Standards are often cited as important, helping to prevent vendor lock-in and allow for interoperability between heterogeneous implementations. However, as Steve Jones points out recently, it is still common for many SOA practitioners to ignore selecting standards at the start of the SOA lifecyle. In this article he outlines where standards should fit in and how REST is no exception to this rule.
IBM, Oracle, Red Hat and others have just announced the formation of the Web Services Test Forum, a venue for continuous testing of interoperability for heterogeneous Web Services implementations as well as a flexible way for vendors and customers to define the interoperability scenarios that are important for them. But how does this relate to WS-I and why has Microsoft not signed up to it yet?
In this interview, recorded at QCon London 2008, Red Hat Director of Standards and Technical Development Manager for the SOA platform Mark Little talks about extended transaction models, the history of transaction standardization, their role for web services and loosely coupled systems, and the possibility of an end to the Web services vs. REST debate.
In a recent discussion Mark Little and Greg Pavlik discuss whether transaction coordinators and transaction protocols are necessary in the context of widely distributed units of work. Isn't the knowledge of state alignment patterns enough?
In his blog, Nick Allen, program manager in the Connected Systems Division at Microsoft, went on to collect several definitions and clarify Microsoft's position on the question.
This question prompted a heated debate on MSDN in the wake of the release of the first web service transaction standard last May. Juval Löwy from IDesign, Arnon Rotem-Gal-Oz and others exchange their arguments as they answer the question.
Sun's latest Project Tango release includes WS-AtomicTransaction and WS-Coordination support. They also have demonstrated interoperability .NET 3.0 clients.
OASIS approves the Web Services Transactions committee specifications as a new standard and the TC co-chair blogs about its history.
ACID transactions don't work for long-lived use cases. This article documents historic approaches taken in the CORBA and J2EE communities toward extended transactions, how SOA is a more natural fit, and why WS-TX & WS-CAF may finally hold the answer.