Under several regards, various of the recently proposed computational paradigms are open-ended, i.e. they may comprise components whose behaviour is not or cannot be fully specified. For instance, applications can be distributed across different administration domains that do not fully disclose their internal business processes to each other, or the dynamics of the system may allow reconfigurations and dynamic bindings whose specification is not available at design time. While a large set of mature design and analysis techniques for closed systems have been developed, their lifting to the open case is not always straightforward. Some existing approaches in the process calculi community are based on the need of proving properties for components that may hold in any, or significantly many, execution environments. Dually, frameworks describing the dynamics of systems with unspecified components have also been presented. In this paper we lay some preliminary ideas on how to extend a symbolic semantics model for open systems in order to deal with name-based calculi. Moreover, we also discuss how the use of a simple type system based on name-decoration for unknown components can improve the expressiveness of the framework. The approach is illustrated on a simple, paradigmatic calculus of web crawlers, which can be understood as a term representation of a simple class of graphs.
On symbolic semantics for name-decorated contexts
BRUNI, ROBERTO;
2009-01-01
Abstract
Under several regards, various of the recently proposed computational paradigms are open-ended, i.e. they may comprise components whose behaviour is not or cannot be fully specified. For instance, applications can be distributed across different administration domains that do not fully disclose their internal business processes to each other, or the dynamics of the system may allow reconfigurations and dynamic bindings whose specification is not available at design time. While a large set of mature design and analysis techniques for closed systems have been developed, their lifting to the open case is not always straightforward. Some existing approaches in the process calculi community are based on the need of proving properties for components that may hold in any, or significantly many, execution environments. Dually, frameworks describing the dynamics of systems with unspecified components have also been presented. In this paper we lay some preliminary ideas on how to extend a symbolic semantics model for open systems in order to deal with name-based calculi. Moreover, we also discuss how the use of a simple type system based on name-decoration for unknown components can improve the expressiveness of the framework. The approach is illustrated on a simple, paradigmatic calculus of web crawlers, which can be understood as a term representation of a simple class of graphs.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
