[DO NOT MERGE] Release 3.3.2

[StrikerRUS]

This is unusual release because it doesn't include all changes made since v3.3.1 release in master, but only contains urgent maintenance change to make LightGBM compatible with R 4.2. Refer to #4923 for more details.

Tags and artifacts are made directly from release_332 branch instead of master as usual, and this PR shouldn't be ever merged.

Release checklist:

• Update VERSION.txt number.
• Update version in Appveyor config file.
• Update version in configure file of R-package: /gha run r-configure.
• Change development.mode from unreleased to release in pkgdown config file.
• [ ] Run R valgrind checks after all PRs are merged: /gha run r-valgrind. (cannot be run due to conflicts with master)
• [ ] Run R Solaris checks after all PRs are merged: /gha run r-solaris. (cannot be run due to conflicts with master)
• Push commit with v* tag to trigger GitHubRelease action at Azure Pipelines.
• Convert automatic GitHub release from Draft to normal one.
• Update stable tag at GitHub.
• Trigger RTD build for new tag.
• Upload release to PyPI.
• Upload release to CRAN.
• Update cran-comments.md when new release is accepted on CRAN.
• Upload release to NuGet.
• Update version and commit hash in Homebrew formula.

old changelog might be needed in the future

Changes

💡 New Features

• [R-package] added argument eval_train_metric to lgb.cv() (fixes [R-package] add flag of displaying train loss for lgb.cv() #4911) @mayer79 ([R-package] added argument eval_train_metric to lgb.cv() (fixes #4911) #4918)
• Add support for Visual Studio 2022 @StrikerRUS (Add support for Visual Studio 2022 #4889)
• Add C API function that returns all parameter names with their aliases @StrikerRUS (Add C API function that returns all parameter names with their aliases #4829)
• [python][sklearn] respect parameters for predictions in init() and set_params() methods @StrikerRUS ([python][sklearn] respect parameters for predictions in init() and set_params() methods #4822)
• Add customized parser support @chjinche (Add customized parser support #4782)
• [R-package] Add print() and summary() methods for Booster @david-cortes ([R-package] Add print() and summary() methods for Booster #4686)
• Add 'nrounds' as an alias for 'num_iterations' (fixes [R-package] Add 'nrounds' as an alias for 'num_iterations' #4743) @mikemahoney218 (Add 'nrounds' as an alias for 'num_iterations' (fixes #4743) #4746)
• [python-package] early stopping min_delta (fixes Feature Request: customizable early_stopping_tolerance #2526) @jmoralez ([python-package] early stopping min_delta (fixes #2526) #4580)
• [python][sklearn] respect objective aliases @StrikerRUS ([python][sklearn] respect objective aliases #4758)
• [python][sklearn] add n_estimators_ and n_iter_ post-fit attributes @StrikerRUS ([python][sklearn] add n_estimators_ and n_iter_ post-fit attributes #4753)

