From 1c1dc3e8485dc91d3ab675ae6821c51aaf082d75 Mon Sep 17 00:00:00 2001 From: Frank Maritato Date: Tue, 14 Dec 2010 17:53:44 +0000 Subject: [PATCH] release instructions and added notes from other releases that werent added git-svn-id: https://lwes.svn.sourceforge.net/svnroot/lwes/lwes-java/trunk@629 a2f82657-cdd2-4550-bd36-68a8e7111808 --- README | 31 +++++++++++++++++++++++++++++-- 1 file changed, 29 insertions(+), 2 deletions(-) diff --git a/README b/README index 3194601..016c249 100644 --- a/README +++ b/README @@ -6,9 +6,12 @@ Prerequisites - Maven 2.2.x (http://apache.maven.org/) *** -How to build +How to build: % mvn clean package +To create a release: +% mvn --batch-mode release:prepare release:perform + *** To increase the read buffer size on the multicast socket, set the system property MulticastReceiveBufferSize. For example, @@ -23,6 +26,30 @@ java -DMulticastReceiveBufferSize=8388608 ... - Size limitations *** +0.2.5 +Changed license to BSD. + +*** +0.2.4 +- There is a memory leak in jdk 1.6.0_20 and earlier in +java.util.concurrent.LinkedBlockingQueue so I made a copy of that class and +fixed the memory leak. + +*** +0.2.3 +- Allow one to override the size check that is done as part of the Event class. Size +really only matters if one is intending on emitting the created event not if they +are reading it from a file. Also allow introspection into Event to get the size in +bytes. + +*** +0.2.1 -> 0.2.2 Changes +- Added the ability to set a maximum capacity on the internal queue used by the +ThreadedProcessor. If you set the queue size, you can avoid out of memory exceptions, +but you run the risk of dropping events on the floor (it would depend on how fast +you are processing events and how big your network buffer is). + +*** 0.1.1 -> 0.2.0 Changes - I removed the files that are autogenerated by javacc and made it part of the build. The .java files will end up in target/generated-sources/javacc if you are interested in @@ -38,4 +65,4 @@ maven wins. :) - Use LinkedBlockingQueues where possible to retain event order and help synchronization. - I removed AttributeNotSetException because throwing an exception if an attribute isn't set -in the event should not be the proper behavior (after all, HashMap and the like do not do this). \ No newline at end of file +in the event should not be the proper behavior (after all, HashMap and the like do not do this). -- 2.11.4.GIT