In strongly typed languages, there are serialization/deserialization libraries that don't require you to type the entire JSON object - just the fields you care about.

I’m not a very skilled Haskell programmer, but every time I’ve tried to parse JSON in it I’ve felt a strong urge to switch back to Python or JavaScript. There are libraries that make it easier but I think it’s always going to be hardly to get started with JSON APIs in a language with a powerful and strict type system. Of course, that extra up-front effort might be worth it in the long run due to the benefits of type safety, but as the grandparent said, sometimes programmers need the freedom to incur technical debt.

I've really come to love strict JSON parsing, because everytime an API call was not as expected, I get the error straight at it's source, not somewhere 20 layers deeps in my views or models, way too late, and only with this one specific combination of events.

Aeson parsing seems straightforward to me, not sure it could be easier in any language. Serialization from your endpoint is automatic, based on your data types.

The real benefit of using a well typed language here is that you only need to parse the data once at this point, after that the type system makes sure the data is of correct format everywhere else in your codebase.

Aeson is not straightforward at all. Just try doing what the GP is saying he wants to do, you will discover you can't.

It is a very good library if you control both sides of the communication, and have fit for purpose interfaces. Otherwise, you are better with any lower level library.

Besides, what is it with the strictness of Haskell JSON libraries? There's nothing that will even accept Window's BOM.

You don't have to define a type for Json data in Haskell if you don't want to. Defining a full type is only norm/preference.

With Aeson, You always have the option work with a map directly and pull out properties by name.

That is also especially useful when you can't predict what keys or shape the data will have

I don't want to detract from Aeson. It's a great library. If you control both ends of your channel, it can be more convenient than the equivalent encoding and decoding on Javascript, and making a JSON library that is more convenient than JS is no small feat.

But it's a very complex library. It may not show at your code interface, but that complexity is there at the documentation and type errors. It is also famous for generating bad runtime messages. There is even another library for fixing this, at the cost of even more complexity.

If you fall outside of its optimum usage scenario, it won't afford you more functionality than a low level parser, so you are much better saving that complexity and going with the parser.

Ok, so use a parser then? Not sure what this has to do with comparison to dynlang/js.

OP was complaining that it's hard to type some kinds of payloads.

Sure you could write a custom parser for it.

With fromJSON in fact this is what you write. Maybe if you've only been driving Json instances generically, you never actually written a fromJSON instance in Aeson? I only use generic deriving in the simplest cases.. otherwise just write the fromJSON instance parser

My point was that you can defer parsing up front and just return a map of you don't want deal with the shape.

But sure you could write a custom parser as well, that falls more into the typed philosophy though than what js folks are used to

Fair enough, you're probably right, I have only used it in contexts where I build both sides of the application.

Out of curiousity, what exactly was the difficulty with decoding JSON in Haskell/

I was trying to use Aeson, which as noted in this excellent tutorial [1] (which didn’t exist when I first tried it) is “hopelessly magical for people who try to learn it by looking at provided examples, and existing tutorials don’t help that much.“ The rest of the tutorial should give you some idea of why I say that parsing JSON in Haskell is complex, at least if you’re used to doing it in JavaScript.

[1]: https://artyom.me/aeson

I'm interested in specifically what issues you had with JSON.

I was replying to someone who said they had trouble.

I found that by explicitly having to write types, the structures become a lot less convoluted. You split things apart, give them names, spot patterns more easily, and encode all sorts of information in a type system. And knowing that all fields are there when you need them is a relief.

Of course a code review can catch it. Of course careful programming can as well. But those are external constraints.

Having internal constraints greatly reduces the amount of mess one can produce.

Admittedly, sometimes to a point to which you’ve typed yourself in to a corner.

I have a library that is much like every other library in javascript. I decided to try TypeScript on it once as a test. The idea was to implement all the correct interfaces so that people couldn't call the wrong chain of methods. The interface list was almost the same size as my entire codebase.

I think TypeScript is essentially for retards. It is like "we need to lock everything down so the mediocre people we have to hire on mass can't break anything".

Typescript has the any type when you need it. You can write your code well typed but treat api returns as any if you wish. Although entering a type definition even for 100 types isn’t a big hassle, compared to day the code that has to do something with those objects. And will probably save the odd mistake so probably comes out ahead in terms of efficiency.

Counterpoint: https://lexi-lambda.github.io/blog/2020/01/19/no-dynamic-typ...