Re: [pgsql-performance] [GENERAL] Large databases, performance - Mailing list pgsql-hackers

From Shridhar Daithankar
Subject Re: [pgsql-performance] [GENERAL] Large databases, performance
Date
Msg-id 3DA2BE03.3674.13CF7DD1@localhost
Whole thread Raw
In response to Re: [pgsql-performance] [GENERAL] Large databases, performance  (Tom Lane <tgl@sss.pgh.pa.us>)
Responses Re: [pgsql-performance] [GENERAL] Large databases, performance
List pgsql-hackers
On 7 Oct 2002 at 11:21, Tom Lane wrote:

> "Shridhar Daithankar" <shridhar_daithankar@persistent.co.in> writes:
> > I say if it's a char field, there should be no indicator of length as
> > it's not required. Just store those many characters straight ahead..
>
> Your assumption fails when considering UNICODE or other multibyte
> character encodings.

Correct but is it possible to have real char string when database is not
unicode or when locale defines size of char, to be exact?

In my case varchar does not make sense as all strings are guaranteed to be of
defined length. While the argument you have put is correct, it's causing a disk
space leak, to say so.

Bye
 Shridhar

--
Boucher's Observation:    He who blows his own horn always plays the music    several
octaves higher than originally written.


pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: Dirty Buffer Writing [was Proposed LogWriter Scheme]
Next
From: "Shridhar Daithankar"
Date:
Subject: Re: [GENERAL] Hot Backup