[
https://issues.apache.org/jira/browse/ARROW-6112?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16974875#comment-16974875
]
Micah Kornfield edited comment on ARROW-6112 at 11/15/19 7:19 AM:
------------------------------------------------------------------
based on discussion on mailing list, there was a request for a rebase on the
original PR that accommodated all vector APIs. I haven't and probably won't
have time to do this. It was mentioned that just redoing ArrowBuf to use
64-bit address space might be more palatable. So I'm going to focus on making
that happen.
This will make it possible to support LargeBinary and LargeString (with the
limitation that string length will still probably be practically limited to 2GB
for most APIs).
However, for LargeArray child arrays will still be limited to 2 billion entries
so this would be of limited utility.
was (Author: [email protected]):
based on discussion on mailing list, there was a request for a rebase on the
original PR that accommodated all vector APIs. I haven't and probably won't
have time to do this. It was mentioned that just redoing ArrowBuf to use
64-bit address space might be more palatable. So I'm going to focus on making
that happen.
> [Java] Update APIs to support 64-bit address space
> --------------------------------------------------
>
> Key: ARROW-6112
> URL: https://issues.apache.org/jira/browse/ARROW-6112
> Project: Apache Arrow
> Issue Type: Improvement
> Components: Java
> Reporter: Micah Kornfield
> Assignee: Micah Kornfield
> Priority: Major
> Labels: pull-request-available
> Time Spent: 2.5h
> Remaining Estimate: 0h
>
> The arrow spec allows for 64 bit address range for buffers (and arrays) we
> should support this at the API level in Java even if the current Netty
> backing buffers don't support it.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
