Re: UTF-8 and LIKE vs =

On Thu, 26 Aug 2004 09:36:20 -0700
David Wheeler <david@kineticode.com> wrote

> On Aug 24, 2004, at 6:58 PM, Joel wrote:
>
> > Sorry, I guess I wasn't clear. It's almost directly corollary to
> > ignoring case. (Almost, but not quite.) Kind of like querying with
> > "TEST"
> > and getting records with "test".
>
> How interesting!
>
> > Which means that the correct solution now is beyond the question of
> > what
> > did the vendor think you would want. No way the vendor is going to be
> > willing to think you really want that, in spite of the number of people
> > who want precisely that. We don't yet have a mechanism that works for
> > general multiple language solutions.
>
> Right, except for "no locale" (is C locale the same thing?).

Nobody's answered this yet, but I'll stick my neck out and so I don't
remember. But it should be hidden in the manual somewhere.

(I guess I should say my memories are conflicting, which means I should
look it up myself some time.)

> > At this point in time, "no locale" is probably close to best practice,
> > but it leaves you a lot of work.
>
> It does?

Well, there are just a lot of things that you want to be sort of
automatic that you end up having to do by hand.

--
Joel <rees@ddcom.co.jp>