Re^2: OO concepts and relational databases

[dragonchild]

I believe that we are talking in the same direction, but in different ways. My OP was discussing the fact that many OO developers choose to use tools that design and use schemas that are very inefficient.

I am saying the same things you are:

• Schema design should be left in the hands of those who know how to do it
• A well-designed schema is so because it's focused on how the data relates to itself, not how any given application might use the data
• Access to a well-designed schema needs to be controlled, through any number of mechanisms. These might include:
  • Views
  • Stored Procedures / Triggers
  • Some middle-tier in code, such as XML-RPC and/or other web services
• Updates to that schema should be controlled by those who designed it

Now, I don't go as far as to say that one method is preferable to another. The primary goal is to ensure the integrity, structure, and safety of the data. How that happens can almost be viewed as a matter of personal taste.

[BrowserUk]

I apologise. I lost sight of the purpose and content of your OP as the thread progressed over time.

There are one or two sub-threads where you are counter-arguing assertions made by others in which you appeared to be arguing the case for things like Class::DBI and Class::Tables etc., and condeming the use of SPs, triggers and views.

Re-reading the entire thread in sequence and context, I now realise that whilst you did argue the case against the latter, it was not in support of the former.

My accusation was unfounded, based on my memory of what I thought you had said over several posts over several days, instead of what you actually said and the context in which you said it.

For allowing my own prejudices on the subject, to cause me to confuse the leaves of thread, for the trunk of the argument. Again, I apologise.