An important piece of wisdom I have to frequently remind myself of is that “perfect is the enemy of good”. What this means is that even something good (enough) will never get finished if one strives for prohibitively expensive perfection.
I mention this because on last week’s devlog I wrote in regards to the new take-off performance formulas:
Initially I thought this would be a trivial application of standard curve fitting algorithms, given that the curves an operator of my tool can “guess” by fiddling with a few parameters are already near-perfect. Turns out that wasn’t the case and I wasted a lot of hours on getting it to work before I had to admit to myself I won’t be able to do so within an acceptable timeframe. While my formulas are a good (enough) approximation, they likely ignore several input factors relevant to making them universally applicable, which makes it tricky/impossible to automatically optimise their inputs.
So for the moment, I’ll shelve my curve fitting ideas and focus on a manual approach that actually works well (enough) for the majority of cases.
In other news I finally completed the last remaining paperwork for our last Prosperous Universe funding project and I spent some time and money with a notary to trigger the legal move of simulogics, which unfortunately has to follow my private one. I also rolled out a substantial data patch, but as mentioned before, I can’t take credit for that…most of the work was @koenc86’s 