[Main Page]

One Sentence and Three Issues for the Spring School

(Difference between revisions)

Main Page | Recent changes | Log in / create account |

Printable version | Disclaimers | Privacy policy | Current revision
(→Student Research Descriptions)
Line 191: Line 191:
'''Mike Mortensen'''
'''Mike Mortensen'''
 +
I am exploring the effectiveness of using aspect-oriented
 +
libraries with object-oriented frameworks. A library of aspects will be developed
 +
that are associated with the framework and used by framework-base applications,
 +
which will be refactored to use the aspect library. My research hypothesis is:
 +
Framework-based aspect libraries will benefit framework-based applications by improving
 +
integrability, maintainability, validation and defect removal without performance impacts or initial adoption effort being prohibitive.
 +
 +
* As new ideas and tools for identifying and refactoring to aspects emerge, how should they be incorporated into on-going research and pilot studies?
 +
* What are the `best practices' for evaluating the impact of new software engineering techniques on software evolution based on change history?
 +
* How much focus should differences between my two pilot studies (one for each framework) receive?

Revision as of 18:33, 17 March 2006

Go back to Spring School.

The "entrance fee" for getting into the Student Extravaganza Spring School is One Sentence with which you describe your research. This sentence states the assumption(s) you make, which you are going to address in your work. Note that for beginning researchers this sentence can be rather vague and broad, whereas for students nearing the deadline of their thesis this is probably a very precise and concrete statement. Additionally you need to have thought about Three Issues that are important to you, these can be questions, worries, pitfalls, ... Fill in your One Sentence and Three Issues next to your name at the bottom of this page.

The Spring School takes place on Sunday from 9h30 until 18h. There will be around 40 students participating and a bunch of AOSD experts!

We will be at the "Bonn-Aachen International Center for Information Technology" (Dahlmannstr. 2, 53113 Bonn, http://www.b-it-center.de/).

Between 10h and 12h Awais Rachid, Shmuel Katz, Mira Mezini and Eddy Truyen will give presentations on sub fields in AOSD in the morning.

At 12h we have a walking lunch.

After lunch, we warm up the interactive afternoon session by holding an informal panel discussion of around 45 min. with Gregor Kiczales, Eric Eide, Theo D'Hondt and whoever else we can get a hold of by Sunday. Ideally, the issues we ask them to discuss are based on your issues, so be sure to prepare them!

After the panel, everyone is invited to join the discussions. We propose a format in which you get the chance to state and defend your opinion, raise the issues that you think are important, but most importantly, get to know everyone.

Example issues are:

  • When have I done enough of a literature study?
  • How do I identify and formulate precisely my contributions?
  • Is my timing and planning feasible?
  • Are negative results also contributions?
  • How do I evaluate and validate my work?
  • What's the difference between foundational work and related work?
  • Am I doing engineering research or scientific research?
  • How can I get useful and constructive feedback from the people who are guiding or advising me?

Student Research Descriptions

Andrea Schauerhuber

Customization, i.e., the adaptation of a web application's services according to its context of use (e.g., user, device, and location), represents a crosscutting concern in web application development. In my thesis I apply aspect-oriented concepts to the model-driven development of ubiquitous Web applications (UWA) and specifically focus on the design level.

  1. How do I evaluate and validate my work? Is one case study enough?

Andrew Eisenberg Using a Metaobject protocol to make programming editors extensible. MOPs are extensible program processors that are used to add semantics to programs. I am looking at adding editing semantics to programs. It's not really AOP, but that shouldn't matter too much, should it? (Oops, that was 4 sentences, now 5)

Issues:

  1. How can I show that the enhanced editor is better than the standard editor? What is validation?
  2. (As Bram below also says) How do I balance engineering and research?

Andrew Jackson Improved reuse through integrating AOM and MDD.

  1. How do I evaluate and validate my work?
  2. How do I identify and formulate precisely my contributions?
  3. How do I know when the thesis project is enough concretized and mature to be finished?


Antonia M. Reina Quintero Separation of Concerns and Model Driven Development Applied to Web Systems

Issues:

  1. How do I evaluate and validate my work?
  2. How do I identify and formulate precisely my contributions?
  3. How do I know when the thesis project is enough concretized and mature to be finished?

