Open krlmlr opened 9 years ago
I don't want to do that in this release as it would require a lot of downstream changes, but it seems like a reasonable idea.
Can we do it right after CRAN release then?
Yeah, although we'll need a plan for informing downstream maintainers. It would be helpful if you could estimate the amount of work by deprecating those functions and then running the revdep checks
How about adding now compatibility symbols in this release (rename
-> prename
et al.), to faciliate later migration?
I'll do the revdep checks.
Hmmmm, anything that requires deprecating existing (and popular) plyr functions is going to cause a lot of pain. I wonder if there's another approach
I guess you don't want to rename the corresponding dplyr
verbs either. We could as well keep everything as is, and explicitly state this incompatibility. Do you think it's worth pursuing the compatibility package I started? I think such a separate glue is preferable to including this into dplyr
.
By the same token -- there seem to be many requests for convenience functions in dplyr
that can be implemented using the API. Do we already have a dplyrAlly
?
revdep_check
is flaky on my system, I'm trying for the third time now. ~There's https://github.com/skranz/dplyrExtras.~
for full (!) forward compatibility with
dplyr
. See http://krlmlr.github.io/pdlyr for an analysis.