Designing Delivery

Foreword

In the late 1990s, I had the honor of working at Forte Software, a company that produced a development and operations platform for what at the time passed for “large scale distributed systems.” The brainchild of Paul Butterworth—a man I believe gets less credit than he deserves in the annuls of distributed systems history—Forte allowed developers to develop applications in a scalable software services model, and then easily deploy the resulting user interfaces and services.

Auto-scaling was built into the platform, though you had to architect smartly to use it well. Redundancy and failure recovery were also selectable by checking a box and entering the number of instances you wanted of any given service. To this day, I have yet to see another developer environment that provides the experience Forte offered for distributed systems—modern Platform-as-a-Service offerings included.

This was my first experience with an environment that bridged development and operations functions, and—though I didn’t recognize it at the time—it created a different dynamic in the teams that built and deployed applications. I remember frequent conversations between the operations, quality assurance, and development teams; three or four people sitting at a computer discussing the best way to organize, test, scale, and replicate services to maximize availability and performance while minimizing cost—both financially and psychologically.

These days, multidisciplinary approaches to software are ...