These days I am working on some 20 year old code used in a few dozen manufacturing plants around the globe; the reason I asked to be allowed to fix it and the r

These days I am working on some 20 year old code used in a few dozen manufacturing plants around the globe; the reason I asked to be allowed to fix it and the reason my manager allowed me to do it is that we have performance issues in some of the biggest plants (by the number of production lines) and this code is part of the problem. If that would not be the case, that code would continue to run as is for another 10 years. Code modernization in some circumstances does not bring business value. In the plants we have some hardware that is decades old and it works as well as the one built last year, more modern software would bring no difference as physical limits (ex: how many bottles you can fill on a line) makes a line capable to run on a smartphone with MS-DOS and Turbo Pascal on it (we don't use that, of course). If it runs and you cannot improve it more than the cost of the improvement, leave it as is.