Hi,
Good find. The test is indeed entirely correct. But I'd like to add that
spec 329 is new since JSF 2.3 too and that this specific case is a rather
unique JSF component implementation (the <h:selectOneRadio group>) which I
don't expect to see elsewhere in the world. I'd say that the implementation
needs at least to be adjusted in order to support the new context param
too. I will take this job on me.
Cheers, B
On Wed, Dec 21, 2016 at 4:42 AM, Edward Burns <edward.burns_at_oracle.com>
wrote:
> >>>>> On Tue, 6 Dec 2016 15:43:27 -0500, Neil Griffin <
> neil.griffin_at_portletfaces.org> said:
> NG> If JAVASERVERFACES_SPEC_PUBLIC-1433 is adopted
>
> >>>>> On Tue, 20 Dec 2016 14:51:05 -0800, Edward Burns <
> edward.burns_at_oracle.com> said:
>
> EB> I've created a feature branch with the above name and the fix is in
> EB> there. Please try it out and let me know what you think by COB
> EB> Thursday.
>
> Turns out I found some evidence to support not turning this on by
> default. When I did, I found that the test Bauke wrote for
> JAVASERVERFACES_SPEC_PUBLIC-329 started failing because it asserted that
> the required validator message only appeared once, but with this on it
> appears three times.
>
> This illustrates a general practice we adhere to with maintaining JSF.
>
> If a change breaks an existing test we err on the side of not changing
> the test, assuming the test is correct. Now, sure, it's possible to
> rewrite the test so that it passes, but in this case Bauke had included
> this comment:
>
> // It should appear only once!
>
> That's a pretty strong indicator that the test should not be changed.
>
> So, I'm going to isolate the test for 1433-UIInputRequiredTrue into its
> own webapp, so the context-param will not interfere.
>
> Ed
>
> --
> | edward.burns_at_oracle.com | office: +1 407 458 0017
> | 13 business days until planned start of JSF 2.3 Public Review
> | 33 business days until DevNexus 2017
> | 58 business days until JavaLand 2017
>