Re: [HACKERS] SELECT BUG

  <p>Tom Lane ha scritto: <blockquote type="CITE">José Soares <jose@sferacarta.com> writes: <br />>> When
I"explain" this, I see that I am getting a mergejoin plan. <br />>> Are you getting a hash join, perhaps? <p>>
Yes.<p>> prova=> explain select  mcode, dcode from master m, detail d where <br />> mcode=dcode; <br />>
NOTICE: QUERY PLAN: <br />> <br />> Hash Join  (cost=156.00 rows=1001 width=24) <br />>   -> Seq Scan on
detaild  (cost=43.00 rows=1000 width=12) <br />>   -> Hash  (cost=43.00 rows=1000 width=12) <br />>      
->Seq Scan on master m  (cost=43.00 rows=1000 width=12) <br />> <br />> EXPLAIN <p>OK, do this: <p>update
pg_operatorset oprcanhash = 'f' where oid = 1054; <p>and I think you'll be OK.  I will put that change into the
sources.<p>                        regards, tom lane <p><tt>************</tt></blockquote><tt>Yes, Tom, now it works,
but...</tt><br/><tt>Informix gives me a different result. Who is right ?</tt><br /><tt></tt> <tt></tt><p><tt>prova=>
select mcode, dcode from master m, detail d where mcode=dcode;</tt><br /><tt>mcode|dcode</tt><br
/><tt>-----+-----</tt><br/><tt>(0 rows)</tt><tt></tt><p><tt>prova=> update pg_operator set oprcanhash = 'f' where
oid= 1054;</tt><br /><tt>UPDATE 1</tt><br /><tt>prova=> select  mcode, dcode from master m, detail d where
mcode=dcode;</tt><br/><tt>mcode      |dcode</tt><br /><tt>-----------+----------------</tt><br /><tt>a1        
|a1</tt><br/><tt>a13        |a13</tt><br /><tt>a13        |a13</tt><br /><tt>(3 rows)</tt><br
/><tt></tt> <tt></tt><p><tt>INFORMIX:</tt><br/><tt>SQL:   New  Run  Modify  Use-editor  Output  Choose  Save  Info 
Drop Exit</tt><br /><tt>Run the current SQL statements.</tt><br /><tt>----------------------- hygea@hygea ------------
PressCTRL-W for Help --------</tt><br /><tt>mcode dcode</tt><br /><tt>a1    a1</tt><br /><tt>a13   a13</tt><br /> 
<p>José<br />  <br />