On Sat, Jun 25, 2011 at 11:20 PM, bob <aj...@hotmail.com> wrote: > i'm getting an issue relating to the use of the random() function in a > SQL query. > > i run 2 PHP scripts very quickly (less than 1 second apart). > > they use an SQL query like SELECT * FROM table ORDER BY random() LIMIT 5; > > because the 2 scripts run so close together (in time) i get the same > values as a result.
Since sqlite initialize the seed only once you probably mean the case when your two queries used on separately opened and closed connections, right? I tried to simulate this with a test table containing numbers from 1 to 5 and query Select value from testtable order by random(). Sqlite gave different results most of the times. This was on a windows machine where the time between connection was more than 1ms and also on a machine when the time was < 1 ms. My tests can be wrong if sqlite initializes the seed once for the library, not for single connection. I tried to track this with loading sqlite dll for every connection, but unfortunately I could not find a windows machine that can do all these tasks in less than a 1 ms. But for tests > 1 ms the numbers were still different So you probably have to look in a different layer Max _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
