On Tue, Dec 13, 2016 at 10:26:24AM +0900, Michael Paquier wrote:
> On Mon, Dec 12, 2016 at 10:22 PM, Greg Stark <stark@mit.edu> wrote:
> > One option might be for Postgres to define duplicate operator names
> > using ¿ or something else. I think ¿ is a good choice because it's a
> > common punctuation mark in spanish so it's probably not hard to find
> > on a lot of keyboards or hard to find instructions on how to type one.
>
> Are you sure that using a non-ASCII character is a good idea for an
> in-core operator? I would think no.
Eventually language designers will cross that Rubicon in mainstream
languages. And why not? It sure would be convenient... from the
designer's p.o.v. Of course, _users_ would be annoyed, as most users
in the English-speaking world will have no idea how to type such
characters, most others also will not know how to, and there will be
users still using non-Unicode locales who will be unable to type such
characters at all. Cut-n-paste will save the day, not doubt, though
mostly/only for users using Unicode locales.
But it is tempting.
Using non-ASCII Unicode characters for _alternatives_ seems like a
possible starting point though, since that leaves users with a
universally- available ASCII alternative. Still, now users would then
have to recognize multiple equivalent forms... ugh.
Nico
--