>>>>> On Fri, 21 Nov 2014 16:22:04 +0100, arjan tijms <arjan.tijms_at_gmail.com> said:
AT> On Fri, Nov 21, 2014 at 4:03 PM, Edward Burns <edward.burns_at_oracle.com> wrote:
EB> Ajran, while your observations are accurate, the backward compatibility
EB> issues raised by Stuart and seconded by Greg are showstoppers for this
EB> change in my opinion at this point.
AT> Well, there's the CDI mechanism to set a priority (@Priority) and
AT> provide alternatives (@Alternative) in case of multiple producers on
AT> the classpath for the same type. I recently wrote a small article
AT> about that, see
AT>
http://jdevelopment.nl/providing-alternatives-jsf-23s-injected-artifacts
AT> I agree that Stuarts concern is valid, but it might not necessarily be
AT> a showstopper. There are even additional mechanisms beyond the CDI
AT> alternatives that might be investigated, such as the deployment
AT> descriptor version check that we use in JSF as well.
I like the idea of using the descriptor version check to handle this
case.
Ed
--
| edward.burns_at_oracle.com | office: +1 407 458 0017
| 56 days til DevNexus 2015
| 66 days til JavaLand 2015
| 76 days til CONFESS 2015