dweiss commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1107374164
> Currently, it can't be marked @Slow because the distribution tests do not
use Lucene's test framework.
@mocobeta you can use an assumption and check if a system property is set
("tes
[
https://issues.apache.org/jira/browse/LUCENE-10528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526769#comment-17526769
]
Dawid Weiss commented on LUCENE-10528:
--
I agree - let's mark it slow. I run slow t
[
https://issues.apache.org/jira/browse/LUCENE-10528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526767#comment-17526767
]
Tomoko Uchida commented on LUCENE-10528:
Maybe my explanation wouldn't be good;
[
https://issues.apache.org/jira/browse/LUCENE-10528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526763#comment-17526763
]
Tomoko Uchida commented on LUCENE-10528:
bq. xvfb hack only applies for *nix, n
[
https://issues.apache.org/jira/browse/LUCENE-10528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526761#comment-17526761
]
Robert Muir edited comment on LUCENE-10528 at 4/23/22 4:08 AM:
--
[
https://issues.apache.org/jira/browse/LUCENE-10528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526761#comment-17526761
]
Robert Muir commented on LUCENE-10528:
--
I don't think we need to change the test t
[
https://issues.apache.org/jira/browse/LUCENE-10528?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526756#comment-17526756
]
Tomoko Uchida commented on LUCENE-10528:
As I said in the PR, I'm fine with run
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1107342012
LOL, I figured I would just try to see if you would bite. I want to make it
better, my windows knowledge is just less, so it takes me longer. Also spinning
up VMs is somehow painful for me.
sonatype-lift[bot] commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856825108
##
lucene/distribution/src/binary-release/bin/luke.sh:
##
@@ -26,6 +26,11 @@ if [ -z "$LAUNCH_CMD" ]; then
else
# We are integration-testing. Force UTF-8 as
rmuir commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1107283010
I would stick with the `yum` personally. if you have dnf, 'yum' still works
as expected. so its backwards compatible like that. and centos 7 is still
kicking.
--
This is an automated messag
mocobeta commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856809167
##
lucene/distribution/src/binary-release/bin/luke.sh:
##
@@ -26,6 +26,11 @@ if [ -z "$LAUNCH_CMD" ]; then
else
# We are integration-testing. Force UTF-8 as the encodi
sonatype-lift[bot] commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856778736
##
lucene/distribution/src/binary-release/bin/luke.sh:
##
@@ -26,6 +26,11 @@ if [ -z "$LAUNCH_CMD" ]; then
else
# We are integration-testing. Force UTF-8 as
mocobeta commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856765168
##
help/tests.txt:
##
@@ -128,6 +128,24 @@ specifying the project and test task or a fully qualified
task path. Example:
gradlew -p lucene/core test -Ptests.verbose=true
rmuir commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1107175458
Yes text looks good. thank you!
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific com
rmuir commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1107174374
> I also tried to make it work on MaxOS then found that it needs additional
tweaks or setup since XQuartz (X Window that runs on MacOS) supports Xvfb but
does not include xvfb-run. I think we
mocobeta commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856754362
##
lucene/distribution/src/binary-release/bin/luke.sh:
##
@@ -26,6 +26,11 @@ if [ -z "$LAUNCH_CMD" ]; then
else
# We are integration-testing. Force UTF-8 as the encodi
sonatype-lift[bot] commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856754366
##
lucene/distribution/src/binary-release/bin/luke.sh:
##
@@ -26,6 +26,11 @@ if [ -z "$LAUNCH_CMD" ]; then
else
# We are integration-testing. Force UTF-8 as
mocobeta commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1107168191
I run this on Linus with and without Xvfb, it works fine for me. I added
some notes about running the test headlessly to help/tests.txt - does it make
sense?
I also tried to make it
sonatype-lift[bot] commented on code in PR #828:
URL: https://github.com/apache/lucene/pull/828#discussion_r856745164
##
lucene/distribution/src/binary-release/bin/luke.sh:
##
@@ -26,6 +26,11 @@ if [ -z "$LAUNCH_CMD" ]; then
else
# We are integration-testing. Force UTF-8 as
Vigya Sharma created LUCENE-10530:
-
Summary: TestTaxonomyFacetAssociations test failure
Key: LUCENE-10530
URL: https://issues.apache.org/jira/browse/LUCENE-10530
Project: Lucene - Core
Issue
dweiss commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106884413
What a time-waster but uh... interesting! :) Added to the pile of todos.
