the STORE block in the proxy
ZzZombo m: class C { has Str $.s;multi method s is raw{return-rw $!s};multi method s(Str:D \s) is raw{$!s:=s};submethod TWEAK(:$s){ self.s($s)} };my $a=C.new(:s<a>);my $s='aaaaaaaaaa';$a.s($s.substr-rw(2,3));$a.s('asd');say ($a,$s) 12:11
camelia 5===SORRY!5===
Trait 'is raw' needs whitespace before block
at <tmp>:1
------> 3$.s;multi method s is raw{return-rw $!s}7⏏5;multi method s(Str:D \s) is raw{$!s:=s}
Missing block (apparently claimed by 'is raw')
at <tmp>:1
------>…
ZzZombo m: class C { has Str $.s;multi method s is raw {return-rw $!s};multi method s(Str:D \s) is raw {$!s:=s};submethod TWEAK(:$s){ self.s($s)} };my $a=C.new(:s<a>);my $s='aaaaaaaaaa';$a.s($s.substr-rw(2,3));$a.s('asd');say ($a,$s)
camelia (C.new(s => "asd") aaaaaaaaaa)
timotimo ah, you're already using a sigilless parameter, that's equivalent to "is raw" 12:12
oh, it looks like it worked what you wanted there? 12:13
like, what's your goal here?
ZzZombo no? the original string didn't change. I want to be able to change the provided part of a string from my objects.
timotimo then what you want is assignment, not binding 12:14
you're binding "asd" into $!s, at which point the result from substr-rw is replaced with "asd" 12:15
you need two different methods here, one to bind so you can get the substr-rw result in, and one to assign so you can assign into the substr-rw result
tyil how can I check if a variable contains an object that does Iterable? 12:23
timotimo smart match against Iterable
tyil thanks :>
timotimo np 12:24
buggable New CPAN upload: Config-1.3.0.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...3.0.tar.gz 12:43
Zoffix m: &say.wrap: sub (|c) { unless $*x { my $*x = 1; samewith 42; }; nextsame }; say 2 13:55
camelia 2
Zoffix m: &say.wrap: sub (|c) { unless $*x { my $*x = 1; say 42; }; nextsame }; say 2
camelia 42
2
Zoffix Why does using `samewith` instead of the routine name directly give different results?
moritz is "because wrapping is evil" a satisfactory answer? 13:56
probably not...
Zoffix No, it isn't.
It's probably because it's same-ing using the wrapper sub instead of reaching out to the wrappee 13:58
m: &say.wrap: sub (|c) { unless $*x { my $*x = 1; samewith 42; }; dd nextcallee }; say 2
camelia Sub say = sub say (| is raw) { #`(Sub|55633040) ... }
Nil
Zoffix m: &say.wrap: sub (|c) { unless $*x { my $*x = 1; say 42; }; dd nextcallee }; say 2
camelia Sub say = sub say (| is raw) { #`(Sub|58302032) ... }
Sub say = sub say (| is raw) { #`(Sub|58302032) ... }
Zoffix m: multi z (|) {}; &say.wrap: multi z($) { unless $*x { my $*x = 1; samewith 42; }; dd nextcallee }; say 2 14:00
camelia Sub z = sub z (| is raw) { #`(Sub|63117312) ... }
Sub say = sub say (| is raw) { #`(Sub|63127496) ... }
Zoffix Yeah. OK. Good enough for me. Thanks. 14:01
lucs How do I get Inline::Perl5 to use my just installed more recent version of Perl 5? 14:15
(the more recent version is the one found in the path where I am using Inline, but it sees my previously installed perl5) 14:17
lucs (uninstalling and reinstalling Inline changed nothing) 14:20
timotimo lucs: i'd say inline::perl5 compiling a helper .so during its installation could have something to do with which perl5 actually gets used 14:26
lucs Makes sense, trying to delete relevant stuff (but not sure exactly what) to reattempt build. 14:28
timotimo i'd zef install --force Inline::Perl5, that ought to do it 14:29
ilmari lucs: presumably it builds against whichever perl is first in your path
lucs timotimo: Tried that, no luck.
