| Article: |
Flawed Understanding of JDO Leads to FUD | |
| Subject: | You definitely got the gist of my article | |
| Date: | 2002-05-31 12:24:51 | |
| From: | balesd | |
|
Response to: Missed point, perhaps
|
||
|
Bravo! You got it! Rather than spend a lot of time trying to find flaws in JDO, which there are really few as a Java centric technology, I was trying to point out in my article the flaw in the approach of continuing to map attributes to relational databases when object-relational technology allows us to define our classes, that is attributes and behavior, in a persistence layer.
|
||
Showing messages 1 through 2 of 2.
-
You definitely got the gist of my article
2002-05-31 12:38:11 tpherndon [View]
-
You definitely got the gist of my article
2002-05-31 12:50:17 Donald Bales |
[View]
With Oracle8i v8.1.6, Oracle's implementation of the SQL:1999 standard was mature enough to use productively with a few work-arounds. But when it came to class (type) evolution, it left a great deal of burden on the developer/DBA to handle evolution manually. With Oracle9i a mature implementation of SQL:1999 is finally available. It supports type definition, inheritance, collections more than one level deep, and makes class evolution easy to name just a few of the improvements. That's why I decided it was time to write an article that focused on shifting paradigms and moving the object model into the objectbase. We are poised on a moment of a great shift in capability and thought, we just need to let go of our old ways and embrace the future. That, as David points out in his article, is hard to do when we have a vested (financial) interest in the present.
As for the defect in my class listing, if someone needs to focus that much on something that petty, then they are grasping for straws.



I'm glad to hear I did understand correctly. What are your thoughts regarding my contention that O/R technology is, as yet, not mature enough for more than basic use? Am I correct, and to what degree? I suspect that you, as the domain expert, know fifty-dozen ways to get around the limitations (and I should, of course, read your book...), but are they "easy enough" yet? What are the flaws in the current implementations, and what would you change if you could?
Thanks for your input.