Re: different sort order in windows and linux version

Поиск
Список
Период
Сортировка
От Martijn van Oosterhout
Тема Re: different sort order in windows and linux version
Дата
Msg-id 20060702212656.GC8316@svana.org
обсуждение исходный текст
Ответ на Re: different sort order in windows and linux version  (Agent M <agentm@themactionfaction.com>)
Список pgsql-general
On Sun, Jul 02, 2006 at 12:25:43PM -0400, Agent M wrote:
> On Jul 2, 2006, at 6:13 AM, Martijn van Oosterhout wrote:
> >But I don't think anyone is actually considering importing ICU into the
> >postgres source tree, are they?
> Why not?

Because it's a project of similar size to postgres and probably nearly
as old and I don't think anyone here actually wants to maintain it.

I mean, we could incorporate the source for readline, openssl,
kerberos, the C library but why. That project has maintainers already
and we only wan to use it, not fork it.

> >If you drop the conversion stuff (because postgres already has that)
> >you're down to about 4MB.
> Why would you drop the ICU transcoding support instead of the existing
> postgres functions? Why the duplicated effort?

Because we would want to be bug-for-bug compatable to previous
releases. I suppose it would be possible if someone checked that the
end result is the same.

> Certain Japanese characters cannot make a reliable round-trip through
> Unicode. ICU uses UTF-16 as its store, so the Japanese folks won't be
> happy with an ICU-only solution. However, it would still be of great
> benefit to allow ICU to handle as much as possible, leaving the string
> encodings to the encoding experts.

We don't need round-trip through unicode, since we're only doing one
way conversions for the purpose of collation.

BTW, this site seems to have a good discussion of Japanese characters
and Unicode.

http://www.jbrowse.com/text/unij.html

> At the very least, it would be great to have ICU to handle encoding on
> a per-column basis (perhaps extending the text datatype with encoding
> info). Perhaps this would be a decent stopgap solution? The backend
> protocol would also need a version bump- currently, it converts all
> strings to a single encoding.

That's called SQL COLLATE support and that's an order of magnitude
harder than adding support for ICU. See previous dicussion on -hackers.

Have a nice day,
--
Martijn van Oosterhout   <kleptog@svana.org>   http://svana.org/kleptog/
> From each according to his ability. To each according to his ability to litigate.

Вложения

В списке pgsql-general по дате отправления:

Предыдущее
От: "Alexander Farber"
Дата:
Сообщение: libpq: bind message supplies 2 parameters, but prepared statement requires 1
Следующее
От: Martijn van Oosterhout
Дата:
Сообщение: Re: libpq: bind message supplies 2 parameters, but prepared statement requires 1