Oops, meeting 14:30
TTYL
ilmari 'zef uninstall Inline::Perl5' first?
lucs ilmari: Yep, did that
El_Che rm -rf ~/.perl6 should fix it, but you'll loose other home installed modules 14:32
DrForr waves to lucs. 14:33
Zoffix Along with `samewith` weirdness, there's also this thing, where you have to pass `self` to nextwith/callwith: 14:44
m: $*OUT.WHAT.^lookup('print').wrap: my method (|) { nextwith 42 }; say 42
camelia Type check failed in binding to parameter '<anon>'; expected IO::Handle but got Int (42)
in method <anon> at <tmp> line 1
in block <unit> at <tmp> line 1
Zoffix Despite `nextsame` knowing what to do... Was this really the design of how this thing was meant to work?
Feels like the wrapper is in want of having its dispatcher tweaked or something 14:45
ugexe zef nuke StoreDir, or zef look Inline::Perl5 && make clean && zef install . 14:46
Zoffix tyil: didn't see anyone mention: the backslashes in `"test\@domain.org"` come from `.perl`. It escapes the `@` because depending on what follows, it could be interpreted as an interpolated Array. Even though in this particular case it wouldn't by escaping all the `@` it doesn't need to know what follows it. 14:47
tyil I saw it before in Config::TOML too, and back then I believed it to be a bug in their lib, but couldnt figure out why 14:48
lucs Hiya DrForr :)
tyil jnthn: also said it was probably the .perl part 14:49
which seems to also be used in Config::TOML, so now I can try to fix another issue as well
and the test I had my problem with earlier today now works as well :>
so everything is fine again
lucs Okay, getting rid of ~/.perl6
tyil if I can fix the Config::TOML bug I can resolve another small bug in cpan6, and then continue on writing more tests 14:50
but that'll have to wait until I get home
Zoffix retracts attempts to mess with dispatcher 14:58
It actually makes sense as is, if you… .wrap… your mind around it ^_^
[Coke] ~~ to Zoffix 15:20
ugexe lucs: zef look Inline::Perl5 && make clean && zef install . 16:08
El_Che ugexe: it looks like zef could use a clean-install command :) 16:11
ugexe perhaps, or alternatively makefiles could run a make clean step first since they will be more familiar with their build system 16:13
im open to suggestions though
perlpilot zef caches module downloads? 16:14
ugexe yes
perlpilot Hmm. if this were only cpan, I'd say only cache the tarballs and unpack them each time. 16:15
El_Che why the caching actually? 16:16
ugexe yeah managing a transparent handling of directory based repo vs archive based repo is not fun
El_Che I can imagine
perlpilot ugexe: you could make tarballs out of the repos :) 16:17
ugexe a couple of reasons. one is because a name/auth/ver/api combo is meant to represent that dist no matter what so caching, in theory, should be perfectly fine
another was so there was a work directory to see what actually happened in a build 16:18
ugexe and finally because it is faster (maybe not compared to tarballs, but to git cloning yeah) 16:19
well, it should be faster than tarballs but I mean its probably not fast enough to make much difference
pmurias what is more correct 'Return 1 if ..., 0 otherwise' or 'Return 1 if ..., otherwise 0' 16:26
moritz return ... ?? 1 !! 0 16:28
perlpilot I don't know about "correct", but I like the second one
moritz oh, this is about language, not code
sorry
perlpilot moritz: yeah, it's hard to tell these days sometimes :)
ilmari prefers the second form too 16:29
ugexe perlpilot: then i have to do cross-os adapters for creating tar archives, not just reading them
perlpilot yeah, that would suck. 16:30
Better would be to give a --no-cache option or something 16:31
--refetch? 16:32
ugexe you can disable the cache with --/cache
--refetch would not help because the distribution is still the same 16:33
er --/cached 16:34
but this also means it can be disabled in config.json 16:35
El_Che ugexe: I am curious where did you get the --/ syntax 16:39
little weird
ugexe its built in to perl6
El_Che ok, now I am curious where perl6 got it :)
moritz you could ask [particle]1 16:43
ugexe not sure, but whats the alternative? if you prefix with `no-` or some such then your MAIN signature no longer works like most other subs because you cant have a `no-whatever` param in the signature 16:44
moritz ugexe: no, the preprocessor could do that for you 16:45
ugexe and if you have two flags (true and false) that means you could also set them both which would conflict
moritz that's the same with --/foo and --no-foo 16:46
El_Che before lizmat fixed MAIN params where pretty useless
ag options before commands stuff.
