[ 
https://issues.apache.org/jira/browse/LUCENE-7186?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15243505#comment-15243505
 ] 

Nicholas Knize commented on LUCENE-7186:
----------------------------------------

bq. I'm asking again Isn't this just doing the same thing as the xor of the 
sign bit that NumericUtils does? If so, I think the xor (and the comment Uwe 
wrote above it!) is cleaner?

Both ways work. Translation seemed more natural to me but if the bit flip makes 
more sense for others I'll make the change.

bq. ...only later, once we have a 1D points based morton encoded geo field, can 
we then think about making these available in core?

Sure. We can carry two GeoEncodingUtils classes for the time being and refactor 
this later.

> Add numerically stable morton encoding to GeoEncodingUtils
> ----------------------------------------------------------
>
>                 Key: LUCENE-7186
>                 URL: https://issues.apache.org/jira/browse/LUCENE-7186
>             Project: Lucene - Core
>          Issue Type: Improvement
>            Reporter: Nicholas Knize
>         Attachments: LUCENE-7186.patch, LUCENE-7186.patch, LUCENE-7186.patch, 
> LUCENE-7186.patch
>
>
> This is the follow on to LUCENE-7184.  It adds a numerically stable morton 
> encoding method to {{o.a.l.geo.GeoEncodingUtils}} that can be reused by 
> {{GeoPointField}} and a new {{GeoPointField}} based on {{Point}} encoding.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]

Reply via email to