[
https://issues.jboss.org/browse/SEAMCRON-22?page=com.atlassian.jira.plugi...
]
Peter Royle updated SEAMCRON-22:
--------------------------------
Description: Notice in the tests since switching to Weld Arquillian and when
redeploying on JBoss AS there is a NPE during undeployment. I have been relying on
provider implementations being @ApplicationScoped which seems to work in OWB but not in
Weld. Think it's to do with the way in which I'm looking up the instances using
CdiUtils.getInstanceByType(). That's probably returning a different version to the one
returned during shutdown, so the shutdown provider's aren't properly initialised,
and the ones which are properly initialised never get shutdown. (was: Notice in the tests
since switching to Weld Arquillian and when redeploying on JBoss AS there is a NPE during
undeployment. I have been incorrectly relying on providers being @ApplicationScoped which
seems to work in OWB but not in Weld. I'm not sure what position the spec takes on
that. Will simply remote the scoping and rework so dependencies are injected into each
observer of the Extension.)
NPE stopping scheduler under Weld
---------------------------------
Key: SEAMCRON-22
URL:
https://issues.jboss.org/browse/SEAMCRON-22
Project: Seam Cron
Issue Type: Bug
Reporter: Peter Royle
Assignee: Peter Royle
Fix For: 3.0.0.Alpha1
Notice in the tests since switching to Weld Arquillian and when redeploying on JBoss AS
there is a NPE during undeployment. I have been relying on provider implementations being
@ApplicationScoped which seems to work in OWB but not in Weld. Think it's to do with
the way in which I'm looking up the instances using CdiUtils.getInstanceByType().
That's probably returning a different version to the one returned during shutdown, so
the shutdown provider's aren't properly initialised, and the ones which are
properly initialised never get shutdown.
--
This message is automatically generated by JIRA.
For more information on JIRA, see:
http://www.atlassian.com/software/jira