This is too facile. First, in terms of capability maturity, management is not the goal of a fully-realized line of industry. Instead, the end is optimization, a situation where everything is already repeatable, defined, and managed; in this situation, our goal is to increase, improve, and simplify our processes. In stark contrast, management happens prior to those goals; the goal of management is to predict, control, and normalize processes.

Second, management is the only portion of a business which is legible to the government. The purpose of management is to be taxable, accountable, and liable, not to handle the day-to-day labors of the business. The Iron Law insists that the business will divide all employees into the two camps of manager and non-manager based solely on whether they are employed in pursuit of this legibility.

Third, consider labor as prior to employment; after all, sometimes people do things of their own cognizance without any manager telling them what to do. So, everybody is actually a non-manager at first! It’s only in the presence of businesses that we have management, and only in the presence of capitalism that we have owners. Consider that management inherits the same issues of top-down command-and-control hierarchy as ownership or landlording.

I’m most familiar with the now-defunct Oregon University System in the USA. The topics I listed off are all covered under extras that aren’t included in a standard four-year degree; some of them are taught at an honors-only level and others are only available for graduate students. Every class in the core was either teaching a language, applying a language, or discrete maths; and the selections were industry-driven: C, Java, Python, and Haskell were all standard teaching languages, and I also recall courses in x86 assembly, C++, and Scheme.

The typical holder of a four-year degree from a decent university, whether it’s in “computer science”, “datalogy”, “data science”, or “informatics”, learns about 3-5 programming languages at an introductory level and knows about programs, algorithms, data structures, and software engineering. Degrees usually require a bit of discrete maths too: sets, graphs, groups, and basic number theory. They do not necessarily know about computability theory: models & limits of computation; information theory: thresholds, tolerances, entropy, compression, machine learning; foundations for graphics, parsing, cryptography, or other essentials for the modern desktop.

For a taste of the difference, consider English WP’s take on computability vs my recent rewrite of the esoteric-languages page, computable. Or compare WP’s page on Conway’s law to the nLab page which I wrote on Conway’s law; it’s kind of jaw-dropping that WP has the wrong quote for the law itself and gets the consequences wrong.

Yeah, writing your own squeeblerizer sucks, but there’s no better option. GNU Scrimble can be used off-the-shelf as a passthrough, so the only real tasks are implementing Squeeb’s algorithm and a sprongler; then, your entire pipeline is merely something like:

$ gscrimble --passthrough --args -- ./your_squeeb | ./your_sprongler

Edit: Whoops! Forgot to mention, GNU Scrimble also has Snorble support out-of-the-box, and Scrimble clients have content auto-negotiation, so your_squeeb can just take JSON on stdin. GNU Scrimble is really nice for this sort of thing, just…big.

And if you want to sprongle directly into a database or etc. then you can write your_sprongler to taste. Full disclosure: I have a fairly fast implementation of Squeeb’s algorithm in rpypkgs. However, I’d really recommend writing your own; it’s like twenty lines of code you can copy from Wikipedia and it’ll give you a good basis for extending it with your own desired changes later.

You can read snorblite’s code if you need to figure out a specific sprongling technique, but it’s way easier to just go look up the original SprongCode from SprongReg. Use a search engine to get around the university’s paywall. This gets you the SprongCode UUID and you don’t have to read code written by a batshit fascist.

I was going to bring this one up. The least realistic part of Antitrust is how the antagonist is defeated, but the parts where somebody is impatiently waiting for javac to finish so that they can pack their .class files into a JAR, or typing in a list of IPv4 addresses one-by-one to see which one works, were painfully plausible.

Extension modules are implemented in C because the interpreter is written in C. If it were written in another language, folks would write extension modules for that language instead. Also, it would be less relevant if people used portable C bindings like cffi, which are portable to PyPy and other interpreters… but they don’t.

By induction, that will only make your user base stupider.

They have two. If the complaint is that neither wiki is as rich as the Gentoo or Arch wiki, consider that perhaps NixOS users don’t need as much supplementary advice for configuring their systems.

It’s been two decades. What kinks do you think NixOS has yet to iron out?

Other way around, actually; C was one of several languages proposed to model UNIX without having to write assembly on every line, and has steadily increased in abstraction. Today, C is specified relative to a high-level abstract machine and doesn’t really resemble any modern processing units’ capabilities.

Incidentally, coming to understand this is precisely what the OP meme is about.

Nailed it. I think about this a lot: a sysadmin is basically a manager of dozens, hundreds, or even thousands of computers. But management is a poor way of orchestrating human labor; small teams usually operate better without management. So, is there a better way to administer computer systems as well?