[
https://issues.apache.org/jira/browse/SPARK-15086?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15325084#comment-15325084
]
Sean Owen commented on SPARK-15086:
-----------------------------------
Good news this might not be an issue:
- Deprecate intAccumulator but leave it as-is
- Deprecate the extra method like accumulator(int) in the JavaSparkContext API;
not sure why they are there
- Deprecate longAccumulator / doubleAccumulator because in the Java API, the
current versions all take an initial value unlike Scala
- Deprecate the existing other accumulator() methods just like the Scala API
and then
- Keep the Java API for Accumulators -- OK that's obvious, even if it's mostly
a pass-through wrapper now
- Add longAccumulator / doubleAccumulator for consistency with Scala, with new
API; no conflict because these don't take an initial value
- Add listAccumulator
The alternative to these last points is just to have people use the Scala API,
since I think there's no longer a primitive vs object problem (? haven't
thought that through) but maybe fine to just keep the extra redundancy.
Let me investigate whether this is actually, well, true.
> Update Java API once the Scala one is finalized
> -----------------------------------------------
>
> Key: SPARK-15086
> URL: https://issues.apache.org/jira/browse/SPARK-15086
> Project: Spark
> Issue Type: Sub-task
> Components: Spark Core
> Reporter: Reynold Xin
> Priority: Blocker
>
> We should make sure we update the Java API once the Scala one is finalized.
> This includes adding the equivalent API in Java as well as deprecating the
> old ones.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscr...@spark.apache.org
For additional commands, e-mail: issues-h...@spark.apache.org
