jpountz commented on PR #12741: URL: https://github.com/apache/lucene/pull/12741#issuecomment-3083673070
This change trades higher disk/memory usage in favor of faster queries, it's a trade-off indeed. I don't think there's an ongoing discussion about it anymore, plus we made many efficiency improvements to other areas of query evaluation, so reverting this change would likely result in a noticeable slowdown. FWIW the higher disk usage will be especially visible on datasets that have dense postings lists (because going from 8 bits per value to 9 is a 12.5% increase while going from 16 to 17 is only a 6.25% increase) and don't have positions (because positions tend dominate disk usage when indexed anyway). IMO, the current trade-off is quite sensible: FOR for doc IDs, and PFOR for frequency and position blocks. -- 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
