mikemccand commented on PR #959: URL: https://github.com/apache/lucene/pull/959#issuecomment-1156474464
> Can the returned order of documents from the (parallel) searcher be different? For score-sorted cases it may not be a problem but whenever there are ties, and the assertions somehow depend on document order, things will be more problematic. I agree this is a risk, but I would expect whether you search concurrently or serially, Lucene really should return the same top hits nomatter the query sort or docid tie-breaking etc. But likely there are bugs lurking :) Hopefully we catch them with this change! The one thing I think (?) can change is the estimated total hit count if BMW kicked in. That can change even if you search the same segments, serially, but in a different order, I think. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
