No, CDI-1.0 was specified to also work in SE already. See the already mentioned
last paragraph in chapter 1. Remember that back in those times there was a
political pressure to not market this feature too much ;)
We also discussed whether we need to change the wording back in CDI-1.1. But
afair
FYI it seems it's more an arquillian-weld-ee-embedded-1.1 issue.
See also https://issues.jboss.org/browse/WELD-2065
Martin
Dne 19.11.2015 v 08:54 Martin Kouba napsal(a):
Ok...
1. Up to 1.2 CDI itself is still an EE technology, and it's usage in
Java SE is completely undefined
2. Why should CD
Ok...
1. Up to 1.2 CDI itself is still an EE technology, and it's usage in
Java SE is completely undefined
2. Why should CDI in Java SE care about EE resources? For the same
reason an application is using (declares a dependency on) EE resources
in Java SE.
So I still think it is correct to f
Martin, it seems you missed that the whole Resource section in the CDI spec
ONLY applies if you run in an EE environment.
This paragraph btw also mentions EJBs as a ‚resource‘ („For a remote EJB, @EJB
must be specified…“). But I assume that you will agree with me that the CDI
container shall N
Well, I wouldn't say it's a Weld regression, more like a change in
behaviour which follows the way CDI handles unsatisfied dependencies. On
the other hand the error message should be definitely more clear (NPE is
not really user-friendly ;-).
Martin
Dne 15.11.2015 v 21:48 John D. Ament napsal
hi john,
since it's a regression in weld, you should be able to upgrade ds
(the resource-injection-point was just moved and not added).
regards,
gerhard
2015-11-15 21:48 GMT+01:00 John D. Ament :
> Not really feasible to upgrade right now, due to regression issues in the
> data module (previo
Not really feasible to upgrade right now, due to regression issues in the
data module (previously discussed on list).
However, your tip was useful. Unfortunately, it is a weld regression
problem. Previously, these resource injected fields were just left as null
(which is more appropriate from my
short addition:
please upgrade to the latest snapshot (the next release is due quite soon).
regards,
gerhard
2015-11-15 21:22 GMT+01:00 Gerhard Petracek :
> hi john,
>
> we just moved the resource-injection-point one "level" back to support
> atomikos (see ManagedUserTransactionResolver).
> th
hi john,
we just moved the resource-injection-point one "level" back to support
atomikos (see ManagedUserTransactionResolver).
the first jndi lookup afterwards (which is used as a fallback), is already
what we had before.
regards,
gerhard
2015-11-15 18:29 GMT+01:00 John D. Ament :
> All,
>
>
All,
I just tried upgrading my app locally to Weld 2.3.1 (in prep for a wildfly
10 upgrade). After upgrading Weld to 2.3.1, some of my repository tests
started failing with the following output.
java.lang.NullPointerException
at org.jboss.weld.util.Preconditions.checkNotNull(Preconditions.java:
10 matches
Mail list logo