EPLlab Project

[HOME] [DUTIES] [WORKING HOURS] [PROJECTS] [PUBLICATION] [RESTRICTED]

Download
Licensing

EPLlab Thumbnail

(Please click on image for a larger picture)
Please note that GIMP appears in the screenshot and is not part of EPLlab.

 

Project Status

16 Oct 2004 : Update on Third Prototype to add IoP's S Series File processing.

08 Oct 2004 : Third Prototype (Legenda) Completed.

 

 

Powered by:

Eclipse Workbench

 

 

Notes:
This is just the temporary home for the project while its permanent home is constructed.

 

 

 

(6 June 2005) News update...

Name changed to BAMMfx. Website moved to http://www-staff.psychiatry.cam.ac.uk/~co224/projects/bammfx

 

What's new in the third prototype (Legenda)

  • Much nicer display.
  • Input Table now accepts mutliple datatypes, e.g., String/File/Directory/Selection. This makes data entry easier and that your input data is more likely to be valid.
  • The data hierarchy and data editor are implemented. You can see it in action when you choose to "Design Matrix" and "fMRI Dataset" in FBAMM's input table: They are both File type, but "Design Matrix" allows you to define "Block/Event Related Design Matrix" and "fMRI Dataset" give you a dialog box that can automatically convert your Bruker Data to UNC Data.
  • Eclipse Extension Mechanism is used to define the data hierarchy and data-editor.

Eclipse PipeLine Laboratory (EPLlab)

Eclipse Pipeline Laboratory, or EPLlab for short, is a GUI specifically build to chain program together. Every program in the GUI is known as a module. The modules are chained up to form a pipeline.

It is designed to be a generic GUI which can be customized for different usage scenario. It is designed as both a development tool, and an end user GUI application. (See Why use this GUI?)

Why use this GUI?

There are a lot of GUI that offers the almost the same capabilities, why should you consider using this GUI?

The proof of the pudding is always in the eating. It always had been and always will be. But we will need to convince you to pick our pudding and not others. These are the important features of our software: our competitors:

  • As long as there is a way of invoking the program via command line, you can put it into the GUI. No programming necessary!
  • The dependency of one module on another is clear: They are connected by wires.
  • Batch capability: Define once, run many times.
  • Because it is written in Java, it is portable across all platforms.

For Users

  • EPLlab offers a "one-click" installation process on all platforms, something most of our rival does not offer.
  • Instant batch-processing capability. As long as you can get all your programs into a pipeline, there is no reason why you cannot apply the same pipeline to all your datasets.
  • Define your own modules/pipeline without having to know how to program the computer.

For Developers

  • The software is build on the Eclipse Platform. This is an open source project with strong support from IBM and other industrial heavy weight. Lets face it: The involvement of these guys with experience in software development puts the software architecture on a stronger foundation from the one you (and I) develop from scratch.
  • Unique to EPLlab, you can impose your own data structure on to EPLlab. Every fields has its own particular data requirement, with most off-the-shelf software you have to hammer/distort your data requirement to "fit" that of your GUI. Ouuch! With EPLlab, we contort the GUI to fit your needs! Transfering all the "Ouuch!" to us. (Don't worry, we have a high pain threshold.) Moreover, we introduces the concept of data hierarchy where a child datatype can be placed where the parent datatype is expected but not vice-versa. For example, a fMRI dataset is just an generic-scanner-generated dataset (images with physical dimension attached to each pixels) with fMRI-sepecific information such as Magnetic Field Strength. Hence, where your program expects generic-scanner-generated dataset, it should and will accept fMRI dataset. However, the opposite is not true and therefore program expecting fMRI dataset should rejects generic-scanner-generated dataset. It is surprising that a lot of GUI software cannot make this distinction!
  • As a "plug-in", EPLlab coexists happily with the Java development environment, C/C++ development environment, and team-support features (CVS) in the Eclipse Platform, allowing you to perform all your tasks in one program!
  • Using Eclipse platform's philosophy, "EPLlab is a kind of universal GUI platform - an open extensible GUI for anything and nothing in particular". For nothing because EPLlab itself does not come with any "useable" program functionality and is not even designed for one. Its for anything because you can customized it easily for your application. Read on...
  • We are going to offer a few ways of integrating your apps with the GUI from using the GUI to configure the command-line call, to full integration at source code level where the GUI interrogate your applications to find out what it needs. Choose the level of integration you are comfortable with.
  • Need to give your command-line-based program a GUI, look no further!
  • Every significant data format is done in XML. This makes it easy to transfer our pipeline to another program.
  • Open source. You will never be held ransom by software suppliers.
  • You can exploits other interesting features of Eclipse Platform such as OnLine Help, automatic updates. Other interesting plugin being developed for this platform by independent developers include "chat-rooms", "instant messaging" and development environment for PERL, php etc. Users love these features ... and you get them for free!

Still interested?

Great! Now its time to get your hand dirty. Checkout our download section.