dweiss commented on PR #14800:
URL: https://github.com/apache/lucene/pull/14800#issuecomment-2979163597
I'll let myself merge this in to proceed with other things. If anybody would
like to change anything here, please comment on the patch and I'll follow-up.
--
This is an automated messag
dweiss opened a new pull request, #14800:
URL: https://github.com/apache/lucene/pull/14800
Fixes #14796
--
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-ma
github-actions[bot] commented on PR #14800:
URL: https://github.com/apache/lucene/pull/14800#issuecomment-2979154910
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
uschindler commented on issue #14785:
URL: https://github.com/apache/lucene/issues/14785#issuecomment-2979121217
Hi,
We can keep it as is, all fine.
The "working copy clean" check was faster and better implemented with jgit,
but for normal stuff like checking out/exporting using na
dweiss commented on PR #14740:
URL: https://github.com/apache/lucene/pull/14740#issuecomment-2979065537
Just FYI - I noticed we already have an editorconfig file in scripts -
https://github.com/apache/lucene/blob/main/dev-tools/scripts/.editorconfig
is this complementary? Could it
dweiss commented on issue #14785:
URL: https://github.com/apache/lucene/issues/14785#issuecomment-2979061860
@uschindler asked for it - simpler to manage on Windows vm boxes, I guess. I
think we can make it an option to use native or jgit... but I'd rather simplify
than complicate so I'll w
HUSTERGS commented on PR #14772:
URL: https://github.com/apache/lucene/pull/14772#issuecomment-2978687251
@gf2121 Thanks for your suggestion!
> I wonder if we can benchmark the following implementation to confirm the
source of speedup?
I'v done the benchmark according to your sugge
dsmiley commented on issue #14785:
URL: https://github.com/apache/lucene/issues/14785#issuecomment-2978673604
Why revert back to jgit?
--
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
github-actions[bot] commented on PR #14799:
URL: https://github.com/apache/lucene/pull/14799#issuecomment-2978185513
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
github-actions[bot] commented on PR #14716:
URL: https://github.com/apache/lucene/pull/14716#issuecomment-2978548505
This PR has not had activity in the past 2 weeks, labeling it as stale. If
the PR is waiting for review, notify the d...@lucene.apache.org list. Thank you
for your contributi
rmuir merged PR #14799:
URL: https://github.com/apache/lucene/pull/14799
--
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.apach
yaser-aj commented on issue #13883:
URL: https://github.com/apache/lucene/issues/13883#issuecomment-2978379030
Me, @lukewilner, @atharvkashyap, and @N624-debu are students from Carnegie
Mellon University, and we’ll be working on this issue as part of a mentored
summer course focused on coll
uschindler commented on issue #14787:
URL: https://github.com/apache/lucene/issues/14787#issuecomment-2978315117
It is working, but spams the log with the above stack traces. I tested it by
modifying a Gradle file. It was reformatted successfully.
--
This is an automated message from the
dweiss commented on issue #14796:
URL: https://github.com/apache/lucene/issues/14796#issuecomment-2978198994
Smoke tester also fails with this - which is probably related:
```
GIT rev: 07013dcf0e9d4f927888c1dd7bfee3753c0f8750
2025-06-16 21:21:27.322717: RUN: ./gradlew --s
rmuir commented on PR #14798:
URL: https://github.com/apache/lucene/pull/14798#issuecomment-2978184994
#14799 to make the messaging better when `make lint` fails.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
rmuir opened a new pull request, #14799:
URL: https://github.com/apache/lucene/pull/14799
if the linter fails, often many of the problems can be safely autofixed.
That's because many of the rules are opinionated / conventional / cosmetic and
can be annoying to deal with manually.
dweiss commented on PR #14798:
URL: https://github.com/apache/lucene/pull/14798#issuecomment-2978159953
Thanks. I think I'll go back to the regexp-scan instead of using the
javaproperties module. I don't think we import the requirements in github
workflow that runs the tests... Will return
rmuir commented on PR #14798:
URL: https://github.com/apache/lucene/pull/14798#issuecomment-2978163695
For now you can try `make autofix` which is like `make reformat` but will
also (safely) fix any linter issues it can too.
--
This is an automated message from the Apache Git Service.
