Alpha2 Metadata Error on Start

[Tom Zeller]

On Mon, Aug 4, 2014 at 8:54 AM, Cantor, Scott <cantor.2 at osu.edu> wrote:
> On 8/4/14, 9:51 AM, "Rod Widdowson" <rdw at steadingsoftware.com> wrote:
>
>>> Note that on the call we alluded to the fact that we probably should
>>> eliminate all the implicit calls to setId, because these keep biting
>>>us.
>>
>>Do you mean the bean post processor thing?  That has always made me
>>nervous.
>
> Mainly using the code, but I agree that we're living dangerously with the
> post processor. That's ok as an airbag of sorts, but I think it should
> warn loudly so we can find cases we have to fix. Or as I say, stop
> requiring it.
>
>>I'd agree with that too.  As a mild 'on the other hand' they are handy for
>>logging, but we should get that fixed by fixing logging, not by tying
>>ourselves in knots with overly restrictive practices.
>
> Yes. Helpful is one thing, causing crazy bugs every other week is not.

I'm not sure how to make the IdentifiableBeanPostProcessor warn as
suggested above, meaning to identify cases we need to fix. I thought
the point of it was to allow for brevity of Spring wiring, for
example, <bean id="foo" /> instead of <bean id="foo" p:ID="foo" />,
especially for non-Spring aware beans (i.e. OpenSAML) for which the
BeanNameAware interface is not available.

Since IdentifiedComponent.getId() is @Nullable, or even empty,
removing the @Nonnull @NotEmpty constraint from
IdentifiableComponent.setId() seems consistent, but not that great
either.