William Vassilis Karageorgos is a research assistant at the IASA inter-university research institute. While pursuing a doctoral degree on High Energy Physics, William has been working on the Applications Database - a service provided by the European Grid Initiative (EGI). This service could help to prevent software decay, so we asked William to give us an overview.

For those of you who are not familiar with it, the Applications Database (or AppDB, as we call it for short) is a service that stores information about tailor-made computing tools, and the programmers and scientists who have developed them. AppDB embraces all scientific fields, from resources to simulate exotic excitation modes in physics, to applications for complex protein sequences analysis. The information stored in the database  is available to everyone. Nearly a year after the project's start-up, and having successfully completed the initial testing phases, we are proud to say that it features complete write access for all registered users, integration with Virtual Organisation data from the EGI Operations Portal, a beta read-only RESTful Web-API, and a demo platform of web widgets.

Storing pre-made applications means that scientists don't have to spend research time developing their own software. Our main aim for AppDB is to avoid duplication of effort across the Distributed Computing Infrastructure (DCI) user communities, and to inspire scientists less familiar with programming into using the European DCI.

Scientists who have developed software and use the system to publish their work may also benefit from the database. Software tends to become stale with time, especially if the its user pool stagnates, and can become unstable if new requirements are added without sufficient planning. The AppDB can help ensure that published work will be sustainable by providing an environment with a vibrant user pool, and experts who may share knowledge across projects of similar interests.

At beginning of the project, our team was primarily focused on migrating existing data from EGEE (Enabling Grids for e-Science: the precursor organisation to EGI), and performing the modifications needed to capture the spirit of EGI and improve its service. We added support for all EGI-endorsed middlewares and introduced the concept of tools, by initially porting the entries of the RESPECT program. We also started storing personal profiles for each application developer so that it would be easy to search for an expert.

It's important for us to provide an extensible, consistent, and responsive end result to our end-users. We based our software design on architectures, patterns, and techniques such as Multitier Architecture, Model-View-Controller (MVC), and Asynchronous JavaScript and XML (AJAX).

The service is under constant development, and we are working hard to address bugs and feature requests, and providing more functionality. We are working to integrate National Grid Initiatives with data from the Grid
Operations Centre Database, to integrate with the forthcoming EGI Community Software Repository, and to create a subsystem that will help ensure quality of information.

Needless to say, if you are a scientist with high computing demands and haven't checked out the EGI Applications Database yet, now would be a good time. If on the other hand you are already familiar with it, and you have found it to be useful - as we hope you would - stay tuned, as more is to come.

If you would like to find out more, visit the Applications Database.