Re: A Growing Dislike for SQL-OOP Mappers

[jk2addict]

All very good information. I'm glad we don't all agree here. That's means thought is going on. The one thing I've realized is that my gripe about current sql->oop practices in perl is that the current popular mappers assume the stance that the data access object is the same as my public object/interface. That may be true in some cases; not true or desirable in others....and both sides are OK.

[herveus]

Howdy!

Interesting...

I had to write a term paper on an "advanced Java server technology" for a Java programming course. I chose to look at Object-Relational Mapping, and focused on something called SimpleORM.

Some ORMs try to make persistent objects indistinguishable from non-persistent objects, while others force you to directly acknowledge the difference in the code. SimpleORM did the latter (although you could then put a wrapper/facade around DAOs to hide that down the line. Now, one reason I liked SimpleORM was that it doesn't use XML anywhere, and doesn't try to hide the fact that some classes are different because they have a backing store for persistence.