🔨 Breaking

• [python] remove early_stopping_rounds argument of train() and cv() functions @StrikerRUS ([python] remove early_stopping_rounds argument of train() and cv() functions #4908)
• [python] remove evals_result argument of train() function @StrikerRUS ([python] remove evals_result argument of train() function #4882)
• [python][sklearn] do not replace empty dict with None for evals_result_ @StrikerRUS ([python][sklearn] do not replace empty dict with None for evals_result_ #4884)
• [python] Drop Python 3.6 support @StrikerRUS ([python] Drop Python 3.6 support #4891)
• [python] remove verbose_eval argument of train() and cv() functions @StrikerRUS ([python] remove verbose_eval argument of train() and cv() functions #4878)
• [python] remove verbose argument of model_from_string() method of Booster class @StrikerRUS ([python] remove verbose argument of model_from_string() method of Booster class #4877)
• [python][sklearn] Remove early_stopping_rounds argument of fit() method @StrikerRUS ([python][sklearn] Remove early_stopping_rounds argument of fit() method #4846)
• [R-package] remove support for '...' in slice() @jameslamb ([R-package] remove support for '...' in slice() #4872)
• [R-package] remove support for '...' in lgb.Dataset() @jameslamb ([R-package] remove support for '...' in lgb.Dataset() #4874)
• [R-package] remove support for '...' in dim.lgb.Dataset() @jameslamb ([R-package] remove support for '...' in dim.lgb.Dataset() #4873)
• [R-package] remove support for '...' in lgb.train() @jameslamb ([R-package] remove support for '...' in lgb.train() #4863)
• [R-package] remove support for '...' in create_valid() @jameslamb ([R-package] remove support for '...' in create_valid() #4865)
• [R-package] remove support for 'info' in Dataset @jameslamb ([R-package] remove support for 'info' in Dataset #4866)
• [R-package] remove Dataset getinfo() @jameslamb ([R-package] remove Dataset getinfo() #4864)
• [R-package] remove support for '...' in lgb.cv() @jameslamb ([R-package] remove support for '...' in lgb.cv() #4860)
• [R-package] remove Dataset setinfo() @jameslamb ([R-package] remove Dataset setinfo() #4854)
• [R-package] remove support for '...' in predict() @jameslamb ([R-package] remove support for '...' in predict() #4857)
• [R-package] remove support for '...' in Booster reset_parameter() @jameslamb ([R-package] remove support for '...' in Booster reset_parameter() #4856)
• [python][sklearn] unify values of best_iteration for sklearn and standard APIs @StrikerRUS ([python][sklearn] unify values of best_iteration for sklearn and standard APIs #4845)
• [ci] migrate CI from macOS 10.14 to 10.15 and drop support of Mojave @StrikerRUS ([ci] migrate CI from macOS 10.14 to 10.15 and drop support of Mojave #4849)
• [R-package] enable saving Booster with saveRDS() and loading it with readRDS() (fixes [R-package] Request: work with R serialization functions #4296) @david-cortes ([R-package] enable saving Booster with saveRDS() and loading it with readRDS() (fixes #4296) #4685)
• [python][sklearn] remove verbose argument from fit() method @StrikerRUS ([python][sklearn] remove verbose argument from fit() method #4832)
• [python] remove learning_rates argument of train() function @StrikerRUS ([python] remove learning_rates argument of train() function #4831)
• [python] remove "auto" value of ylabel argument of plot_metric() function @StrikerRUS ([python] remove "auto" value of ylabel argument of plot_metric() function #4818)
• [python] Remove print_evaluation() function @StrikerRUS ([python] Remove print_evaluation() function #4819)
• [python] Remove silent argument @StrikerRUS ([python] Remove silent argument #4800)

🚀 Efficiency Improvement

• clear memory allocated for sampled data when constructing Dataset from text file @xuchuanyin (clear memory allocated for sampled data when constructing Dataset from text file #4890)
• [python] Faster categorical column names selection @Neronuser ([python] Faster categorical column names selection #4787)
• [R-package] parallelize compilation in CMake-based builds @jameslamb ([R-package] parallelize compilation in CMake-based builds #4525)

🐛 Bug Fixes

• [R-package] respect 'verbose' argument in lgb.cv() (fixes [R-package] lgb.cv ignores verbose argument #4667) @jameslamb ([R-package] respect 'verbose' argument in lgb.cv() (fixes #4667) #4903)
• [R-package] Apply patch for R4.2 on Windows @shiyu1994 ([R-package] Apply patch for R4.2 on Windows #4923)
• [R-package] respect aliases for objective and metric and lgb.train() and lgb.cv() @jameslamb ([R-package] respect aliases for objective and metric and lgb.train() and lgb.cv() #4913)
• [python] raise an informative error instead of segfaulting when custom objective produces incorrect output @yaxxie ([python] raise an informative error instead of segfaulting when custom objective produces incorrect output #4815)
• [R-package] fix handling of duplicate parameters (fixes [R-package] Possibly-incorrect handling of duplicate parameters #4521) @jameslamb ([R-package] fix handling of duplicate parameters (fixes #4521) #4914)
• [R-package] update parameter 'verbosity' based on keyword arg 'verbose' @jameslamb ([R-package] update parameter 'verbosity' based on keyword arg 'verbose' #4899)
• [R-package] fix CVBooster reset_parameter() method (fixes [R-package] CVBooster$reset_parameter() is broken #4900) @jameslamb ([R-package] fix CVBooster reset_parameter() method (fixes #4900) #4901)
• [python] reset storage in record evaluation callback each time before starting training @StrikerRUS ([python] reset storage in record evaluation callback each time before starting training #4885)
• [python] reset storages in early stopping callback after finishing training @StrikerRUS ([python] reset storages in early stopping callback after finishing training #4868)
• [R-package] fix --no-build-vignettes option for build-cran-package.sh @jameslamb ([R-package] fix --no-build-vignettes option for build-cran-package.sh #4848)
• [python][docs] fix type hints for custom functions and remove vague array-like wording @StrikerRUS ([python][docs] fix type hints for custom functions and remove vague array-like wording #4816)
• Always respect forced splits, even when feature_fraction < 1.0 (fixes Using feature_fraction and forced splits together can cause training failures #4601) @tongwu-msft (Always respect forced splits, even when feature_fraction < 1.0 (fixes #4601) #4725)
• Reset OpenMP thread number if num_threads <= 0 @hzy46 (Reset OpenMP thread number if num_threads <= 0 #4704)

