paleolimbot commented on PR #34825: URL: https://github.com/apache/arrow/pull/34825#issuecomment-1503313833
I can also attest to having to `as.data.frame(as.data.frame(arrow_stuff))` and having it be really annoying. This usually comes up in tests where `expect_identical()` complains or in benchmarks where `bench::mark()` complains about non-equality. If it was okay for `as.data.frame()` to return a tibble, then `as.data.frame()` on a tibble would return a tibble (it does not). I also do not see the point of lossless roundtrip to/from a file by default. (The option should certainly exist to the extent we have the capacity to support it). We are as a package in a place where we need to move towards simplicity to reflect the fact that we have very, very, very few contributors. I do not think that having the end result of this PR be "it was too hard so we didn't fix it" is a sustainable path. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
