Disjointed tools have inundated the application lifecycle. At its roots, tool diversity is a good thing. Over the past few years, it has transformed the way software is built via Agile methods, open source tools and differentiating vendors. But it has also wreaked havoc on the decade-old promise of Application Lifecycle Management (ALM). We need a new kind of infrastructure to deliver on that promise in a world where developers get to choose the tools that make them most productive. The time has come for an integrated lifecycle bus that allows information to flow freely between developers, testers, business analysts and administrators. This infrastructure will enable us to connect the best offerings in ALM in order to create a Lean Software Lifecycle, and implement a build-measure-learn loop that connects business idea to app store, and back again. We need a set of common data models and architectural patterns. Most importantly, we need to establish the common artifact that will connect the lifecycle. We are calling the category of infrastructure that will enable this Software Lifecycle Integration (SLI).
When outlined on a whiteboard or diagram, the architecture of today’s ALM stack resembles a half-eaten bowl of spaghetti, with meatballs corresponding to the various tools selected by stakeholders. We have two choices, either find a way back to the homogenous and single vendor solution of the 1990s, or embrace heterogeneity and come up with an infrastructure that provides end-to-end collaboration and traceability across best-of-breed tools.
Not long ago, we witnessed a similar transformation of the app dev stack. Once the services, databases and app server space enabled heterogeneity, the middleware category materialized and the Enterprise Service Bus (ESB) emerged, along with the new title of “Enterprise Architect”. History doesn’t repeat, but it does rhyme. It’s now time to create the role of the Lifecycle Architect, and to define an architectural framework for connecting the software lifecycle. Just as the notion of services was key to enabling the ESB, and file and documents abstractions were to sharing data, we now need an abstraction to connect the software lifecycle and to create a Software Lifecycle Bus. That abstraction is the social task.
Today Tasktop is a kicking off an effort to bootstrap the SLI discipline, with a series of whitepapers discussing the technical architecture, common data model, and technical tools. We are proposing the Eclipse Mylyn m4 open source project as a home for collaborating on a de facto implementation of the SLI data model, which will leverage what has been learned from the adoption of Mylyn, integrations that build on it, and new efforts around the open standards of Linked Data and OSLC. Later this week we are also launching an SLI integration patterns catalog, based on existing input from our customers, and open for all to contribute to. By the end of the week, with input from key thought leaders present at the ALM Connect subconference of EclipseCon, we plan to release a first draft of the SLI manifesto. To learn more and to participate, see:
||Whitepaper: Building the Business Case for Software Lifecycle Integration|
||Whitepaper: Software Lifecycle Integration Architecture|
||Software Lifecycle Integration landing page|
||Eclipse Mylyn m4 Project Proposal Draft|
||SLI Patterns Catalog Wiki (to come)|
About Mik Kersten
Dr. Mik Kersten is the CEO of Tasktop Technologies, creator of the Eclipse Mylyn open source project and inventor of the task-focused interface. At Tasktop, Mik sets the strategic direction of the company as well as drives many of Tasktop's key partnerships and key customers accounts. He created Mylyn and the task-focused interface during his PhD in Computer Science at the University of British Columbia. Mik has been an Eclipse committer since 2002, is a 3-time elected member of the Eclipse Board of Directors and serves on the Eclipse Architecture Council. Mik's thought leadership on task-focused collaboration and improving the software economy makes him a popular speaker at software conferences, and he was voted a JavaOne Rock Star speaker in 2008 and 2009.