[R-package] [ci] update to R 4.0.2 for Windows CI jobs (fixes #3191) #3193
Conversation
| @@ -165,7 +165,7 @@ if ($env:COMPILER -ne "MSVC") { | |||
| } else { | |||
| $env:TMPDIR = $env:USERPROFILE # to avoid warnings about incremental builds inside a temp directory | |||
| $INSTALL_LOG_FILE_NAME = "$env:BUILD_SOURCESDIRECTORY\00install_out.txt" | |||
| Run-R-Code-Redirect-Stderr "source('build_r.R')" *> $INSTALL_LOG_FILE_NAME ; $install_succeeded = $? | |||
| Run-R-Code-Redirect-Stderr "source('build_r.R')" 1> $INSTALL_LOG_FILE_NAME ; $install_succeeded = $? | |||
There was a problem hiding this comment.
Oh, wait, does it mean that we lose all not success logs that are written not to stdout?
There was a problem hiding this comment.
We already lost any of them by the changes from #3168 , I think. This step will still break if something is broken, but to get better logs we'd have to re-run with this set to Rscript build_r.R instead of using Run-R-Code-Redirect-Stderr.
I know that's bad, but the main thing is that we know the build will fail if a chance breaks the R package. I hope that PowerShell/PowerShell#12823 will get resolved and GitHub Actions will be updated soon, then maybe we can remove the workaround to suppress stderr
|
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. |
This PR attempts to fix #3191 , and issue that is blocking other PRs. I'm not sure why, but updating to R 4.0.2 and changing the output redirection strategy from builds seems to be enough to fix the issues documented there.
The R 4.0.2 release notes are here, if you're curious: https://cran.r-project.org/doc/manuals/r-devel/NEWS.html
I tried this on my fork, and got 3 successful builds of R 4.0 + MSVC and 3 successful builds of R 4.0 + MINGW: https://github.com/jameslamb/LightGBM/pull/32/checks?check_run_id=813328837