From @masak

<masak> r​: my $s = set; my $e = 5; $s = $s (|) $e; say $s.perl
<p6eval> rakudo cf2761​: OUTPUT«set()␤»
<masak> huh!
<masak> not how I expect (|) to work... :/
<masak> r​: my $s = set; my $e = 5; say ($s (|) $e).perl
<p6eval> rakudo cf2761​: OUTPUT«set("5")␤»
<masak> something is wrong when doing both '=' and '(|)'
* masak submits rakudobug
<jnthn> masak​: iirc, colomon++ mentioned he copied the precedence from STD
<jnthn> masak​: But it's suspect.
<jnthn> Or something like...
<colomon> Not from STD.
<jnthn> Ah, where from?
<colomon> I think that was the precedence TimToady had on it.
<jnthn> OK
* colomon wonders were the original Set implementation is on his hard drive.
<jnthn> Seems a bit surprising to me.
<colomon> definitely feels wrong to me after having tried it.
<jnthn> yeah

From @jnthn

On Sat Dec 15 09​:46​:04 2012, masak wrote​:

<masak> r​: my $s = set; my $e = 5; $s = $s (|) $e; say $s.perl
<p6eval> rakudo cf2761​: OUTPUT«set()␤»
<masak> huh!
<masak> not how I expect (|) to work... :/
<masak> r​: my $s = set; my $e = 5; say ($s (|) $e).perl
<p6eval> rakudo cf2761​: OUTPUT«set("5")␤»
<masak> something is wrong when doing both '=' and '(|)'
* masak submits rakudobug
<jnthn> masak​: iirc, colomon++ mentioned he copied the precedence from STD
<jnthn> masak​: But it's suspect.
<jnthn> Or something like...
<colomon> Not from STD.
<jnthn> Ah, where from?
<colomon> I think that was the precedence TimToady had on it.
<jnthn> OK
* colomon wonders were the original Set implementation is on his hard
drive.
<jnthn> Seems a bit surprising to me.
<colomon> definitely feels wrong to me after having tried it.
<jnthn> yeah

The spec and STD got its precedence for these updated, and Rakudo has
now followed suit, putting them at junctive precedences.

Tagging testneeded.

/jnthn

The RT System itself - Status changed from 'new' to 'open'

From @peschwa

Am Fr 11. Jan 2013, 13​:56​:40, jnthn@​jnthn.net schrieb​:

On Sat Dec 15 09​:46​:04 2012, masak wrote​:

<masak> r​: my $s = set; my $e = 5; $s = $s (|) $e; say $s.perl
<p6eval> rakudo cf2761​: OUTPUT«set()␤»
<masak> huh!
<masak> not how I expect (|) to work... :/
<masak> r​: my $s = set; my $e = 5; say ($s (|) $e).perl
<p6eval> rakudo cf2761​: OUTPUT«set("5")␤»
<masak> something is wrong when doing both '=' and '(|)'
* masak submits rakudobug
<jnthn> masak​: iirc, colomon++ mentioned he copied the precedence from STD
<jnthn> masak​: But it's suspect.
<jnthn> Or something like...
<colomon> Not from STD.
<jnthn> Ah, where from?
<colomon> I think that was the precedence TimToady had on it.
<jnthn> OK
* colomon wonders were the original Set implementation is on his hard
drive.
<jnthn> Seems a bit surprising to me.
<colomon> definitely feels wrong to me after having tried it.
<jnthn> yeah

The spec and STD got its precedence for these updated, and Rakudo has
now followed suit, putting them at junctive precedences.

Tagging testneeded.

/jnthn

Added test to roast in commit db7936e as per the example above.

From @coke

jnthn++ added a passing test to roast; closing ticket.

--
Will "Coke" Coleda

@coke - Status changed from 'open' to 'resolved'

From @coke

On Thu Feb 20 07​:54​:07 2014, coke wrote​:

jnthn++ added a passing test to roast; closing ticket.

Whoops, was peschwa++

--
Will "Coke" Coleda