Re: encoding: ODBC, createdb - Mailing list pgsql-hackers

From Hiroshi Inoue
Subject Re: encoding: ODBC, createdb
Date
Msg-id 3B81B2A5.D9BEA715@tpf.co.jp
Whole thread Raw
In response to Re: [PATCHES] encoding names  (Tatsuo Ishii <t-ishii@sra.co.jp>)
Responses Re: encoding: ODBC, createdb
List pgsql-hackers
Karel Zak wrote:
> 
>  I found some other things:
> 
> - ODBC -- here is some multibyte stuff too. Why ODBC code don't use
>   pg_wchar.h where is all defined? In odbc/multibyte.h is again defined
>   all encoding identificators.
> 
>   IMHO we can use for ODBC same solution as for libpq and compile it
>   with encname.c file too.

ODBC under Windows needs no source/header files in PostgreSQL
other than in src/interfaces/odbc. It's not preferable for 
psqlodbc driver to be sensitive about other PostgreSQL changes
because the driver has to be able to talk multiple versions of
PostgreSQL servers. In fact the current driver could talk to 
any server whose version >= 6.2(according to a person).
As for pg_wchar.h I'm not sure if it could be an exception
and we could expect for the maintainer to take care of ODBC.
If I were he, I would hate it.

regards,
Hiroshi Inoue


pgsql-hackers by date:

Previous
From: "Dave Cramer"
Date:
Subject: Fully qualified column names
Next
From: Tom Lane
Date:
Subject: Re: Status of ipcclean