Wharton Computing’s Software Development Technology Assessment Process (DevTAP) provides a repeatable approach for measuring both existing and new technologies for their ability to deliver solutions that meet the mission of all Wharton departments, no matter how diverse. Designed to iterate on a six-month schedule, DevTAP will be used to reassess existing technologies in use on a regular basis.
Wharton Computing was facing a problem common to most IT organizations:
When an institution’s existing technology is reaching obsolescence, what is the best way to select and integrate its replacement?
devTAP was launched in 2013 to address this need and, more specifically, to find a new software development language that could replace ColdFusion. As it’s difficult to build a consensus for this type of decision —given the complexity of weighing so many options and opinions —the authors broke down the decision-making process into logical areas that were assessed using performance indicators. The primary strategic goal for the team, however, was to ensure that the process was proactive; thus, DevTAP’s assessment framework was designed to be iterative, with decisions revisited and reassessed on a regular basis.
As a result of DevTAP’s first iteration, which concluded in March 2014, Python/Django was selected as the replacement for ColdFusion, with C#/.NET designated for use with other business-specific requirements as needed.
DevTAP committee members developed a variety of metrics to assess and score nearly a dozen new programming languages for Wharton. The final decision came down to:
- Long-term viability: Python/Django has the best chance of keeping Web development at Wharton Computing on stable footing for the foreseeable future.
- Collaboration: Python has the largest penetration as a language across Wharton Computing, as it is already used in research programming, server-side scripting and some Web applications. In choosing Python/Django, existing knowledge can be better leveraged across IT teams and expand the use of a single language, rather than add another to the mix.
- Prevalence of use/Resources/Training: Python/Django is extremely popular in the external development community (and is rapidly gaining traction at Penn), so it was determined that an ample amount of training resources are readily available to support the transition.