ugexe --no-foo is a different name, it would (should) show up twice
moritz there's a step that translates @*ARGS into a capture. It could turn --no-foo into :foo(False)
ugexe --foo and --/foo would not
well yeah, but now your MAIN signature handles named items significantly different than other subs 16:47
El_Che I have no associaciation in my head with / as a negation
moritz ugexe: it's not MAIN that does it
El_Che hence the question
ugexe when i program a MAIN sub i'm not thinking about the preprocessor, and hence not thinking if I can add a :no-whatever option or not 16:49
if I already have a :whatever, then I cannot add a :no-whatever
moritz but, how often do you want to add both a :whatever and a :no-whatever option? 16:50
this sounds like a very theoretical option to me
s/option/problem/
perlpilot You have: no tea.
moritz Getopt::Long in Perl 5 does it
argparse in python does it too (but without the dash) 16:51
ugexe and yet people love perl6's MAIN handling and i never hear praise about either of those
ilmari ugexe: if you have a Bool :whatever you can do --/whatever to disable it 16:52
moritz ugexe: but it's not because of the weird negation thing that people love Perl 6's MAIN handling
El_Che should you just give default values (false) and add 1 bool to disabled if needed? 16:53
that's what's most CLI's use
ugexe ilmari: right
El_Che so you'll end up with no --cache option, but a --no-cache is cache is the default
ugexe and what if the user chooses the defaults? 16:54
El_Che if cache is the defaut
ugexe: he'll use a wrapper?
ugexe in zef you can set any of the plugins to any command you want
flag rather
moritz well, i'd proppose --cache calls MAIN(:cache), and --no-cache calls MAIN(:!cache)
and MAIN can do a $cache // $users-default 16:55
and MAIN would be declared as sub MAIN(Bool :$cache) { }
ugexe still must deal with short name, -c -/c 17:11
well not must, but its a thing 17:12
ugexe but as for it being a theoretical option, i'd point you to github.com/ugexe/zef/issues/147 17:20
ugexe which is about the rakudo bin installation wrapper adding a named option that it handles 17:20
well, not quite the same. it does expose issues about what naming you can use in MAIN though 17:24
ryn1x_ How can I make the values of a hash bindings to a variable when creating it? I can do %a<1> := $a after the hash is created, but is there a way to use the my %a = 1 => ???, 2=> ??? syntax? 17:35
ugexe m: sub MAIN(:f(:$foo)) { say << $foo = $foo.Bool() >> }; @*ARGS=<--/foo=bar>; $ = &MAIN; # also this :o 17:36
camelia (bar = False)
raschipi m: my %a := {1 => '???', 2=> '???'}; say %a 17:37
camelia {1 => ???, 2 => ???}
Zoffix ryn1x: no, but you could alias the hash to $_ and save on variable typing, so at least length-wise the code is comparatively the same 17:45
m: my $a = 111; my $b = 222; with my %a {.<1> := $a; .<2> := $b }; $a = 33; $b = 44; say %a 17:46
camelia {1 => 33, 2 => 44}
ryn1x_ Zoffix: Your solution works perfectly. Thanks! 17:50
Zoffix m: my $a = 111; my $b = 222; sub circumfix:<raw{ }> (@_ is raw) { my %h; for @_ -> (:key($k), :value($v) is raw) { %h{$k} := $v }; %h }; my %a := raw{ 1 => $a, 2 => $b }; $a = 33; $b = 44; say %a 17:51
camelia {1 => 33, 2 => 44}
Zoffix ryn1x_: ^ or you could make your own raw-hash-making op
ryn1x_ That is really cool. I have not messed with making operators yet. 17:54
Zoffix m: sub infix:<¯\(°_o)/¯> { "@_[] not a big deal; just subs".say }; "They" ¯\(°_o)/¯ "are" 17:55
camelia They are not a big deal; just subs
Zoffix m: my $a = 111; my $b = 222; sub circumfix:<raw{ }> { my %h; %h{.key} := .value for @_; %h }; my %a := raw{ 1 => $a, 2 => $b }; $a = 33; $b = 44; say %a 17:57
camelia {1 => 33, 2 => 44}
Zoffix (a bit shorter version) 17:58
lucs ugexe: Hi, just got back to the computer. 18:54
lucs re. 'zef look Inline::Perl5 && make clean && zef install .', looks like the found directory has no Makefile yet (only the Makefile.in). 18:56
Um, will try the install...
