Re: PL/pgSQL bug?

Tom Lane wrote:
>
> Tatsuo Ishii <t-ishii(at)sra(dot)co(dot)jp> writes:
> > NOTICE: ctid (0,5) xmin 645188 xmax 645190 cmin 2 cmax 2
> > This is odd too, since xmax > 0 or cmax > 0 should never happen with
> > visible tuples, in my understanding.
>
> That's what the docs presently say, but they're in error --- nonzero
> xmax could represent a not-yet-committed deleting xact (or one that
> did commit, but not in your snapshot); or it could be from a deleting
> xact that rolled back.

or it can come from referential integrity triggers:

hannu=# create table parent(parid integer primary key);
NOTICE: CREATE TABLE/PRIMARY KEY will create implicit index
'parent_pkey' for table 'parent'
CREATE
hannu=# create table child(cldid integer references parent on update
cascade);
NOTICE: CREATE TABLE will create implicit trigger(s) for FOREIGN KEY
check(s)
CREATE
hannu=# insert into parent values(1);
INSERT 20652 1
hannu=# insert into child values(1);
INSERT 20653 1
hannu=# update parent set parid=2;
UPDATE 1
hannu=# select xmin,xmax,cmin,cmax,parid from parent;
xmin | xmax | cmin | cmax | parid
------+------+------+------+-------
731 | 731 | 0 | 4 | 2
(1 row)

Now I have a question: if xmax is not used in determining tuple
visibility
(as I had assumed earlier) then what is ? How does postgres decide that
a
tuple is deleted ?

--------------------
Hannu