Validate if the @AggregateIdentifier field implements toString() function

When instantiating the meta-model of the aggregate, more specifically on retrieving the @AggregateIdentifier annotated field, we should verify that the annotated parameter implements a more precise toString() function than the default Object#toString() function.

This enforces users to some extent to ensure that the identifier they created returns a reasonable String version of their identifier.

[smcvb]

I would consider both to be fine, @vab2048. Axon is not biased to the exact format of the toString() result, as long as it isn’t the Object#toString() method which is being used. Using an UUID is definitely recommended; appending the type might be useful if you would want to reuse the UUID between distinct aggregate/entities.

So:

Does it matter which one is used?

No it does not, both would be fine.

[vab2048]

I have come across this exception when testing my aggregates.

What exactly would constitute good practice for the implementation of toString?

Consider the following two:

1. PortfolioId(id=bfb21d6e-7587-4dc4-8dbe-3ad4f3d758b5)
2. bfb21d6e-7587-4dc4-8dbe-3ad4f3d758b5

The first is a default toString as would be implemented by lombok. The second a custom toString which just returns the id.

Does it matter which one is used?