Am 22.01.12 14:22, schrieb Giuseppe Sucameli:
That is not a bug, but a feature. See section 5.4 of the documentation
"System Columns":

"Every table has several system columns that are implicitly defined by
the system. Therefore, these names cannot be used as names of
user-defined columns. (Note that these restrictions are separate from
whether the name is a key word or not; quoting a name will not allow you
to escape these restrictions.) You do not really need to be concerned
about these columns; just know they exist."

and further down:

"xmin

The identity (transaction ID) of the inserting transaction for this row
version. (A row version is an individual state of a row; each update of
a row creates a new row version for the same logical row.)"


--
Sent via pgsql-bugs mailing list (pgsql-bugs (AT) postgresql (DOT) org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs

On Mon, Jan 23, 2012 at 11:25, Marc Balmer <marc (AT) msys (DOT) ch> wrote:

I see it as a message bug. Why wouldn't ALTER TABLE also tell us that xmin
is a system column? It makes things much more clear for newbies who don't
see the column yet are told it exists if they're also told it's a system
column.

I would try to cook up a patch but I have no skills :-(

Hi Marc,

On Mon, Jan 23, 2012 at 3:04 PM, Vik Reykja <vikreykja (AT) gmail (DOT) com> wrote:
I agree with Vik, the CREATE TABLE tells "xmin" is a system column,
why wouldn't ALTER TABLE do the same?

This would be a feature if CREATE TABLE didn't tell us that
xmin is a system column, otherwise is a bug.

I'm going to write a patch to fix this problem.
Regards.

--
Giuseppe Sucameli

--
Sent via pgsql-bugs mailing list (pgsql-bugs (AT) postgresql (DOT) org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs

On Tue, Jan 24, 2012 at 14:41, Giuseppe Sucameli <brush.tyler (AT) gmail (DOT) com>wrote:


I managed to put something together and have posted it on hackers.

Hi Vik,

On Wed, Jan 25, 2012 at 12:28 AM, Vik Reykja <vikreykja (AT) gmail (DOT) com> wrote:
thinking you have no skills, I wrote a patch too and posted it to
hackers ML about 2 days ago, but I didn't subscribe that list so
it is stalled, waiting for someone's approval.

For this reason my email is not displayed in the archives.
I'm so sorry both have wasted our time in writing 2 patches for
the same bug.

Regards.

--
Giuseppe Sucameli

--
Sent via pgsql-bugs mailing list (pgsql-bugs (AT) postgresql (DOT) org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs

On Wed, Jan 25, 2012 at 22:19, Giuseppe Sucameli <brush.tyler (AT) gmail (DOT) com>wrote:

I think your patch is more complete than mine so it's definitely not
wasted. I spent several hours on mine and now I have a little bit of
skills :-)

Thank you for your effort; I'm sorry if you feel you have wasted your time.

On Thu, Jan 26, 2012 at 5:43 AM, Vik Reykja <vikreykja (AT) gmail (DOT) com> wrote:
I think Vik's patch is better, because it doesn't rely as much on
things that happen to be true today - like which relkinds have system
columns associated with them. But I do agree we ought to handle both
the ADD COLUMN and RENAME COLUMN cases, and as such have posted an
update of Vik's patch on the other thread.

--
Robert Haas
EnterpriseDB: http://www.enterprisedb.com
The Enterprise PostgreSQL Compan

--
Sent via pgsql-bugs mailing list (pgsql-bugs (AT) postgresql (DOT) org)
To make changes to your subscription:
http://www.postgresql.org/mailpref/pgsql-bugs