From @zoffixznet

I think this is related to https://rt-archive.perl.org/perl6/Ticket/Display.html?id=127134

The code below shows three scenarious​:
1) Subcall with two named args of same name without a comma between them, with the latter overriding the former
2) Subcall with a slipped hash and a named parameter of same name as the one in the hash, NOT separated by comma​: the named parameter does NOT override the value
3) Subcall with a slipped hash and a named parameter of same name as the one in the hash, SEPARATED by comma​: the named parameter DOES override the value

<Zoffix> m​: sub SUB(​:$foo) {say "[$foo]" }; my %args = :foo<foo>; SUB :foo<foo> :foo<newfoo>; SUB |%args :foo<newfoo>; SUB |%args, :foo<newfoo>;
<camelia> rakudo-moar 7d4cdb​: OUTPUT«[newfoo]␤[foo]␤[newfoo]␤»

The Expected Thing™ would be that slipped hash behaves same as regular named args or at least an error/warning is shown.

From @skids

Without the comma, the colonpair is in infix position, which means
it applies to the prefix unary | operator.

So whether this should be an error or "work" depends on what we can teach the |
operator to do with fake_infix adverbs it is passed.

Note if we fix | to use fake infix adverbs, then we create the problem
that |(1,1),(2,2),​:f :g will be 1,1,​:g,(2,2),​:f.

It might be best to teach | to throw when it is handed such things.

There are several other operators that will find themselves in the same
situation... especially we have other tickets about = which gets tricky
due to the difference between list and scalar assignment.

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

From @skids

On Mon, 21 May 2018 09​:13​:15 -0700, bri@​abrij.org wrote​:

Note if we fix | to use fake infix adverbs, then we create the problem
that |(1,1),(2,2),​:f :g will be 1,1,​:g,(2,2),​:f.

...sorry that might be OK due to the "first pair in term position"...
but

|(1,1),(2,2),1 :g will be 1,1,​:g,(2,2),1.