[
https://issues.apache.org/jira/browse/LUCENE-3345?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13071711#comment-13071711
]
Robert Muir commented on LUCENE-3345:
-------------------------------------
This flag can be per-segment right?
Another idea is, if the flag is important to docvalues, we write it into
fieldinfos.
we decided to make docvalues a part of core, and I think this would make better
sense.
I don't like the way this works today, we are out of normal field bits (I used
the last one for LUCENE-2048), but docvalues adds a whole 'type byte' that is
always written, when in the majority of causes there won't even be a docvalues.
I think instead, we should add a second 'type byte', and the first flag is
'hasDocValues', if this is true, then there is follow-on byte for docvalues
flags, which could include this one.
I will open an issue.
> docvalues FNFE
> --------------
>
> Key: LUCENE-3345
> URL: https://issues.apache.org/jira/browse/LUCENE-3345
> Project: Lucene - Java
> Issue Type: Bug
> Reporter: Robert Muir
> Assignee: Simon Willnauer
>
> I created a test for LUCENE-3335, and it found an unrelated bug in docvalues.
--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