📖 Documentation

• [R-package] [docs] fix calculation of R test coverage (fixes [R-package] [docs] instructions for generating code coverage produce incorrect results #4919) @jameslamb ([R-package] [docs] fix calculation of R test coverage (fixes #4919) #4922)
• [docs] Update link to FLAML code example @qingyun-wu ([docs] Update link to FLAML code example #4892)
• [R-package] [docs] add Michael Mayer to DESCRIPTION @jameslamb ([R-package] [docs] add Michael Mayer to DESCRIPTION #4867)
• [docs] document that pred_early_stop can be used only in normal and raw scores prediction @StrikerRUS ([docs] document that pred_early_stop can be used only in normal and raw scores prediction #4823)
• [R-package] [docs] add intro vignette ([R-package] [docs] add intro vignette #3946) @jameslamb ([R-package] [docs] add intro vignette (#3946) #4775)
• [docs] fix broken SynapseML link @StrikerRUS ([docs] fix broken SynapseML link #4795)
• [docs] [dask] Add return information to Dask fit() docs (fixes [python] Add Returns section to docstrings of fit() methods of Dask estimators #4402) @jameslamb ([docs] [dask] Add return information to Dask fit() docs (fixes #4402) #4716)
• [docs] add specific estimator names to sklearn fit() docs @jameslamb ([docs] add specific estimator names to sklearn fit() docs #4774)
• [docs] [R-package] update cran-comments for v3.3.1 release @jameslamb ([docs] [R-package] update cran-comments for v3.3.1 release #4738)
• [ci] only use conda-forge when installing R packages in docs builds @jameslamb ([ci] only use conda-forge when installing R packages in docs builds #4767)
• [docs] improve docs about nthreads parameter @StrikerRUS ([docs] improve docs about nthreads parameter #4756)
• [docs] update link to LightGBM Ruby @StrikerRUS ([docs] update link to LightGBM Ruby #4740)
• [docs] fix broken SynapseML link @jameslamb ([docs] fix broken SynapseML link #4712)

