22 MARCH 2023 WorldWide Drilling Resource® The Downside of Updates by Britt Storkson Owner, P2FlowLLC Not all software updates are created better than the old stuff. Sometimes new software updates can actually degrade the product, not improve it. There is no guarantee “new and improved” really is “new and improved.” At my “day job” at a retail store, we use a portable scanning device to track stock status by providing product pricing and location information. We’ve been using these devices for something like 15 years now and I’m surprised at how little in the way of improved performance the numerous updates have provided over that time period. The device still locks up and needs to be rebooted on a regular basis, resulting in considerable wasted time. The newer version is smaller, more compact, and weighs less than the old one, but it still exhibits some of the same old problems that -apparently - still haven’t been solved in 15 years. Enter the Tesla electric car. Nora Naughton at Insider.com reported Tesla’s “unexpected braking” problem caused a nine-car pileup in California. “The crash occurred on Thanksgiving, hours after Tesla CEO Elon Musk announced on Twitter the car company's “Full Self-Driving Beta” would be available to all Tesla drivers in North America who have purchased the option.” Problems with self-driving software are not new or isolated incidents. The article goes on to say the National Traffic Safety Board is investigating 758 reports from Tesla owners for what’s been dubbed “phantom braking.” That’s where the car stops for some unknown (to the driver) reason. Most of us for one reason or another have had vehicles we were driving simply stop running - not a pleasant experience, but usually not fatal either. The Tesla vehicles, it has been reported, actively applied the brakes which greatly increases the chances of having a collision with something. The article goes on to say: “The new software appears to still have some bugs to work through.” Well, duh! That’s kind of like saying the surgery was successful, but the patient died. If the surgery was really successful, the patient would not have died if the objective of the surgery was to save the life of the patient. No matter how great the programmer is, the program must still be tested - and fully tested, not just the part the programmer thinks he or she “fixed.” That’s because by “fixing” one area of the program it is possible to alter other areas of the programming and not even be aware of the changes. Only diligent testing will reveal whether or not the program and computer are functioning properly. And this includes testing under ALL conditions, not just in the lab or test facility. The corporate bean counters don’t like to spend money on testing because they see no immediate return on investment, but it must be done if you want a reliable product. Britt Britt Storkson may be contacted via e-mail to michele@worldwidedrillingresource.com
RkJQdWJsaXNoZXIy NDk4Mzk=