[
https://issues.apache.org/jira/browse/ARROW-1710?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16215113#comment-16215113
]
Wes McKinney commented on ARROW-1710:
-------------------------------------
As the saying goes, "you only have a performance problem if you can prove it".
So we shouldn't make any technical decisions about performance unless we have
hard numbers to back them up. Since you always have the option to ignore the
validity bitmap if the null count is 0, I am not sure why not having
non-nullable containers would preclude performance optimizations for data
without nulls.
It would be helpful to understand the reason for the bad performance that
you're talking about, whether it's something about the Java class structure or
something lower level (e.g. ByteBuffer vs. netty's AbstractByteBuf)
> [Java] Decide what to do with non-nullable vectors in new vector class
> hierarchy
> ---------------------------------------------------------------------------------
>
> Key: ARROW-1710
> URL: https://issues.apache.org/jira/browse/ARROW-1710
> Project: Apache Arrow
> Issue Type: Sub-task
> Components: Java - Vectors
> Reporter: Li Jin
> Fix For: 0.8.0
>
>
> So far the consensus seems to be remove all non-nullable vectors.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
