[Save this search]

Status
All
   Fixed (7187)
  Closed (5023)
✓Open (2608)
   Won't Fix (545)
   Duplicate (297)
   Invalid (217)
   Not A Problem (195)
Issue type
✓All
  Issue (2271)
  PR (337)
Author relation
✓All
  None (2057)
  Member (644)
  Contributor (469)
  New contributor (69)
Created
✓All
  Past day (0)
  Past 2 days (0)
  Past 3 days (5)
  Past week (10)
  Past month (46)
  Past 3 months (116)
  Past 6 months (171)
  Past year (275)
Updated
✓All
  Past day (0)
  Past 2 days (0)
  Past 3 days (13)
  Past week (49)
  Past month (113)
  Past 3 months (156)
  Past 6 months (217)
  Past year (507)
Updated ago
✓All
  > 1 day ago (2608)
  > 2 days ago (2608)
  > 3 days ago (2595)
  > 1 week ago (2559)
  > 1 month ago (2495)
  > 3 months ago (2452)
  > 1 year ago (2101)
Comment count
✓All
  0 (581)
  1 (369)
  2 - 5 (805)
  6 - 10 (442)
  10 - 20 (314)
  > 20 (171)
Reaction count
✓All
  0 (2325)
  1 (170)
  2 - 5 (91)
  6 - 10 (16)
  10 - 20 (5)
  > 20 (1)
Review Requested
✓All
  jpountz (13)
  mikemccand (9)
  rmuir (5)
  benwtrent (5)
  iverase (4)
  dweiss (3)
  msokolov (3)

See all 26...
Mentioned
✓All
  jpountz (93)
  mikemccand (77)
  rmuir (54)
  benwtrent (54)
  msokolov (41)
  uschindler (38)
  romseygeek (21)

See all 207...
Reviewed
✓All
  mikemccand (25)
  jpountz (23)
  benwtrent (14)
  rmuir (13)
  msokolov (12)
  dweiss (11)
  uschindler (11)

See all 65...
Commented
✓All
  asfimport (1368)
  github-actions[bot] (254)
  jpountz (140)
  mikemccand (126)
  rmuir (96)
  msokolov (85)
  benwtrent (84)

See all 304...
User
✓All
  asfimport (1810)
  github-actions[bot] (304)
  mikemccand (211)
  jpountz (201)
  rmuir (130)
  benwtrent (122)
  msokolov (105)

See all 456...
Last comment user
✓All
  asfimport (1324)
  github-actions[bot] (245)
  mikemccand (33)
  jpountz (32)
  rmuir (23)
  msokolov (20)
  benwtrent (17)

See all 162...
Draft
✓All
  No (280)
  Yes (57)
Component
✓All
  core (622)
  analysis (155)
  highlighter (48)
  spatial (41)
  facet (41)
  queryparser (27)
  test-framework (24)

See all 23...
Type
✓All
  enhancement (1194)
  bug (724)
  task (202)
  test (80)
  documentation (22)
Labels
✓All
  Stale (245)
  legacy-jira-fix-versio... (214)
  legacy-jira-fix-versio... (169)
  affects-version:4.0-ALPHA (81)
  tool:build (54)
  vector-based-search (48)
  affects-version:6.0 (37)

See all 157...
Commits?
✓All
  No (2608)
Reporter
✓All
  rmuir (270)
  mikemccand (154)
  jpountz (126)
  dsmiley (67)
  uschindler (52)
  iverase (44)
  dweiss (41)

See all 752...
Assignee
✓All
  Unassigned (2367)
  mikemccand (35)
  uschindler (31)
  romseygeek (27)
  dsmiley (22)
  rmuir (15)
  jpountz (11)

See all 44...
  
Filters: Status (Open),  Issue type,  Author relation,  Created,  Updated,  Updated ago,  Comment count,  Reaction count,  Review Requested,  Mentioned,  Reviewed,  Commented,  User,  Last comment user,  Draft,  Component,  Type,  Labels,  Commits?,  Reporter,  Assignee

#16383 PR: Add fp16 (float16) vector encoding support - no quantization support
2.2 days ago  82 comments  0 votes  0 watches  Pulkitg64, danmuzi, github-actions[bot], jeho-rpls, msokolov
Description Introduce a new FLOAT16 VectorEncoding that stores vectors at half precision (16 ... This introduces the raw fp16 encoding only; scalar quantization of fp16 vectors is intentionally ... Will raise another PR with quantization support.
    Pulkitg64 2.2 days ago:  Fixed in the next revision
    Pulkitg64 2.2 days ago:  These functions were unused so I have removed them in the next revision.

