From @moritz

proto sub a($x) is export { say "in proto"; nextsame }
sub a(Str $x) { say "Str $x" };
sub a(Int $x) { say "Int $x" };

a([]);

causes

in proto
Null PMC access in clone()
  in 'a' at line 1
  in main program body at line 5​:Foo.pm

Note that even when the nextsame() dispatch should succeed (in case the
proto is exported, and on calls a(3) from a calling module) the same
error occurs

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

From @coke

On Wed Jun 09 10​:50​:59 2010, moritz wrote​:

proto sub a($x) is export { say "in proto"; nextsame }
sub a(Str $x) { say "Str $x" };
sub a(Int $x) { say "Int $x" };

a([]);

causes

in proto
Null PMC access in clone()
in 'a' at line 1
in main program body at line 5​:Foo.pm

Note that even when the nextsame() dispatch should succeed (in case
the
proto is exported, and on calls a(3) from a calling module) the same
error occurs

This now dies with​:

$ ./perl6 foo
===SORRY!===
Illegal redeclaration of routine 'a' at line 2, near ";\nsub a(In"

--
Will "Coke" Coleda

From @moritz

On Mon Oct 03 11​:40​:26 2011, coke wrote​:

On Wed Jun 09 10​:50​:59 2010, moritz wrote​:

proto sub a($x) is export { say "in proto"; nextsame }
sub a(Str $x) { say "Str $x" };
sub a(Int $x) { say "Int $x" };

a([]);

causes

in proto
Null PMC access in clone()
in 'a' at line 1
in main program body at line 5​:Foo.pm

Note that even when the nextsame() dispatch should succeed (in case
the
proto is exported, and on calls a(3) from a calling module) the same
error occurs

This now dies with​:

$ ./perl6 foo
===SORRY!===
Illegal redeclaration of routine 'a' at line 2, near ";\nsub a(In"

There's the separate issue (already documented elsewhere) that a proto
doesn't auto-promote subs in the same lexical scope to multis. If one
works around this problem by writing

proto sub a($x) is export { say "in proto"; nextsame }
sub a(Str $x) { say "Str $x" };
sub a(Int $x) { say "Int $x" };

a([]);

the output with current Rakudo is

in proto
nextsame is not in the dynamic scope of a dispatcher

which I believe is correct, since dispatching from the proto to its
candidates is now done with {*} and not with nextsame.

So closable with tests.

From @moritz

On Wed Oct 03 13​:14​:00 2012, moritz wrote​:

There's the separate issue (already documented elsewhere) that a proto
doesn't auto-promote subs in the same lexical scope to multis. If one
works around this problem by writing

proto sub a($x) is export { say "in proto"; nextsame }
sub a(Str $x) { say "Str $x" };
sub a(Int $x) { say "Int $x" };

sorry, I meant
multi sub a(Str $x) { ... }
multi sub a(Int $x) { ... }

Cheers,
Moritz

From @moritz

tested in S32-exceptions/misc.t.

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