Re: [PATCH] Make byte-compile-error-on-warn a safe variable for file compilation

[Noam Postavsky <npostavs@users.sourceforge.net>]

On Wed, Jan 10, 2018 at 2:36 PM, Eli Zaretskii <eliz@gnu.org> wrote:
>> From: Noam Postavsky <npostavs@users.sourceforge.net>
>> Date: Wed, 10 Jan 2018 11:29:45 -0500
>> Cc: Stefan Monnier <monnier@iro.umontreal.ca>, Emacs developers <emacs-devel@gnu.org>
>>
>> >> Since this autoload vs :safe property consideration applies more
>> >> generally, shouldn't we rather explain in it in the manual?
>> >
>> > I think we should do both.  The fact that some unusual technique is in
>> > the manual doesn't yet mean commentary about it will be redundant.
>>
>> But there are many other cases of this autoloaded put, should we
>> comment on all of them?
>
> Maybe we should do that on more than just this one.  But even a
> thousand-mile journey begins with a single step.
>
> Did you never have this moment of staring at a piece of code that
> doesn't explain itself and has no comments, and wondering why the heck
> did they do it that way?  It is not easy to find answers to such
> questions, even if they are in some manual.

To me, this looks like it should be considered a standard technique,
that we shouldn't need to explain over and over again (but currently
it's not even explained once in the manual, which is not good).