I understand whis is pretty much WIP, but still, it's too unorganized to be anything useful. I thought the most interesting to be features page[1], which is nea

I understand whis is pretty much WIP, but still, it's too unorganized to be anything useful. I thought the most interesting to be features page[1], which is nearly empty, and this effort in taxonomy is rather too complicated to be crowd-sourced without supervision. For example, let's take a look at traits[2] and mixins[3]. There are a couple of issues here. First off, why it's 2 different pages? There's no real difference between a trait in PHP, and mixin in… well, no languages except for Racket actually have a syntactic construct called "mixin", but I guess modules in Ruby or Julia are close enough. Scala also has something that's called "traits", and it's also basically the same thing, but with caveats. On the other hand, D has both "mixins" and "traits", but these are completely different features, and these "traits" have nothing to do with traits in Scala or PHP. So if somebody were to make a comprehensive list of features of D in this DB, should these "traits" appear on the same page as PHP and Scala traits (which are mixins)? Furthermore, unlike PHP, Scala, Ruby or Julia — Python's "mixins" aren't just mixins with a different name. It's not even clear if it has mixins at all. There's something people call a "mixin" in Python, but these are just classes, so you cannot really say "yes". However, Python has multiple inheritance, which makes "mixins" borderline pointless: classes are (or can be used as) mixins, if you have multiple inheritance! Templates in some languages can be used this way as well. Which brings us to the next issue — it's not clear, if a language should be marked as having a feature if it comes built-in, explicitly, or if a feature can be implemented in it. Does every language have a semaphore? I cannot remember any where it couldn't be implemented (that would be weird), but I cannot remember any where it's an explicit feature construct either (well, arguably, maybe some SQL-extensions?). All this isn't to say that the current list is bad. All the questions above can be answered in any way, and it's up to a "researcher" which definition to use in order to actually get a useful taxonomy. It's a non-trivial job. [1] - https://pldb.com/lists/features.html https://pldb.com/lists/features.html [2] - https://pldb.com/languages/traits-feature.html https://pldb.com/languages/traits-feature.html [3] - https://pldb.com/languages/mixin-feature.html https://pldb.com/languages/mixin-feature.html [4] - https://pldb.com/languages/semaphores-feature.html https://pldb.com/languages/semaphores-feature.html