A running list of sources of potential technical debt within ORCA and its environment.
The code is stored in Rational ClearCase, a revision control system from the early 1990s. It doesn't integrate with any modern DevOps tools, which functionally blocks any modernization of the deployment process.
The deploy process is run on a custom system. It works just fine, but, like the use of ClearCase, this is an obstacle to integrating with modern DevOps tools.
The repository has a fair amount of COBOL in it. This is said to have been replaced with Java-based functionality in ~2013, but it remains in the repo.
The repository has a fair number of batch scripts throughout, for the testing and deploy process. This is quite normal, but apparently these were replaced with Java-based functionality sometime in the past few years. Leaving them in the repo provides the illusion of functionality in the wrong place.
The front-end of ORCA was written in the mid-00s, targeting Internet Explorer, and it has the shortcomings that one would expect from that. Reliance on IE-specific functionality (e.g., window.showModalDialog) renders the site almost completely unusable in other browsers.
There is no API layer separating the front end from the back end. This makes it quite difficult to modularize the site.
The front-end was created in an era when it wasn’t standard to use semantic HTML, restricting design to CSS. Instead, layout is done in the HTML itself. Most block elements are divs. This makes it difficult to e.g. move to a responsive design.