Re: Writing an Operating System in Ada

["Dmitry A. Kazakov" <mailbox@dmitry-kazakov.de>]

On Mon, 18 Jan 2010 18:17:20 +0100, Georg Bauhaus wrote:

> Dmitry A. Kazakov schrieb:
> 
>> There is a common ancestor for both types. What you probably had in mind
>> was a conversion is between types from independent hierarchies rooted on
>> different hosts. That normally should not be needed.
> 
> Well, I learned from the discussion that independent roots of
> blobs' types it is a quite common case.

No. It is rather like

Node A:  Numeric_Type <-- A_Integer_Type
              ||
Node B:  Numeric_Type <-- B_Integer_Type (implementation)

B_Integer_Type is an invisible ad-hoc implementation of A_Integer_Type.
Node B uses the object from A as Numeric_Type'Class known to both nodes.

If some type is unknown to B, there is no way to write a program in terms
of this type. If you had a program for some unrelated type, you could not
convert any object to this type. It would be a type error.

An ability for example to convert Float to Integer is based on their
relationship, each can be considered a subtype of another. This
relationship is given, known on each node. It is possible to introduce
ad-hoc subtypes, but then it is the programmer's responsibility to provide
conversions, because there is no way to deduce them. This is the reason why
a powerful type system is needed. You will need to describe richer
relationships between types than contemporary OOPL can. Without these
relationships the system would become unusable or else degrade to the
low-level read/write interfaces.

-- 
Regards,
Dmitry A. Kazakov
http://www.dmitry-kazakov.de