#modernise — Public Fediverse posts

Many large organisations are in a #crisis to #modernise their old, rotting codebases of OG #programming languages, FORTRAN and COBOL being the most prominent. IT has run out of its reserve of old programmers.

In general, #FORTRAN programmers were scientists and engineers, who were experts in their particular STEM field and they usually picked up other languages with relative ease. So, the transition from FORTRAN to C, MATLAB, Java, Python, etc., was comparatively smoother. And FORTRAN programmes remained isolated in highly specialised STEM fields, so the total codebase was far smaller. Moreover, there remains relatively fewer number of old FORTRAN systems still in use, because the transition began in earnest in the early 1980s, when C became the Darling of IT and when FORTRAN-obsessed supercomputers began to be replaced with modern hardware.

But #COBOL presents a far greater, and more urgent, problem. There are many billions of lines of COBOL code still in heavy use, especially in insurance, financial, logistics, transportation, utilities, and other institutions that form the fabric of society. The modern world runs on old COBOL code, as it were. And unlike FORTRAN programmers, COBOL coders lacked domain knowledge, because they were IT specialists. Also, the original bad design of COBOL and its slow adoption of modern design practices induced programmers to do shoddy, cut-and-paste job. Documentation and institutional knowledge about old COBOL systems have been lost to time, yet new ones still keep popping up in the business sector, due to inertia. The awareness of this impending disaster began about 20 years ago, but there is, as yet, no viable strategy to displace COBOL. Maybe it's not killable.