We're back from our Bunny Day break (to stay in Animal Crossing New Horizons lingo) and talked about (perceived) dead technologies.
We recorded this episode already about a week ago and it was due to me (Kai) stressing around giving a virtual conference talk and then getting on with life in general that it just is being published now, sorry for that, peeps!
Why did we decide to talk about what people perceived as dead tech? The topic got kind of triggered by the recent demand in COBOL developers due to a wide range of changes in countries' social security or tax systems all over the world. It turns out that many of these environments run on mainframe systems that were originally built in the 1960s-1980s.
After a brief look at the language we talk about the variety of risk environments organisations operate in. Consumer products get iterated over much more rapidly than bank or government systems dealing with fundamental societal infrastructure and that's part of the reason why we still find a lot of COBOL-based mainframe applications in these kind of organisations.
But there are many other technologies that are perceived dead. One of them is ColdFusion - a commercial web application back end platform (nowadays owned by Adobe), which is a very unusual business model for web app back end technology in 2020. Kai has, among other technologies, been using CFML (the language behind ColdFusion) since the late 1990s and Miguel talks about his personal recent experiences (the good AND the bad) with CFML. He was using Lucee though, an LGPL-based and open-source CFML spec implementation. CFML is certainly a niche language and has still a place for product development in certain environments.
From there we move on to talk about a bunch of other things: Prolog, Visual Basic and also about Java. Will Java ever be perceived dead? In Android-circles it certainly already is and nearly everyone has moved on to using Kotlin. But part of the problem there is that Android's supported Java version is Java 8, which actually lacks of lot of useful and expected features these days.
What defines a technology as dead and can it ever truly die? Certainly technologies come and go with the Gartner-quadrant-what-ever-hype cycle, but eventually a technology will probably become used less and less and end up in a long tail. There's also the question of available producers and consumers of libraries and the overall developer ecosystem. The question then is - what's economically and technologically better: stay with a chosen platform or at some point rebuild from scratch? Or maybe a middle ground is the way to go?
Many things to consider and we hope you enjoy us talking about some of these considerations.