From @masak

<masak> hooray for runnable, concrete examples!
<PerlJam> indeed, moritz_++
<moritz_> don't praise me until you actually tested it :-)
<PerlJam> my rakudo barfs on it, but that's my rakudo's fault
<PerlJam> (perhaps rakudo in general)
<moritz_> yes, I just tested it
<moritz_> and I don't understand why
<moritz_> we have that as a test case even
<PerlJam> with anonymous scalars?
<moritz_> don't think so
* PerlJam bets not
<PerlJam> yeah, if I give the scalars names it works fine
<PerlJam> if I give *one* of the scalars a name, it works fine
* masak submits rakudobug
<moritz_> masak​: you are being called to du... d'oh, I'm too slow
<masak> hah!
<masak> here's a minimal example​:
<masak> rakudo​: class Scissors {}; class Paper {}; multi wins(Scissors
$, Paper $) {}; multi wins($, $) {}; wins(Scissors, Paper)
<p6eval> rakudo d5a2ee​: OUTPUT«Parameter type check failed; expected
Scissors, but got Paper for $ in call to wins [...]
<masak> rakudo​: class Scissors {}; class Paper {}; multi wins(Scissors
$a, Paper $a) {}; multi wins($, $) {}; wins(Scissors, Paper)
<p6eval> rakudo d5a2ee​: OUTPUT«Parameter type check failed; expected
Scissors, but got Paper for $a in call to wins [...]
<masak> std​: sub foo($a, $a) {}
<p6eval> std 28796​: OUTPUT«Potential difficulties​:␤ Useless
redeclaration of variabl [...] ok 00​:02 105m␤»
<masak> so it's a warning, not an error?
<masak> then the binder must know what to do when a call to such a
routine actually happens.

From @kyleha

This is an automatically generated mail to inform you that tests are now available in t/spec/S06-multi/type-based.t

commit fff5be9d99e2e6824395bb2183062e57c746336c
Author​: moritz <moritz@​c213334d-75ef-0310-aa23-eaa082d1ae64>
Date​: Thu Oct 15 17​:19​:52 2009 +0000

  [t/spec] tests for RT #​69798
 
  git-svn-id​: http://svn.pugscode.org/pugs@&#8203;28798 c213334d-75ef-0310-aa23-eaa082d1ae64

diff --git a/t/spec/S06-multi/type-based.t b/t/spec/S06-multi/type-based.t
index 0027d4d..251a90c 100644
--- a/t/spec/S06-multi/type-based.t
+++ b/t/spec/S06-multi/type-based.t
@@ -1,6 +1,6 @@
 use v6;
 use Test;
-plan 45;
+plan *;
 
 # type based dispatching
 #
@@ -83,8 +83,22 @@ is(mmd(1..3), 2, 'Slurpy MMD to listop via list');
     is wins(Scissor.new, Paper.new),   1,  'Basic sanity 2';
     is wins(Paper.new,   Paper.new),   0,  'multi dispatch with faked generics';
     is wins(Paper.new,   Scissor.new), -1, 'fallback if there is a faked generic';
+
+    # now try again with anonymous parameters (see RT #69798)
+    multi wins_anon(Scissor $, Paper   $) { 1  }
+    multi wins_anon(Paper   $, Stone   $) { 1  }
+    multi wins_anon(Stone   $, Scissor $) { 1  }
+    multi wins_anon(::T     $, T       $) { 0  }
+    multi wins_anon(        $,         $) { -1 }
+
+    #?rakudo 3 skip 'RT 69798'
+    is wins_anon(Scissor, Paper),  1, 'MMD with anonymous parameters (1)';
+    is wins_anon(Paper,   Paper),  0, 'MMD with anonymous parameters (2)';
+    is wins_anon(Stone,   Paper), -1, 'MMD with anonymous parameters (3)';
+
 }
 
+
 {
     multi m($x,$y where { $x==$y }) { 0 };
     multi m($x,$y) { 1 };
@@ -164,4 +178,6 @@ is(mmd(1..3), 2, 'Slurpy MMD to listop via list');
     is y(1, 2.5), 2, 'generics in multis (-)';
 }
 
+done_testing;
+
 # vim: ft=perl6

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

From @moritz

Fixed with the new signature binder, tests unfudged

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