Re: Mixing OO and DB

On 2008-02-29, Patrick May <pjm_at_spe.com> wrote:
> Eric <eric_at_deptj.demon.co.uk> writes:
>> On 2008-02-28, Patrick May <pjm_at_spe.com> wrote:
>>> ... traversal is done by object
>>> reference rather than by some subset of object state
>>
>> So whatever you think you are saying your subconscious is
>> following pointers.
>
> I tried to be very clear with what I was saying. In the large OO
> systems that I've worked on, there was no problem with proliferation
> of finder methods in practice. Typically, once a core set of objects
> have been instantiated, access to related objects is via reference
> rather than repeated, explicit database access.

I understood your argument, I might even believe it, but I was commenting on your choice of words, which seemed to put you precisely in that category of object proponents who do not understand why database people have problems with them (either the objects or the proponents).

>
>> The trouble is you think you own the data, and also you can only
>> look at it in your own way. You do not understand how much work the
>> database could do for you, and you discourage proper database design
>> because you think you can do better at your (inappropriate) level,
>> so that the unknown future uses of the data will be made more
>> difficult.
>
> I do no such thing. When I started out in this industry, lo
> those many years ago, I was working for a CASE tool company, doing
> both internal and external development primarily based around
> relational databases. From there I moved on to consulting engagements
> with customers including Oracle, Ingres, and DEC, again all related to
> database-centric development. I know proper database design quite
> well, thank you.

This followed from putting you in a category as I said above. You may consider the remarks addressed to members of the category rather than to you personally if you like, but I am not altogether sure that you _have_ excluded yourself from the category.

>
>> You are like a man with a hammer, seeing only nails.
>
> Not at all. If you check my posting history in comp.object
> you'll find that I argue for a full toolbox, including OO, relational,
> procedural, functional, and any other techniques that prove useful in
> developing quality software.

I don't know that I am obliged to check on the poster's history before responding to a single post, but anyway I have already explained how I came to make the statement.

>
> There are those in comp.object at the moment arguing for one true
> paradigm, but those people aren't saying it should be OO.

Just because many people saying that OO shouldn't be the one true paradigm seem to come from comp.databases.theory, you can't assume that all of them think relational is the one true paradigm. There are people on both sides who don't believe in a single true paradigm - as well as some who do (who are all wrong!).

>
>> Objects are a useful tool but they are not universal, you do not
>> have to use them for everything, and you must _not_ campaign to have
>> them used for everything. Stop, and realise that there is a wider
>> picture.
>
> The same can be said for any approach, including relational. I
> know, I've said it. I suggest you take your own advice regarding the
> wider picture.

Is that addressed to me specifically, or to a category of people. If the former, perhaps you can suggest one thing that I should add to my picture.

Eric Received on Fri Feb 29 2008 - 19:56:47 CET