Hi there,

All IO methods (and Commons) have been sorted in AB order for years. It is
the simplest and easiest to mainatin and understand IMO.

If you keep it that way in your PR, then no post merge edits are needed ;-)

Gary


On Wed, Oct 15, 2025, 05:03 Piotr P. Karwasz <[email protected]>
wrote:

> Hi all,
>
> On 14.10.2025 21:02, [email protected] wrote:
> > The following commit(s) were added to refs/heads/master by this push:
> >      new f51d19cab Sort members
> > f51d19cab is described below
> >
> > commit f51d19cabad7b8c67e2992a2ab22465658249485
> > Author: Gary Gregory <[email protected]>
> > AuthorDate: Tue Oct 14 15:02:23 2025 -0400
> >
> >     Sort members
>
> Sorting class members can be quite a disruptive change, as it breaks
> most history and diff comparison algorithms. Could we please avoid such
> commits directly on the `master` branch?
>
> I’m fully in favor of following clear conventions, but I’d prefer if we
> discuss and document them before. The default alphabetical sorting used
> by Eclipse is only one of several possible approaches, and not always
> the most meaningful one. For instance, I often prefer ordering private
> helper methods according to the call graph for better readability.
>
> In this particular commit [1], the sorting order (static methods
> alphabetically followed by instance methods alphabetically) feels
> counterproductive, especially for unit test classes.
>
> For example, when working with parameterized tests, I typically use a
> pattern like:
>
>     static Stream<Arguments> testFoo() {
>         ...
>     }
>
>     @ParameterizedTest
>     @MethodSource
>     void testFoo(...) {
>         ...
>     }
>
> This grouping keeps related data providers and test methods together and
> visible on one screen. Moving all static methods before instance methods
> breaks that logical pairing and reduces readability.
>
> Piotr
>
> References:
> [1]
>
> https://github.com/apache/commons-io/commit/f51d19cabad7b8c67e2992a2ab22465658249485
>
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
>

Reply via email to