Say we go a conservative approach and we only load custom-file automatically if set by the user.
We wait to see if people like it and then include a message when custom-file was already loaded by the user.
And finally we attempt the 'revolution' setting a default value for custom-file and requiring users to set it to nil if they want the original emacs behaviour.

Attached is an attempt at a patch for the first step:

/PA

On Wed, 5 Jan 2022 at 11:54, Colin Baxter 😺 <m43cap@yandex.com> wrote:
>>>>> Robert Pluim <rpluim@gmail.com> writes:

>>>>> On Wed, 05 Jan 2022 17:17:12 +0800, LdBeth <andpuke@foxmail.com> said:
>>>>> In <87bl0q8vfa.fsf@yahoo.com>
    >>>>>>> Po Lu <luangruo@yahoo.com> wrote:

    >>> Most people were certainly happy for at least the past decade.

    LdBeth> I think saving custom set variables to init file somehow
    LdBeth> prevents using byte compiled init.el file effectively
    LdBeth> (unless the user hooks auto compile whenever it is changed
    LdBeth> by emacs). From that perspective, I'm happy to see that this
    LdBeth> behavior is to be changed.

    > This is one of the two things that people do with Emacs that I
    > just donʼt understand. My init file contains setq and
    > custom-set-variables and key bindings. Any actual code that would
    > benefit from byte-compilation is stored in separate files. So why
    > do people byte-compile their init files?

Well, I can't speak for "people" but I can say why I byte compile my
~/.emacs, which I've done for many years. I have found it to be a useful
check on my lisp in finding silly errors. It has also on more than one
occasion allowed me to run emacs when my ~/.emacs was either corrupted
or "helpfully" over-written by some application. I wouldn't tell other
users to do the same - what they do is up to them.

Colin



--
Fragen sind nicht da um beantwortet zu werden,
Fragen sind da um gestellt zu werden
Georg Kreisler