Re: Java persistence


"Dan Nuttle" <d_nuttle@xxxxxxxxxxx> wrote in message
news:wStYd.1543$qf2.351@xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> Unless one of the truly world-class propeller heads can prove me wrong,
> I'm
> going out on a limb and saying that, if your class definition changes in
> your code, then you can't load an instance of an earlier version of the
> class into the JVM running the new version.

It may be possible using reflection, but the problem is ill-defined. If
for example, you have a field called "foo", and then you rename the field
"bar", is the persistence framework supposed to be able to guess that "bar"
is a renaming of "foo" and not just an entirely new field (which
coincidentally appear just when the field "foo" disappeared)?

There's no technical reason why this wouldn't be possible, only
pragmatic/business-rule reasons.

- Oliver


.

Relevant Pages

  • Re: Syntax checker?
    ... > Regardless of what the example does, my point is, how should it warn you? ... By saying "foo might be undefined at line 2351". ... complex case except for such annoyances, use it, but invoke the "Yeah, ...
    (comp.lang.ruby)
  • Re: Is linux/unix case sensitive or case insensitive?
    ... Indeed somebody was saying his linux operating ... > I think you are talking about file systems. ... all of the other applictions (e.g. foo -f and foo -F will not ...
    (comp.os.linux.misc)
  • Re: macros
    ... > the standard wasn't clear as to what happens if FOO ... But others were saying ... C99 allows empty macro arguments. ... FOO is defined to expect one argument, ...
    (comp.lang.c)
  • Re: #returning and #tap
    ... On Fri, 17 Nov 2006, Joel VanderWerf wrote: ... going to a new line and saying: ... class Foo ... #then is just a way of using arbitrary classes in this way. ...
    (comp.lang.ruby)