Re: Package hierarchy. Extensibility.

[Niklas Holsti <niklas.holsti@tidorum.invalid>]

Jarno Neniu wrote:
> Hello. I'm not able to decide about which design is the best
> (extensibility, maintainability and recompilation friendliness are
> important), so I'm demanding for any humble and not so humble
> opinions. As an example of hierarchy consider representation of
> geometrical points. One of representations is preferable and
> considered to be the main representation (Cartesian 2D and 3D points),
> but there is unspecified amount of other representations, which could
> be added to the system later (points in polar, spherical, cylindrical
> coordinate systems). Conversion between representations should be
> somehow available and it is desirable for representations to be
> generic. Which is the best way to design the hierarchy? Should it be
> something like this:
> 
> package Points is
>    --  Preferable representation. Cartesian point.
>    type Point is ...
> end Points;
> 
> package Points.Spherical is
>    type Spherical_Point is ...
> end Points.Spherical;
> 
> package Points.Cylindrical is
>    type Cylindrical_Point is ...
> end Points.Cylindrical;
> 
> Or should it be dummy package �Points� to which other representations
> are connected:
> 
> package Points is
> end Points;
> 
> package Points.Cartesian is
>    type Cartesian_Point is ...
> end Points.Cartesian;
> 
> package Points.Spherical is
>    type Spherical_Point is ...
> end Points.Spherical;
> 
> Something else?

An interface type that defines the common view of "Points" of 
whatever representation? That would allow non-hierarchical and 
independent implementations of the representations as long as every 
representation implements this interface.


-- 
Niklas Holsti
Tidorum Ltd
niklas holsti tidorum fi
       .      @       .