Actually, looking at the Gump messages in the Hivemind mailing list suggest that the problem was all along the lack of an activateOptions call before adding an appender and was never the addition of isClosed(), isActive() and activate.

Please apply the following patch, this should resolve the problem and still be compatible with log4j 1.2.x. I'll repost the patch to the hivemind-dev mailing list if there is not quick response to see if I can get another committer to get it in before tonights Gump run.



Index: framework/src/java/org/apache/hivemind/test/HiveMindTestCase.java
===================================================================
RCS file: /home/cvspublic/jakarta-hivemind/framework/src/java/org/apache/ hivemind/test/HiveMindTestCase.java,v
retrieving revision 1.23
diff -u -r1.23 HiveMindTestCase.java
--- framework/src/java/org/apache/hivemind/test/HiveMindTestCase.java 19 Feb 2005 02:40:55 -0000 1.23
+++ framework/src/java/org/apache/hivemind/test/HiveMindTestCase.java 24 Feb 2005 18:44:56 -0000
@@ -183,6 +183,7 @@


         _interceptedLoggerName = loggerName;
         _appender = new StoreAppender();
+        _appender.activateOptions();

         logger.setLevel(Level.DEBUG);
         logger.setAdditivity(false);


On Feb 24, 2005, at 12:26 PM, Ceki G�lc� wrote:


Hello Knut,

We'd like to reintroduce the isClosed(), isActive() and activate() methods back into the Appender interface. Since these methods are already implemented by AppenderSkeleton, existing appenders should remain compile-time compatible. Do you remember if there were any compile time issues with these new methods affecting Hivemind?

Would you care to check if and why these changes break compile-time compatibility in Hivemind? Curt intends to put them back in Appender tonight and the next Gump run should tell us about any failures.


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]



Reply via email to