Re: [HACKERS] ordering RH6.1 - Mailing list pgsql-hackers

From Tom Lane
Subject Re: [HACKERS] ordering RH6.1
Date
Msg-id 16598.945214137@sss.pgh.pa.us
Whole thread Raw
In response to ordering RH6.1  (Frans Van Elsacker <fve@atbib.be>)
Responses Re: [HACKERS] ordering RH6.1
List pgsql-hackers
Frans Van Elsacker <fve@atbib.be> writes:
> When i do the following
> CREATE TABLE BLANK (column1 varchar(5));
> INSERT  INTO BLANK (column1) VALUES ('    1');
> INSERT  INTO BLANK (column1) VALUES ('   11');
> INSERT  INTO BLANK (column1) VALUES ('  100');
> INSERT  INTO BLANK (column1) VALUES ('    2');
> then:
> SELECT * FROM BLANK order by column1;

> I received
>   1
>  100
>  11
>   2        --> mark also a not aligned output.

> and I expected
>     1
>     2
>    11
>   100

> Anybody has an idea?? 

Bizarre.  I see the expected results under both 6.5.3 and current
development sources:

play=> SELECT * FROM BLANK order by column1;
column1
-------     1     2    11   100
(4 rows)

I wonder if this could be a LOCALE or MULTIBYTE issue.  Do you have
either feature enabled in your copy, and if so what locale/encoding
do you use?  (I'm running plain vanilla no-USE_LOCALE, no-MULTIBYTE
code, so that might be why I don't see anything funny...)
        regards, tom lane


pgsql-hackers by date:

Previous
From: Matthew Hagerty
Date:
Subject: Backend core dump, different server!
Next
From: Tom Lane
Date:
Subject: Re: [HACKERS] Re: [PATCHES] createdb/dropdb fixes