Re: THe OverRelational Manifesto (ORM)

Bob Badour wrote:
> JOG wrote:
>
> > Alfredo Novoa wrote:
> >
> > [snip]
> >
> >>But of course there are more relationships between OO and the
> >>Relational Model.
> >>
> >>For instance:
> >>
> >>operator = method
> >>pointer = OID
> >>scalar value = object
> >>scalar variable = object
> >>scalar value = instance
> >>
> >>Etc.
> >
> > [snip]
> >
> > While I agree with virtually everything you say Alfred, I never liked
> > the "pointer=OID comparison.". Pointers are variable, and OID's are far
> > more analagous to a memory address. Of course, that doesn't diminish
> > the disadvantages of navigational databases, but the sofware engineer
> > in me always cringes slightly when I see the comparison.
>
> A pointer variable has a memory address value (or a storage location
> value). An OID variable has an OID value. I don't see how that makes the
> comparison any less valid.

As far as I know OID's are not variable - however I thankfully do not use OODBMS, so this is based only on peripheral knowledge. (If I need to persist OO code in this way I see no reason not to just boost::serialize). I did point out that this doesn't diminish the comparison anyhow Bob, it just makes the grievance more precise, which is always a good thing. Received on Wed Apr 19 2006 - 22:18:26 CEST