Filters

2012 - 20192
Reset filters

Settings

COAR Access Rights: open accessSubject: Web Services

Search Results

Now showing 1 - 2 of 2
  • Thumbnail Image
    Article
    Citation - WoS: 8
    Citation - Scopus: 10
    Cobol Systems Migration To Soa: Assessing Antipatterns and Complexity
    (Kaunas Univ Technology, 2019) Mateos, Cristian; Zunino, Alejandro; Flores, Andres; Misra, Sanjay
    SOA and Web Services allow users to easily expose business functions to build larger distributed systems. However, legacy systems - mostly in COBOL - are left aside unless applying a migration approach. The main approaches are direct and indirect migration. The former implies wrapping COBOL programs with a thin layer of a Web Service oriented language/platform. The latter needs reengineering COBOL functions to a modern language/platform. In our previous work, we presented an intermediate approach based on direct migration where developed Web Services are later refactored to improve the quality of their interfaces. Refactorings mainly capture good practices inherent to indirect migration. For this, antipatterns for WSDL documents (common bad practices) are detected to prevent issues related to WSDLs understanding and discoverability. In this paper, we assess antipatterns of Web Services' WSDL documents generated upon the three migration approaches. In addition, generated Web Services' interfaces are measured in complexity to attend both comprehension and interoperability. We apply a metric suite (by Baski & Misra) to measure complexity on services interfaces - i.e., WSDL documents. Migrations of two real COBOL systems upon the three approaches were assessed on antipatterns evidences and the complexity level of the generated SOA frontiers - a total of 431 WSDL documents.
  • Thumbnail Image
    Conference Object
    Citation - Scopus: 13
    Predicting Web Service Maintainability Via Object-Oriented Metrics: a Statistics-Based Approach
    (2012) Coscia,J.L.O.; Crasso,M.; Mateos,C.; Zunino,A.; Misra,S.
    The Service-Oriented Computing paradigm enables the construction of distributed systems by assembling loosely coupled pieces of software called services, which have clear interfaces to their functionalities. Service interface descriptions have many aspects, such as complexity and quality, all of which can be measured. This paper presents empirical evidence showing that services interfaces maintainability can be predicted by applying traditional software metrics in service implementations. A total of 11 source code level metrics and 5 service interface metrics have been statistically correlated using 154 real world services. © 2012 Springer-Verlag.