Slight tangent on this and something I've always wondered: if something as basic yet extremely crucial as scheduling software is outdated (SWA might very well be on a mainframe or transmitting CSV files in an FTP server and that server could've acted up), how on earth are they expected to scale and adopt new tech? It seems like a ticking time bomb and gross mismanagement.
Fortunately I'm not flying this week, but I've already seen and experienced numerous meltdowns from SWA in the past. If this doesn't signal a wake-up call to invest in tech infrastructure for any company I don't know what else will.
There are a multitude of different ways to bridge from mainframe (IBM Websphere MQ/Host integration server) systems or to make file based systems look mostly online (small, frequent batch files). So technically it is possible and done pretty reliablly for banks and the like.
Extending use cases to cover behaviours that were not covered by the original software is where things get exciting. You will end up with a hodge podge of systems & processes that try to implement multi-step work flows and deal with failures through manual or equally complex automated methods to compensate.
Generally most industries are pretty terrible at evolving the solutions they have or completely replacing them. So you will see a multi-generational set of systems that are in place to provide an acceptably current set of behaviours.
This situation can arise due to varying commercial pressures, however Developer and Engineer's inability to communicate what the system currently does and what is easy and hard to change in no small way contributes to following paths that result in poor system outcomes over the lifetimeof systems.
Fortunately I'm not flying this week, but I've already seen and experienced numerous meltdowns from SWA in the past. If this doesn't signal a wake-up call to invest in tech infrastructure for any company I don't know what else will.