🧰 Maintenance

• [R-package] [tests] remove uses of testthat::context() @jameslamb ([R-package] [tests] remove uses of testthat::context() #4915)
• [R-package][tests] remove uses of testthat::expect_is() @jameslamb ([R-package][tests] remove uses of testthat::expect_is() #4916)
• [R-package] remove unnecessary comments and fix typos in comments @jameslamb ([R-package] remove unnecessary comments and fix typos in comments #4902)
• [R-package] reduce verbosity in tests using lgb.train() @jameslamb ([R-package] reduce verbosity in tests using lgb.train() #4896)
• [tests][python-package] change boston dataset to synthetic dataset in tests that don't check score @jmoralez ([tests][python-package] change boston dataset to synthetic dataset in tests that don't check score #4895)
• [R-package] reduce verbosity in some unit tests @jameslamb ([R-package] reduce verbosity in some unit tests #4879)
• [R-package] remove unused callback cb.reset.parameter @jameslamb ([R-package] remove unused callback cb.reset.parameter #4871)
• [python-package][dask] handle failures parsing worker host names @jameslamb ([python-package][dask] handle failures parsing worker host names #4852)
• [python-package] [dask] fix mypy error about worker_addresses @jameslamb ([python-package] [dask] fix mypy error about worker_addresses #4851)
• [python] remove workaround for UnboundLocalError in early stopping callback for Python 2 @StrikerRUS ([python] remove workaround for UnboundLocalError in early stopping callback for Python 2 #4855)
• [python-package] remove unused imports @jameslamb ([python-package] remove unused imports #4850)
• [python-package] fix mypy errors in engine.py @jameslamb ([python-package] fix mypy errors in engine.py #4839)
• [python-package] [dask] fix mypy errors from dask.py docstrings @jameslamb ([python-package] [dask] fix mypy errors from dask.py docstrings #4844)
• [python-package] fix mypy errors in sklearn.py @jameslamb ([python-package] fix mypy errors in sklearn.py #4837)
• [tests][dask] fix argument names in custom eval function in Dask test @StrikerRUS ([tests][dask] fix argument names in custom eval function in Dask test #4833)
• [python-package] fix mypy errors in plotting.py @jameslamb ([python-package] fix mypy errors in plotting.py #4838)
• [python-package] fix mypy error about missing type hint in dask.py @jameslamb ([python-package] fix mypy error about missing type hint in dask.py #4840)
• [python-package] fix mypy errors in compat.py and setup.py @jameslamb ([python-package] fix mypy errors in compat.py and setup.py #4836)
• [docs][python] simplify mocking in docs @StrikerRUS ([docs][python] simplify mocking in docs #4830)
• [python] add type hints to _compare_params_for_warning() and make it reusable @StrikerRUS ([python] add type hints to _compare_params_for_warning() and make it reusable #4824)
• [R-package] Move R6 to Imports @david-cortes ([R-package] Move R6 to Imports #4812)
• [ci] Fix CI job for R artifact @StrikerRUS ([ci] Fix CI job for R artifact #4811)
• [ci] fix getting status of optional workflows in PRs with a lot of comments @StrikerRUS ([ci] fix getting status of optional workflows in PRs with a lot of comments #4806)
• [python] add type hints for custom objective and metric functions in scikit-learn interface @jameslamb ([python] add type hints for custom objective and metric functions in scikit-learn interface #4547)
• [c_api] Improve ANSI compatibility by avoiding <stdbool.h> @drewmiller ([c_api] Improve ANSI compatibility by avoiding <stdbool.h> #4697)
• [ci] ignore CMakeLint errors related to package names @StrikerRUS ([ci] ignore CMakeLint errors related to package names #4801)
• [ci] fix CMakeLint errors related to function naming case @StrikerRUS ([ci] fix CMakeLint errors related to function naming case #4794)
• [ci] simplify processing of CMAKE_CUDA_FLAGS variable in CMakeLists.txt @StrikerRUS ([ci] simplify processing of CMAKE_CUDA_FLAGS variable in CMakeLists.txt #4799)
• [ci] fix CMakeLint linelength errors @StrikerRUS ([ci] fix CMakeLint linelength errors #4796)
• [ci] use pure endif() and endfunction() without expression inside in CMakeLists.txt @StrikerRUS ([ci] use pure endif() and endfunction() without expression inside in CMakeLists.txt #4798)
• cmake: use object library to avoid duplicate compilation. @cyfdecyf (cmake: use object library to avoid duplicate compilation. #4489)
• Suppress categorical warning (fixes [python] suppress the warning about categorical feature override #3379) @hzy46 (Suppress categorical warning (fixes #3379) #4768)
• [ci][tests][python] remove assertion for filename that is no longer true with new version of graphviz @StrikerRUS ([ci][tests][python] remove assertion for filename that is no longer true with new version of graphviz #4778)
• [ci] Revert temp workaround for freetype version @StrikerRUS ([ci] Revert temp workaround for freetype version #4776)
• [ci] increase timeout for valgrind job to 240 minutes @jameslamb ([ci] increase timeout for valgrind job to 240 minutes #4773)
• [python] improve warning message about aliases in cv() function @StrikerRUS ([python] improve warning message about aliases in cv() function #4766)
• [ci] Temporary pin dask version at CI @StrikerRUS ([ci] Temporary pin dask version at CI #4770)
• [ci] use wch1/r-debug image in Solaris tests @jameslamb ([ci] use wch1/r-debug image in Solaris tests #4765)
• [ci] upgrade actions/checkout to v2.4.0 @jameslamb ([ci] upgrade actions/checkout to v2.4.0 #4763)
• [docs][python] add datatable to the mocked modules during docs building process and sort them alphabetically @StrikerRUS ([docs][python] add datatable to the mocked modules during docs building process and sort them alphabetically #4750)
• [R-package] allow use of custom R executable when building CRAN package @jameslamb ([R-package] allow use of custom R executable when building CRAN package #4754)
• [R-package] remove temporary files created in configure.win @jameslamb ([R-package] remove temporary files created in configure.win #4752)
• [ci] indicate support of Monterey @StrikerRUS ([ci] indicate support of Monterey #4732)
• [python] Make dummy classes constructible with any arguments @StrikerRUS ([python] Make dummy classes constructible with any arguments #4749)
• [ci] Temporary pin freetype version at Windows CI @StrikerRUS ([ci] Temporary pin freetype version at Windows CI #4747)
• Remove checks for label when loading dataset from binary file because label is ignored in that case @StrikerRUS (Remove checks for label when loading dataset from binary file because label is ignored in that case #4737)
• [tests][python] add test for non-serializable callback @StrikerRUS ([tests][python] add test for non-serializable callback #4741)
• Improve warning wordings @StrikerRUS (Improve warning wordings #4731)
• [ci] Bump version for development @StrikerRUS ([ci] Bump version for development #4730)
• Add some warnings when loading dataset from binary file @StrikerRUS (Add some warnings when loading dataset from binary file #4724)

