I've been getting sucked hard into the Lua universe in recent weeks. Some links. Introductory magazine article: https://cacm.acm.org/magazines/2018/11/232214-a-look-at-the-design-of-lua/fulltext Very cool technical paper on how they manage to implement an efficient interpreter and register-based VM in 12k LoC: https://www.lua.org/doc/jucs05.pdf A very interesting "federated monorepo" approach to a batteries-included Lua distribution: https://luapower.com. In particular check out https://luapower.com/philosophy. I particularly feel like the author and I were separated at birth for this passage:

NO [SUB-]DIRECTORIES

This may be a hard sell but I stand by it. [Sub-]directories are evil. Not so much because of semantics, but because of the tools we use suck at working with them. [No global overview;] instead you have to navigate them.

---

While you're digging into Lua I suggest you check out other implementations that are open source

Oh, I totally missed that PICO-8 is programmed in a subset of Lua!

Then you no doubt have read "The Evolution of Lua" https://dl.acm.org/doi/10.1145/1238844.1238846. Favorite quotes... Most successful languages are raised rather than designed. ... We include a new feature in Lua only when all three of us agree. Lua provides mechanism not policy. ... Instead of littering the language with lots of features, we provided ways for users to program the features themselves, in the way they wanted them, and only for those features they needed. Since its first version Lua has had real users, that is, users others than ourselves, who care not about the language itself but only about how to use it productively. Lua is best described as a closed-development, open-source project. Lua is not developed in a collaborative way. We accept user suggestions, but never their code verbatim. We always try to do our own implementation.

The "design of Lua" article is very nice. I'd love to see more descriptions of languages (and other infrastructure software) at that level. Also, the specific form of minimalism adopted by Lua looks very interesting. More pragmatic than e.g. Scheme. What do people use Lua for in real life nowadays? Given the existence of multiple implementations, there are probably different important use cases. @Scott Anderson mentioned games. Are there other big application domains?

It's used as configuration for a variety of different software, which is kind of it's original intended use case

Embedded use cases make a lot of sense, also because there is little competition in that space, with Tcl mostly dead and Python having moved on elsewhere. But what about the heavier machinery I see, such as LuaJIT or Terra? Terra in particular looks like a replacement for C, so the code structure should be Lua outside and Terra inside, rather than Lua inside something else.