Auditing support -- UpdateBatchQuery needs accessor for old snapshot values

[Mike Kienenberger <mk...@gmail.com>]

Andrus,

I've almost finished the collecting of values from the various BatchQueries.
The one that remains is getting the old value for a column from
UpdateBatchQuery.   I know that these values are available as part of
the qualifier when optimistic locking is used, but I don't know if
that's the best way to get them.

What do you think of having the old value snapshot stored directly on
UpdateBatchQuery and made publically accessible?   I'm not sure if
this might make dealing with optimistic locking easier or not in the
process.

Re: Auditing support -- UpdateBatchQuery needs accessor for old snapshot values

[Andrus Adamchik <an...@objectstyle.org>]

I am tentatively +1 as I don't see a problem with this extra  
redundant data being attached to the query. Let's see how it works  
out and whether it affects the commit algorithm too much.

Andrus

On Aug 27, 2006, at 1:31 AM, Mike Kienenberger wrote:

> Andrus,
>
> I've almost finished the collecting of values from the various  
> BatchQueries.
> The one that remains is getting the old value for a column from
> UpdateBatchQuery.   I know that these values are available as part of
> the qualifier when optimistic locking is used, but I don't know if
> that's the best way to get them.
>
> What do you think of having the old value snapshot stored directly on
> UpdateBatchQuery and made publically accessible?   I'm not sure if
> this might make dealing with optimistic locking easier or not in the
> process.
>