[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] [List Home]
Re: [aspectj-dev] Progress towards 1.2 release candidate (and some actions...)

For LTW main class, I meant in addition to the script, not
instead-of.  I understand the benefits of the scripts.


Wes Isberg wrote:

On the features front,

- We probably want to see why JDK 1.5 beta
falls down when running .class files produced by AspectJ.
1.5 is not released, but we would want to submit a bug to Sun
if our implementation techniques are valid, so we don't have to
change those techniques.  For 1.2, we should at document
if we don't fix, since many people are using 1.5.

- For LTW, I still haven't heard why having a separate jar and
main class that creates a classloader and delegate wouldn't
be a good interface.  That should work in Java 1.2+ and creates
a testable interface.  I understand that it's not sufficient
for some more complex J2EE deployments, but it should work
fine for simple programs.  (I believe I submitted this as
a patch...)

On the build front,

- the ajdoc/tools.jar should work like the taskdefs/ant.jar
relationship, where ant/lib/ant.jar is not sucked in.  I updated
the (unfortunately-hard-wired) code in Module.java for this;
see if that works for you.

- the images are probably getting munged on copy when the build
scripts filter files during copy for release variables.  I
noticed docs/build.xml variable ${binary.pattern} does not
include **/*.jpg or capitalized versions of other variables;
you might adjust for your images.


Adrian Colyer wrote:

We're getting very close to a 1.2 release candidate now....

* I've created initial versions of README-12, the changes file and the porting guide, which are in docs/dist/docs. Please could the committers take a read through and check that I've represented your features correctly, and that you are happy with the content, style, etc. etc..Any minor changes you want to make please just go ahead and commit them.

* Load-time weaving support is checked in

* I've patched ajdoc (Mik - the method you were calling to launch javadoc does a System.exit() on JDK 1.3, which rather stops things in their tracks. I've added a test to only proceed under the new execute method in 1.4, and to fail gracefully under 1.3. Not sure what your intentions were for 1.3 support? I'm ok with a statement that says you need 1.4 or later to run ajdoc).

* George is wrestling with the build system to get ajdoc to appear in all the right places. Mostly working following your instructions Wes, but we've found that aspectjtools.jar was picking up all sorts of stuff - including the entire contents of (Sun's) tools.jar. George has a workaround for now, will need to be looked at more closely soon.

* There's some issue with the packaging / installing I haven't had a chance to look into yet - README-12 refers to a couple of images in an images subdirectory. These are fine in CVS, and make it into the distribution, but appear to be corrupted at some point during the packaging or installation process, such that the ones you see in an installed distribution are no good.

* I've trawled through the bug DB again, and the only remaining bug I would really like to see fixed before a release candidate is 57238. Erik, Jim we need your help on this one please (Erik, this is the bug we talked briefly about at AOSD 2004 last week). I know of at least one project team using AspectJ for which this is a blocker, but it is not straightforward to determine what the correct behaviour should be, nor how to implement it. See some comments in the bug db.

* Please take a look through the bug db for yourselves to see if there are any other bugs you consider must-do's for a 1.2 release candidate. Mik, what about 54238?
-- Adrian