Re: Searching OO Associations with RDBMS Persistence Models
Date: 6 Jun 2006 03:29:31 -0700
I do not quite understand the following sentencesR Martin used...
"They were semantic elements that had identity"
"but they weren't database entities, and so the concept of "key" was superfluous"
How did you define *identity* on these semantic elements without a key? What other mechanism of identification have you used to coherently identify these semantic elements? Do you suppose they is a need to distinguish them or not?
> > semantic elements that had identity
Mikito Harakiri wrote:
> Robert Martin wrote:
> > It's not a matter of preference. Application objects are often complex
> > assemblies of many different tables and relationships, all of which
> > have their own keys. Which of those keys is the identity of the
> > application object?
> > For example, I once worked on a network management system in which
> > circuits were terminated by CSUs. There was a CSU table and a circuit
> > table. Logical connections between parties could use many different
> > circuits. The application objects that embodied these logical
> > connections were complex assemblies of circuits and CSUs. They were
> > semantic elements that had identity; but they weren't database
> > entities, and so the concept of "key" was superfluous.
> > Sorry, what I mean is that the database keys are not unique identifiers
> > for the logical application entities. For example the key for a
> > circuit could not be used to identify a logical connection. A logical
> > connection is a semantic entity, composed of elements from the
> > database. There is a need for these semantic entities to have
> > identities which are not just database keys.
> Could you be more specific, please? You have
> table Circuits (
> table CSVs (
> What important attributes are there? Sample data too (couple of records
> to make the matter clear would suffice).
> Then, what application object looks like?
> (Crossposted c.d.t, as I quit following c.o).
Received on Tue Jun 06 2006 - 12:29:31 CEST