I plan on readdressing the issue in the future.
To summarize here for you however, the greatest complaints about my article
was that I compared custom entities/collections to UNTYPED datasets, as
opposed to Typed Dataset. I think those are totally valid complaints,
however, they missed the point of the article. In fairness, I think I wrote
it in such a way that it was _very_ easy to miss the point. I was trying to
make people think of their domain model, that is, creating a model in their
code that represents the particular domain they are working on. I don't
care what mechanism people use to do this, but I strongly feel that _a lot_
of developers still need to THINK about their code that way. That was my
goal, to try and change how they look at programming.
There are differences between typed datasets and custom entities (and the
other solutions) I have my preference, which is obvious
But I'm a
happy guy if I got people thinking about a new dimension of their code, and
a VERY happy guy if I ever have to maintain/fix code that was built with
that viewpoint in mind - whatever specific implementation was used.
If you understand both implementations, and more important understand what
they are trying to do (which is the same thing), then you have all the
knowledge to pick the right one for your project/team!
As far as example, I actually don't know of any great example. As far as I
can remember, none of the start kits I've looked at use Typed DataSets (I'm
not saying there aren't any, but the 3 or 4 I've gone through didn't).
Karl