[StrikerRUS]

/gha run r-configure

[StrikerRUS]

3.3.2 release is drafted and ready for the review. Seems that all artifacts have been generated successfully.
https://github.com/microsoft/LightGBM/releases/tag/untagged-63c8f1e7a23d1a0a3cf2
https://github.com/microsoft/LightGBM/commits/v3.3.2
v3.3.1...v3.3.2

@jameslamb @shiyu1994 @jmoralez

[jameslamb]

Thanks for doing this!

Ok with me that we don't have the valgrind and solaris checks on this PR, since they ran successfully on #4923. I pull this branch today just to double-check, and can confirm (from running git log -n 10) that this only contains the new patch applied on top of v3.3.1.

One thing I'm unsure about is the release tag. Will we have to permanently keep this release_332 branch in the repo from this point forward? Since the tagged commit (dce7e58) will never be on master.

If so I think that's fine, but we should add a branch protection on release_332 to prevent deleting it or pushing changes to it without a pull request (GitHub docs). I don't have sufficient permissions to do that on this repo, so it will have to be you (if you have the permissions) or @shiyu1994.

[jameslamb]

Once @shiyu1994 approves as well, I can handle uploading this release to CRAN.

[StrikerRUS]

@jameslamb

Will we have to permanently keep this release_332 branch in the repo from this point forward?

I'll check this.

[shiyu1994]

LGTM. @StrikerRUS @jameslamb Thanks you!

[StrikerRUS]

OK, thanks for the feedback, guys!

New GitHub release is out now and available for public access.

[jameslamb]

Excellent! I'll submit to CRAN shortly and let you both know here when I do.

I think we can close this pull request.

[StrikerRUS]

Yeah, thanks!

Everything is done according to the release checklist, except CRAN upload.

Homebrew formula is under review right now: Homebrew/homebrew-core#92706.

I'll submit to CRAN shortly and let you both know here when I do.

I believe we should remind them to remove auto-patching in comments during upload.

[jameslamb]

I believe we should remind them to remove auto-patching in comments during upload.

Good point! I'll do that.

[StrikerRUS]

Will we have to permanently keep this release_332 branch in the repo from this point forward?

I'll check this.

I've checked this and it seems we can remove underlying branch. I created test branch with test commit. Then I pushed test tag referring test commit. After that I removed entire test branch. And here's what I've got:

[jameslamb]

I've checked this and it seems we can remove underlying branch

I'm nervous about this. I've seen that when you delete a branch in the UI, there is a period of time where GitHub keeps it around and you can click "restore branch" to bring it back. Then, after some time (I do not know how long, but on the order of hours or days, not weeks or months), that's no longer possible. I wonder if there is the same "keep around for a short time and then permanently delete" behavior with commits not attached to a branch.

I have two requests:

1. Could we wait a day and see if that commit is still visible?
2. Can you check that it's actually possible to checkout to that tag in a local git client (not the GitHub UI)?

   git fetch origin --tags
   git checkout test_tag

[StrikerRUS]

OK, I'll do this.

