Nov 13, release middleware 1.1

It is so happy that Frank, who is the developer from Gintel starts to use the middleware to build applications. Applications add the meaning to the middleware. ^_^

Development integration work between partners starts!
(working...)

Nov 8, release middleware 1.0

This is how new SIMS middleware looks like. With a component(application) browser, middleware manages the life cycle of the applications. Finally, user can "see" the middleware instead of asking where is the middleware. Middleware is instantiated before applications, takes charge of the resources and behaves like a simple OS. Sexy!

Since the limitation of the Midp, for example we can only run one Midlet at one time, I put a lot effect to fulfill the multi-applications. I will talk more details of Midp and Midlet later.

Web Services is also employed to be the communication protocol from this release! It's easy to generate WS mobile client using WSDL. All the sending and receiving between server and clients are done by WS. Instead of converting XML to objects by using KXml, which is I did in previous versions, WS takes care of the work. All I need to do is making and maintaining good WSDLs.

It is also good that I can apply a lot of knowleages I learned from my master's courses.

Oct 23, New middleware 1.0 Design

Here is some pictures about the design of messaging(APIs, algorithms), which I worked out with Attakorn Wongwattanarat. I like these scratches very much! I will write more about what a mobile middleware does, and how to design such a middleware both in my thesis and this blog.

Oct 8-10, SIMS Meeting in Stockholm

It is so nice to meet all the SIMS people. They are professors, researchers and developers, and we work as SIMS family!

Before this meeting, I worked for the weekend to prepare the meeting. These days are Erik's last days Erik in Appear. We worked closely and figured out some very important issues showing blow. I will write more about this meeting and the discussion with Erik.

Improved Thesis Topic

SIMS Middleware Architecture, Design and Implementation

In SIMS, Appearnetworks develops middleware support for service discovery, session management and messaging. The middleware is also designed to support finding service opportunities and downloading to prove complex SIMS theories. The user of the middleware is SIMS application developers. The thesis project is to architecture the middleware, design the APIs and implement the software running on mobile devices. A primary developed implementation in C# can be used as reference. The new implementation is based on J2ME and J2EE, crossing distributed and asynchronous systems.

Tasks

1. Understand SIMS theory
2. Architecture Middleware, design APIs, implement software
3. Deploy applications on mobile phones and test
4. Make example applications based on middleware and document the software
5. Integrate with other SIMS results from other partners
   

Result

The work will result in a running prototype for the SIMS project, to be used by Gintel for further evaluation of the SIMS technology. At project end, the person working in the project should have acquired good understanding of SIMS technology, middleware design and distributed system programming.

======== Old Proposal (by Erik) ========
A Semantic Interface Middleware Implementation on Smart Phones

Description: In SIMS, Appear has developed middleware support for semantic interface powered service discovery. The middleware is based on the Appear IQ and is implemented as an extension/customization to the proxy and to the client of Appear IQ. The task is to port the existing C# client code to MIDP and design and develop a new version of the SIMS mw.

Tasks:

1. Make a project plan
2. Read-up on SIMS
3. Port SIMS C# client to MIDP
4. Improve the SIMS mw (closer connection to Appear IQ).
5. Prepare Deliverable D5.3

Result: The work will result in a running prototype for the SIMS project, to be used by Gintel for further evaluation of the SIMS technology. The prototype should be integrated with the MIDP library of MAGIC, context enabling the SIMS system. At project end, the person working in the project should have acquired good understanding of both SIMS and the Appear IQ system.

Oct 2, Go for a New Architecture

After reading and analysis, I found some problems existing in the current middleware:

1. It does not support all the requirements.
2. the middleware mechanism is not reasonable. Applications instantiates the middleware. How to make sure we run several applications simultaneously but only one middleware instance? The appropriate way could be the middleware instantiates the applications, manage a branch of applications, and has a browser -like user interface.
   
3. It is not flexible. SIMS is a research project, the theoretical part is continually developing, bringing changes to implementation. For example, how to find a compatible player when joining a service is still discussing. From a developer 's perspective, what I can do is to architecture the software well to deal with the changes.
4. Integration in SIMS project is also a big problem. We have 7 partners in the project. The overall SIMS architecture for runtime applications is still not clear. To make well-designed middleware is also a contribution to the overall architecture.
   

I decided to re-architecture and build the middleware again. Only keep some design from the previous versions, but abandon all the code, both server and client side thoroughly.

At the same time, I have found what I can really contribute to this project and decided to change my thesis topic to a more challenging one!

18 Sep - 1st Oct, Reading List & Requirement Analysis

To get more into SIMS project, both theory and technology, Vedran and I made a reading list for ourselves, I stopped development but only reading.

• D 1.1 Scenarios & Requirements for Methods, Tools and Middleware
• D 2.2 Language and Method Guidelines
• D 5.1 Middleware Support, initial version
• TF 01 - Service Example
• TF 02 - Service Development Process
• TF 03 - Runtime Example (& minutes)
• TF 11 - ODA Usage at Design-time and Run-time (to be read more)
• TF 20 - Service Discovery API Specification
• TF 22 - Code Generation Result

I read most of them carefully, and got a basic overall picture of SIMS. I try to analysis the requirements to the middleware, also use SIMS theory as reference. Does the current API satisfy the requirements? Will the middleware support SIMS applications well? What knowledges, such as Ontology, do I still lack of?

Due to some undetermined issues in SIMS, to have an extendable and flexible middleware is under my consideration.

Sep 17, release Midp middleware 0.2

To show the usage of the middleware, I built a chat application on top of the middleware. The caller and callee is the first example for interpreting SIMS theory in documents.

To bring some buzz to SIMS consortium, we took Erik's suggestion, and we made a desktop video together with Vedran, who is the new EU Project Team Leader and my new thesis supervisor at Appearnetworks.

>>Play Demo Video

After implementing the first version of the middleware, I have acquainted myself with SIMS theory.

(Figure from SIMS Deliverable Document D1.1)

Sep 5, release Midp Middleware 0.1

After a big reading list, I rush to the code. The APIs for this release is designed by Mrs. Åsa Pehrsson. She is great person, loves spicy recipes, and gives me a lot of tips to make things easy. She took the first SEDS program at KTH, so we had a lot to talk even at the first day. She strongly recommend me to ask Pro. Misha as my examiner, since she also got some advices from him when SIMS project started. Åsa left Appear the next day, wish her a happy life with her new career. This is the first movement I showed to SIMS consortium. Memorable!