[ 
https://issues.apache.org/jira/browse/LUCENE-6578?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14603190#comment-14603190
 ] 

Karl Wright commented on LUCENE-6578:
-------------------------------------

bq. the choice of "ArcDistance.INSTANCE" as a means of referring to the a 
distance. It's fine but I prefer the approach in Lucene's FilteredQuery in 
which the choices are referred via named constants, like 
FilteredQuery.RANDOM_ACCESS_FILTER_STRATEGY. This avoids seeing "INSTANCE". 
Taking that approach here would mean having a DistanceStyle.ARC_DISTANCE or 
perhaps just "DistanceStyle.ARC". What do you think?

Why not support both?  The current strategy is eminently extensible, which is 
what I like about it.

bq. GeoPath.distance() has an implementation that confused me until I realized 
what it was doing... and I realized that wasn't what came to mind when I saw 
the distance method. Perhaps this method should be named distanceAlongPath and 
further document that it's from the beginning?

I construed this as a generic "internal distance" method, which is how it's 
used in my employer's application.  Removing the genericism would bust that 
application.  So I prefer to leave this the way it is, thanks.


> Geo3d: arcDistanceToShape() method may be useful
> ------------------------------------------------
>
>                 Key: LUCENE-6578
>                 URL: https://issues.apache.org/jira/browse/LUCENE-6578
>             Project: Lucene - Core
>          Issue Type: Bug
>          Components: modules/spatial
>            Reporter: Karl Wright
>         Attachments: LUCENE-6578.patch, LUCENE-6578.patch
>
>
> I've got an application that seems like it may need the ability to compute a 
> new kind of arc distance, from a GeoPoint to the nearest edge/point of a 
> GeoShape.  Adding this method to the interface, and corresponding 
> implementations, would increase the utility of the package for ranking 
> purposes.



--
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