On 10 Oct 2009, at 9:27pm, Jay A. Kreibich wrote: > On Sat, Oct 10, 2009 at 07:38:08PM +0100, Simon Slavin scratched on > the wall: >> > >> Don't forget to use transactions, even for when you are just doing >> SELECTs without changing any data. > > Using transactions speeds up a long series of SELECTs because it > eliminates the need to re-acquire a read-only file-lock for each > individual SELECT. > > Since in-memory databases have no file locks, I'm not sure that is > relevant to this specific case.
I wasn't sure about that. It could still be slower, even if the individual transactions don't eventually result in any calls to file- handling routines. At what point does the code decide whether a lock is needed ? Does it get dropped almost immediately, or does it go through multiple levels of calls before some low-level code says "iff disk-database then lock" ? Sorry, I've never looked at the source. Simon. _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
