Hi Igor, No, no... I really meant a roundtrips to disk. So, while sqlite3_get_table() seemed like a faster but memory hungry mechanism to retrieve the result set, sqlite3_step() seemed that it would take longer, especially in disk-based systems, but more memory-efficient.
Thanks for pointing to me that sqlite3_get_table() is not faster than sqlite3_step(). -- Tito On 21 Nov 2010, at 14:14, Igor Tandetnik <itandet...@mvps.org> wrote: > Tito Ciuro <tci...@mac.com> wrote: >> Let me start by saying that I'm aware that sqlite3_get_table() should not be >> used (as per the documentation). I'm curious about >> one thing though: if the computer/device has sufficient memory to hold the >> result set returned by sqlite3_get_table(), wouldn't >> it be more optimized performance-wise than calling sqlite3_step N times to >> obtain the same result set? > > sqlite3_get_table is just a wrapper around sqlite3_prepare/sqlite3_step. > >> Sounds like sqlite3_get_table() would take less time to access the storage >> subsystem as opposed to sqlite_step() with multiple >> roundtrips > > You say "roundtrip" as if it's a network request to some remote machine or > something. Here, a "roundtrip" is a function call within the same process. In > this sense, a call to malloc is a roundtrip to memory allocation subsystem - > you want to avoid those at all costs, right? > -- > Igor Tandetnik > > _______________________________________________ > sqlite-users mailing list > sqlite-users@sqlite.org > http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users _______________________________________________ sqlite-users mailing list sqlite-users@sqlite.org http://sqlite.org:8080/cgi-bin/mailman/listinfo/sqlite-users
