[
https://issues.apache.org/jira/browse/BEAM-111?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15689507#comment-15689507
]
Amit Sela commented on BEAM-111:
--------------------------------
Seems like {{org.apache.beam.sdk.io.hdfs.WritableCoder}} is (almost) identical
to {{org.apache.beam.runners.spark.coders.WritableCoder}}. Both authored by Tom
White.
I'll go ahead and change this ticket to remove
{{org.apache.beam.runners.spark.coders.WritableCoder}} and tag with {{starter}}.
> Use SDK implementation of WritableCoder
> ---------------------------------------
>
> Key: BEAM-111
> URL: https://issues.apache.org/jira/browse/BEAM-111
> Project: Beam
> Issue Type: Improvement
> Components: runner-spark
> Reporter: Amit Sela
>
> The Spark runner currently uses it's own implementation of WritableCoder,
> should use the one in contrib/hadoop
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
