paleolimbot commented on PR #34825: URL: https://github.com/apache/arrow/pull/34825#issuecomment-1496474390
> I believe we should avoid adding a hard dependency on tibble if we can avoid it We already depend on all of tibble's dependencies because we have a hard dependency on vctrs...I don't think this is an issue. > and provide a way to return a plain old base R data.frame for users who prefer that Users that need a plain data.frame can do `as.data.frame(read_whatever())`? (They need to do that today, anyway, because whoever wrote the data *might* have written a tibble). We adopt the tidyverse paradigm almost everywhere else in an attempt to minimize the number of things users have to get used to to use Arrow (e.g., our function arguments intentionally mirror readr's). It's not that wanting a base R data frame is a bad thing or that users shouldn't want it, it's that it's an inconsistent design choice with respect to the rest of the package. ...or if returning a tibble really *is* a problem, we should commit to returning only data.frames, because returning a tibble only sometimes is the worst of all worlds. -- 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]
