8. RELATED WORK
SOAP vs. REST has been an ongoing discussion for some time
on the blogosphere and has also recently gained attention in the academic
community. None of the existing work employs a structured
and detailed comparison method based on architectural decisions.
For example, the ECOWS 2005 keynote [17] focused on the reconciliation
of WS-* with REST, whereas [55] gives a comparison
of the two approaches from the point of view of their application
to workflow systems. A good discussion on whether the Web (and
in particular RESTful Web services) can fulfill the requirements of
enterprise computing can be found in a recent W3C workshop [23].
A comparison of RESTful Web services and so-called “Big Web
Services” is also found in Chapter 10 of [33]. In it, a critical look
to the WS-* stack is given in terms of how it does not fit with
the “resource-oriented” paradigm of the Web. The chapter also attempts
to show how simpler RESTful techniques can be used to
replace the corresponding WS-* technologies. The distinction between
“resource-oriented” and “service-oriented” architectures was
first introduced by [36]. Unfortunately, the book does not provide
a clear definition of the terms services and resources, and its technology
comparison is not based on measurable, objective criteria
such as software quality attributes, design and development effort,
technical risk, and QoS characteristics.
Even if HTTP is a synchronous protocol, the comparison presented
in [22] argues that RESTful calls are asynchronous from an
application layer perspective. Thus, REST can be seen as favorable
solution for simple integration scenarios. Additional architectural
concerns such as the URL design and payload format are not discussed.