[
https://issues.apache.org/jira/browse/ARROW-10733?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Neal Richardson resolved ARROW-10733.
-------------------------------------
Resolution: Fixed
Issue resolved by pull request 9034
[https://github.com/apache/arrow/pull/9034]
> [R] Improvements to Linux installation troubleshooting
> ------------------------------------------------------
>
> Key: ARROW-10733
> URL: https://issues.apache.org/jira/browse/ARROW-10733
> Project: Apache Arrow
> Issue Type: New Feature
> Components: R
> Reporter: Neal Richardson
> Assignee: Neal Richardson
> Priority: Major
> Labels: pull-request-available
> Fix For: 3.0.0
>
> Time Spent: 1h 10m
> Remaining Estimate: 0h
>
> * Some people think that they need to install the arrow package and then call
> install_arrow... which just installs the package again. A number of packages
> with external dependencies (python, etc.) have this pattern, so it's
> understandable why people are confused.
> * If installation fails, you get the arrow-without-arrow package, which tells
> you to run install_arrow. This was reasonable advice when we first got on
> CRAN and the CRAN version did not build the C++ library but install_arrow
> would. But now that installation from CRAN will try to build the C++ library,
> if installation failed the first time, it will probably fail the second time
> too. Some possible improvements: (1) don't recommend install_arrow in that
> error message; (2) set ARROW_R_DEV=true in install_arrow for verbosity,
> detect if installation failed, and direct the user to report the full logs if
> it did; (3) if the user is on centos-7, print a special message pointing to
> the installation troubleshooting.
> * The linux installation vignette's troubleshooting section should
> prominently recommend retrying with ARROW_R_DEV=true and reporting the logs
> in your bug report. It's usually the first thing I ask in response to a bug
> report.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)