"All candidates are currently installed" 18:57
ugexe uninstall it or use --force install
lucs I believe I did that, but, trying again. 18:58
lucs Still appears to use 5.22.1 :/ 19:03
I think I'll nuke very .precomp directory in sight...
ugexe well you can go into its direcoty with `zef look Inline::Perl5` and do `perl6 -e 'use Inline::Perl5; ...'` to avoid installation altogether 19:04
to rule that out
er, perl -Ilib -e 'use Inline::Perl5; ...'
perl6^ 19:05
lucs Okay, trying...
ctilmes m: my $proc = run 'cat', :in, :out; $proc.in.say: "test\n"; say $proc.out.get; $proc.in.close; 19:17
camelia run is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in sub run at src/RESTRICTED.setting line 14
in block <unit> at <tmp> line 1
ctilmes m: my $proc = run 'cat', :in, :out; say $proc.out.native-descriptor; $proc.in.say: "test\n"; say $proc.out.get; $proc.in.close; 19:18
camelia run is disallowed in restricted setting
in sub restricted at src/RESTRICTED.setting line 1
in sub run at src/RESTRICTED.setting line 14
in block <unit> at <tmp> line 1
ctilmes If I print out the native-descriptor, it breaks this.
lucs ugexe: This worked: zef uninstall Inline::Perl5; ｢manually removed the "found" directory｣; zef look Inline::Perl5 ｢(Maybe just 'zef install Inline::Perl5' would have worked, but there it is. This recreated the "found" directory and shelled into it.)｣; zef install . 19:24
Now Inline::Perl5 sees my more recent perl 5. 19:25
lucs But, this Inline::Perl5 problem remains: gist.github.com/lucs/43aa5fa13f686...4a7b45d811 19:27
lucs Is Inline::Perl5 brokened or what? 19:28
moritz lucs: try using ${ some_attr => 42 } 19:29
lucs Okay, trying... 19:29
moritz lucs: Inline::Perl5 has this infuriating thing where it flattens out things that aren't supposed to be flattened out
and putting it into a scalar (usually by prepeding $) works around that 19:30
lucs That indeed appears to work.
Thanks 19:31
lucs I wonder if the synopsis example shown at github.com/niner/Inline-Perl5 works... 19:32
perlawhirl Hi perlers
python's itertools.permutations takes a second argument that can give you permutations of length N 19:34
ie. list(permutations('ABC', 2)) --> [(A B) (A C) (B A) (B C) (C A) (C B)]
trying to find a nice way to do it in perl and this is what i got
my @x = <A B C>; my $n = 2; say ([X] ^@x xx $n).map({ @x[|$_] if set($_) == $n })
m: my @x = <A B C>; my $n = 2; say ([X] ^@x xx $n).map({ @x[|$_] if set($_) == $n })
camelia ((A B) (A C) (B A) (B C) (C A) (C B))
moritz m: say permutations([<a b c>], 2) 19:35
camelia Too many positionals passed; expected 1 argument but got 2
in block <unit> at <tmp> line 1
moritz m: say combinations([<a b c>], 2)
camelia ((0 1) (0 2) (1 2))
moritz m: say combinations([<a b c>], 2).map: *.permutations
camelia (((0 1) (1 0)) ((0 2) (2 0)) ((1 2) (2 1)))
moritz m: say combinations([<a b c>], 2).map(*.permutations).flat
camelia (0 1 1 0 0 2 2 0 1 2 2 1)
moritz too flat :-) 19:36
perlawhirl moritz: the combinations trick works... albeit in a different order 19:38
m: my x = <A B C>; say combinations(@x, 2).map({ |@x[|.permutations] })
camelia 5===SORRY!5=== Error while compiling <tmp>
Malformed my (did you mean to declare a sigilless \x or $x?)
