[
https://issues.apache.org/jira/browse/SPARK-17679?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Sean Owen updated SPARK-17679:
------------------------------
Assignee: Jason White
> Remove unnecessary Py4J ListConverter patch
> -------------------------------------------
>
> Key: SPARK-17679
> URL: https://issues.apache.org/jira/browse/SPARK-17679
> Project: Spark
> Issue Type: Bug
> Components: PySpark
> Affects Versions: 2.0.0
> Reporter: Jason White
> Assignee: Jason White
> Priority: Minor
> Fix For: 2.1.0
>
>
> In SPARK-6949 davies documented a couple of bugs with Py4J that prevented
> Spark from registering a converter for date and datetime objects. Patched in
> https://github.com/apache/spark/pull/5570.
> Specifically https://github.com/bartdag/py4j/issues/160 dealt with
> ListConverter automatically converting bytearrays into ArrayList instead of
> leaving it alone.
> Py4J #160 has since been fixed in Py4J, since the 0.9 release a couple of
> months after Spark #5570. According to spark-core's pom.xml, we're using
> 0.10.3.
> We should remove this patch on ListConverter since the upstream package no
> longer has this issue.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
