« "Besöksrekord" - Grattis på mig | Main | Skånsk bloggaremiddag/-lunch »
november 30, 2004
Nathan LaBelle, Eugene Wallingford: Inter-Package Dependency Networks in Open-Source Software
Nathan LaBelle, Eugene Wallingford: Inter-Package Dependency Networks in Open-Source Software
Abstract:
This research analyzes complex networks in open-source software at the inter-package level, where package dependencies often span across projects and between development groups. We review complex networks identified at "lower" levels of abstraction, and then formulate a description of interacting software components at the package level, a relatively "high" level of abstraction. By mining open-source software repositories from two sources, we empirically show that the coupling of modules at this granularity creates a small-world and scale-free network in both instances.
Man avslutar med följande konklusion och diskussion:
This research has shown that package dependency networks mined from two open-source software repositories share the following properties typical to other real-world networks:
• The small-world effect: short geodesic path lengths and high clustering.
• Near power-law distribution of edges.
• The presence of a giant component, [....]
There are many directions for future research in the study of software networks. Currently, there is no model of network formation that takes software dynamics (reuse, refactoring, addition of new packages) in to account. Also, the impact of the network structure on software dynamics should be investigated. Future research should identify other networks in software and move towards formulating a theory of networks and their value to software engineering. Additional dependency networks can be constructed on Windows computers using memory profiling tools, and determining interactions based on shared .DLL (Dynamic Library Link) files and Active-X controls.
Jämför t.ex. med Komplexitet i mjukvaruarkitektur.
Posted by hakank at november 30, 2004 07:30 FM Posted to Social Network Analysis/Complex Networks