Re: [HACKERS] GSoC - proposal - Materialized Views in PostgreSQL

2010-04-09 Thread Robert Haas
2010/4/9 pavelbaros baro...@seznam.cz: Implementation:  could be divided to few steps: 1) create materialized view - modify grammar (parser): CREATE MATERIALIZED VIEW mv_name AS SELECT ... - change executor, so that it will create physical table defined by select statement This basically

Re: [HACKERS] GSoC - proposal - Materialized Views in PostgreSQL

2010-04-09 Thread Greg Smith
pavelbaros wrote: I am also waiting for approval for my repository named materialized_view on git.postgresql.org, so I could publish completed parts. Presuming that you're going to wander there and get assigned what looks like an official repo name for this project is a bit...optimistic. I

Re: [HACKERS] GSoC - proposal - Materialized Views in PostgreSQL

2010-04-09 Thread Robert Haas
2010/4/9 Greg Smith g...@2ndquadrant.com: The main hidden complexity in this particular project relates to handling view refreshes.  The non-obvious problem is that when the view updates, you need something like a SQL MERGE to really handle that in a robust way that doesn't conflict with

<    1   2