[
https://issues.apache.org/jira/browse/FLINK-2779?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14941160#comment-14941160
]
ASF GitHub Bot commented on FLINK-2779:
---------------------------------------
Github user senorcarbone commented on the pull request:
https://github.com/apache/flink/pull/1208#issuecomment-145028978
From a quick read on the documentation (and prior knowledge from google
dataflow) it is easy to get a full picture of the new semantics. Even though I
like it from a engineering perspective, It needs to be pointed out though that
the addition of the *Assigner* restricts our previous windowing semantics to
periodic windows, i.e. at the moment you receive a record you have to know all
the possible buckets this record will fall into while before we did not have
such a constraint. In other words we no longer support dynamic windows where
range and slide can change dynamically.
> Update documentation to reflect new Stream/Window API
> -----------------------------------------------------
>
> Key: FLINK-2779
> URL: https://issues.apache.org/jira/browse/FLINK-2779
> Project: Flink
> Issue Type: Sub-task
> Components: Streaming
> Reporter: Aljoscha Krettek
> Assignee: Kostas Tzoumas
> Fix For: 0.10
>
>
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
