Soft deprecate type_sum once vctrs is on CRAN

[hadley]

And replace type_sum.default with a call to vctrs::vec_ptype_abbr()

I think this makes sense because (unless you want to control precise details of printing) then you should only need to define methods for functions in vctrs when you create a new vector class.

Technically, vec_ptype_abbr() has more limited scope than type_sum(), since it only applies to vectors, but this shouldn't be a problem for pillar/tibble since we're only concerned with things that can go in data frames.

[krlmlr]

We could remove all specializations and only implement type_sum.default() as forwarder to vec_ptype_abbr() . Then, contact the maintainers who have implemented type_sum() in one of their packages, and remove eventually.

Is vec_ptype_abbr() stable, or do you prefer waiting a bit?

[hadley]

Yeah, I think that's right.

I don't think vec_ptype_abbr() is quite stable yet — we need to think a bit more about the family of functions that generate type names for use in different locations — r-lib/vctrs#139

[github-actions]

This old thread has been automatically locked. If you think you have found something related to this, please open a new issue and link to this old issue if necessary.