Over the course of a series of tutorials, we explored the booming ecosystem of low-code tools. We built mobile, desktop, and admin applications. We also built workflow automation to help manage the flow of data.
- Mobile app development with Draftbit.
- Web-app development with Bubble.
- Internal tooling with Retool.
In surveying the current ecosystem, one that is estimated to reach $13 billion by the end of 2021, there are some key observations to make.
As with any tool that increases the speed of change, more thought is required for getting the intended results. Low-code is no different. If you are able to add a table to an interface, attach some buttons, and create reactive listeners within a small handful of clicks, knowing what you intend them to do becomes the majority of effort applied.
Well-documented application flows lead to well-communicated production timelines, lead to well-defined outcomes.
While the widgets abstract the implementation details of the code written, thinking like a programmer (logic handlers, algorithm implementation, reactivity) is still very much required. Low-code lends itself very well to long-term maintenance by someone with technical acumen and logical thinking, without needing to learn the specific syntax of the programming language du jour.
Of the observations that stood out, one that stood out the most was just how well these tools play with something like Hasura. When coupled with a hosted tool like Hasura Cloud, you have a closed circuit that’s able to iterate quickly on both the data and presentation layer – delivering more "app interfaces" for teams needing to act on and respond to your businesses changing data environment.
The permissions layer allows you to determine which tools can access which data and be very specific about your data security.
Being able to leverage the dominance of REST with the descriptive nature of GraphQL lets you both fine-tune exactly what data any specific client will be requesting, as well as provide flawless interoperability without the need for a proxy wrapper.
Using the monitoring tab it was simple to see if the queries were malformed, the database was under too high of load, or whatever else may be the cause of an API-side error. Sometimes the low-code tools would emit opaque messages about failing to load data, and the Hasura Cloud functionality to see what was failing was helpful.
Many of the low-code tools are not optimized for query performance, specifically around call frequency. Being able to sprinkle cache on as needed allowed for quick query performance, but more importantly, helped keep these tools from DDoSing our server.
The low-code space is one worth watching if not outright embracing. With significant offerings from major name brands in the enterprise system, down to the classic “prosumer” product offerings, virtually any company could benefit from the efficiency and improved time-to-value that low-code offers.