at <tmp>:1
------> 3my x7⏏5 = <A B C>; say combinations(@x, 2).map(
perlawhirl m: my @x = <A B C>; say combinations(@x, 2).map({ |@x[|.permutations] })
camelia ((A B) (B A) (A C) (C A) (B C) (C B))
perlawhirl tho it is much nicer, thanks
perlawhirl lolfixed 19:42
m: my @x = <A B C>; say combinations(@x, 2).map({ |@x[ [Z] |.permutations ] })
camelia ((A B) (B A) (A C) (C A) (B C) (C B))
perlawhirl oh wait... still not the same order as python... almost
can just .sort it after the fact, i suppose 19:43
moritz both orders make sense on some level
perlawhirl yeah
moritz m: say 'abc' ~~ /$<0>=[.]/ 19:44
camelia ｢a｣
0 => ｢a｣
moritz m: say 'abc' ~~ /$<0>=[.](.)/
camelia ｢ab｣
0 => ｢a｣
1 => ｢b｣
moritz I half-expected this to make $0 a list 19:45
m: say 'abc' ~~ /$<0>=[.](.)$<1>=[.]/
camelia ｢abc｣
0 => ｢a｣
1 => ｢b｣
1 => ｢c｣
moritz m: say 'abc' ~~ /$<0>=[.](.)$<1>=[.]/; say $1.^name
camelia ｢abc｣
0 => ｢a｣
1 => ｢b｣
1 => ｢c｣
Array
moritz fascinating 19:46
perlawhirl ahh... this is most idiomatic... 19:50
m: my @x = <A B C>; say @x.combinations(2).map(|*.permutations)
camelia ((A B) (B A) (A C) (C A) (B C) (C B))
moritz random fun fact: my regex book will have an index entry for "Secret Magic Sauce™" :-) 19:55
buggable New CPAN upload: Font-FreeType-0.0.1.tar.gz by WARRINGD cpan.metacpan.org/authors/id/W/WA/...0.1.tar.gz 20:13
s0me0ne-unkn0wn Guys, can someone please explain me that NETWORK_TESTING thing? Is it described somewhere in docs? If I listen/connect on a loopback interface, is it NETWORK_TESTING or not? 20:27
comborico1611 Sorry, i don't know. 20:40
ugexe is there somewhere in the core you see NETWORK_TESTING used? 20:44
or are you asking about externals module use of environmental variables?
moritz github/azawawi/perl6-net-curl and github/perl6/perl6-lwp-simple use it 20:47
also http-useragent and openssl
mr_ron rakudo: { my int ($a, $one, $three) = (42, 1, 3); for ^10_000_000 { $a += $one + $a%$three }; say now - ENTER now; say $a } { my int ($a, $one, $three) = (42, 1, 3); for ^10_000_000 -> int --> Nil { $a += $one + $a%$three }; say now - ENTER now; say $a }
camelia 5===SORRY!5=== Error while compiling <tmp>
Unexpected block in infix position (missing statement control word before the expression?)
