Confessions of an Automationeer, Part 30: Unreal Problems

Confessions of an Automationeer, Part 30: Unreal Problems

CSR60 marked a milestone for the long-running competition, for it was the first one to use the UE4 version after 59 straight rounds in the original Kee version. It seemed like a good idea at the time, but... there were teething troubles, more on which later. In the meantime, here's the premise for this ground-breaking round.

Rather than take the safe route by making a round about affordable stuff like economy or family cars, the host requested the entrants to build and submit premium sporting convertibles from the years 2018 to 2020. Given that performance cars were my forte, entering this round was a no-brainer. I opted for a fairly simple approach - a big normally aspirated engine up front, driving the rear wheels via a six-speed manual gearbox, all wrapped up in a two-seat body. The resulting car, the LAP LE52, was submitted using the exact same method as used in Kee - by attaching its associated export file to a PM for the host. We assumed there would be nothing wrong with this feature, as had always been the case previously...

Unfortunately, the import function for UE4 Automation was bugged to the point of uselessness, so the host requested an alternate method: sending a zip file containing screenshots from the trim (design, overview, wheels/tires, brakes, safety and suspension) and engine (one showing an overview of the engine variant, and one showing its dyno sheet) tabs, with exactly one additional screenshot from the photo mode included. I duly obliged, but was miffed about being used as a guinea pig for a new, unproven take on such an established competition. In fact, this round is not likely to be remembered as fondly as it should have been for that glaring issue alone. And no matter what the results are, I won't be as satisfied as I would usually be after wrapping up a round of CSR.

One major consequence of this botched round is that until the UE4 version has been properly sorted, CSR is highly likely to revert to Kee for the time being, starting with the next round. I have mixed feelings about this. While the previous round used the Kee engine, its obtuse scoring and pricing systems put many users off, me included. So returning to a more conventional scoring method made perfect sense. On the other hand, I was giddy with anticipation after finding out that this round would be a leap of faith into the unknown world that was UE4 Automation, and only now, just before the deadline, have I realized that it is still very much half-baked in comparison to Kee; thankfully, the host has also realized his mistake, and is unlikely to repeat it again. Here's hoping that CSR61 takes place as normal, in the Kee engine, without anything as farcical as what happened this round, or the previous one for that matter.