As for Windows - I agree with your statement from a long time ago that it's
one of the slowest filesystems out
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106882377
and just related to this, in case we wanted to fix it one day. I think, it
made sense to use "fileKey" in the beginning when this tracking was simple. But
somehow its now very complicated (ren
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106870648
Thanks @dweiss! Currently, we can't actually use fake-windows (WindowsFS) on
an actual Windows machine. It only works everywhere else.
The problem is the mechanism we use to track file h
dweiss commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106863200
(late to the party - if you need me to check something on a physical Windows
machine, let me know.)
--
This is an automated message from the Apache Git Service.
To respond to the message, p
rmuir commented on code in PR #827:
URL: https://github.com/apache/lucene/pull/827#discussion_r856521224
##
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java:
##
@@ -1112,12 +1112,17 @@ public void seekExact(long ord) throws IOException {
jpountz commented on code in PR #827:
URL: https://github.com/apache/lucene/pull/827#discussion_r856517717
##
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java:
##
@@ -1112,12 +1112,17 @@ public void seekExact(long ord) throws IOException {
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106782689
I opened https://issues.apache.org/jira/browse/LUCENE-10529 for the random
facets failure that was triggered here while testing. It is not related to your
changes.
--
This is an automated m
Robert Muir created LUCENE-10529:
Summary: TestTaxonomyFacetAssociations may have floating point
issues
Key: LUCENE-10529
URL: https://issues.apache.org/jira/browse/LUCENE-10529
Project: Lucene - Core
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106767024
k, tests are passing now by cleaning up the `FilterPath` abstraction. Now
you don't have to worry about separators or anything, because the default
methods work just fine for that.
I ad
rmuir commented on code in PR #829:
URL: https://github.com/apache/lucene/pull/829#discussion_r856481822
##
lucene/test-framework/src/test/org/apache/lucene/tests/mockfile/TestWindowsFS.java:
##
@@ -180,4 +186,22 @@ public void testMove() throws IOException {
assertEquals
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106757373
It is also bad that your subclass has to override 4 methods and deal with
strings at all. These already have default implementations.
Let's make the string-based methods `final`. they al
[
https://issues.apache.org/jira/browse/LUCENE-10521?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526611#comment-17526611
]
Dawid Weiss commented on LUCENE-10521:
--
You can use a custom IndexDeletionPolicy -
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106745369
Yeah looks like the checking has to handle the separators properly. because
a lot of stuff is allowed:
```
jshell> Path p = Paths.get("/home/rmuir")
p ==> /home/rmuir
jshell> p
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106741520
Looks like a bug in the checking? We may have to "normalize" the path first
or something before doing the check?
I see fails like this:
```
java.nio.file.InvalidPathException: File
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106739760
It's not the only test. see my temporary commit to crank up WindowsFS. Just
want to flush them all out at once so we can see what's happening.
--
This is an automated message from the Apache
[
https://issues.apache.org/jira/browse/LUCENE-10521?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526607#comment-17526607
]
Gautam Worah commented on LUCENE-10521:
---
Thanks for taking a look Dawid. The test
rmuir commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106736469
Lemme see, ill help look. i wouldn't trust that it is just one test. ill try
your branch, and hack it locally to use windowsfs as much as possible, see what
trips :)
--
This is an automated
gautamworah96 commented on PR #829:
URL: https://github.com/apache/lucene/pull/829#issuecomment-1106733480
A gradle test on WindowsFS is failing because of the check we are
introducing in this test. I'll fire up my Windows 10 machine and see if the
test fails on the machine as well. If yes,
[
https://issues.apache.org/jira/browse/LUCENE-1383?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17526499#comment-17526499
]
Michael McCandless commented on LUCENE-1383:
> Even if the issue is closed,
JarvisCraft opened a new pull request, #830:
URL: https://github.com/apache/lucene/pull/830
# Description
This PR adds a `lucene-bom` to list other Lucene modules in its
`dependencyManagement` POM segment.
This allows external projects to only specify BOM-module's version while
mocobeta commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1106534628
ah, ok, I haven't used it but I know there are various window managers.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and
rmuir commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1106513352
> I've been guessing there are people who dislike - or wonder - that the
window suddenly pops up and disappears (I myself can live with it since I know
why this is needed). We could add some n
[
https://issues.apache.org/jira/browse/LUCENE-10386?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Petr Portnov updated LUCENE-10386:
--
Description:
h1. Short description
Add a Bill-of-Materials (BOM) module to Lucene to allow f
mikemccand commented on PR #812:
URL: https://github.com/apache/lucene/pull/812#issuecomment-1106493050
> Thanks @mikemccand. It's good to see that some got faster. Strange that
some got a little slower too though. I'll take a look and try to reproduce.