To
rmuir commented on PR #14798:
URL: https://github.com/apache/lucene/pull/14798#issuecomment-2978151874
Try a `make reformat`. Not sure what editor you use, but eg for vscode if
you can install ruff and basedpyright extensions, and enable format-on-save and
organize-imports-on-save too, you
github-actions[bot] commented on PR #14798:
URL: https://github.com/apache/lucene/pull/14798#issuecomment-2978119425
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
dweiss opened a new pull request, #14798:
URL: https://github.com/apache/lucene/pull/14798
#14786
### 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 c
dweiss closed issue #14789: Add an option to turn off gradle/groovy spotless
checks/validations
URL: https://github.com/apache/lucene/issues/14789
--
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 th
dweiss merged PR #14791:
URL: https://github.com/apache/lucene/pull/14791
--
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.apac
rmuir commented on issue #14785:
URL: https://github.com/apache/lucene/issues/14785#issuecomment-2977983045
My biggest complaint about jgit is that it doesn't truly match git behavior.
But I already set special options (e.g. `pull.twohead=recursive`) to work
around bugs in jgit, so I don't
RKSPD commented on issue #14681:
URL: https://github.com/apache/lucene/issues/14681#issuecomment-2977958739
I have a working JVector implementation for Lucene from a while ago, (no
leaford handling, RandomAccessVectorsWriter, etc) and I have benchmarks for
that version. There are issues lik
benwtrent commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2977941605
@Pulkitg64
Basically, I don't think callers should "know" directly if they are hitting
quantized vectors or raw.
Requiring the user to pick the right thing seems unneces
benwtrent commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2977917618
> Returning the vectors in search results
Why would you need to do this?
Generally, I would assume that any access to the vector would be "Give me
what I gave you", and th
github-actions[bot] commented on PR #14797:
URL: https://github.com/apache/lucene/pull/14797#issuecomment-2977901473
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
benwtrent opened a new pull request, #14797:
URL: https://github.com/apache/lucene/pull/14797
assertOffHeapByteSize makes a ton of assumptions and these don't really work
well for custom formats.
Changing to `protected` instead of `static` to allow simple overriding by
custom format
dweiss commented on issue #14785:
URL: https://github.com/apache/lucene/issues/14785#issuecomment-2977896283
I'd like to note that we use native git for other things within the build -
assembling a source tgz, fetching repos for hunspell... I can add jgit support
for this status fetching bu
breskeby commented on code in PR #14788:
URL: https://github.com/apache/lucene/pull/14788#discussion_r2150602765
##
build-tools/build-infra/src/main/groovy/lucene.java.core.mrjar.gradle:
##
@@ -32,9 +32,14 @@ configure(project(":lucene:core")) {
tasks.named("compileMain${
dweiss opened a new issue, #14796:
URL: https://github.com/apache/lucene/issues/14796
### Description
https://ci-builds.apache.org/job/Lucene/job/Lucene-Artifacts-main/1663/
### Version and environment details
_No response_
--
This is an automated message from the Apac
dweiss commented on code in PR #14788:
URL: https://github.com/apache/lucene/pull/14788#discussion_r2150729480
##
build-tools/build-infra/src/main/groovy/lucene.java.core.mrjar.gradle:
##
@@ -32,9 +32,14 @@ configure(project(":lucene:core")) {
tasks.named("compileMain${jd
Pulkitg64 commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2977859146
We are experimenting with large vector indexes, and since (raw unquantized)
vectors consume significant disk space (4x more than quantized vectors), we
want to drop the raw vectors fro
benwtrent commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2977492907
> As for the usefulness of accessing quantized bytes directly - we have
specific use cases, such as returning the vectors themselves when requested in
a query.
I would assume th
Pulkitg64 commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2977454253
Thanks @benwtrent for the quick review and comments.
Regarding your comment about how this relates to issue #13158 - I agree in a
way this PR doesn't directly help create a "read
gf2121 commented on PR #14772:
URL: https://github.com/apache/lucene/pull/14772#issuecomment-2977389388
Thanks for contribution, it's a nice speedup!
