Sure I will answer them. Thanks for help in identifying this On Mon, Sep 15, 2014 at 10:18 AM, Lance Java <lance.j...@googlemail.com> wrote:
> I've created an issue here: > https://github.com/uklance/tapestry-cometd/issues/62 > > Not sure of an eta yet. You could help by answering the questions in the > issue. > On 14 Sep 2014 23:27, "Boris Horvat" <horvat.z.bo...@gmail.com> wrote: > > > Brilliant at least we figured out the issue here :) > > > > Any eta on the fix? I would be happy to test the fix :) > > > > On Sun, Sep 14, 2014 at 11:49 PM, Lance Java <lance.j...@googlemail.com> > > wrote: > > > > > Yes, looks like a bug. I assumed the ArrayEventContext would it's > > > TypeCoercer reference to do the type coercion. Seems I was wrong. > > > > > > The workaround for now is to convert from string yourself. > > > On 14 Sep 2014 22:17, "Boris Horvat" <horvat.z.bo...@gmail.com> > wrote: > > > > > > > You are right, the implementation is different between the two cases. > > > > > > > > - eventContext = > > > > {org.apache.tapestry5.internal.*URLEventContext*@10311}"EventContext: > > > > null" > > > > - valueEncoder = {$ContextValueEncoder_9660c5cdf431@10618}"" > > > > - values = {java.lang.String[1]@10619} > > > > - values = null > > > > > > > > > > > > - eventContext = {org.apache.tapestry5.internal.services. > > > > *ArrayEventContext*@10795}"EventContext: null" > > > > - typeCoercer = {$TypeCoercer_9660c5cdf3af@10797}"" > > > > - values = {java.lang.Object[1]@10798} > > > > - values = null > > > > > > > > So this is a bug? > > > > > > > > Thanks :) > > > > > > > > > > > > On Sun, Sep 14, 2014 at 11:08 PM, Lance Java < > > lance.j...@googlemail.com> > > > > wrote: > > > > > > > > > Hmm... Perhaps this is a bug in the way tapestry-cometd fires the > > > > > onActivate. > > > > > > > > > > Can you fire up a debugger and inspect the EventContext object > passed > > > to > > > > > onActivate() in > > > > > 1. the normal page render > > > > > 2. the tapestry-cometd component > > > > > > > > > > I think there's something preventing type coercion (which passes > > > through > > > > to > > > > > ValueEncoder). I'm guessing it's a different EventContext > > > implementation > > > > in > > > > > the 2 cases. > > > > > > > > > > > > > > > > > > > > > -- > > > > Sincerely > > > > *Boris Horvat* > > > > > > > > > > > > > > > -- > > Sincerely > > *Boris Horvat* > > > -- Sincerely *Boris Horvat*
