Re: euro symbol

[Eli Zaretskii]

> From: Francesco Potorti` <address@hidden>
> Date: Fri, 26 Oct 2001 12:07:32 +0200
> 
> In fact it *is* a trivial change  (it just adds a new input method for a
> single  character in  a single  national language)  that does  not break
> anything that  existed before.  I  think it is  also a good  change, and
> that  it is better  than before,  given that  no Italian  latin-9 method
> curently exists.

I think it introduces a bug: you promise the user to insert a Euro,
but in fact insert a different character.

> By   the   way,   it's    me   who   wrote   the   italian-postfix   and
> italian-alt-postfix methods,  which together  with the reasons  above is
> why I dared to commit the changes without asking.

>From my experience, Mule-related changes are seldom simple, and it's
best to discuss them first.  YMMV.

>                    I've just tried one input method you've changed
>    (italian-postfix), and the change does the wrong thing: typing "E="
>    inserts the latin-1 currency character, not the Euro (which doesn't
>    exist in Latin-1).
> 
> It has the  same character code.

No, it doesn't; not inside Emacs, anyway.  Try "C-u C-x =" on a Euro
(inserted with the latin-9-prefix) and on the character produced by
"E=" in italian-postfix, and you will see that the characters are
different.

> So, on a console,  it depends on which
> font you have loaded.  On X, it shows the international currency symbol.

Yes, I know that by reshuffling fonts and lying to Emacs about the
character sets supported by each font, you can get Emacs to display
almost any character as any other character.  I can, for example,
display Hebrew text while Emacs would think it's Latin-1 text.  But
it's hardly the right thing to introduce something like this into
Emacs.  We will have a flood of FAQs and will leave gobs of users
confused for life.  I think Mule is complicated enough even without
this.

> However, having  a simple  way to  input the
> international currency symbol is good, in my opinion.  You cannot always
> have a latin-9 font, while latin-1  is much more common.  If you have to
> do with latin-1, the international currency is a good substitute for the
> euro symbol.  That's why I think that change should be left alone.

I disagree; I think we should remove this change.  If you are not
convinced, I guess Gerd and Richard will have to decide.

> More than that, I again ask if people think that the general latin-1 and
> national  input  methods  should  have  a similarly  simple  method  for
> entering the international currency symbol.

I don't think I understand what do you mean by that.  Could you please
elaborate?

>    It certainly does: there's the latin-9-prefix input method (you get
>    the Euro if you type "~e").  Maybe we need more Latin-9 support; if
>    so, let's add more input methods, but let's do it right.
> 
> Hm.  So I  should add an Italian style for  latin-9.

Yes, I think so.  In general, Latin-9 should have the same assortment
of input methods as Latin-1.

> However, the tilde was maybe an unfortunate choice.  It certainly is
> for Italian, as Italian keyboards do not have a tilde symbol.

We could add "=E" there; I don't see any problems with that.  And if
you are going to add an italian input method, you can drop the "~e"
combination if you like.