Fractal ObjectWeb project and further Fractal developments

[Jean-Bernard Stefani <Jean-Bernard.Stefani@xxxxxxxxxxxx>]

Dear all,

The Fractal community, and the Fractal related 
developments are growing nicely, so it is perhaps 
a good time to think about the Fractal project, 
and to plan the next steps. I believe there are 3 
main issues that need to be considered:
1/ Fractal v3 and the evolution of the Julia technology
2/ Making Fractal libraries and frameworks available for Fractal users at 
large
3/ Consolidating a development plan for the Fractal technology at large

Issue 1/ has to do with the development of a new 
version of the Fractal specification, and the 
further development fo the Julia code base. A new 
version of the Fractal specification seems to be 
warranted on a number of grounds:
- clearing out various uncertainties or 
ambiguities in the specification (e.g. making it 
clear that Fractal is not dependent on a a 
synchronous method call semantics);
- improving on areas of the specification that 
remain under-specified (e.g. clarifying the 
constraints applying to configurations with 
shared components);
- extending the catalog of useful component 
controllers, based on several works in the 
Fractal community (this would probably imply a 
re-structuring of the specs so as to ease the 
evolution of this catalog in the future without 
waiting between major releases of the 
specification).

Julia warrants to be extended or evolved on similar grounds:
- allowing different ways to program component 
controllers (eg using aspectoriented programming 
tools)
- consolidating a controler & interceptor 
library, in line with the catalog mentioned above
- allowing different back-ends for the Julia tool 
chain (eg to deal with different target languages)
- evolving, and consolidating a module catalog for the Fractal ADL

Issue 2/ has to do with making better known, and 
properly documenting and advertising  several 
developments based on Fractal, and consolidating
a library of Fractal-based or Fractal-related 
software frameworks and tools, which can be 
re-used by the Fractal community at large, often 
in different settings than previously imagined by 
the framework developers. In effect, this means 
trying to assemble a reuse library of Fractal 
components, frameworks, and tools. This would 
complement the controller catalog alluded to 
above. Examples of frameworks are the Perseus 
framework, or the Dream framework. Examples of 
useful tools are the Fractal Explorer, and the 
Fractal Eclipse plug-in.

issue 3/ is self-explanatory, but it involves 
also thinking about the way we structure and 
document the Fractal project, clearly identifying 
mature things from more experimental stuff, 
organizing and maintaing the catalog and library, 
etc.

To do all these things requires some coordination 
and participation from the Fractal community at 
large. I suggest as a first step to organize a 
Fractal workshop at the next ObjectWeb 
architecture meeting, where we
can dedicate one or two days to discuss these 
issues, and learn about ongoing Fractal related 
developments.
One of the goal of this workshop would be to come 
up with some decisions for action on each of the 
three
issues I mentioned above.

If you are interested in participating to this 
workshop or lending a hand in dealing the issues 
above, or if you just want to comment on this 
message, please drop me a word or comment back on 
the Fractal mailing list.

Best regards,

Jean-Bernard
--

*************************************************************
Jean-Bernard STEFANI
Research Director, SARDES Project
INRIA Rhône-Alpes
655, avenue de l'Europe
Montbonnot
38334 St Ismier Cedex
France
tel : +33 (0)4 76 61 52 57
fax : +33 (0)4 76 61 52 52
email : Jean-Bernard.Stefani@xxxxxxxx
*************************************************************