Lance,
On 9/6/2012 5:40 AM, Lance Andersen - Oracle wrote:
Here is the updated webrev http://cr.openjdk.java.net/~lancea/7192302/webrev.01
It's good to see this change that jdbc rowset doesn't depend on java.beans.
This looks okay to me as you have explained why you remove the call to
connect() in one of the constructors. Alan and you discussed if
setConcurrency could be removed in this fix. setType() is another one
that could be removed. I'm fine with leaving these cleanup in another
fix and have this fix to focus on removing the dependency of
PropertyChangeSupport. It'd be good to file a CR as a follow-up of this
work.
Thanks
Mandy
I know there is more clean-up that can be done to remove other Rave added code
(such as the removal of set/getPreparedStatement/Connection/ResultSet), I want
to keep the focus to just removing PropertyChangeSupport. SQE and RowSet TCKs
continue to pass with these changes.
Best
Lance
On Sep 5, 2012, at 5:17 PM, Alan Bateman wrote:
On 05/09/2012 22:04, Lance Andersen - Oracle wrote:
Hi all,
Looking for a reviewer for the removal of PropertyChangeSupport from
JDBCRowSetImpl that was originally going to be used by the EOL Rave product.
As it is no longer needed the code has been removed. The SQE and RowSet TCK
tests all continue to run without regression.
The webrev can be found at http://cr.openjdk.java.net/~lancea/7192302/webrev.00
Thanks Lance, it's good to remove this dependency. In both commit and rollback
then it looks to me that the setting of oldVal can be removed. Otherwise looks
good to me.
-Alan
Lance Andersen| Principal Member of Technical Staff | +1.781.442.2037
Oracle Java Engineering
1 Network Drive
Burlington, MA 01803
lance.ander...@oracle.com
