On 3/30/06, Michael Bayer <[EMAIL PROTECTED]> wrote:
> so, better to use the version id
>option, which is very simple and easy to make efficient...the developer
>just adds a proper index to the version ID column. works with BLOBs etc
STM that you don't even need an index on the version column, the update is already indexed by the primary key in this case. You're very likely to have very large % of dup values in the version col anyway.
Rick
- Re: [Sqlalchemy-users] hibernate Michael Bayer
- Re: [Sqlalchemy-users] hibernate Florian Boesch
- Re: [Sqlalchemy-users] hibernate Michael Bayer
- Re: [Sqlalchemy-users] hibernate Michael Bayer
- Re: [Sqlalchemy-users] optimistic concurrency WAS hibe... Daniel Miller
- Re: [Sqlalchemy-users] optimistic concurrency WAS ... Michael Bayer
- Re: [Sqlalchemy-users] optimistic concurrency ... Jonathan Ellis
- Re: [Sqlalchemy-users] optimistic concurre... Michael Bayer
- Re: [Sqlalchemy-users] optimistic con... Huy Do
- Re: [Sqlalchemy-users] optimistic... Jonathan Ellis
- Re: [Sqlalchemy-users] optimistic con... Rick Morrison
- Re: [Sqlalchemy-users] optimistic concurrency ... Florian Boesch
- Re: [Sqlalchemy-users] optimistic concurre... Florian Boesch