Low Code Development

My previous experience with a low-code framework left much to be desired. Building basic MVC applications was straightforward, but anything beyond that presented significant challenges. While theoretically possible to extend functionality with Java and Javascript code, the process was incredibly cumbersome.

Scalability proved to be a major issue. The visual representation of code as boxes, while initially appealing, became a hindrance as the application grew. Reading, understanding, and refactoring the increasingly complex web of interconnected boxes became a nightmare.

Vendor lock-in was another significant drawback. Exporting the application to standard code for migration to another platform was simply not possible. The only escape route was a complete rewrite. Even performance profiling required vendor intervention, as no third-party tools were compatible. This dependence became acutely problematic when performance issues arose, requiring on-site visits from the vendor to diagnose the problem with their proprietary profiler. Furthermore, the vendor's version control system, or lack thereof, made collaborative development difficult. Changes were frequently overwritten, and the available merging and diffing tools were inadequate.

Community support was virtually nonexistent. There was no readily available resource for troubleshooting and knowledge sharing. This isolation compounded the challenges faced during development.

Finally, finding developers willing to work with the low-code platform was a constant struggle. The reality is that low-code development still requires programming skills. Non-programmers lacked the expertise to maintain or enhance the application, especially given the complexity of the business logic. Those who did acquire the necessary skills often preferred to work on more traditional projects.