Re: [HACKERS] Using the same condition twice

2002-11-24 Thread Tom Lane
I wrote: > Interesting. We seem to recognize the fact that the extra clause is > redundant in nearly all places ... but not in indexscan plan generation. > I tried this simplified test case: > create table t_wert(werttypid int); > create table t_werttyp(id int); > create index idx_wert_werttypid

Re: [HACKERS] Using the same condition twice

2002-10-24 Thread Tom Lane
=?ISO-8859-1?Q?Hans-J=FCrgen_Sch=F6nig?= <[EMAIL PROTECTED]> writes: > I came across a quite interesting issue I don't really understand but > maybe Tom will know. Interesting. We seem to recognize the fact that the extra clause is redundant in nearly all places ... but not in indexscan plan gen

[HACKERS] Using the same condition twice

2002-10-24 Thread Hans-Jürgen Schönig
I came across a quite interesting issue I don't really understand but maybe Tom will know. This happened rather accidentally. I have a rather complex query which executes efficiently. There is one interesting thing - let's have a look at the query: SELECT t_struktur.id, t_text.code, COUNT(t_we