You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@commons.apache.org by tc...@apache.org on 2005/01/29 16:43:41 UTC
svn commit: r149051 - /jakarta/commons/sandbox/javaflow/trunk/TODO
Author: tcurdt
Date: Sat Jan 29 07:43:40 2005
New Revision: 149051
URL: http://svn.apache.org/viewcvs?view=rev&rev=149051
Log:
updated what's on the table
Modified:
jakarta/commons/sandbox/javaflow/trunk/TODO
Modified: jakarta/commons/sandbox/javaflow/trunk/TODO
Url: http://svn.apache.org/viewcvs/jakarta/commons/sandbox/javaflow/trunk/TODO?view=diff&rev=149051&p1=jakarta/commons/sandbox/javaflow/trunk/TODO&r1=149050&p2=jakarta/commons/sandbox/javaflow/trunk/TODO&r2=149051
==============================================================================
--- jakarta/commons/sandbox/javaflow/trunk/TODO (original)
+++ jakarta/commons/sandbox/javaflow/trunk/TODO Sat Jan 29 07:43:40 2005
@@ -1,8 +1,44 @@
+In no particular order:
+
o integrate just4log into the build system
o website
o testcases!!!!
-o documentation of the rewriting process
-o research: asm implementation instead of BCEL
+o documentation of the rewriting process.
+ maybe re-evaluate and talk to the RIFE
+ guys about joining forces
+o make the Stack class use a hierarchical
+ approach to support differential continuations
+o make the Stack and Continuation classes
+ serializable. (make sure all dynamics are
+ handled through the context object. what
+ about logging?)
+o get rid of the Continuable and ContinuationCapable
+ marker interfaces
o fix bugs:
o use of "suspend" in constructors
o inner classes
+o implement an ANT task for build time class
+ rewriting (already have something on my disk)
+o addition to async ContinuationCompilingClassLoader
+ also re-implement the old sync behaviour in the
+ ContinuationClassLoader
+o change the ClassTransform interface to
+ be byte-code library independend
+
+ byte[] transform( final InputStream is ) or
+ byte[] transform( byte[] clazz )
+
+ o check whether we need two passes
+ 1. parsing
+ 2. rewriting
+ o maybe only wrap the function calls
+ that may lead to a suspend (control
+ flow analysis)
+ o asm implementation instead of BCEL
+
+o BCEL
+ o push some bug fixes in BCEL
+ o get rid of the static repository approach
+
+o talk to the aspectwerkz and aspectj folks
+ about working together
---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org