bug#25030: 25.1; Unexpected indentation and syntax-highlighting in `emacs-lisp-mode'

[Drew Adams]

> >> This may not be a bug, but it is certainly a mis-feature.
> >>
> >> Warning should be reserved for syntax which may have unintended or
> >> surprising semantics. Indentation that does not follow a convention is
> >> not wrong either systacically or semantically.
> >
> > I'm not convinced by this.  Code with unconventional indendation has
> > surprising syntax to a human reader (or from another perspective, when
> > I'm writing code which indents strangely, that clues me in that I've
> > written some unintended syntax), therefore, it seems a warning is
> > exactly appropriate.
> 
> I disagree. The interpreter and byte compiler do not care about the
> indentation style that you choose for your code: the syntax and
> semantics are unaffected.
> 
> Style choices should not produce warnings. An indication that code layout
> is following an unusual style may be useful, but it should be optional, and
> it should not use the warning face (it should have a separate face that can
> be customised independently of the warning face).

What Andy said.  This has nothing to do with byte-compiling
(or interpreting, for that matter).

There is nothing wrong with having optional (especially opt-in)
indications of flouting conventional style.  And even then we
should not use, or inherit from, the warning face.

It should be easy for users to give the face(s) used for
stylistic highlighting different appearance(s) from standard
Emacs faces that have other meanings.

Error and warning faces are to be avoided for anything that is
not an error or warning.  And any faces used by the byte-compiler
should be about something relevant to byte-compiling.