But please note a yellow note of the screenshot above that indicates that the branch was removed.

[StrikerRUS]

Also, our scenario is described here:

The tag and commit would still exist if the branch is deleted.
https://github.saobby.my.eu.orgmunity/t/what-happens-to-a-tag-on-a-deleted-branch/195995/2

If you delete a branch that a tag was created from, this will have no effect on the tag.
https://stackoverflow.com/a/33283539

[StrikerRUS]

One thing that doesn't work with such type of releases is release auto-changelog.

We are lacking all items since v3.3.1. Luckily, I was expecting this and preserved old changelog draft here #4930 (comment). So, we just need manually merge that preserved changelog with new items that will be added since today for v4.0.0 release.

[jameslamb]

I just uploaded v3.3.2 to CRAN.

@shiyu1994 you should receive an email from CRAN asking you to approve this submission.

I included the following note with this submission (will add this to cran-comments.md in a PR)

I am submitting this new version of {lightgbm} on behalf of its maintainer, Yu Shi (yushi2@microsoft.com), with his permission.

This release, v3.3.2, is identical to the existing release on CRAN (v3.3.1) except in one way: it includes the patch at https://svn.r-project.org/R-dev-web/trunk/WindowsBuilds/winutf8/ucrt3/r_packages/patches/CRAN/lightgbm.diff, which Tomas Kalibera and Uwe Ligges emailed us about.

If this submission is approved, please inform Tomas and Uwe so that patch can be removed.

[shiyu1994]

@jameslamb Upload is now confirmed. Thank you!

[shiyu1994]

The newly submitted CRAN package did not pass the pretest. Following is the email from CRAN. This seems because CRAN still automatically patches socket_wrapper.hpp and causes compilation error:
https://win-builder.r-project.org/incoming_pretest/lightgbm_3.3.2_20220108_034245/Windows/00install.out
Following is the full text of the email from CRAN.

Dear maintainer,

package lightgbm_3.3.2.tar.gz does not pass the incoming checks automatically, please see the following pre-tests:
Windows: https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwin-builder.r-project.org%2Fincoming_pretest%2Flightgbm_3.3.2_20220108_034245%2FWindows%2F00check.log&data=04%7C01%7Cyushi2%40microsoft.com%7C18b33ff99dec4905461308d9d28a30da%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637772314174652416%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=gunCgJBSltmLhxRc3HrUXEALn9NLCetD4dE7sI5pd7k%3D&reserved=0
Status: 1 ERROR
Debian: https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwin-builder.r-project.org%2Fincoming_pretest%2Flightgbm_3.3.2_20220108_034245%2FDebian%2F00check.log&data=04%7C01%7Cyushi2%40microsoft.com%7C18b33ff99dec4905461308d9d28a30da%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637772314174652416%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=KWmSnTXea%2BnyMVj2hiZu5qzfd8kLh6wSxf7taT0TAuQ%3D&reserved=0
Status: OK

Last released version's CRAN status: OK: 7, NOTE: 6
See: https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fcran.r-project.org%2Fweb%2Fchecks%2Fcheck_results_lightgbm.html&data=04%7C01%7Cyushi2%40microsoft.com%7C18b33ff99dec4905461308d9d28a30da%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637772314174652416%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=SH4ek74lElca%2FJBnpUbkNuFv00FHaCNxaPuyLXiX3Q4%3D&reserved=0

CRAN Web: https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fcran.r-project.org%2Fpackage%3Dlightgbm&data=04%7C01%7Cyushi2%40microsoft.com%7C18b33ff99dec4905461308d9d28a30da%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637772314174652416%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=w6RPk93QHKKKGhT9%2FhzDB5A8kRDYtUA%2B2mYbxAGRkZQ%3D&reserved=0

Please fix all problems and resubmit a fixed version via the webform.
If you are not sure how to fix the problems shown, please ask for help on the R-package-devel mailing list:
https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fstat.ethz.ch%2Fmailman%2Flistinfo%2Fr-package-devel&data=04%7C01%7Cyushi2%40microsoft.com%7C18b33ff99dec4905461308d9d28a30da%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637772314174652416%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=jtjtXAyY0DEUeCDjzDNsbewXpkW8x%2BG6TAFaXrQQOS0%3D&reserved=0
If you are fairly certain the rejection is a false positive, please reply-all to this message and explain.

