O/R mismatch: What is the Problem?
August 28, 2007
There has been quite a discussion recently on the so called “O/R mismatch”.
This is a quite interesting discussion. The bottom line is that after so many years, still object persistence does not seem to have a fully adequate solution.
This is ackward, bringing programming languages and databases seems still a rather diffcult task…!
There are a number of interesting resources I have recently published on this subject on ODBMS.ORG.
In cooperation with FranklinsNet, ODBMS.ORG has published the transcript of the panel discussion “ORM Smackdown” between Ted Neward and Oren “Ayende” Eini on different viewpoints on Object-Relational Mapping (ORM) systems.
It is an interesting reading. Pls check: ORM Smackdown
I have also published Ted Neward’s follow on essay discussing solutions to the problems
of Object/Relational-Mapping titled “Avoiding the Quagmire”.
This new essay is a follow on to Neward’s “The Vietnam of Computer Science” , which compared
the inherent problems of object/relational mapping to the quagmire in the Vietnam war.
The initial “Vietnam” essay was first published in 2006 and widely discussed in the industry.
“Avoiding the Quagmire” discusses the impact of choosing to integrate object concepts into the database as opposed to using relational concepts or object/relational mappers.
Neward states that while using an object oriented database management system (ODBMS) will not completely eliminate all of the problems described in the intial “Vietnam” essay, it does address some of the more egregious problems. ODBMS thus frequently provide the developer a better chance of avoiding the quagmire and allowing them to focus more clearly on the problem at hand.
Pls check: Avoiding the Quagmire
I published a copy of Ted Neward’s “The Vietnam of Computer Science”.
Neward argues that the O/R mismatch is a quagmire where current approaches including object-relational mappers (ORMs) are subject to decreasing marginal returns. He lists the abandonment of objects (as a programming paradigm) or of relational data structures (as a database paradigm) as the only wholehearted solutions, while living with the pain or full integration of ORMs into languages or databases are other approaches.
I personally do not like the analogy with Vietnam… but the article has a number of interesting points. The article as you may immagine has received a mix feedback from the readers….
Here is the reference: The Vietnam of Computer Science