[
https://issues.apache.org/jira/browse/FLINK-7442?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16133184#comment-16133184
]
ASF GitHub Bot commented on FLINK-7442:
---------------------------------------
Github user StephanEwen commented on the issue:
https://github.com/apache/flink/pull/4564
Thanks, this looks better to me!
I would suggest to not pass the configuration into the library cache
manager.
I think passing configuration objects into specialized / dedicated
components is an anti-pattern. It makes testing complicated, signatures
inexplicit, etc.
For the second task of the classloader, loading resources: Do we have a
test that validates the resource resolution? I think this passes the tests of
the previous pull request because it behaves for resources like a *parent-first
classloader*, which seems inconsistent. Admittedly the existing implementation
of the child-first classloader was for resources a child-only classloader,
which was also not correct.
> Add option for using a child-first classloader for loading user code
> --------------------------------------------------------------------
>
> Key: FLINK-7442
> URL: https://issues.apache.org/jira/browse/FLINK-7442
> Project: Flink
> Issue Type: Improvement
> Components: Local Runtime
> Reporter: Aljoscha Krettek
> Assignee: Aljoscha Krettek
>
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