Bram Adams Re(verse)-engineering legacy C systems is a killer application for AOP.

  1. How to find a good balance between engineering and scientific research?
  2. A couple of times, what had seemed to be an original idea turned out to have been explored a bit earlier by others. How does one cope with this?
  3. Papers about AOP are not only published at the AOSD conference, but also appear at ECOOP, ICSM, . . . As time and money are limited, how can one keep up-to-date on the latest developments? Is there a central repository where one can find announcements of new publications (by RSS e.g.)?


Brecht Desmet "My work proposes that a synergy between mixin layers (collaborations) and an expressive pointcut language provide a powerful means to support the development of context-aware and self-adaptable systems."

  1. A couple of times, what had seemed to be an original idea turned out to have been explored a bit earlier by others. How does one cope with this?
  2. How do I identify and formulate precisely my contributions?
  3. How do I evaluate and validate my work?


Carla Silva "My work proposes an aspect-oriented approach to support separation and modularization of social patterns concerns in multi-agent systems."

  1. Can I use the OO metrics to evaluate if my approach has brought benefits to the system design?
  2. How many case studies I have to develop in my thesis in order to illustrate my approach ?
  3. How many related work I have to compare my approach to in order to show its advantages?


Carlos Noguera Behavioural AOP and it relationship with static/structural AOP

  1. As I am starting my phd, how should I know when to take a bird's view of the topic, and when to dive into the details?
  2. Related to the one before, how do I know if I am bitting more than I can chew, that is, if the scope of the work is too hard/large?

Celina Gibbs


Charlotte Herzeel


Chris Mattews My research deals with the structure and transplantability of operating systems components in the context of a SMP locality management system.

  1. how can you obtain test results that will satisfy your critics without testing everything?
  2. how can you make meaningful comparisons between already existing (and probably more complex) systems and the things you build?
  3. how can you know which of your future work will be "best" to pursue?


Christoph Bockisch


Chinthaka Induruwana My research interests are in the field of aspect oriented software development and service oriented architecture (http://www.cs.man.ac.uk/~induruwc/).


David Stauch I am developing Larissa, an aspect language for the family of synchronous programming languages (which are used to program safety-critical embedded systems), with a special emphasis on a semantic definition of aspects.

  1. How do I integrate well into the AOP community?
  2. How do I identify and formulate precisely my contributions?
  3. How do I evaluate and validate my work?


Elisa Gonzalez Boix

My research deals with distributed garbage collection in mobile adhoc networks. I'm proposing a semi-automatic approach where the collector is guided by the developed using an annotation system.

1. How do I evaluate and validate my work?

2. When do we have explored enough of the current state-of-the-art approaches (resembles the second question of Silvia)

Grigoreta Moldovan

My research interests are aspect mining, aspect oriented refactoring and aspect composition.


Jennifer Baldwin

My research entails reimplementing distribution from a patch file to aspects on an open-source JVM and seeing how each approach affects software engineering principles.

  1. Due to the motivation for distribution, I am concerned that performance degradation may cause aspects to be impratical in this particular example. What are the best ways to measure performance?
  2. When is it appropriate to release code? When it "works" or when it has been properly tested? Perhaps in stages?
  3. As Bram said earlier, "A couple of times, what had seemed to be an original idea turned out to have been explored a bit earlier by others. How does one cope with this?". In tying in with my above question, has premature posting of code/results ever led to other researchers expanding on research that has not yet been completed by the individual that started it, therefore causing that individual's future planned research to be irrelevant?

Jenny Munnelly We are working on a domain-specific aspect language for ubiquitous computing.

Issues:

  1. When do we have explored enough of the current state-of-the-art approaches (resembles the second question of Silvia)
  2. How to find a goog splitting point so that we can make two theses out of the topic
  3. Last question resembles the one from Bram about how to deal with the fact that others have already expored the same research question


José Manuel Magno Lopes

My research goal is to develop a framework to derive an Aspect-Oriented Architecture from an Aspect-Oriented Requirements Specification.

Issues:

  1. Can we construct a software architecture by iteratively composing concerns at the architectural level?
  2. What is the best process to select an architectural representation of a concern?
  3. Does it make sense to talk about plug-in of concerns?


Lukasz Szala


Marc Bartsch

My research interest is on aspect-oriented programming and measurement, in particular the validation of ao measures.

