Rajesh Kartha (JIRA) wrote:
I also think the title of the issue also need to change to reflect that self
trigger does not throw ERROR 54038 in some cases.
A few questions/concerns on this issue:
1) CORRECT TRIGGER DEFINITION:
To get the semantics the user seems to want, which is to update the
timestamp when TESTID changes, is this how the trigger should be defined?
CREATE TRIGGER UPDATE_TEST
AFTER UPDATE ON TEST
REFERENCING OLD AS OLD
FOR EACH ROW MODE DB2SQL
UPDATE TEST SET TIMESTAMP = CURRENT_TIMESTAMP WHERE
TESTID = OLD.TESTID AND OLD.TESTID != NEW.TESTID;
2) CORRECT FILING OF THE ISSUE
Since the summary, and description are really wrong for this issue, I
think we perhaps file a separate issue and close DERBY-1603 as invalid.
I will do that.
3) USER IMPACT
This one is a bit worrisome. I think the mistake made by this user
would be a fairly common one and users may hit this on upgrade, but that
is an existing condition. I think all we can really do is have a Release
Note, mark the new issue with Existing Application Impact , and get the
fix to 10.1 as soon as possible.
Kathey
