JÃ¼rgen FÃ¤lchle - stock.adobe.c
Software developers in 2021 can look forward to increased support for low-code/no-code platforms, simplified Kubernetes development and the beginning of artificial intelligence systems building software for themselves.
No outlook on software development would be complete without a nod to the emergence of low-code/no-code application development platforms that enable developers, particularly folks with little to no programming experience, to easily build applications using a variety of templates or drag-and-drop interfaces.
"The rise of these low-code/no-code platforms will continue, and while they are never going to threaten the developer community, they will muddy the waters," said Roy Illsley, an analyst at London-based Omdia. "Because of the speed and simplicity of these platforms, business leaders can become lulled into a false sense of simplicity," he said.
As a former developer, Illsley said he understands that low-code/no-code systems are attractive for the time they can save and the productivity they can provide. But the complexity of code and the challenges of getting it bug-free will require "autonomous assurance," which is essentially automating the testing phase of the CI/CD pipeline, he said.
Yet low-code development will become part of the software strategy of the majority of enterprises -- up to 75% of them, according to recent research from Forrester Research, said John Bratincevic, an analyst at the firm.
"For developers -- the coders -- this means they may be asked to learn a low-code platform or will have to work with low-code developers to do extensions and customizations," he said. Meanwhile, citizen development will continue to increase and that may start to affect developers.
"For developers, I expect this means that at least some of their work will be offloaded to businesspeople and some of these developers will be tapped to be mentors, evangelists and coaches for citizen developers," Bratincevic said. "This will be a gradual thing, of course."
"While it will not be as quick as low/no-code it will be functionally richer and a better user experience and delivered faster than traditional methods," he said. "The other key [app dev] accelerator is the move to microservices and the use of Kubernetes and serverless; the result of these technologies will be to make the code generation simpler but the management, orchestration and correlation of these microservices more complex. That is where Kubernetes, service mesh and API gateways will come in."
Moreover, serverless is going to gain more traction with the developers as it minimizes the DevOps overhead, said Krishnan Subramanian, an analyst at Rishidot Research in Redmond, Wash. Not just function as a service but also serverless containers.
"Google Cloud Run is already a favorite of developers and we can expect other cloud providers to offer an equivalent abstraction," he said.
Working at the edge
Another key area of focus for developers will be to develop apps for edge computing and IoT environments.
"As edge computing and IoT gain traction, developers will be moving from developing applications at the cloud toward the edge with real-time applications," Subramanian said. "It will be interesting to watch for any new development patterns to emerge as more applications are created at the edge."
Then there is the trend of hyperscalers moving public cloud services closer to and within customers' on-premises data centers. This will prompt a growing need during 2021 for developers to easily navigate the subtleties of notions like serverless computing and emerging services such as Google Anthos, said Bradley Shimmin, an analyst at Omdia.
Edge development will evolve around use case-specific scenarios. While the platforms and infrastructure layers share common components and can be used to run the applications, the apps themselves will be very specific to the use case and will probably evolve in silos based on languages and tools that meet the developers' specific needs.
"For example, I would expect the retail sector to be very interested in video, the coding of applications for video processing and analysis will be very different to another use case such as customer experience in stadiums, which will be location and GPS," Illsley said.
Kubernetes is king
Meanwhile, the second generation of Kubernetes development offerings will make container-based projects easier than ever to build, said Holger Mueller, an analyst at Constellation Research in Cupertino, Calif.
"All hail the king, Kubernetes," Shimmin said. "With Kubernetes now the preferred means of orchestrating containers of any sort, Omdia recommends that developers invest in this technology as it will serve as a driving force behind many important technology trends such as the rise of operationalized software development within areas like data for DataOps and data science for MLOps."
For its part, Docker is on the way out, he said. But Omdia anticipates that this change is both positive and presages some important work on a standardized container interface like the Linux Foundation's Open Container Initiative, Shimmin noted.
Software that builds itself
Meanwhile, in a prediction that may not be fully realized until three to five years from now, in some cases software tools will learn to automatically build more software, said Mike Gualtieri, an analyst at Forrester. Despite the technology not being fully mature yet, the foundation for this capability is being laid today in 2021 via AI tools, he said.
AI technologies have enabled a category of software development tools that Forrester calls TuringBots, in honor of famed computer scientist Alan Turing. These bots will automatically generate code to create applications based on the design artifacts, target platform, and architectural qualities that software teams define.
"Together, software teams and TuringBots will speed development; in some cases, the bots will replace developers to build, change, test, debug and refactor applications orders of magnitude faster and with fewer resources," Gualtieri said. "This will be a big leap forward in how AI will ultimately learn to code, enabling digital businesses to respond faster to customers, competitors and market changes."
Jason BloombergAnalyst, Intellyx
Monitoring and observability
Another developer technology that will become more available this year is observability systems, which provide DevOps teams with greater insight into application performance, efficiencies, governance and security.
"Monitoring and observability technologies such as application performance management have taken a backseat in the cloud's evolution in recent years to more important breakthroughs such as Kubernetes containerization, cluster management, and low-code platforms," said Charlotte Dunlap, an analyst at GlobalData in Santa Cruz, Calif.
However, DevOps initiatives recognize that monitoring, tracing and visibility are highly relevant parts of the cloud's value chain, she said.
Work from home persists
Even if the arrival of vaccines signals an end to the COVID-19 pandemic, development organizations will continue moving forward to a new context for working from home (WFH). This new environment will affect developers' day-to-day work and require additional software solutions to support the new hybrid WFH/in-office model across an organization, said Jason Bloomberg, an analyst at Intellyx in Suffolk, Va.
"Developers need to be ready for 'work from home: the sequel,'" he said. "In 2020, they'd already navigated WFH for themselves and their teams, as well as supporting their organizations as they move to WFH across the enterprise."
These new work from home environments will require "remote-first" software development tools and approaches.