Hi ...
In my quest to get rid of the oid dependency, i have made some new low level code with the help from many nice people from this community (thanks for that), but I still have one somewhat big problem.
I am running PG 7.4.6, btw.
I have a "sale" table that have a BIGSERIAL as primary key, but explain keeps telling me (and so does the performance) that it will perform a "seq scan" of my table when performing this statement (not using the pkey index) :
select * from sale where id = currval( 'sale_id_seq' );
I tried this too :
select * from sale where id = currval( 'sale_id_seq' )::bigint;
But this still did not work (still using seq scan) :-(
At last I did a :
explain select * from sale where id = 42::bigint;
Just to make sure, and this made the proper optimizations (used the pkey index).
What have I done wrong, or do PG still have some casting problems in the optimizer ?
/BL
---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match
