msokolov commented on PR #13542: URL: https://github.com/apache/lucene/pull/13542#issuecomment-2329642279
> I don't entirely follow how we may have painted ourselves in a corner and how running benchmarks relates to this. They're not related; perhaps I should have broken my comment into multiple posts. My main concern is that we should be able, on a single thread, to evaluate matches from portions of multiple segments, and it sounds as if this API will allow for that. Regarding the benchmarks, I understand these changes are not intended to have any impact on existing search code paths and need to be invoked in a special way. I still think testing should be done to verify that. And this relates to a different concern - that we are proposing to launch a large amount of dead code that will have no real purpose until further work is done. Again, we might learn something from implementing changes that actually *do* move benchmarks. I guess an underlying concern is that we are trying to get stuff in before an arbitrary (Lucene 10) release date, and assuming that no breaking changes will later be required. Isn't it conceivable that we will need to make a change to the Scorer API to support cloning? Maybe "painted into a corner" was too strong, but there could certainly be future changes we might need. To me it would make more sense to hold on off on merging (even if it means missing the impending Lucene 10 release) until we have some positive results. -- 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