> I don't believe that being able to use Arrays#fill helps much, but maybe
the fact that this change helps reduce polymorphism does?
github-actions[bot] commented on PR #14791:
URL: https://github.com/apache/lucene/pull/14791#issuecomment-2977323935
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
vigyasharma merged PR #14793:
URL: https://github.com/apache/lucene/pull/14793
--
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
dweiss merged PR #14795:
URL: https://github.com/apache/lucene/pull/14795
--
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.apac
dweiss opened a new pull request, #14795:
URL: https://github.com/apache/lucene/pull/14795
This adds logical (well) build option groups for the 'allOptions' task.
Looks like this:
--
This
github-actions[bot] commented on PR #14795:
URL: https://github.com/apache/lucene/pull/14795#issuecomment-2977171162
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
benwtrent commented on issue #14694:
URL: https://github.com/apache/lucene/issues/14694#issuecomment-2977130359
@tteofili it depends on how many vectors are in the segment. Given the
`Lucene99ScalarQuantizedVectorScorer` scorer, I have seen issues where uniform
vectors cause strange scoring
prudhvigodithi commented on PR #14784:
URL: https://github.com/apache/lucene/pull/14784#issuecomment-2977045873
Thanks Adrien, updated the CHANGES.txt moving to 10.3.0.
@getsaurabh02
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
dweiss merged PR #14794:
URL: https://github.com/apache/lucene/pull/14794
--
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.apac
tteofili commented on issue #14694:
URL: https://github.com/apache/lucene/issues/14694#issuecomment-2976969898
this is a bit weird, basically we have the correct resulting docs in the
array, but their sorting order is wrong.
```
scoreDocs[2] = {ScoreDoc@4520} "doc=1 score=0.5 shard
dsmiley commented on PR #14740:
URL: https://github.com/apache/lucene/pull/14740#issuecomment-2976920502
> please remove the python completely as we already have a python
autoformatter and verifier
I think you're missing a point of the value of this. It's _complementary_
with Spotle
github-actions[bot] commented on PR #14794:
URL: https://github.com/apache/lucene/pull/14794#issuecomment-2976818462
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
dweiss opened a new pull request, #14794:
URL: https://github.com/apache/lucene/pull/14794
Use direct dependencies on dependencychecks plugins and build options.
Remove some leftover files.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log
github-actions[bot] commented on PR #14793:
URL: https://github.com/apache/lucene/pull/14793#issuecomment-2976551678
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
ChrisHegarty opened a new pull request, #14793:
URL: https://github.com/apache/lucene/pull/14793
This commit fixes an intermittent failure of TestDoubleValuesSourceRescorer,
where the test sometimes randomly selects too few documents to index leading to
no matches in the search. Failure exa
rmuir commented on PR #14740:
URL: https://github.com/apache/lucene/pull/14740#issuecomment-2976483811
please remove the python completely as we already have a python
autoformatter and verifier (like spotless, except not slow as hell). We don't
need conflicting editor configuration around i
rmuir commented on code in PR #14740:
URL: https://github.com/apache/lucene/pull/14740#discussion_r2149895516
##
.editorconfig:
##
@@ -0,0 +1,918 @@
+# EditorConfig: https://editorconfig.org
+# for consistent code style configuration across editors/IDEs
+
+# top-most EditorConfi
benwtrent commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2976430071
@Pulkitg64 I don't understand how this is part of
https://github.com/apache/lucene/issues/13158
I would have thought the APIs stay the same. Quantization should be able to
"rehy
jpountz commented on PR #14775:
URL: https://github.com/apache/lucene/pull/14775#issuecomment-2976435614
Thinking a bit more about it, adding `searchBefore` sounds like it could
work. Would you like to give it a try?
--
This is an automated message from the Apache Git Service.
To respond
jpountz commented on PR #14784:
URL: https://github.com/apache/lucene/pull/14784#issuecomment-2976386308
I'm good with the change, but I'd put the change in 10.3 instead of 10.2.2
since it's a new feature rather than a bug fix.
--
This is an automated message from the Apache Git Service.
dsmiley commented on PR #14654:
URL: https://github.com/apache/lucene/pull/14654#issuecomment-2976307246
Done.
