The statement in this meme is false. There are many programming languages which can be written by humans but which are intended primarily to be generated by other programs (such as compilers for higher-level languages).
The distinction can sometimes be missed even by people who are successfully writing code in these languages; this comment from Jeffrey Friedl (author of the book Mastering Regular Expressions) stuck with me:
I’ve written full-fledged applications in PostScript – it can be done – but it’s important to remember that PostScript has been designed for machine-generated scripts. A human does not normally code in PostScript directly, but rather, they write a program in another language that produces PostScript to do what they want. (I realized this after having written said applications :-)) —Jeffrey
(there is a lot of fascinating history in that thread on his blog...)
I've kinda noticed this block when working with non-developers attempting low-code and no-code platforms. Anecdotally, non-coders tend to assume that knowing how to code is the hard part of software development. It's really not though, there's tons of resources to learn any language you want for free, and cs students cover all of the basics in their first year. The actually hard part (well one of them) is knowing what to code: the data structures and algorithms. Pro_code, low-code, or no-code, there's just no way around not knowing how to design a working, efficient algorithm or a clean, scalable database schema. Ironically, for anything but the most trivial problems, the lack of maturity in low-code platforms tends to only make the algorithm harder to implement.
A lot of people really have difficulty with maths and programming.
The way i imagine it, programming is something non-real, something metaphysical, or how you want to call it. And a lot of people even plainly reject that such a thing meaningfully exists. Think about how many people reject the existence of "spirits", "demons", or "god", based on nothing else but the argument that it is not tangible. Something similar is going on with maths and programming.
I heard there was a programming language where you programmed a tree, that you could only manipulate manipulate in a "IDE" that looked a bit like Microsoft Word and saved the "source code" as a binary file.
I've recently had to help the wife with some VRChat "Udon" language.
I mean I get it, all the stuff is like the underlying shit in a parser I wrote years ago to speed up execution. And looking up the name for that, it's an abstract syntax tree.
It's just I don't know why you would try to write stuff in it directly. All the tutorials have this mass of on screen spaghetti for "if a=45 then b.visible=false".
It's like everyone gets this idea that coding is hard and a bunch of text, and then they spit it out on screen so no none of us can understand it at first glance.
Maybe I'm an old fogey, but I usually hear more pushback against visual languages as being too finicky to actually create anything with and I usually advocate for a blending of them, like working in Godot and having nodes to organize behaviour but written scripts to implement it.
I really appreciate the talks from Bret Victor, like Inventing on Principle (https://youtu.be/PUv66718DII), where he makes some great points about what sorts of things our tooling, in addition to the language, could do to offload some of the cognitive load while coding. I think it's a great direction to be thinking, where it's feasible anyways.
Also, one reason folks new to programming at least struggle with text code is that they don't have the patterns built up. When you're experienced and look at a block of code, you usually don't see each keyword, you see the concept. You see a list comprehension in Python and instantly go "Oh it's a filter", or you see a nested loop and go "Oh it's doing a row/column traversal of a 2d matrix". A newbie just sees symbols and keywords and pieces each one together individually.
like working in Godot and having nodes to organize behaviour but written scripts to implement it
That was the intent with Godot's (3.X) implementation of VS (Visual Scripting) but I think most people didn't like it (thus why it was gone in 4.X). The major flaw with that idea is that programmers probably don't want to work on VS and... is it really better than just components with exported script variables and either way well-documented code (especially with gdscript)? Also communication on desired effects.
VS should be easy for beginners, if it fails at that a huge amount of people who aren't in a team will find it to be useless. For comparison, UE's Blueprints are usually what people point as better than Godot's VS (which failed at discoverability due to lower-level workflow and IIRC wasn't fleshed out with organization either), so this wasn't strictly a problem with the idea of VS.
There are 3rd-party things now (Orchestrator, also Block Coding which generates gdscript) that might work better, though I don't know.
Yeah, I wasn't a fan of the visual scripting, but I do consider composing nodes in the editor, connecting signals, modifying field values with sliders, having global variables in a separate editor, visual curve editors, file managers, etc. to be a form of visual scripting by a different name, and I do quite like that.
I've been curious how this sort of editor would work for non-game code, like making a CLI in C, C++, Kotlin, etc. Where you primarily interact with nodes and inspectors for data organization and scripts for behaviour implementation. I need to go back to Smalltalk to see some of the ideas there for alternative code organization structures.
Like much of math, people are often eager to talk about the cool stuff and make it sound hard because they are proud they understood it. For a newcomer, this is just a brick in the face.
People who know math make math sound hard? Eehh, that does not seem to track. Math can become incredibly difficult, and even the cryptic terminologies are a way to alleviate the difficulty.
Literally everything we ever came up with is comprehensible by humans, and is likely to be comprehensible by a layman given enough time and making sure prerequisites are filled.
In fact, it takes a good explanation that would click with a given person's experience and level of expertise to make anyone understand anything.
It's just that sometimes people need that specific thing X, and normally it's needed to those who have some knowledge in another specific thing Y, and it gets expected that a person needing X knows Y (which is not necessarily true)
This is especially common in the world of computers. Everyone uses them, everyone has to troubleshoot them, but not everyone is the system administrator, to which 85% of the guides often seem to be addressed.
When I took over programming for my robotics team in highschool I switched from whatever visual flowchart bullshit they were using to robotc. I can't make heads or tails of programming without actual words that literally say what the program does.
It took me looking at unfamiliar programming languages and realizing that I could read most of them without really knowing them for me to realize I probably could learn to at least read another language.
It's been years since then and I'm still probably shit at Spanish, but just like programming languages regular languages were made by humans to communicate with other humans, you're capable of understanding any of them given a reasonable amount of time and guidance.
I wish I understood this point of view better. I crunch through information, so I want it to be densely packed. I'd love to know why and how this helps you so I can better help my peers that are like you?
It... Kind of was though, IR gives us a way to translate higher level concepts to lower (but not the lowest) level representation. It also gives us a way to optimize before machine translation.
For me, I think it's that most common-language things that I happen to look at are 500-line+ with non-obvious short names (initialisms? might be an issue with low-level). Some of it might be down to optimization or language features/requirements, or not using libraries. Though I also don't hate whitespace so it may just be my brain.
The other side of the coin is that interpreted languages (being more readable) are slower(+single-threaded) and have other limitations/issues. I have some hope that Python's update with JIT and no-GIL may change that, but integrating it into other tools is still an issue so I haven't looked into it.
The one language that has clicked for me is Nim-lang (compiles-to-C, interop). I haven't done enough real projects, but I like the syntactic sugar and UFCS. Not sure if that's the best way to say it, but it's like the options that exist can be used to make code more concise. Something that seems small like how you can write conditions or loops can make a big difference.