On 10/06/2010 07:38 PM, Tatu Saloranta wrote:
> On Wed, Oct 6, 2010 at 9:56 AM, Tim McCune<tim_at_mccune.name> wrote:
>
>> Hmm, I haven't made any changes to my application to enable Jackson. Jersey
>> just started using Jackson when I upgraded from 1.1.5.1 to 1.4. So maybe
>> there is some unintended side-effect of the way that I've always used Jersey
>> that is enabling Jackson? At any rate, with Jackson doing the marshalling,
>> I'd still expect it to not output the "null"s, which it seems to be doing
>> when it's turned on in Jersey, regardless of whether or not it's actually
>> supposed to be turned on. :)
>>
> It really depends on who is enabling support -- Jackson has no
> knowledge of Jersey, so it uses its defaults unless explicitly told
> otherwise.
> This behavior is not mandated by JAX-RS, either; so both approaches
> are equally valid from general viewpoint. I agree in that for
> backwards compatibility changes in behavior do not make sense.
>
As i wrote in the other post, Jersey has not changed the default.
It is either a bug in Jersey or a misconfiguration on the application side.
In the previous post, i have also attached a test case application showing
that in the default case, the null values are not written out.
My guess is Tim wants to utilize Jackson for non-JAXB classes
and JAXB otherwise, and this scenario is indeed not supported
at the moment in Jersey. But it is definitely a change in the default
behavior.
Tim, could you please confirm?
Thanks,
~Jakub
> -+ Tatu +-
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe_at_jersey.dev.java.net
> For additional commands, e-mail: users-help_at_jersey.dev.java.net
>
>
>