mikemccand commented on PR #13149: URL: https://github.com/apache/lucene/pull/13149#issuecomment-2025041318
> Do we believe that we can merge this PR and then continue with changing the BKD visit API in a later change, or should we try to change the abstraction in this PR? +1 -- PNP! > * In luceneutil, I believe that this corresponds to setting `taskCountPerCat` high. The nightly benchy uses `taskCountPerCat=5` -- is that enough to reveal the impact of this optimization? But, the nightly benchy does not `forceMerge` down to one segment, so, we may not have the right number of docs to use 16 or 32 bit int encoding in the BKD file. Actually, I don't think we should block this change on fixing nightly benchy to reveal its gains? Your above `luceneutil` repro is sufficient? I'm wary of allowing benchmarking to unduly hold up otherwise good Lucene changes in general... it's a helpful tool, not a crutch ;) -- 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
