Homework #14 - 10/3

      The article we were to read for today was all about legacy systems and how to go about transferring from their use to a new system. Some of the reasons for transferring include being able to take advantage of new technologies that the legacy system cannot incorporate, high costs of maintaining the legacy system, and investing in the business. This whole article has much familiarity in my life. The reason being, one of my jobs is being a pizza driver, and the system we use to take orders is incredibly old. I'm not talking windows 95 old, I'm talking green-screen-monitor-and-tower-combined-with-only-keyboard-input old. This consistently blows me away, why are we stuck in the 80s? Don't ask me how they replace these dinosaurs, because they have broken before and we end up getting "new" ones, which are the exact same model. It's also strange, because this isn't some local small business either, this is an international chain corporation! That being said, the store I work for is franchise-owned, and so it may be possible that only the franchise has yet to update its technology, and the corporate stores already have. Either way, the franchise isn't that small either, so you'd think they could benefit from upgrading.
     I've often thought about how I would program a new system to take orders for our store, and what kinds of things I would do differently and/or inherit from the old system. Upgrading the legacy system for me, seems like an easy enough task, but then when you read this article and actually imagine everything involved with the legacy system and all of its different components, it quickly becomes a daunting task and I can now understand why the pizza joint has yet to upgrade. I was trying to imagine how I would go about doing it if somehow they ended up hiring me to do the job, and I would definitely prefer to use the butterfly method discussed in the article rather than the chicken little method. All the extra complexity of writing components to make the new and legacy systems able to interact just doesn't seem worth it. I wonder, though, if there are other methods not discussed in this article as it was published in 1999. They speak toward the end of the paper about the research field of migrating legacy systems and I find myself curious if there has been any and what the results of it were.