Re: [PR] Compute facets while collecting [lucene]

epotyom commented on code in PR #13568: URL: https://github.com/apache/lucene/pull/13568#discussion_r1724471801 ## lucene/core/src/java/org/apache/lucene/search/CollectorOwner.java: ## @@ -0,0 +1,75 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more +

[PR] Modernize list get first element [lucene]

mrhbj opened a new pull request, #13677: URL: https://github.com/apache/lucene/pull/13677 ### Description -- 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 uns

Re: [PR] Take advantage of the doc value skipper when it is primary sort [lucene]

gsmiller commented on code in PR #13592: URL: https://github.com/apache/lucene/pull/13592#discussion_r1723543404 ## lucene/core/src/java/org/apache/lucene/index/DocValuesSkipper.java: ## @@ -98,4 +98,29 @@ public abstract class DocValuesSkipper { /** Return the global numbe

Re: [PR] Display frame types when analyzing top frames. [lucene]

jpountz merged PR #13670: URL: https://github.com/apache/lucene/pull/13670 -- 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.apa

Re: [I] Try applying bipartite graph reordering to KNN graph node ids [lucene]

jpountz commented on issue #13565: URL: https://github.com/apache/lucene/issues/13565#issuecomment-2299807259 Sorry that this change is being annoying for BP, the goal of this change was to simplify the API, which I found a bit messy at the time with both a sequential and random-access API.

[I] Eclipse - one or more cycles were detected [lucene]

plutext opened a new issue, #13676: URL: https://github.com/apache/lucene/issues/13676 ### Description I followed the instructions at https://github.com/apache/lucene/blob/main/help/IDEs.txt#L7 to set up Eclipse project files: ``` ./gradlew eclipse ``` and then import

Re: [PR] Override single byte writes to OutputStreamIndexOutput to remove locking [lucene]

msfroh commented on PR #13543: URL: https://github.com/apache/lucene/pull/13543#issuecomment-2299646044 I was just looking at a flame graph for an indexing performance regression with JDK21 and noticed a lot of time spent in `BufferedOutputStream#growIfNeeded`. Without virtual threads, it l

Re: [I] Try applying bipartite graph reordering to KNN graph node ids [lucene]

msokolov commented on issue #13565: URL: https://github.com/apache/lucene/issues/13565#issuecomment-2299601054 Another thing I stumbled on that surprised me is that SortingCodecReader now requires loading vectors into RAM in order to do its work. It used to rely on random access to avoid th

Re: [I] Try applying bipartite graph reordering to KNN graph node ids [lucene]

msokolov commented on issue #13565: URL: https://github.com/apache/lucene/issues/13565#issuecomment-2299587209 Yes, that seems incontrovertibly correct. I was planning to include it along with some other changes, but let's fix the easy stuff first. https://github.com/apache/lucene/pull/1367

[PR] import definition of default parameter values from HnswGraphBuilder [lucene]

msokolov opened a new pull request, #13674: URL: https://github.com/apache/lucene/pull/13674 (no comment) -- 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-

[I] Link and Code ANN algorithm from FaceBook in Lucene? [lucene]

tanyaroosta opened a new issue, #13673: URL: https://github.com/apache/lucene/issues/13673 ### Description Hi, I was wondering if anyone has looked into implementing the link and code algorithm from FaceBook in Lucene. It seems the performance is really good based on the benchmar

Re: [I] Try applying bipartite graph reordering to KNN graph node ids [lucene]

mikemccand commented on issue #13565: URL: https://github.com/apache/lucene/issues/13565#issuecomment-2298787050 > update: I found we set this `DEFAULT_MAX_CONN` in two different places! Once in `Lucene99HnswVectorsWriter` and also in `HnswGraphBuilder`. In my case it's the latter one that

Re: [PR] Introduce new encoding of BPV 21 for DocIdsWriter used in BKD Tree [lucene]

expani commented on PR #13521: URL: https://github.com/apache/lucene/pull/13521#issuecomment-2298718372 While trying out 2 different ways ( one a subset of other ) I found that `Bit21With3StepsEncoder` is better than `Bit21With2StepsEncoder` in aarch64 platforms with OpenJDK 21/22 whereas t

Re: [PR] Convert BKDConfig to a record [lucene]

uschindler commented on PR #13668: URL: https://github.com/apache/lucene/pull/13668#issuecomment-2298573177 Thanks. All fine now. 👍 -- 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 co

Re: [PR] Convert BKDConfig to a record [lucene]

iverase merged PR #13668: URL: https://github.com/apache/lucene/pull/13668 -- 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.apa