Re: SQL_ASCII vs. 7-bit ASCII encodings - Mailing list pgsql-hackers

From Alvaro Herrera
Subject Re: SQL_ASCII vs. 7-bit ASCII encodings
Date
Msg-id 20050513223220.GB24203@surnet.cl
Whole thread Raw
In response to Re: SQL_ASCII vs. 7-bit ASCII encodings  (Tom Lane <tgl@sss.pgh.pa.us>)
List pgsql-hackers
On Fri, May 13, 2005 at 10:22:06AM -0400, Tom Lane wrote:
> Alvaro Herrera <alvherre@surnet.cl> writes:
> > In fact I've seen many more people with this problem after 8.0 was
> > released, at least in pgsql-es-ayuda.
> 
> Which problem exactly?  Most of the 8.0 complaints I can recall seemed
> to come from people who were trying to dump from a SQL_ASCII database
> and reload into a UTF8 database --- that is, their problems were
> specifically *caused by* our decision to shift to a more restrictive
> default.  Which is why I'm not eager to make it even more restrictive.

Well, of all kinds really.  There are some of those.  Also I've seen
several questions from people using, say, a Python environment and then
they have to add a Windows/ODBC application.  The Python apps have been
using SQL_ASCII all along, but ODBC apparently wants UTF8 or some sort
of recode.  So non-7-bit chars display funny.

Anyway now that I look at them ISTM that most complaints actually come
from misunderstanding of the whole encoding issue.  They just create a
database and an app and start coding and throwing data.  At first it all
seems fine.  Later they find out that the client is using, say,
win-1250, and has been using client_encoding latin1; so it has been
wrong all around and they only find out when a new, different app shows
something strange.

So maybe it's not about 7 bits or not, it's just lack of knowledge.
But I've certainly seen much more problems reported after 8.0 was
released.

-- 
Alvaro Herrera (<alvherre[a]surnet.cl>)
"Pensar que el espectro que vemos es ilusorio no lo despoja de espanto,
sólo le suma el nuevo terror de la locura" (Perelandra, CSLewis)


pgsql-hackers by date:

Previous
From: Alvaro Herrera
Date:
Subject: PREPARE TRANSACTION and friends, separate node?
Next
From: Josh Berkus
Date:
Subject: Re: Do you have any objections on contributing your improvements of the PostgreSQL core back into the product?