Kirill Müller

Suggest finalize_version(push = TRUE) after tag_version()

1

for patch releases in non-CRAN packages.

pre_release(): preflight check

1

- [ ] On main branch - [x] `GITHUB_PAT` available - [ ] Branch with new version doesn't exist yet - [ ] Remote branch with new version doesn't exist...

pre_release(): Check for "Remotes" field?

1

It's picked up by winbuilder, I sometimes wish it'd be picked up earlier. Let's fail when a `"Remotes:"` field is present, with instructions how to proceed.

Run devtools::document() as part of bump_version

5

because it's forgotten too often.

There are a few ways where `pre_release()` can fail: - [ ] `GITHUB_PAT` not set (should we use gitcreds)? - [ ] The remote is not a GitHub remote (we...

Had to merge manually with RSQLite 2.2.1.

similarly to `fct_lump()` . ``` r library(forcats) x [1] 4 1 3 1 fct_infreq(x) #> [1] a b c b #> Levels: b a c fct_infreq(x, w = w) #>...

Improve fct_lump() behavior for factors with many rare levels

3

Example: ``` r library(magrittr) N

For https://github.com/cynkra/dm/issues/1187. We could take a look at #897 as part of it too. The current implementation never forwards to `pillar::tbl_sum()`, which makes it difficult to implement mixins that contribute...

MSSQL has no `TRUE` literal, but it's returned from translations: https://stackoverflow.com/a/7171264/946850. dm uses a home-grown routine to insert records into MS SQL Server, via `dbplyr::escape()` . This breaks with dbplyr...