Re: Objects and Relations

[Snipped]
> >> I am inclined to regard the data as the testimony of the users about the
> > >real world.
> > I understand your point. If you don't mind, I propose to replace
> > *testimony* by *description* a term that more domain neutral and that
> > seems safer candidate for generalization of reasonning.
>
> I am afraid I still prefer the concept of testimony. We are used to the
> idea of people giving testimony; we know it is a description of *their*
> perception, not a description of reality. Testimony can be recorded
> faithfully without conceding that it is accurate. Since a database system
> cannot warrant accuracy (only, at most, consistency) we should remind
> ourselves that what we find in a database is only provisionally true.
I see your point. I have the impression you use *testimony* because it somehow implies the notion of responsibility for accurateness of description. This is quite interesting. For the moment, I think you convinced me on this point, even though I am afraid that such direction may lead to some form of deception because I do not believe in such thing at collective level but solely at individual level. Neverthless, I thank you for triggering some further thought.

> >> The keys that exist about which the users can say things are
> >> ultimately synthetic as you say, but they are pre-existing in the real
> >> world.
> > I am curious about that last sentence. Would you mind telling me
> > which *pre existing keys* are you thinking of ? And most of all what
> >constitutes the criteria which makes it pre existing ?
>
> As I went on to say in the next paragraph, a pre-existing key is a key whose
> value is known before the application software invents one. It already
> exists with respect to some enterprise of interest. Examples would be your
> name or mine. Those are synthetic because our parents invented them, but to
> our employers they are pre-existing. My credit card number is synthetic
> because the credit card company invented it, but to Amazon it is
> pre-existing. I took this to be precisely what you meant when you said all
> keys are surrogates.
I understand now what you are getting at by *pre-existence*. It seems you are using a relative temporal referential with each new key generation cycle.

> >>The surrogates that we generate in the software during the process
> >> of recording the testimony are surrogates that spring into existence
> >> *after*
> >> the user assesses the real world and forms his testimony. Therefore they
> >> cannot have anything to do with the real world. They *are* different, if
> >> no
> >> more synthetic, and they are not strictly necessary--merely convenient.
> > I understand. Thank you for this description.
> >>> In my perspective,
> >>> a natural key is nothing but a socially accepted surrogate key.
> >> I agree, but a basic reason it is capable of being socially accepted is
> >> that
> >> it is pre-existing in the real world of interest.
> > Mmm...I am not convinced that the acceptance necessarily implies
> > preexistence of some natural key. For instance, practicality of use
> > and/or familiarity are sufficient self sustaining causes for
> > explaining acceptance.
>
> I'm not sure I was making a useful point so let's park that discussion.
Well, you obviously were but not directly. You made me realize that there maybe a definition of keys that is proper and relative to some development cycle. It is a different valuable perspective from the fundamental approach that analyzes the process of natural keys *independently* from any development cycle. Such distinction is useful and may help characterize both comparatively.

[Snipped]

> Given my explanation of what I understand by the term "natural key" above,
> would you be willing to consider that precedence can indeed be established?
I believed this was obvious when I asked you to ellaborate.

My first conclusion over this exchange is that the the framework of analysis is either development cycle-dependent or development cycleindependent  and that such distinction is what basically defines and characterizes surrogate, natural and the relationship of precedence between the two identifiers.

It is good to know that once in a blue moon, this NG teaches me something useful and that my time is not completely wasted here. Thank you. Received on Thu Feb 15 2007 - 10:21:22 CET