Re: GNU Emacs raison d'etre

[Dmitry Gutov]

On 20.05.2020 22:53, address@hidden wrote:
> Sorry, I don't quite understand. You are proposing that message-fun
> should not only take over messages proper but also the repl's print?

Yes, why not?

> Because (message "foo") displays foo twice (well once it's foo, as
> a side effect, then it's "foo", as the repl's "print" part, as the
> result of the S-expression's evaluation.

When it is evaluated in the REPL, yes. Because you're seeing both the 
result of 'message' itself, as well as the result of eval-expression.

> FWIW, I'd expect setting set-message-function (strange name for a
> var, BTW) to just take over the side effect, not the repl's "print".

Do you have an example of set-message-function where it would be a bad idea?