Bug #71640 for Moose: Moose->init_meta may fail to reinitialize properly in some cases

Wed Oct 12 17:39:46 2011 timk [...] jtse.com - Ticket created

Observed via code inspection, I believe the following is a typo, in Moose.pm line 165 ff: unless ( $meta->isa("Moose::Meta::Class") ) { my $error_message = "$class already has a metaclass, but it does not inherit $metaclass ($meta)."; I believe the condition in the first line ought to read: $meta->isa($metaclass). Usually won't cause a problem, since usually $metaclass eq "Moose::Meta::Class". The failure mode: If the class already has a $meta that does not inherit from the new $metaclass, the current code will happily ignore the new $metaclass. For example, if My::Meta and My::Meta2 both extend Moose::Meta::Class, and if I create a class with My::Meta, then call init_meta with metaclass => 'My::Meta2', init_meta will silently ignore my request to initialize the metaclass.

Fri Oct 14 17:12:21 2011 DROLSKY [...] cpan.org - Correspondence added

On Wed Oct 12 17:39:46 2011, timk wrote: Show quoted text

> Observed via code inspection, I believe the following is a typo, in > Moose.pm line 165 ff: > > unless ( $meta->isa("Moose::Meta::Class") ) { > my $error_message = "$class already has a metaclass, but > it does not inherit > $metaclass ($meta)."; > > I believe the condition in the first line ought to read: $meta-

> >isa($metaclass). Usually won't

> cause a problem, since usually $metaclass eq "Moose::Meta::Class".

If you could write a test for this, that'd be fantastic.

Fri Oct 14 17:12:23 2011 The RT System itself - Status changed from 'new' to 'open'

Fri Oct 14 17:16:29 2011 DROLSKY [...] cpan.org - Subject changed from (no value) to 'Moose->init_meta may fail to reinitialize properly in some cases'

Sun Sep 11 12:44:29 2016 DROLSKY [...] cpan.org - Status changed from 'open' to 'stalled'