Release ggplot2 3.2.0

[thomasp85]

Prepare for release:

• devtools::check()
• devtools::check_win_devel()
• rhub::check_for_cran()
• revdepcheck::revdep_check(num_workers = 4)
• Polish NEWS
• Polish pkgdown reference index
• Draft blog post

Submit to CRAN:

• usethis::use_version('minor')
• Update cran-comments.md
• devtools::submit_cran()
• Approve email

Wait for CRAN...

• Accepted 🎉
• usethis::use_github_release()
• usethis::use_dev_version()
• Finish blog post
• Tweet
• Add link to blog post in pkgdown news menu

[thomasp85]

@clauswilke, @yutannihilation, @karawoo, @dpseidel, and @hadley - Thanks for the work on finishing the release. We got all milestone issues merged just in time. We'll have a feature freeze for now so please don't merge any PR's in until 3.2.0 is released.

We'll start revdepcheck and I'll get back to you if help is needed here. @paleolimbot is interning with RStudio right now so I'll probably try to handle most of the issues together with him

[clauswilke]

Updated revdeps results are currently in #3303, in case anybody is looking for them.

[hadley]

When is ggplot2 scheduled to be released?

[clauswilke]

I don't think anybody has made a concrete schedule yet.

We still need to do a little more triaging of the revdeps. Let's say that is completed by Monday, May 13. Then we need to give maintainers of affected packages at least 2 additional weeks to fix things. So how about Monday, May 27?

[clauswilke]

I wanted to summarize the current status of where we are with the revdep check. We may need to delay our schedule until @thomasp85 is fully back.

• Release notes for 3.2.0 need to mention possible failures of downstream visual tests #3304 Improved release notes. I have a PR (Improved release notes #3324).
• pathGrob usage failure #3312 pathGrob() failure. @thomasp85 wanted to look into this.
• facet_wrap() and facet_grid() with a date variable is broken #3313 facet_wrap() and facet_grid() with a date variable is broken. We don't have a good course of action for this one yet. The primary cause seems to be differences between plyr functions and their ggplot2 replacements. See also Differences between plyr functions and functions in compat-plyr.R are causing failures #3315.
• Violin plots with too few observations are broken #3314, Annotations with zero-length aesthetics are broken #3317 Issues with zero-length data. I have a PR (Allow empty annotations. #3320). It would be good to get @thomasp85's review, since this touches his code.
• Differences between plyr functions and functions in compat-plyr.R are causing failures #3315 Differences between plyr functions and their ggplot2 replacements. It would be good if @thomasp85 could take a closer look.

[thomasp85]

I hope to be able to begin a very limited amount of work sometime next week, but experience have thought me that I’m always overly optimistic (I kept saying next week for 3 month last time — this is not as bad though)

[smouksassi]

Get well soon thomasp85 number one recommendation is to reduce screen time and stop work. We want you back soon take the time for full recovery. I have dealt with several colleagues and family who had concussion. It might be wiser for the team to push back the release. ( have seen many cases where concussion recovery was not understood and underestimated especially by the patient).

https://www.healthline.com/health/concussion-recovery#recovery-tips

(I am concerned as a health care professional pharmacist) I hope you will not even read this reply thomasp58 feel free to delete this comment.

[thomasp85]

I've got a blog post draft at tidyverse/tidyverse.org#289 - can people give it a rundown and see if I've missed something important..?

[thomasp85]

All further release development is now happening on the v3.2.0-rc branch so please make sure to merge things that should be part of 3.2.0 into that branch instead of master. This also means that master is again open for merging of code that will not be part of 3.2.0

[thomasp85]

We will release in a week (Monday next week). All revdeps have been fixed (except for those we've decided not to), but one package hadn't been contacted about breaking changes so we'll give them a week to fix before releasing

[paleolimbot]

Just contacted the last developer!

[paleolimbot]

FYI, the last developer just released a new version of his package to CRAN and he's fixed the boxplot that was previously failing.

[thomasp85]

ggplot2 3.2.0 just submitted 🤞

[clauswilke]

Thanks for all your effort on this!

[thomasp85]

Haha — this has been 90% @paleolimbot and you, due to my strategically well-scheduled concussion. Thanks to both of you

[lock]

This old issue has been automatically locked. If you believe you have found a related problem, please file a new issue (with reprex) and link to this issue. https://reprex.tidyverse.org/