On Tue, 26 Mar 2024 16:38:07 GMT, Andy Goryachev <ango...@openjdk.org> wrote:
> > We should replace this with the recommended sort order. > > My 2 cents would be on the other side: I would strongly recommend **against** > enforcing a specific order, especially since it causes no issues. We should > avoid wildcards, yes (except for static imports), but the order - we should > follow the current recommendation of ignoring the order. The main reason it > is adding zero value and costs some time (just like whitespace warnings in > jcheck). I agree that we shouldn't _enforce_ the order (and didn't mean to imply otherwise). I do see some value in having a "soft" recommendation, though, and encouraging IDE users to configure their IDEs with those recommendations. Otherwise, it can be a source of merge conflicts if one IDE sorts it one way, and another IDE sorts it a different way, and two different people are proposing changes to the same file. Maybe unlikely, but it has happened. So the questions are: 1. Should we have at least a soft recommendation or leave it unspecified? 2. If the answer to 1 is "yes", what should that recommendation be? ------------- PR Comment: https://git.openjdk.org/jfx/pull/1420#issuecomment-2020970308