Track: Architectures for Concurrent and Distributed Systems

 

Most of today's programming languages were designed when computers typically had a single processor and a megabyte of RAM was a luxury. It is not surprising that they do not match current hardware. Nor is it surprising that it is difficult to design and implement efficient and reliable software systems that run on multicore, networked computers.

 

One approach is to start with a successful, sequential paradigm -- object-oriented programming -- and to allow objects to execute multiple threads. Many believe that this approach leads to unmanageable complexity. Functional programming is often cited as a panacea, but it is making its way into mainstream software engineering only with slow, small steps. Almost sixty years after McCarthy introduced LISP, we have reached the point ast which everyone wants 'lambdas'.

 

For this track of C3S2E'2013, we are interested in new approaches that treat concurrent and distributed programming as fundamental features. Processes that communicate by exchanging messages seem to be more promising than objects that share memory. However, the higher-level abstractions that we need for building scalable software have yet to be designed. Although there is a large body of theory -- much of it dating back thirty years or more -- we still lack practical methods for verification and validation of large concurrent systems except, perhaps, in a few specialized application areas.

 

The objective of this track is to encourage more research in this topic by providing a forum for researchers to advertize their results, compare techniques, and exchange ideas.

 

Possible topics of interest include but are not limited to: process specification; modelling and model-checking techniques for concurrent systems; high-level abstractions; plug-and-play systems; abstract interpretation; development tools, including checkers and debuggers; performance studies.