Re: [HACKERS] ALTER SET DISTINCT vs. Oracle-like DBMS_STATS

2009-07-06 Thread Itagaki Takahiro
Euler Taveira de Oliveira wrote: > >> INSERT INTO dbms_stats.columns(starelid, ataattnum, stadistinct) > >> VALUES ('tablename'::regclass, 3, 100); > > > > Why wouldn't you implement this through reloptions? > > > Because it is column-based and not table-based? In this case, we need

Re: [HACKERS] ALTER SET DISTINCT vs. Oracle-like DBMS_STATS

2009-07-06 Thread Alvaro Herrera
Itagaki Takahiro escribió: > It is just similar to Oracle's DBMS_STATS package. > http://download.oracle.com/docs/cd/B28359_01/appdev.111/b28419/d_stats.htm > If it were, "ALTER TABLE tablename ALTER COLUMN 3rd-column SET DISITNCT 100" > could be written as: > > INSERT INTO dbms_stats.columns

[HACKERS] ALTER SET DISTINCT vs. Oracle-like DBMS_STATS

2009-07-06 Thread Itagaki Takahiro
Hello, A new feature "ALTER TABLE ... ALTER COLUMN ... SET DISTINCT" is submitted to the next commetfest: http://archives.postgresql.org/message-id/603c8f070905041913r667b3f32oa068d758ba5f1...@mail.gmail.com but I have another approach for the plan stability issues. It might conflict ALTER SET DI