[ http://issues.apache.org/jira/browse/BEEHIVE-351?page=history ]
Kyle Marvin updated BEEHIVE-351:
--------------------------------
Fix Version: V1
(was: V1Beta)
This would be easy enough to fix for beta; however, its is going to require
massive changes in test cases and other places which likely don't declare
control instances to be serializable.
One kindler, gentler approach would be to just give an APT *warning* if the
implementation class is not serializable. This will a) keep from breaking
existing impls, and b) allow users to decide whether they care or not.
If we migrate to a warning, though, we may want to provide a way to make the
warning go away (w/out adding java.io.Serializable), perhaps something like:
@ControlImplementation(transient=true) to indicate that a control impl instance
should never be persisted, just its properties (and a new impl instance created
on deserialization).
I'm interested in feedback on this....
> JCS annotation processor should check serializability
> -----------------------------------------------------
>
> Key: BEEHIVE-351
> URL: http://issues.apache.org/jira/browse/BEEHIVE-351
> Project: Beehive
> Type: Bug
> Components: Controls
> Versions: V1Beta
> Reporter: Kyle Marvin
> Assignee: Kyle Marvin
> Priority: Minor
> Fix For: V1
>
> Looking at the tomcat log file after a Controls detailed test run, I see lots
> of errors of the form:
> java.io.NotSerializableException:
> org.apache.beehive.controls.test.controls.property.BoundPropertyControlImpl
> ...
> The JCS annotation processor should enforce that impl classes are marked as
> Serializable.
--
This message is automatically generated by JIRA.
-
If you think it was sent incorrectly contact one of the administrators:
http://issues.apache.org/jira/secure/Administrators.jspa
-
If you want more information on JIRA, or have a bug to report see:
http://www.atlassian.com/software/jira
