Last year Apigee created an Apigee 127 open source project to make production-grade APIs using Node.js and Swagger documentation tools a whole lot easier to develop. In 2015, Jeff West, a product manager for Apigee, says the company plans to extend the reach of that project to include Java and, by extension, a wider range of platforms in and out of the cloud.
Now that Apigee 127 is available on GitHub, West says that going into 2015 Apigee plans to make the development platform that enables developers to create entire applications out of a Swagger file more broadly accessible.
The goal, says West, is not only to make it easier for professional developers to create applications, but also to empower a new generation of “citizen developers.” Apigee 127 accomplishes that goal, says West, because all message validation, OAuth, quota, caching and other services are managed via the metadata defined in the Swagger specifications.
Obviously, Apigee is hoping that developers will deploy those APIs on Apigee Edge. But developers have the option of deploying the APIs on any platform-as-a-service environment that can support Node.js or, in the future, Java.
The degree to which the developers will accept API modeling and automation in 2015 remains to be seen. Most developers clearly believe that APIs need to be optimized for particular use cases. At the same time, however, there are fairly routine usages where automatically generating APIs using modeling tools should be more than sufficient for the task at hand.
It will be up to individual developers to decide what approach makes the most sense at any given time. But as application backlogs continue to build in the new year, it’s apparent that pressure to accelerate the API development and deployment process is mounting.
In fact, as DevOps processes continue to mature, more APIs might start being generated by, for example, an IT operations team looking to consistently expose a specific back-end application or service. It may even become desirable over time for the IT operations team to take over responsibility for the ongoing management of APIs created by both professional and citizen developers.
Whatever the outcome, West says that regardless of the programming language that an API is written in, there will be more reliance on API modeling tools and automation in 2015. To what extent modeling tools and automation take over the API development and deployment process altogether is, of course, still anybody’s guess.