[
https://issues.apache.org/jira/browse/FELIX-5549?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15888472#comment-15888472
]
Alex Soto edited comment on FELIX-5549 at 2/28/17 6:35 PM:
-----------------------------------------------------------
Agree is not perfect right now, but, can the OSGi people (committee members)
fix the specification in the future? In other words, is there some underline
reason for this behavior, or is is an overlook? I believe the need to
programmatically instantiate (and destroy) DS components is valid, but at the
same time, why give up dependency injection (of both configuration and other
components)?
was (Author: alex.soto):
Agree is not perfect right now, but, can the OSGi people fix it in the future?
In other words, is there some underline reason for this behavior, or is is an
overlook? I believe the need to programmatically instantiate and destroy DS
components is valid, but at the same time, why give up dependency injection of
dependency components and configuration?
> Factory component fails to reactivate after config changes
> ----------------------------------------------------------
>
> Key: FELIX-5549
> URL: https://issues.apache.org/jira/browse/FELIX-5549
> Project: Felix
> Issue Type: Bug
> Components: Declarative Services (SCR)
> Environment: Karaf 4.0.8
> Reporter: Alex Soto
>
> A factory component fails to reactive after the configuration changes.
> Initially, the component initializes normally. After it is in Active state,
> the configuration referenced by the component _configurationPid_ changes,
> which causes the component to not activate again.
> A minimal application demonstrating this behavior is available here:
> https://github.com/lexsoto/blueprint-ds-config-reload
--
This message was sent by Atlassian JIRA
(v6.3.15#6346)
