Over time I've come to conclude we are not optimizing for language features, we are optimizing for the community around a programming language or stack. Sure,

Over time I've come to conclude we are not optimizing for language features, we are optimizing for the community around a programming language or stack. Sure, features are important and if critical ones are missing it might be a show stopper. So there is an initial thresshold that all candidates must pass. But problem solving is not a one-off exercise. It tends to be both dynamic (=facing unpredictable challenges) and recurring over long time horizons. Which means having a healthy, engaged, resourced community that will invest in adapting / solving future requirements is essential. So the "optimization" problem includes quite a bit more than the presently known developer team, its software stack its hardware and current problem definition / user requirement. I think you see this dynamic in several cases (including python) where you might not think that it makes rational sense.