#15769: PathHierarchyTokenizer "ancestor search" use case broken in lucene >= 10.0
2.3 days ago  3 comments  0 votes  0 watches  hossman, msokolov, vismaytiwari
Description The javadocs for PathHierarchyTokenizerFactory describe the two main use cases it ... "decendent path" search - by configuring it at index time, and using KeywordTokenizer at query time ... But it will not match documents indexed with values like Books, or Books/Fic 2.
    vismaytiwari 2.3 days ago:  @hossman thanks for the detailed write-up and the test patch — made this easy to pin down. Opened 16402. I went with reverting to positionIncrement 0 on the tokens after the first (plain synonyms) rather ...
    hossman 4.6 months ago:  > Does the class have distinct indexing and query modes? No, it's never needed one before -- you either configure(d) it at index time for descendant search, ...

#16402 PR: GITHUB#15769: Restore PathHierarchyTokenizer synonym positions for ancestor-path search
2.3 days ago  0 comments  0 votes  0 watches  github-actions[bot], hossman, vismaytiwari
Description Fixes 15769. PathHierarchyTokenizer is documented to emit each path prefix at the *same* position, so that query ... That "ancestor path" use case has been broken since 10.0: GITHUB12875 (5d6086e) changed every token ...

#16401 PR: Speed up TestHnswMergeAbort and make its assertion deterministic
2.3 days ago  0 comments  0 votes  0 watches  github-actions[bot], jeho-rpls
Fixes 16397. The tests asserted that rollback() returns within 10 seconds, which needed big segments (4 x 12k ... Changes: * Instead of asserting on elapsed time, each test now asserts through InfoStream that the ...

#16397: Test suite TestHnswMergeAbort slows down build
2.3 days ago  3 comments  0 votes  0 watches  jeho-rpls, kaivalnp, msokolov
Description 16368 recently added a new test suite TestHnswMergeAbort, which is slowing down local ... I wonder if we can speed up the tests or mark them as @Monster? Gradle command to reproduce.
    msokolov 2.3 days ago:  I ilke the InfoStream approach since we signal completion there. We can supply some kind of test-only MockInfoStream to the writer. Please feel free
    jeho-rpls 2.3 days ago:  Hi, I added this test in 16368. The size was deliberate since the timing assertion needs a wide margin, but I did't weigh what that ... How about asserting through InfoStream that the graph build never ran to completion, instead of ...

#16400: HNSW graph reuse during merge is silently disabled under per-field codecs since 10.4.0
2.4 days ago  1 comments  0 votes  0 watches  jeho-rpls, msokolov
Description# Summary IncrementalHnswGraphMerger is meant to reuse existing HNSW graphs from input ... Since 10.4.0 this reuse never triggers under a per-field vectors codec, which is how Lucene's ... As a result every merge rebuilds the full graph by re-inserting every vector, and the cost that ...
    msokolov 2.4 days ago:  Oh my, what a mess! Thanks for finding this; hard to believe it took this long to notice the indexing slowdown. Please feel free to take a stab at fixing.

#16044: Introduce a pread Directory based on Panama-FFI ?
2.5 days ago  11 comments  0 votes  0 watches  gf2121, jilldoty-elastic, jimczi, neoremind, spinscale, stevenschlansker
Description > Following is generally written by LLM but benchmark is run by myself :) # 1. Motivation: NIOFSDirectory is still relevant In recent memory-constrained deployments ( ... Switching to NIOFSDirectory helps us resolve it. # 2.
    neoremind 2.5 days ago:  Thanks @jimczi! I've read 14156. One nuance from my benchmarking though: I actually got a bit different take. isLoaded() itself ...
    jimczi 10.4 days ago:  That's good finding @neoremind. There was an attempt to remove the isLoaded bits in https://github.com/apache/lucene/pull/14156 but ... We should probably resume this effort to get the full power of prefetch on the mmap directory.

[15.1 msec search, 16.1 msec total]