Stephen Boyd wrote:

>If I understand things correctly, doing a COMMIT RETAINING does
>nothing to advance the OAT. If this is in fact the case, how does one
>avoid sticking the OAT during long running sequential reads / mass
>updates of a table? If would be nice to be able to COMMIT every
>hundred records or so to advance the OAT while retaining open cursors
>but there doesn't seem to be any way to accomplish this. I was
>wondering if other people had found a way to circumvent this problem.
>
>
>

Firebird supports an (unlimited?) number of transactions in "paralel",
inside a session. Why not process your updates in another transaction?

Ivan.