at <tmp>:1
------> 3ay now - ENTER now; say $a }7⏏5 { my int ($a, $one, $three…
ugexe right, but that doesn't mean it belongs in the perl6 docs 20:48
moritz comborico1611: I don't know if it's specified that well. I'd personally be bold enough to do localohost-only tests without that env var
and only change that if somebody complains, and has good reason
mr_ron rakudo: { my int ($a, $one, $three) = (42, 1, 3); for ^10_000_000 { $a += $one + $a%$three }; say now - ENTER now; say $a }; {my int ($a, $one, $three) = (42, 1, 3); for ^10_000_000 -> int --> Nil { $a += $one + $a%$three }; say now - ENTER now;say $a } 20:50
camelia 2.491267
15000042
1.96962212
15000042
mr_ron rakudo: { my int ($a, $one, $three, $limit) = (42, 1, 3, 10_000_000); loop (my int $i = 0; $i < $limit; $i++) { $a += $one + $a%$three }; say now - ENTER now; say $a } 20:52
camelia 0.8225151
15000042
mr_ron timotimo: RT 130982 - did I get the int --> Nil signature right or as you intended? 20:53
buggable New CPAN upload: Font-FreeType-0.0.2.tar.gz by WARRINGD cpan.metacpan.org/authors/id/W/WA/...0.2.tar.gz
synopsebot RT#130982 [open]: rt.perl.org/Ticket/Display.html?id=130982 [PERF] "for $a..$b -> $i { ... }" loops are sometimes much slower than c-style loops
s0me0ne-unkn0wn ugexe: asking regarding install-time tests of external modules, I need to write some more tests for IO::Socket::Async:SSL and I want to understand "dos and donts" better before I start :) 20:54
timotimo mr_ron: yeah, that works
i haven't looked closer at why the native loop (...) outperforms the optimizer's result for the for-range-loop by so much 20:55
mr_ron thank your
moritz s0me0ne-unkn0wn: as a data point, roast (the perl 6 test suite) does not require NETWORK_TESTING for tests on the loopback device
s0me0ne-unkn0wn moritz: thanks, that's good news 20:57
mr_ron (sorry, meant thank you)
ugexe you can use set BLARG=1; if %*ENV<BLARG> { <do tests> } if you wish. e.g. its something the test author exposes to allow the end user to give it parameters
ugexe the fact NETWORK_TESTING is used in different modules is a good thing, but it has no defined meaning beyond what each author gives it 20:58
s0me0ne-unkn0wn Ok, than I'll just pray nobody runs those tests on a host without loopback interface :) 21:00
s0me0ne-unkn0wn One won't require IO:Socket::Async::SSL anyway, I suppose 21:00
ugexe `require IO::Socket::SSL` might be the most used require in the ecosystem
s0me0ne-unkn0wn But not on the hosts with no loopback interface, I hope :) 21:01
ugexe if you can detect something exists or not you should and make the tests act accordingly. NETWORK_TESTING also disables it for slow networks, whenyou dont care, etc
buggable New CPAN upload: Config-1.3.1.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...3.1.tar.gz 21:13
ctilmes Test::When uses ONLINE_TESTING instead of NETWORK_TESTING 21:39
comborico1611 New p6 alias idea: 6lerp (pronounced slurp). 22:06
Geth doc: 00e1a3abe7 | (Will "Coke" Coleda)++ | EXAMPLES.md
formatting fix
 22:07
comborico1611 Huh, huh?
++Zoffix, 6lerp?
buggable New CPAN upload: Config-1.3.2.tar.gz by TYIL cpan.metacpan.org/authors/id/T/TY/...3.2.tar.gz 22:43
Herby_ o/ 23:33
timotimo m: sub wiggle($a) { $a.uc }; my @data = { foo => a }, { foo => b }, { foo => c }; for @data { .<foo> .= &wiggle }; say @data 23:34
camelia 5===SORRY!5=== Error while compiling <tmp>
Undeclared routines:
a used at line 1
b used at line 1
c used at line 1
timotimo m: sub wiggle($a) { $a.uc }; my @data = { foo => "a" }, { foo => "b" }, { foo => "c" }; for @data { .<foo> .= &wiggle }; say @data 23:35
camelia [{foo => A} {foo => B} {foo => C}]
timotimo i like this.
timotimo o/ Herby_ 23:35

Powered by Cro and the Raku® Programming Language.

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

Thank you!