More details are given in the directory:
https://nam06.safelinks.protection.outlook.com/?url=https%3A%2F%2Fwin-builder.r-project.org%2Fincoming_pretest%2Flightgbm_3.3.2_20220108_034245%2F&data=04%7C01%7Cyushi2%40microsoft.com%7C18b33ff99dec4905461308d9d28a30da%7C72f988bf86f141af91ab2d7cd011db47%7C1%7C0%7C637772314174652416%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=4jAtdHkrwCHCM1pQAEZ6KArlJuCzt8DnT7WDYZpBqBY%3D&reserved=0
The files will be removed after roughly 7 days.

No strong reverse dependencies to be checked.

Best regards,
CRAN teams' auto-check service
00details.log[33215].log

[StrikerRUS]

This seems because CRAN still automatically patches socket_wrapper.hpp and causes compilation error:

I strongly believe you are right:

** applying installation-time patches
patching file src/network/socket_wrapper.hpp
Hunk #1 FAILED at 60.
Hunk #2 succeeded at 90 with fuzz 1 (offset 4 lines).
1 out of 2 hunks FAILED -- saving rejects to file src/network/socket_wrapper.hpp.rej
WARNING: failed to apply patch patches/CRAN/lightgbm.diff

...

g++  -std=gnu++11 -I"D:/RCompile/recent/R/include" -DNDEBUG -I./include -DEIGEN_MPL2_ONLY -DMM_PREFETCH=1 -DMM_MALLOC=1 -DUSE_SOCKET -DLGB_R_BUILD    -I"d:/rtools42/x86_64-w64-mingw32.static.posix/include"  -fopenmp -pthread   -O2 -Wall  -mfpmath=sse -msse2 -mstackrealign -fexceptions  -c network/linkers_socket.cpp -o network/linkers_socket.o
In file included from network/linkers.h:22,
                 from network/linkers_socket.cpp:20:
network/socket_wrapper.hpp:332:2: error: #endif without #if
  332 | #endif   // LightGBM_NETWORK_SOCKET_WRAPPER_HPP_
      |  ^~~~~
make: *** [D:/RCompile/recent/R/etc/x64/Makeconf:259: network/linkers_socket.o] Error 1
ERROR: compilation failed for package 'lightgbm'

We again need to ask CRAN team somehow to stop patching lightgbm...

[StrikerRUS]

@jameslamb

I have two requests:

1. Could we wait a day and see if that commit is still visible?
2. Can you check that it's actually possible to checkout to that tag in a local git client (not the GitHub UI)?
   shell git fetch origin --tags git checkout test_tag

test_branch was removed ~2022-01-07T17:01:42Z. Right now it's 2022-01-09T15:14:12Z.

1. I still can access target commit at GitHub:
   
2. Locally everything is fine as well:
   

I believe we can remove release_332 branch.

[jameslamb]

Ok thanks for waiting and checking @StrikerRUS . Let's delete it.

@shiyu1994 can you reply to CRAN and ask them to please stop applying that patch?

[shiyu1994]

@shiyu1994 can you reply to CRAN and ask them to please stop applying that patch?

Sure. Done with that.

[StrikerRUS]

@shiyu1994

Sure. Done with that.

Have they answered anything?

On CRAN version is still 3.3.1 but now with r-devel-windows-x86_64-new-UL and r-devel-windows-x86_64-new-TK tests failed with network/socket_wrapper.hpp:241:43: error: call of overloaded 'inet_pton(int, const char*&, in_addr*)' is ambiguous error due to duplicated function definitions.

[shiyu1994]

They've replied that the automatic patch is removed and checks are retriggered. I just saw the status right now too. I'll send another email to ensure that lightgbm 3.3.2 is being checked.

[StrikerRUS]

I hope they have started new checks:

[StrikerRUS]

OK, they have started:

Tests on Fedora have passed already.

[StrikerRUS]

OK, all tests have passed! 🎉
https://cran.r-project.org/web/checks/check_results_lightgbm.html

Not sure about additional tests though.

[github-actions]

This pull request has been automatically locked since there has not been any recent activity since it was closed. To start a new related discussion, open a new issue at https://github.com/microsoft/LightGBM/issues including a reference to this.