For whatever it's worth, I didn't share your expectation. The real lorem ipsum is just a fixed load of text, after all - describing something as "lorem ipsum but for code" shouldn't necessarily imply any functionality.
Given that lorem ipsum is a fixed load of text, as you said, I would have expected a lorem ipsum for code to be a fixed load of code (that's nonsensical but looks real enough from afar).
I am a Russian speaker, and this ambiguous aspect of English has always bugged me to hell! I think english was a bad model as a starting point for programming languages.
I'm guessing they're referring to linguistic features that English doesn't have, such as inflected nouns to unambiguously indicate the subject, direct object and indirect objects.
`bite(dog, man)` is ambiguous. You need to look up the function documentation, make a guess, or ask the library developer to come up with a less ambiguous name.
If you could decline your variables, however, you could write (in terrible fake Latin)
`bite(dog, man-um)` or `bite(dog-um, man)` and they would be interpreted as two different -- and unambiguous -- things.
It might just be me, but isn't that just how you name your attributes within the function definition? I don't see how that would help disambiguate without have to look at the function definition at the very least.
The point is it's a new way of looking at arguments (in this toy example). In Latin, the order of the nouns typically does not matter. Their role in the statement is unambiguously determined by their declension.
So if you wrote a library with the verb "bite", you wouldn't specify that the first bites the second, or the first was bitten by the second. Those are ambiguous, because they're you're choice as a developer. Rather, you'd specify that the noun in the subject case bites the noun in the object case. That's unambigious in all declined languages.
So, as a reader of someone else's code, I can see immediately that `bite(boy-um, dog)` -- in my fake Latin where "um" denotes the object (cf. "Puer - Puerum") -- that the boy was bitten by the dog. I don't need to look up any documentation or look at how the variables are named in a third-party library.
Again, of course this case is a toy example. The article is more in-depth.
34 comments
[ 2.9 ms ] story [ 64.0 ms ] threadhttps://en.m.wikipedia.org/wiki/Metasyntactic_variable
Hackertyper is absolutely using real code.
Specifically, it's using groups.c from the Linux kernel:
https://github.com/torvalds/linux/blob/master/kernel/groups....
Compare that to the code used by Hackertyper:
http://hackertyper.com/code.txt
I am a Russian speaker, and this ambiguous aspect of English has always bugged me to hell! I think english was a bad model as a starting point for programming languages.
`bite(dog, man)` is ambiguous. You need to look up the function documentation, make a guess, or ask the library developer to come up with a less ambiguous name.
If you could decline your variables, however, you could write (in terrible fake Latin)
`bite(dog, man-um)` or `bite(dog-um, man)` and they would be interpreted as two different -- and unambiguous -- things.
...but I'm just repeating GP's article.
So if you wrote a library with the verb "bite", you wouldn't specify that the first bites the second, or the first was bitten by the second. Those are ambiguous, because they're you're choice as a developer. Rather, you'd specify that the noun in the subject case bites the noun in the object case. That's unambigious in all declined languages.
So, as a reader of someone else's code, I can see immediately that `bite(boy-um, dog)` -- in my fake Latin where "um" denotes the object (cf. "Puer - Puerum") -- that the boy was bitten by the dog. I don't need to look up any documentation or look at how the variables are named in a third-party library.
Again, of course this case is a toy example. The article is more in-depth.
I also wonder how large such an obfuscated code fragment would need to be to be able to find back the original code on GitHub/GitLab.
What I was hoping for is a "Lorem Ipsum"-style generator for code for like 10 different languages. So you'd have a Lorem Ipsum for JS, C++, Rust, etc.
This is just find/replacing identifiers with lorem ipsum, which doesn't really mask the underlying code that well.
"SyntaxError: unknown: Unexpected token (1:20)"
on valid code