  1. What is the best approach to discuss and test "half-baked" ideas about one's research area in order to improve on it?
  2. I also support Bram's worry about how to deal with the fact that others have already expored the same research question.


Marco Yuen


Marta S. Tabares It is possible to study the different tracing models for crosscutting concerns and to propose a general model, or domain model, which conforms to the domain elements of the set of models. Issues

  1. Is there a consensus or orientation about the best formal specification language that can be used in the context AOSD?
  2. What kind of relationships should the meta-model have between their components?
  3. Should be possible the meta-model to provide semantics for obliviousness features?


Michael Desmond

A Fluid source code editor in Eclipse, allows exploration of the software space related to a particular document, within the context of the document itself. Uses interactive code technology and code "folding" for exploration.

  • Are negative results still a valid contribution?
  • How can you tell (or even begin to) if your research will be useful in the real world?


Mike Mortensen I am exploring the effectiveness of using aspect-oriented libraries with object-oriented frameworks. A library of aspects will be developed that are associated with the framework and used by framework-base applications, which will be refactored to use the aspect library. My research hypothesis is: Framework-based aspect libraries will benefit framework-based applications by improving integrability, maintainability, validation and defect removal without performance impacts or initial adoption effort being prohibitive.

  • As new ideas and tools for identifying and refactoring to aspects emerge, how should they be incorporated into on-going research and pilot studies?
  • What are the `best practices' for evaluating the impact of new software engineering techniques on software evolution based on change history?
  • How much focus should differences between my two pilot studies (one for each framework) receive?


Mohammed Al-Mansari

I am working in the application of AOSD in middleware services, my current focus is on object persistence.


Nelis Boucké

Mechanisms are needed that explicitly modularize architectural concerns in the architectural description and that support explicit relations between (or composition of) architectural concerns.

  1. What are good and objective metrics to quantify in a scientific way how "goog" a software engineering approach is?
  2. How to get from a good 'idea' to a good paper? Which process to follow to write a high quality paper?
  3. How to validate your research with a reasonable effort when it is situated on the architectural level and intended for large-scale systems?

Pericles Leng Cheng

I am interested in applying AOSD techniques in embedded systems and finding the trade offs that may be required to prove that such an application is beneficial to embedded systems developers.

  1. Where can aspects be used in embedded systems without any negative implications?
  2. If there are negative and positive results from implementing aspects in an embedded system can a trade off be computed so that it proves the effectiveness of using AOSD in that realm?

Pourya Shaker


Sara Forghanizadeh


Sebastian Bergmann

I am working on a lightweight, dynamic implementation of AOP for PHP that does not make changes to the language's syntax.

Issues:

  1. Is bringing AOP to yet another programming language sufficient for a diploma thesis?
  2. What are typical Web Development problems that could be solved (more easy or more elegantly) with AOP?

Serena Fritsch We are working on a domain-specific aspect language for ubiquitous computing.

Issues:

  1. When do we have explored enough of the current state-of-the-art approaches (resembles the second question of Silvia)
  2. How to find a good splitting point so that we can make two theses out of the topic
  3. Last question resembles the one from Bram about how to deal with the fact that others have already expored the same research question.


Silvia Breu

Crosscutting concerns in general are not atomic and therefore may not be expressible as independent aspects.

Issues:

  1. One for the money: There seem to be huge differences in the volume of a PhD thesis between countries and even same-country universities. How do I know that I've done enough work of anything: literature, related work, own ideas, etc. and how do I cope with the 'prejudice' that a PhD from country A is worth more than from country B and doesn't get me a job I want because of that inflexible view on things?
  2. Two for the dough: While there is no doubt that there exist crosscutting concerns, there is the doubt that AOP is the sole solution to cope with them. Are we open enough for other approaches or can we be sure that AOP subsumes these alternatives? How can we make sure to not run blindly in the wrong direction?
  3. Three to get deadly: How does a research student deal with stupid reviews such as 2-sentences-reviews, or reviews where obviously the paper wasn't read?

Thorsten Schäfer


Retrieved from "http://aosd.net/wiki/index.php?title=One_Sentence_and_Three_Issues_for_the_Spring_School"


Wiki

Instant Feedback

Edited by the AOSD Steering Committee.  Maintained by the webmaster