2010/06/23 by Scott Mervyn Walker, Scott Walker, Walker, Scott +7
Computer Science · #Advanced Software Engineering Methodologies #Distributed #Distributed and Parallel Computing Systems #Distributed systems and fault tolerance #FOS: Computer and information sciences #Parallel #and Cluster Computing (cs.DC) #cs.DC
paper · pdf · doi:10.48550/arxiv.1006.4504
Submitted to SAC05
arxiv created 2010/06/23 · openalex publication_date 2010/06/23 · arxiv updated 2010/06/24 · openalex created_date 2025/10/10 · openalex updated_date 2026/08/04
This paper explores technology permitting arbitrary application components to be exposed for remote access from other software. Using this, the application and its constituent components can be written without concern for its distribution. Software running in different address spaces, on different machines, can perform operations on the remotely accessible components. This is of utility in the creation of distributed applications and in permitting tools such as debuggers, component browsers, observers or remote probes access to application components. Current middleware systems do not allow arbitrary exposure of application components: instead, the programmer is forced to decide statically which classes of component will support remote accessibility. In the work described here, arbitrary components of any class can be dynamically exposed via Web Services. Traditional Web Services are extended with a remote reference scheme. This extension permits application components to be invoked using either the traditional pass-by-value semantics supported by Web Services or pass-by-reference semantics. The latter permits the preservation of local call semantics across address space boundaries.