Serena is well-known for cross-platform change management but it has been expanding its offering of late, to cover more of the Application Lifecycle. Its SAFE (Serena Application Framework for Enterprises) vision centres on configuration/change management and version control but now includes business process modelling and automation, requirements management, issue and problem management, IT governance and methodologies for development and support, build management, deployment and production support – but not coding automation or testing (see the product summary here).
So, it is complementary to offerings from people such as Microsoft and Compuware, which are more code-oriented. Its strength is that it can support the enterprise as a whole, including the mainframe. Other lifecycle offerings tend to use third-party offerings (from BMC, for example), for operational support across the enterprise.
However, with the widespread availability and acceptance of standard frameworks such as ITIL and OMG's MDA, all the players in this space are expanding their scope – Telelogic, for example, has acquired System Architect; and Release 10.3 of this now gives its Application Lifecycle Management toolkit a respected capability for enterprise architecture and business process management modelling.
The appearance of such visualisation tools addressing the entire application lifecycle is a response to the increasing complexity of the IT environment and recognition of its importance to the business.
There is now increasing acceptance of abstraction and modelling in general, but this has only been possible because of the emergence of genuine standards, catalysed (we think) by Rational's sterling work in ending arguments over the shape of boxes with the UML (Unified Modelling Language) and RUP (Rational Unified Process). IBM now owns Rational and IBM/Rational's support of the (now independent) Eclipse open-source platform points (we think) towards the future in this area.
Watching these developments, Serena has seen the need for a pragmatic, low cost and open framework for managing the application lifecycle, using a loosely-coupled service-oriented approach.
Point-to-point integrations of different operational technologies are as costly to develop and maintain as point-to-point integrations of business applications, as Kevin Parker, VP, market development at Serena and now Evangelist for what it calls "ALF", explained to me. ALF – Application Lifecycle Framework – is a new Eclipse project sponsored by Serena, although other notable companies are involved including: Catalyst Systems Corporation (with Openmake); Cognizant (IT service delivery model); Compuware (IT governance and QA though to development and QA); Secure Software (application security) and Segue Software (testing and quality optimisation).
ALF, it's promised, will be a service-based architecture providing a platform-agnostic foundation for multi-vendor integration and service orchestration.
Parker told me a story illustrating the need for process in application delivery. Apparently, the rear windshield specification for a certain German car included the ability to withstand impacts well over 120 km/hour. An obvious mistake, thought the commendably thorough engineers, even posh German cars can't travel that fast in reverse, and reduced it to something more reasonable. U
nfortunately, the cars were to be shipped to the South of France at high speed by rail, packed on open tracks and facing backwards, and they all arrived with the rear windshields shattered. A bit of process might have encouraged the engineers to investigate that specification further before making an arbitrary correction.
According to Parker, ALF won't be "high ceremony". It will provide a low-barrier to entry and minimize the need for customising a universal meta-model before participants can get started. Nevertheless, the experience of the OMG with MDA and the Meta-object Facility (MOF) seems to show that building on a properly thought-out meta-model brings long-term benefits (pragmatic initiatives fit together better) and I was pleased to note that Serena is apparently looking at the theoretical underpinnings of ALF too, by sponsoring academic research.
The ALF project is important, and not only because application lifecycle management is an important aspect of delivering the IT Governance companies are expected to have today as part of Corporate Governance generally; and is too important to entirely trust to vendors. Eclipse is successful but it is now expanding its scope beyond the IDE and a community of pure developers – will the Eclipse organisation cope with the change? ALF is, perhaps, a test case.
And can Serena's culture cope with managing an open source community (and, indeed, with its recent take-over by a benevolent VC, which is taking it back to private ownership)? I am optimistic that all will go well and that we're seeing the emergence of a truly independent framework around which to build IT governance – but only time will tell.
Nevertheless, although the competing Microsoft application lifecycle environment is at last getting decent tools and an underlying model, I can't really see the industry accepting Microsoft's platform in the same way that it might accept Eclipse as "an independent open eco-system [built] around royalty-free technology and a universal platform for tools integration" for more than just coding.®
David Norfolk is the author of IT Governance, published by Thorogood. More details here.