Please explain.

[abalter]

I'm so daft, I don't really understand what this package (or bplyr, freebase, tbltools, ...) does?

Exacly what do I gain by having "dependency-free" versions of {dplyr} verbs if I now have this as a dependency?

I am genuinely asking, not trolling.

[barryrowlingson]

By adding any one package to your project as a dependency, you are implicitly adding all of its dependencies too. In the case of poorman thats... nothing. In the case of dplyr that's 9 external dependencies which all have more dependencies of their own. Why does a package for manipulating data frames need a package for formatting output in colour (cli), or pasting strings together (glue) or colouring table columns (pillar), or any of another half-a-dozen things? If those packages introduce incompatible changes, then dplyr breaks. If you trust that the developers will never let this happen then fine, carry on using it, but if all you want to do is select, filter and mutate then its likely that poorman will be stable.

Or remove these calls entirely and use base R functionality, which will be even more stable.

For more on this, the tinyverse philosophy starts here: https://www.tinyverse.org/

[abalter]

Thank you for explaining! I very much like the philosophy. Dependency hell is a real problem, especially in the R-verse.

At some point I think/hope tidyverse may become base R.