Originally Posted by
jerseytom
Speaking as a software engineer, there's an expression "Never change a running system."
As a consumer we're all used to getting the newest / latest / greatest. Want the latest iPhone or whatever? Easy. Turn your old one off, turn your new one on, and off you go. Only affects you and there's really no interruption of anything.
Now consider a system that's the IT backbone responsible for the transport of over 500,000 people every day. Planes in the air 24/7/365, no let up.
If you wanted to wholesale gut and replace that - when do you do it? There's no downtime. Not to mention the tremendous risk of error or failure. And for what? A snazzier UI for a GA? Does that add value to the customer's end experience? No - there's no tangible benefit for them, only the risk to piss people off by having some IT bugs and issues of a new system behind the scenes.
For that reason, a lot of backbone mainframe systems have long lifetimes. If it's running and working, don't change it. Early in my career I was very much the guy that wanted to burn old stuff to the ground and make everything new and better... but after a while you become more risk averse and cognizant of the risk / reward trade off for making big changes.
Agree. And, many of these older terminal like systems have an advantage of being extremely fast and stable. The downside is that they are not as pretty, which usually equals harder to train new people on them, but for those that know them, they work great.
I use applications in my job that have an older terminal interface available for those that know how to get to it. There's also a nice new shiny interface that 99% of our people use. I'm the 1% - and I can do things in seconds that take others minutes because of this.