API Predictions for 2017
From issues involving API copyright to concerns with IoT security and APIs, there are tons of events likely to come from 2017.
Join the DZone community and get the full member experience.Join For Free
APIs have now become a key must-have strategy component for most large companies. We’re seeing widespread deployment of APIs and of tools to help create, design, and manage them.
Still, each year we like to gaze into our crystal ball and see what we think the future holds for APIs. Here is what we and our new Red Hat colleagues dreamt up for 2017.
API Product Management Will Break Through
The notion of treating an API as a fully fledged product and applying product management techniques has been emerging for a number of years. However, as the number of APIs has grown, this has become increasingly important and the concepts have matured. We expect the notion that APIs should be managed as products to become clear best practices in 2017. Expect to see APIs extending to include a full life cycle of prototyping, user feedback, road mapping, user communication, versioning, and retirement to be applied.
Bots and AI Will Drive API Adoption
The speed with which bots, artificial intelligence, and voice interactions have become a fixture in API usage was something we missed in our predictions for 2016 (though our team has been talking about it all year – props to @picsoung). In 2017, we expect this to accelerate even more. New conversational modes for interacting with APIs and data are a clear winner and we expect many organizations to begin experimenting with text and voice-based interfaces to compliment browser and app type experiences for their brands.
API Consumption Will Become Increasingly Important
APIs exist to be used and there is continually growing demand across many organizations to make API usage more accessible to non-developers. IFTTT and Zapier are excellent examples of this, but new tools such as Datafire.io and Apiant are emerging in a similar vein.
API Discovery and Marketplaces Will Advance
In tandem with the increased focus on API consumption, the challenge of how to discover APIs has become more important. New marketplaces such as RapidAPI and Hitch are springing up and technologies such as APIs.json should make progress. In particular, API discovery is now not only related to external public APIs but is often also a challenge for internal APIs. In particular, as microservices become more common, full automation is needed around API publication and discovery to match backend deployment.
Full Lifecycle API Management Will Become More Important
Strongly related to the first prediction, we expect API management and other tools to keep expanding and cover the full life-cycle of an API. Gartner’s Magic Quadrant for API Management is already known as the “Magic Quadrant for Full Life Cycle API Management.” (A download with is available here.) We expect 2017 to see increasing demands for solutions covering the whole spectrum. This will further mean standards such as OAI/OAS will become increasingly important since tools will need to interface more effectively.
Graph QL Will Continue to Grow Quickly
Given its adoption levels in 2016, graphQL seems to be a valuable API pattern and we’ll likely see more applications for search type applications. It seems likely to sit alongside Hypermedia and common REST-like patterns as a highly useful interface pattern.
OAI Open API Specification Will Gain Wider Adoption and Drive New Innovation Around Tools
This is a prediction we’re repeating from 2016 that saw definite progress from API standards in development and adoption, but tool growth did not break out quite as we had expected. With the imminent release of the OAI 3.0 specification (contribute here!) in the next two months, we expect tool growth to rapidly accelerate.
Microservices, DevOps, and APIs Will Become More Closely Acquainted
One of the most exciting things for 3scale in joining Red Hat has been working along side the Fuse, OpenShift teams and Ansible teams and witnessing the convergence many companies are experiencing on a common set of infrastructure techniques. Microservices patterns have genuinely taken hold and we expect microservices and APIs to become significantly more integrated, and deployments and management to become highly automated. Look out for API-powered microservices managed in a fully automated way to be a common enterprise pattern in 2017.
IoT Security Will Shine a Light on API Security
With the proliferation of IoT devices in our homes and business as well as clear indications that many are poorly secured, we expect 2017 to see a major focus on security for IoT devices and APIs in particular. At CES there have already been a number of specialist routers revealed. The problem goes far deeper, however – new design patterns and standards are likely needed to make genuine progress.
API Copyright Will Continue to Be a Critical Debate
2016 saw a temporary resolution in the long-running Google vs. Oracle copyright case. While the resolution found in Google’s favor, the result also suggests that API definitions are potentially copyrightable. See excellent coverage of the result here. It seems likely Oracle will appeal the verdict in 2017. However, either way, copyright is now an important issue for APIs, and API owners will need to start deciding their stance on copyright. Our take on how to proceed can be found in the slideshow, How to Survive the API Copyright Apocalypse.
Since there is so much going on, we left one of the clearest predictions for last.
Serverless Architectures Will Be Pushed Hard by All Major Cloud Providers
Amazon’s AWS Lambda services have been growing rapidly since launch in 2015. Most notable is the new ability to connect lambda functions to the AWS Marketplace via the Amazon API Gateway. This now makes it possible to provide simple, paid for functions as an API service to any developer – creating new potential channels to market and allowing Amazon to learn how popular different API service might be. Microsoft Azure Functions and Google Cloud Functions are both nascent frameworks with similar serverless aims. It seems highly likely serverless will become a must-have feature of the major clouds. Our prediction is that meta-services will begin to evolve, which provide equivalent code in each framework.
We hope these predictions provide some food for thought, though we can’t guarantee how close we’ll come to the mark. It’s been an exciting and interesting 2016, and we’re looking forward to what 2017 brings us.
Published at DZone with permission of Steven Willmott, DZone MVB. See the original article here.
Opinions expressed by DZone contributors are their own.