The source content for blog.juliobiason.me
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

6.0 KiB

+++ title = "Realm of Racket: Lean to Program, One Game at a Time!" date = 2020-12-18

[taxonomies] tags = ["books", "reviews", "it", "racket", "books:2020", "2 stars"] +++

GoodReads Summary: Racket is a descendant of Lisp, a programming language renowned for its elegance, power, and challenging learning curve. But while Racket retains the functional goodness of Lisp, it was designed with beginning programmers in mind. Realm of Racket is your introduction to the Racket language.

{{ stars(stars=2) }}

This is a weird book. Not because its content, or the topic at hand, but the path chosen to explain things. For example, even after reading the whole book, I don't have any idea why I should pick Racket over any other Lisp variant.

So, laundry list of things that put me off:

  • No real explanation on why I should pick Racket over any other Lisp (or any other language, for that matter); they don't say it's faster, or more accurate, or it's syntax makes you write less bugs or the compiler will prevent you from doing it so.

  • There is a strong favouritism for Racket. It isn't that unexpected, but they authors start by bashing other languages, which gives a feeling that they just want to make their favourite language look better (this goes away only in the last chapter, when comparing Racket to Java).

  • They use different words for common things. There isn't and IDE, it is a PDE, "program development environment"; the integrated development environment becomes "interactive development environment"; functions don't "receive" parameters, they "consume" the values, which make it sound more like a Rust thing, in which the function gets the ownership of the value and, on its end, the memory is freed, which doesn't happen in Racket.

  • There is a bunch of "trivial", "simple", and such. This is bad, specially if you want newbies to read the book; if they don't grasp the concept, saying it is "trivial" will make them feel bad about themselves.

  • A lot of explanations are basically "(+ a 1)" -> "adds 1 to a". Sure, one may not understand all the facets of the code, but when you explain "why" you wrote code that way instead of "what" the code is, it is a lot easier to understand.

  • Because the book is too focused on explaining one single library, some things feel like "if everything you have is a hammer". For example, there is one game of attacking monsters and which the authors put the target in the state, and it doesn't feel like this should be part of the state of the game. Sure, the selected target affects the state, but it is not part of the state, at least for me, and there isn't any good explanation on why it should be part of it.

  • Another example uses a complex number to represent points in a 2D plane. Sure, 2D points and complex numbers are composed by two elements, but that doesn't mean that 2D points are complex numbers. Types matter, even in dynamic languages.

So... yeah, I think I learn a bit of Racket, but I still don't know why I should pick it over anything else. Is it just the game building library? Should one pick a language exclusively because of some built in library? Doesn't anything else affects this decision? (All questions are rhetorical, it is obvious that you shouldn't pick a language just because one library).


Highlights:

(foo)*g++.baz(!&qux::zip->ding());

"Let me write a very contrived example, just to make my favourite language look better compared to this one."

(sqrt (+ (sqr 3) (sqr 4)))

This is nowhere near the example in C++. If you want to compare languages, at least write the same thing.

If you’re on a *nix box, you are already a hero and don’t need instructions.

What a load of bull. It sounds more like you don't have an idea on how to do it in Linux.

(the compiler) DrRacket reads this expression, evaluates it, prints the result, and then displays its prompt again.

You just described an interpreter, not a compiler. For all purposes and intents, sure, it compiles the code to make sure it is not invalid, but the concept of "compiler" is used mostly for things that take a source and produce a stand-alone executable. Otherwise, even Python have a compiler.

the function uses a set! expression to change the value of a variable.

Why set!? Why not just set? What the ! means? Nothing? Is it just a convention?

Racket has three kinds of comments

And absolutely no example of those kinds.

(/ 4 6)
2/3```

This is one cool thing: Racket keeps the values in their fractional format.

(student-name freshman1)
'Joe
(student-id# freshman1)
1234```

Creating structs and extracting values from them.

  (cond [(= x 7) 5]
        [(odd? x) 'odd-number]
        [else 'even-number])```

Multiple tests. Also, there is this use of brackets ("[") and parenthesis ("(") which is never properly explained, though.

(when (and file-modified (ask-user-about-saving)) (save-file))

Because and and or return the value if it's not the boolean false (#f), you can use tests to simplify the code.

(define filename "my-first-program.rkt")
(unless (ask-user-whether-to-keep-file filename) (delete-file filename))

unless is a form of if that performs an action only if the value is #f (false).

(struct orc monster (club))
(struct hydra monster ())
(struct slime monster (sliminess))
(struct brigand monster ())

Examples of deriving structs from other structs.

In summary, mutators make programming complex

Just a few paragraphs better, you used a mutator saying that this would make the code better.

raco exe -l hello-world.rkt

Generating a standalone executable from Racket source code.

(define snake%
  (class object%
    (super-new)
    (init-field dir head tail)
    (define/public (slither)
      (set! tail (cons head (all-but-last tail)))
      (set! head (next-head)))

Classes in Racket.