On Tue, 2005-08-23 at 21:46 +0200, tone skoda wrote:
> read (SELECT) has to find physical position of record on disk.
> update too has to find physical position of record on disk.
> this is inefficient if it's done twice in a row, because two searches for
> record have to be performed.
> only one search would be enough and more efficient.
> is there any way SELECT and UPDATE can be done in in one SQL statement?
>
> in fact, why isn't it possible to refer to records by their physicall offset
> on
> disk? as far as i know thay don't change.
>
SQLite supports variable-length records with automatic
compaction and defragmentation. Records are subject to
being moved about on the disk after any change to nearby
records.
--
D. Richard Hipp <[EMAIL PROTECTED]>