Re: Call for 7.5 feature completion

From: Dawid Kuroczko <qnex42(at)gmail(dot)com>
To: Bruce Momjian <pgman(at)candle(dot)pha(dot)pa(dot)us>, pgsql-hackers(at)postgresql(dot)org
Subject: Re: Call for 7.5 feature completion
Date: 2005-08-31 12:21:12
Message-ID: 758d5e7f0508310521fc1e687@mail.gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

On 8/26/05, Alvaro Herrera <alvherre(at)alvh(dot)no-ip(dot)org> wrote:
>
> Or, slightly different, what are people's most wanted features?

One feature, or rather set of features which was missing from the list and I
think
it is important: i18n. :)

I mean, PostgreSQL has a number of good features concerning
internationalization,
like UTF-8 support, transparent charset conversions, etc, but it also is
area where
new users are likely to get bit. One of the most gotcha-prone areas in
PostgreSQL
IMHO.

If you stick with English, its OK. If you want different language, say
Polish, German,
whatever you'll probably careful enough to set a good locale. If you decide
you
want to make a "hybrid" Polish-German database -- you may run into problems,
like
indexes and ordering -- indexes are ordered using only one collation
mechanism,
so you should probably use "C" locale. If you're unlucky -- you have to
recreate
whole database. And then if you intend to use tsearch2, you have to set it
up carefully
for given needs. I'm not saying that mysqlish approach of setting collate
per table
would be a good solution.

Frankly I don't think there is an ideal solution for this.

Some time ago someone suggested using "universal" UTF-8 collation, which is
good for most languages (and not for Turkish :)) -- I believe I've seen a
patch for
this on this list. Having some "one size fits most" solution could be
helpful.

Anyway, the i18n problem is a child-age illness, once you get over with it,
you're
most likely safe from it for the rest of your life. But some newbies may not
get
through it. ;)

Regards,
Dawid

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Martijn van Oosterhout 2005-08-31 13:32:31 On hardcoded type aliases and typmod for user types
Previous Message prasanna mavinakuli 2005-08-31 12:08:10 problem with PQExecParams