Richard Bucker

Lang for the Next 40

Posted at — Sep 11, 2020

I’m half way through “Rust: A Language for the Next 40 Years - Carol Nichols” on youtube and losing my patience. Up until this point we have learned plenty about Carol and Rust but nothing about why rust is in the critical path for the next 40 years.

The problem with rust is that it’s [a] complicated [b] huge.

By complicated I mean that your “work” is no less complicated than having been written in ‘C’. Whether it’s pointers, datatypes, type safety, memory safety and the like. You still have to test errors and while there is no concept of NULL you still have to test Option.

think “least surprise” !

Rust also took a page from the node project adding the npm tools called cargo. So the tooling is complicated too. There is the open source packages which has a certain amount of risk. And so on.

I’m looking for a langauage that allows me to describe “work” in simple terms. And then shim the work to the libraries as needed or implement the missing libs. I call this the pandemic model because I want simple tools that can be built with rudimentary tools.

One thing that has been distracting me is that verbosity and debug print statements all require condition checks at runtime. At least C/C++ and Rust have macros. But then this creates a new set of problems.

PS scratch would have been a great programming language name but it was taken my MIT even though that was more of scratching an itch than “from scratch”.