On Thu, 20 Mar 2003, Carsten Ziegeler wrote:
>
> Stephan Michels wrote:
> > Hmm, I'll think that I'm not the only one, who went into the trap. You're
> > right returning a null validity hurts the contract, this will cause an
> > exception. But a NPE in isValid() doesn't help searching the problem
Stephan Michels wrote:
> Hmm, I'll think that I'm not the only one, who went into the trap. You're
> right returning a null validity hurts the contract, this will cause an
> exception. But a NPE in isValid() doesn't help searching the problem.
>
Yes, but adding it in add().
Carsten
On Thu, 20 Mar 2003, Carsten Ziegeler wrote:
> Hi Stephan,
>
> you don't have to ask for changes (if they make sense). I don't
> own the code :)
I want to be sure ;-)
> So, making the code more understandable is good.
>
> > From: Stephan Michels [mailto:[EMAIL PROTECTED]
> >
> > can I move the
Hi Stephan,
you don't have to ask for changes (if they make sense). I don't
own the code :)
So, making the code more understandable is good.
> From: Stephan Michels [mailto:[EMAIL PROTECTED]
>
> can I move the part where you generate the validity objects
> into a separate method, like generateVa
Hi Carsten,
can I move the part where you generate the validity objects
into a separate method, like generateValidities() ?
I whould like to enhance the profiler components to measure
the setup() time.
Thanks.
BTW, can you add a check for null in AbstractAggregatedValidity.add().
I got sometimes
