Re: [sqlite] Very important performance regression due to totally different query plan

Hi Richard, I found this thread (http://www.mail-archive.com/sqlite-users@sqlite.org/msg76831.html) and applied the same correction. Now, I have good performances on all version of sqlite. Thank you for your help. Regards, Stephane Le 08/05/2013 12:14, Stephane MANKOWSKI a écrit : Hi

Re: [sqlite] Very important performance regression due to totally different query plan

Hi Richard, Yes, I already considered including source code but this is not so easy because I develop in fact with QT and the sqlite driver. Anyway, the main reason of my initial post is to: 1-inform you about this important performance regression 2-know if this is normal or if I have to open

[sqlite] Very important performance regression due to totally different query plan

Hi, I am the main developer of Skrooge (application using sqlite) and since my upgrade to ubuntu 13.04, I have very bad performances due to sqlite. Here is my analysis: All tests are done with the file http://skrooge.org/files/skrooge_bug_performance.sqlite

Re: [sqlite] Views Optimization

Hi Richard, In fact, the real need is "common subexpression elimination". As you can see in attached email, we already had this discussion. To improve performances on SELECT in skrooge (see datamodel in test.sqlite), I created some tables (named vm_xxx) updated after each transaction through