English and Unix are both strong arguments that "good enough" is the best way to communicate, that being more rigid and formal and, well, consistent, is actually counterproductive.

English isn't elegant, but it is very expressive, with plenty of scope for nuance.

But relatively rigid ("strongly typed"?) languages like Latin and ancient Greek were also very expressive, so I'm not sure that's a convincing argument.

As for Unix - it's the Java of operating systems.

I'd love to see the Haskell or the Lisp of operating systems, but we seem stuck around 1975 as far as OS research goes.

I agree, and I suspect a lot of the suggestions given in these comments is post-facto reasoning. At one time as I'm sure you know, while Middle English was quite well established, Latin was the "Lingua Franca" of the "civilized" world due to the institutional networks of the Church and the university. Plenty of flat and embarrassing Latin, which we still have, was written at this time (with a number of its classical features modified), that could provide much fodder for a similar article some 900 years ago. Were it to have gone on, and had English died out, leaving us with few examples other than Beowulf, the memoirs of elite regents, letters between statesmen and the like, we might have another impression altogether.

Instead, Imperial England left on the world a linguistic legacy to which we adhere out of convenience of circumstance, and probably will for the long foreseeable future: "learn once, enjoy commerce and education everywhere!"

Genera is the Lisp of operating systems. :-)

I'm not sure what the Haskell of operating systems would be like.

Genera might be the Common Lisp of operating systems (as well as the, or at least an, operating system of Common Lisp). Is there a Scheme of operating systems?

Genera was originally developed in ZetaLisp. Later Cmmon Lisp was added and many newer parts of the OS were developed in Common Lisp.

Yup. (Was that intended as a correction to something I wrote? If so, I wasn't clear enough and I apologize.)

Regardless, it may be worth adding that ZetaLisp was one of the most important "source" dialects from which Common Lisp emerged, and that Symbolics people were heavily involved in the process. Genera wasn't so far from being "an OS of Common Lisp" even before there was a Common Lisp.

The main difference: ZetaLisp was developed specifically as a language and its implementation for a certain type of Lisp Machines as an OS language.

Common Lisp OTOH is a domain independent Lisp standard with only very few specific OS-related features.

> As for Unix - it's the Java of operating systems.

What do you mean by this? IMHO Java's goal was ubiquitous mediocrity: a cross-platform VM for a language suitable for large teams of interchangeable workers. We had Lisp operating systems (and hardware) back in the day, and it didn't catch on.

I don't think Java actually started that way. It was a lab experiment gone rogue, starting as a way of shoehorning complex application code into the very primitive web browsers of the mid-late 1990s. That never really caught on, but it was fashionable enough for Sun to invest in developing JEE as a way to build sophisticated server-side apps. Back then, we were hacking apps together with CGI and server-side includes - so primitive, so insecure. IBM took to the field with Sun and started selling Websphere so management could feel better spending millions on licenses rather than using that crazy open source stuff, and it all rolled from there.

I like the Latin/strongly typed analogy.

Also in my opinion, word order is easier to deal with instead of tons of endings for parts of speech.

Right, like there's a certain "peak" at which the value of consistency does not justify the learning curve required to maintain that consistency.

No they both are so widely used that its easier to deal with them instead of trying to change the status quo.

The other languages look harder because a) you don't know how to acquire languages easily and b) you don't speak them yet.