[DO NOT MERGE] Release v3.3.3 #5525
Conversation
|
I can work on #5513 |
|
Thanks @jmoralez ! That would be super helpful. |
This comment was marked as outdated.
This comment was marked as outdated.
jameslamb
requested review from
StrikerRUS,
jmoralez,
guolinke and
shiyu1994
as code owners
jameslamb
changed the title
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
jameslamb
force-pushed
the
release_v3.3.3
branch
from
4adafbf
to
f4895db
Compare
This comment was marked as resolved.
This comment was marked as resolved.
jameslamb
force-pushed
the
release_v3.3.3
branch
from
e5ac5ed
to
c766592
Compare
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
|
Happy to say that all of the new builds on win-builder for the current state of this branch came back successful! build statuses and logs (click me)win-builder NOTE: the I've been waiting for about 45 minutes for the R Hub ones to run and it looks like they haven't even started. I think we can be pretty confident in our changes given the win-builder successes + this project's CI, so I'm going to move forward with the release. Thanks everyone for the quick reviews, and especially to @jmoralez for all the help reviewing, testing, and investigating over the last 3 days. |
|
Just submitted to CRAN.
|
|
Closing this pull request, as no additional commits or CI runs should be required. |
|
pushed a new tag git tag v3.3.3
git push upstream v3.3.3(I use |
|
draft release was just created by Azure DevOps: https://github.com/microsoft/LightGBM/releases/tag/untagged-5ec2972aed86f421b3e7 and looks like all artifacts were uploaded successfully 🎉
|
|
release |
|
Added
And it built successfully and is now an option on the docs site. https://lightgbm.readthedocs.io/en/v3.3.3/
|
jameslamb
changed the title
|
@jameslamb @jmoralez Thank you so much for the update. lgb seems already updated on CRAN, superb job! |
|
Thanks @mayer79 ! Sorry if we made you nervous, doing the best we can 😅 |
|
updated the https://github.com/microsoft/LightGBM/tags
|
Haha, not at all :-) |
|
updated Homebrew: Homebrew/homebrew-core#112795 |
|
uploaded to PyPi! I downloaded the 5 Python distributions ( twine upload -u ${PYPI_USER} dist/*
https://pypi.org/project/lightgbm
|
|
CRAN checks are slowly proceeding. There have been 4 so far (3 on Linux, 1 on Windows).
https://cran.r-project.org/web/checks/check_results_lightgbm.html |
|
🎉 All but one CRAN checks are passing on v3.3.3. Including, importantly, the
|
|
@jameslamb Thank you very much for the hard work on this unplanned release! Your fighting with git complications is impressive! I just noticed that #5159 and #4889 were not included in the release's changelog. So, I just added them manually. |
|
New release on NuGet has been created:
|
|
BTW, can we remove |
|
oh great, thank you so much for that @StrikerRUS ! And for your awesome notes on #4930 . They were very very helpful to us in preparing this release. |
|
This pull request has been automatically locked since there has not been any recent activity since it was closed. |
Similar to #4930, this is an unusual release. It is
v3.3.3, but doesn't include all of the changes made sincev3.3.2.Instead, this PR just contains the changes on top of
v3.3.2needed to prevent the R package from being archived by CRAN.THIS PR SHOULD NEVER BE MERGED
Setup
I created a new branch,
release/v3.3.2, which contains only:v3.3.2master(82b7d31)how I did that (click me)
Tags and artifacts should be made directly from this
release_v3.3.3branch instead ofmaster.Release checklist:
Copied from #4930.
configurefile of R-package:/gha run r-configure.development.modefromunreleasedtoreleasein pkgdown config file.v*tag to triggerGitHubReleaseaction at Azure Pipelines.Draftto normal one.stabletag at GitHub.Special items for this release
rhub/debian-clang-develimage (built from fix Debian build errors related to 'useradd' r-hub/rhub-linux-builders#62 if necessary)windows-x86_64-devel(logs)windows-x86_64-oldrel(logs)windows-x86_64-release(logs)debian-clang-devel(logs)debian-gcc-devel(logs)debian-gcc-devel-nold(logs)debian-gcc-patched(logs)Running ‘testthat.R’Build timed out (after 20 minutes)on 3 attemptsdebian-gcc-release(logs)Running ‘testthat.R’Build timed out (after 20 minutes)on 3 attemptsfedora-clang-devel(logs)fedora-gcc-devel(logs)Running ‘testthat.R’Build timed out (after 20 minutes)on 3 attemptslinux-x86_64-rocker-gcc-san(logs)macos-highsierra-release(logs)macos-highsierra-release-cran(logs)macos-m1-bigsur-releaseubuntu-gcc-devel(logs)Running ‘testthat.R’Build timed out (after 20 minutes)on 3 attemptsubuntu-gcc-release(logs)Running ‘testthat.R’Build timed out (after 20 minutes)on 3 attemptsubuntu-rchk(logs)ERROR: too many states (abstraction error?)thatrchkalways raises, which we and CRAN ignore (LightGBM code link)windows-x86_64-devel(logs)windows-x86_64-oldrel(logs)windows-x86_64-patched(logs)windows-x86_64-release(logs)What about the changelog?
As described in #4930 (comment), doing this interferes with the automatic changelog generation we get from the release-draft bot.
So just like @StrikerRUS did in #4930, I've included the current draft changelog here, as well as the one saved in the comments on #4930.
changelog from v3.3.1 to when v3.3.2 was created (click me)
NOTE: I wrapped this in a markdown text block so GitHub stops re-evaluating the links.
changelog from when v3.3.2 was created to today (click me)
NOTE: I wrapped this in a markdown text block so GitHub stops re-evaluating the links.