--
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 unsubscrib
github-actions[bot] commented on PR #14792:
URL: https://github.com/apache/lucene/pull/14792#issuecomment-2976186238
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
Pulkitg64 opened a new pull request, #14792:
URL: https://github.com/apache/lucene/pull/14792
### Description
Introduce `getQuantizedVectorValues` method in `LeafReader` to access
`QuantizedVectorValues`.
In a search architecture where searchers and writer runs on separate
mac
dsmiley commented on PR #14740:
URL: https://github.com/apache/lucene/pull/14740#issuecomment-2976114609
I plan to merge this tonight (~10pm EST) if I don't hear advise to the
contrary.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on t
dweiss closed issue #14790: Fix jacoco/ coverage plugin not found
URL: https://github.com/apache/lucene/issues/14790
--
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 unsubscr
ChrisHegarty commented on PR #14654:
URL: https://github.com/apache/lucene/pull/14654#issuecomment-2976032407
@dsmiley Given that this change has been backported to 9.12.2, which will
release in tandem with 10.2.2, do you want to port this change to `branch_10_2`
also?
--
This is an aut
github-actions[bot] commented on PR #14791:
URL: https://github.com/apache/lucene/pull/14791#issuecomment-2976005719
This PR does not have an entry in lucene/CHANGES.txt. Consider adding one.
If the PR doesn't need a changelog entry, then add the skip-changelog label to
it and you will stop
dweiss opened a new pull request, #14791:
URL: https://github.com/apache/lucene/pull/14791
Fixes #14789. Allows locally skipping the heavy greclipse download and
costly formatting/validation step.
--
This is an automated message from the Apache Git Service.
To respond to the message, ple
dweiss commented on PR #14788:
URL: https://github.com/apache/lucene/pull/14788#issuecomment-2975931132
Yes, I think this can be relied upon. I don't think it's mentioned anywhere,
officially, but I think it's the case.
--
This is an automated message from the Apache Git Service.
To respo
easyice merged PR #14774:
URL: https://github.com/apache/lucene/pull/14774
--
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
uschindler commented on PR #14788:
URL: https://github.com/apache/lucene/pull/14788#issuecomment-2975564900
If I remember: the order of configuration is still predictable by the order
of including them into the main build script. I always felt bad with that, but
it worked. Just want to conf
dweiss merged PR #14788:
URL: https://github.com/apache/lucene/pull/14788
--
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.apac
dweiss commented on PR #14764:
URL: https://github.com/apache/lucene/pull/14764#issuecomment-2975342904
> Please keep this for longer time, as the Jenkins build randomization can
only work on environment variables. There maybe tweaks to pass this on command
line in Jenkins's gradle options,
dweiss closed issue #14782: compileMain24Java fails with JDK25+
URL: https://github.com/apache/lucene/issues/14782
--
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 unsubscrib
uschindler commented on issue #14782:
URL: https://github.com/apache/lucene/issues/14782#issuecomment-2975525126
I updated the JDK bug report.
--
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
dweiss commented on PR #14788:
URL: https://github.com/apache/lucene/pull/14788#issuecomment-2975466304
I will apply and backport, no worries, Uwe. Enjoy a beer in Berlin!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and us
uschindler commented on PR #14788:
URL: https://github.com/apache/lucene/pull/14788#issuecomment-2975442200
Can you backport this change?
--
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 speci
dweiss opened a new issue, #14790:
URL: https://github.com/apache/lucene/issues/14790
### Description
https://ci-builds.apache.org/job/Lucene/job/Lucene-Coverage-main/1468/
org.gradle.api.plugins.UnknownPluginException: Plugin with id
'org.barfuin.gradle.jacocolog' not found.
dweiss commented on PR #14764:
URL: https://github.com/apache/lucene/pull/14764#issuecomment-2975350683
> I'd rather disable this Groovy/Gradle code formatting - sorry for this,
but this took me ages on battery power today in train while debugging the other
windows build failures on Jenkins
dweiss opened a new issue, #14789:
URL: https://github.com/apache/lucene/issues/14789
### Description
These download huge amount of data and are slow to apply. Makes sense to
have the ability to turn them off locally. I would like to keep them on the CI.
--
This is an automated mes
dweiss merged PR #14783:
URL: https://github.com/apache/lucene/pull/14783
--
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.apac
77 matches
Mail list logo
