Re: [aspectj-users] AOP Master Thesis

[Patrick Chanezon <patlist@xxxxxxxxxxxx>]

The outline seems OK.
- For the framework piece I would add:
* AspectWerkz, and BEA Weblogic (see http://aspectwerkz.codehaus.org/ and http://www.jroller.com/page/cbeust/20030409): BEA recently hired Jonas Bonér (see http://www.jroller.com/page/cpurdy?anchor=aop_hard_on) so I guess they'll put more emphasis on AspectWerkz in the future.
You should also take a look at the weblogic aspect framework http://www.jroller.com/page/cbeust/20030722, which lets you advise a few predefined pointcuts in weblogic today.
I guess their first step will be to create an adapter for aspectwerks.
* The new Apache Geronimo J2EE container: http://incubator.apache.org/projects/geronimo.html
It is based on AspectWerk as well.
It is the new JBoss competitor in the open source app server land.
A few JBoss contributors left JBoss and created the core developers network http://www.coredevelopers.net/news/ and joined the Apache Geronimo project.
* JAC, http://jac.aopsys.com/ Renaud Pawlak's aspect based app server.
Don't know if it will succeed in the marketplace but the ideas he develops are interesting.

You should read http://beust.com/aop-thread.html which uncovers a few problems that still need to be solved for AOP to pick up.
You should look at the aosd-discuss mailing list.
The aopi-discuss mailing list had some interesting discussions about standardization of AOP, but it seems dead now.

I recommend the book Generative Programming by Krzysztof Czarnecki & Ulrich Eisenecker (http://www.amazon.com/exec/obidos/tg/detail/-/0201309777/104-0879821-5436754?vi=glance) to get a better idea of AOP in the more general perspective of generative programming techniques.
See my weblog entry about that:
see http://radio.weblogs.com/0105673/2003/04/25.html

Last, I recommend you to contact the 3 guys of the Thesis Weblog, http://www.dalager.com/thesis/
They started a weblog about their PhD on JBoss (see http://dalager.com/thesis/archives/000009.html)
You could join your project as a module of theirs, or at least align your efforts, since you're all from accademia and share a common interest.

Hope this helps.
Good luck: this is a great and very relevant topic for a master thesis !

P@

Steffen Euch wrote:

Hi!

I'm studying "Computer Science and Media" and I'm going to write my master thesis about a AOP topic. I'm planning to concentrate on AOP in connection with J2EE. I would welcome some comments or thoughts what you think about it.

Following is short outline of what I'm planning to write so far:

- Weaknesses of OOP and J2EE
- Introduction of AOP (aspects, join points, advices/interceptors, aspect weaver, domains or problems where AOP can help, overview over the state of the research field)
- Comparison with 1 or 2 related Post OO Techniques (MDSOC or others)
- Comparison of frameworks, that could be used in connection with J2EE (AspectJ, JBoss, ...)
- AO Software Development Techniques and their applicability in connection with J2EE
- Practical evaluation of AOP on a concrete J2EE-task
- Concluding assessment of AOP in connection with J2EE

I started to read about AOP and connected research fields about a month ago and I have realized that the field is wider than I thought at the beginning. Because of this huge field, I have some difficulties to narrow my thesis down on a central question or problem.

Thanks for your comments!

Steffen Euch

_______________________________________________
aspectj-users mailing list
aspectj-users@xxxxxxxxxxx
http://dev.eclipse.org/mailman/listinfo/aspectj-users