It'll also be interesting to see how
ChrisHegarty commented on PR #812:
URL: https://github.com/apache/lucene/pull/812#issuecomment-1106484402
Thanks @mikemccand. It's good to see that some got faster. Strange that some
got a little slower too though. I'll take a look and try to reproduce. It'll
also be interesting to see how
mocobeta commented on PR #828:
URL: https://github.com/apache/lucene/pull/828#issuecomment-1106480769
Sorry I couldn't take time for this today. I'll run it on Linux and would
like to try it also on MacOS this weekend.
https://stackoverflow.com/questions/25451133/xvfb-run-on-os-x
I
mikemccand commented on code in PR #633:
URL: https://github.com/apache/lucene/pull/633#discussion_r856191713
##
lucene/core/src/test/org/apache/lucene/index/TestIndexWriterOnDiskFull.java:
##
@@ -352,15 +352,15 @@ public void testAddIndexOnDiskFull() throws IOException {
rmuir commented on code in PR #827:
URL: https://github.com/apache/lucene/pull/827#discussion_r856173930
##
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java:
##
@@ -1112,12 +1112,17 @@ public void seekExact(long ord) throws IOException {
mikemccand commented on code in PR #633:
URL: https://github.com/apache/lucene/pull/633#discussion_r856173156
##
lucene/CHANGES.txt:
##
@@ -89,6 +89,9 @@ New Features
Improvements
-
+* LUCENE-10216: Use MergePolicy to define and MergeScheduler to trigger
rmuir commented on PR #816:
URL: https://github.com/apache/lucene/pull/816#issuecomment-1106454791
`./gradlew check` still fails. looks like the new test catches
interruptedexception and doesn't use the exception. in our codebase
compiler/linter will fail on this. please annotate the variab
rmuir commented on code in PR #827:
URL: https://github.com/apache/lucene/pull/827#discussion_r856168982
##
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java:
##
@@ -1112,12 +1112,17 @@ public void seekExact(long ord) throws IOException {
mikemccand commented on PR #633:
URL: https://github.com/apache/lucene/pull/633#issuecomment-1106450445
I restarted the `gradle test` job.
--
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 spec
mikemccand commented on PR #663:
URL: https://github.com/apache/lucene/pull/663#issuecomment-1106448433
This change looks good to me, except why did we need to resurrect
`DirectDocValuesFormat`? Was it an accident to include that in this PR maybe?
--
This is an automated message from the
mikemccand commented on PR #633:
URL: https://github.com/apache/lucene/pull/633#issuecomment-1106439994
Thanks for digging into that failure @vigyasharma.
> For the failing random seed, the test also fails on the main branch.
OK, so this is pre-existing! Could you open a separa
mikemccand commented on PR #817:
URL: https://github.com/apache/lucene/pull/817#issuecomment-1106436474
Thank you @rmuir and @dweiss and @rbowen!
--
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 t
mikemccand commented on PR #812:
URL: https://github.com/apache/lucene/pull/812#issuecomment-1106435368
Thanks @ChrisHegarty! What an awesome improvement.
Some of the [nightly faceting
tasks](https://home.apache.org/~mikemccand/lucenebench/) got a nice bump, e.g.
[pure browse hierar
rmuir commented on PR #817:
URL: https://github.com/apache/lucene/pull/817#issuecomment-1106409139
> A patch has been accepted in spotless so we can fix this without any hacks
and in a clean way after the next spotless release.
thank you @dweiss!
--
This is an automated message fro
boicehuang commented on code in PR #816:
URL: https://github.com/apache/lucene/pull/816#discussion_r855931374
##
lucene/core/src/java/org/apache/lucene/util/CloseableThreadLocal.java:
##
@@ -123,12 +122,34 @@ public void close() {
// Clear the hard refs; then, the only rema
boicehuang commented on code in PR #816:
URL: https://github.com/apache/lucene/pull/816#discussion_r855737016
##
lucene/core/src/java/org/apache/lucene/util/CloseableThreadLocal.java:
##
@@ -123,12 +122,34 @@ public void close() {
// Clear the hard refs; then, the only rema
jpountz commented on code in PR #827:
URL: https://github.com/apache/lucene/pull/827#discussion_r855854944
##
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90DocValuesProducer.java:
##
@@ -1112,12 +1112,17 @@ public void seekExact(long ord) throws IOException {
60 matches
Mail list logo
