Abilities

 view release on metacpan or  search on metacpan

README  view on Meta::CPAN

    actions, values will be 1 (for yes/no actions) or a single-item
    array-ref with the name of a constraint (for constrained actions).

UPGRADING FROM v0.2
    Up to version 0.2, "Abilities" required the "roles" and "actions"
    attributes to return objects. While this made it easier to calculate
    abilities, it made this system a bit less flexible.

    In version 0.3, "Abilities" changed the requirement such that both these
    attributes need to return strings (the names of the roles/actions). If
    your implementation has granted roles and actions stored in a database
    by names, this made life a bit easier for you. On other implementations,
    however, this has the potential of requiring you to write a bit more
    code. If that is the case, I apologize, but keep in mind that you can
    still store granted roles and actions any way you want in a database
    (either by names or by references), just as long as you correctly
    provide "roles" and "actions".

    Unfortunately, in both versions 0.3 and 0.4, I made a bit of a mess that
    rendered both versions unusable. While I documented the "roles"
    attribute as requiring role names instead of role objects, the actual
    implementation still required role objects. This has now been fixed, but
    it also meant I had to add a new requirement: consuming classes now have
    to provide a method called "get_role()" that takes the name of a role
    and returns its object. This will probably means loading the role from a
    database and blessing it into your role class that also consumes this
    module.

    I apologize for any inconvenience this might have caused.

AUTHOR
    Ido Perlmuter, "<ido at ido50 dot net>"

BUGS
    Please report any bugs or feature requests to "bug-abilities at
    rt.cpan.org", or through the web interface at

lib/Abilities.pm  view on Meta::CPAN

}

=head1 UPGRADING FROM v0.2

Up to version 0.2, C<Abilities> required the C<roles> and C<actions>
attributes to return objects. While this made it easier to calculate
abilities, it made this system a bit less flexible.

In version 0.3, C<Abilities> changed the requirement such that both these
attributes need to return strings (the names of the roles/actions). If your implementation
has granted roles and actions stored in a database by names, this made life a bit easier
for you. On other implementations, however, this has the potential of
requiring you to write a bit more code. If that is the case, I apologize,
but keep in mind that you can still store granted roles and actions
any way you want in a database (either by names or by references), just
as long as you correctly provide C<roles> and C<actions>.

Unfortunately, in both versions 0.3 and 0.4, I made a bit of a mess
that rendered both versions unusable. While I documented the C<roles>
attribute as requiring role names instead of role objects, the actual
implementation still required role objects. This has now been fixed,
but it also meant I had to add a new requirement: consuming classes
now have to provide a method called C<get_role()> that takes the name
of a role and returns its object. This will probably means loading the
role from a database and blessing it into your role class that also consumes
this module.

I apologize for any inconvenience this might have caused.

=head1 AUTHOR

Ido Perlmuter, C<< <ido at ido50 dot net> >>

=head1 BUGS

lib/Abilities/Features.pm  view on Meta::CPAN

}

=head1 UPGRADING FROM v0.2

Up to version 0.2, C<Abilities::Features> required the C<plans> and C<features>
attributes to return objects. While this made it easier to calculate
available features, it made this system a bit less flexible.

In version 0.3, C<Abilities::Features> changed the requirement such that both these
attributes need to return strings (the names of the plans/features). If your implementation
has granted plans and features stored in a database by names, this made life a bit easier
for you. On other implementations, however, this has the potential of
requiring you to write a bit more code. If that is the case, I apologize,
but keep in mind that you can still store granted plans and features
any way you want in a database (either by names or by references), just
as long as you correctly provide C<plans> and C<features>.

Unfortunately, in both versions 0.3 and 0.4, I made a bit of a mess
that rendered both versions unusable. While I documented the C<plans>
attribute as requiring plan names instead of plan objects, the actual
implementation still required plan objects. This has now been fixed,
but it also meant I had to add a new requirement: consuming classes
now have to provide a method called C<get_plan()> that takes the name
of a plan and returns its object. This will probably means loading the
plan from a database and blessing it into your plan class that also consumes
this module.

I apologize for any inconvenience this might have caused.

=head1 AUTHOR

Ido Perlmuter, C<< <ido at ido50 dot net> >>

=head1 BUGS

 view all matches for this distribution
 view release on metacpan -  search on metacpan