Relational-to-OOP Tax (was: Critique of Robert C. Martin's...)


Patrick May wrote:
"topmind" <topmind@xxxxxxxxxxxxxxxx> writes:
Patrick May wrote:
Let me reword it: Nobody can/will prove it objectively worse.

That's not a rewording, that's an attempt to squirm out of your
burden of proof.

It would be like asking somebody to prove that Santa does not exist.
It is not possible (unless simultaneous video-cams are put on every
spot).

If you can't prove it, retract it and stop making such claims.

I am not implying that something is better. The OO book authors are,
and it is their burden to prove it or apologize and then shut up.


Nobody HAS SO FAR proven that p/r is objectively worse. Why you
expect people to take your word for it and give up their favorite
tools for 60's navigational crap with a new name, I have no idea.

You are the one infesting a forum for the discussion of object
technology with your unsupported claims that object orientation is
flawed.

No, my stance is that people keep implying it is objectively better,
that heavily wrapping the database is always "good", that polymorphism
over IF statements is always "good", that procedural is "messy", etc.
THE BURDEN IS ON YOU GUYS. I think it is largely a subjectively
personal preference. However, translating between paradigms will cost
you. See below.

The burden of proof is on you to prove your claims. The only
participants in this newsgroup who are claiming that one paradigm
suffices for all domains are the relational one trick ponies.

I never claimed that relational was objectively better, only that it
is not objectively worse.

However, it logically follows that if you are using a relational
database, then designing the application close to it rather than
trying to wrap it away in OO classes and conventions will simplify the
design because one is not translating from one paradigm to another.
That is not is not a claim of relational betterment, but rather of
relational *fit*. I don't have solid math to prove such, but it is a
reasonable and rational argument.

The given is that a relational database is being used. Martin didn't
propose using an OODBMS.

Do you think translating between paradigms is cost-free? I expect an
answer. If not, then you would agree that not doing it saves code and
time. Thus, if one goes away from it, then OO must offer some benefit
to *counter* the translation costs. In that case, you are implying
some objective betterment. Otherwise you wouldn't translate.


None of
the participants who are discussing OO from a position of knowledge
are making the claims you seem to wish they were making.

Is the scientific spirit truly dead in software enginering? Are we
back to the days of Alchemy? We are, and people like you seem to
like it that way.

The hypocrisy, it burns!

It's posts like this that swing me from the opinion that you're
simply deluded and inexperienced to the view that you are deliberately
trolling with malice aforethought. You consistently make sweeping
claims with absolutely no support and, when pressed, retreat to "It's
just my opinion, you shouldn't treat it like a real claim."

Well, that is not me. You are characterizing somebody else.

You
squirm and twist to avoid the burden of proof that clearly accrues to
your baseless assertions. You demand code, but refuse (or are unable)
to provide your "superior" version for review.

MY baseless assumptions? The topic is Martin's baseless assumptions.
Can you read titles?


Demonstrate a little intellectual honesty and integrity yourself
before casting aspersions on others.

The problem is you, not me. Look at all the OO books on the shelf that
bash relational and procedural. I don't see the opposite. If you do,
you are fricken delusional.


Sincerely,

Patrick


-T-

.