From @masak

<masak> rakudo​: class Foo does Positional[​::T] {}; say Foo ~~ Positional
<p6eval> rakudo c57fbd​: OUTPUT«Null PMC access in isa()␤in Main [...]
* masak submits rakudobug
<masak> rakudo​: role A {}; class Foo does A[​::T] {}
<p6eval> rakudo c57fbd​: OUTPUT«Null PMC access in isa()␤in Main [...]

From @bbkr

On Atlanta build

[16​:06] <bbkr> rakudo​: class Foo does Positional[​::T] {}; say Foo ~~
Positional
[16​:06] <p6eval> rakudo 9808d7​: OUTPUT«1␤»

[16​:07] <bbkr> rakudo​: role A {}; class Foo does A[​::T] {}
[16​:07] <p6eval> rakudo 9808d7​: OUTPUT«===SORRY!===␤No applicable
candidates found to dispatch to for '_block120'. Available candidates
are​:␤​:()␤␤»

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

From @coke

On Sun Aug 02 13​:38​:45 2009, masak wrote​:

<masak> rakudo​: class Foo does Positional[​::T] {}; say Foo ~~
Positional
<p6eval> rakudo c57fbd​: OUTPUT«Null PMC access in isa()␤in Main [...]
* masak submits rakudobug
<masak> rakudo​: role A {}; class Foo does A[​::T] {}
<p6eval> rakudo c57fbd​: OUTPUT«Null PMC access in isa()␤in Main [...]

15​:39 < [Coke]> rakudo​: class Foo does Positional[​::T] {}; say Foo ~~
Positional
15​:39 <+p6eval> rakudo ce15be​: OUTPUT«===SORRY!===␤Could not locate
  compile-time value for symbol T␤»
15​:39 < jnthn> That looks correct.
15​:39 < jnthn> It needs to know what T is there to use it as a type
argument at
  compile time.
15​:39 < [Coke]> T isn't some magic thing that is a type?
15​:39 < [Coke]> ah, yes, but not magical. OK.

So, Closable with tests.

--
Will "Coke" Coleda

From @moritz

now tested in S14-roles/parameterized-type.t.

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