[user3939382]

I’m really excited about dependent types. I’m expecting we won’t get them for a while though. Dependent Haskell is progressing but apparently it’s hard to retrofit. Idris’ own creator has said he expects it to be a model for other languages, I don’t think it will ever have mainstream adoption. Coq and Agda, F* aren’t really designed to be general purpose.

Although the implementation for the compiler is complex, and the syntax can get complex and verbose, to me my requirement is simple: I want to encode everything about input and output that I know.

Right now in mainstream languages I often know more about my arguments or output than the type system will allow me to specify.

[valyagolev]

I totally share your excitement about dependent types, but it seems that, unlike the type systems we're used to, theorems about the dependent types are much harder to prove, which makes them not very comfortable to use for the whole program.

If only there was some kind of a gradual, perhaps typescript-like approach to adding arbitrary type-level value-limiting information in random places without having to have everything proven everywhere...

[skulk]

Every non-dependent typing relation is also a dependently typed relation so I think things are already the way you want, unless you have a certain example in mind.

[valyagolev]

Sure, in the same sense that every "untyped" program is already typed with some kind of a universal type, but what's the point?

What I want is to be able to specify and have (when possible) automatically proven arbitrary assertions anywhere in the code without necessarily making sure that every possible presupposition is proven from the ground up. Just like in Typescript, I can add a type at any point where there was only "any", and have a small portion of the program typed without typing the whole thing

[dunham]

Not sure if it's what you're asking for, but in lean, you can put `sorry` to skip a proof. It's `believe_me` in Idris2, and I think the Agda version is `trustMe`.

You can also use stuff like ! when you don't want to prove your array indices (it might crash or substitute a dummy value if you're wrong).

[skulk]

> What I want is to be able to specify and have (when possible) automatically proven arbitrary assertions anywhere in the code

This doesn't even exist in TypeScript. If I change

    function foo(a: any) { return baz(a); }

to

    function foo(a: number) { return baz(a); }

whoever calls foo has to still prove (or assert) that the argument is a number.

Is that what you're after, asserting a dependent type? For example being able to change:

    function divByTwo(a: number): number { return a/2; }

to

    function divByTwo(a: even_number): number { return a/2; }

You want every place divByTwo is called to also automatically supply a proof that the argument is even?

[valyagolev]

i don't even need it to go beyond the insides of a function. something like a proof of an invariant that's only relevant inside the function's body would be fine. e.g. in Rust, almost every place where you see something like ".unwrap()" with "// unwrap safety:" comment, this comment could be an assertion easily proven from the few lines above

[mettamage]

Eli5 dependent types?

Chatgpt:

Dependent types are a concept in computer science and programming languages where the type of a variable can depend on the value of another variable. In simpler terms, imagine you have a list of numbers and you also have information about how long that list is. With dependent types, you could create a type for the list that explicitly includes its length, ensuring at compile time that operations respect this length. This makes it possible to catch certain kinds of errors before the code even runs.

For example, in a language with dependent types, you could specify a function to take a list of length 3 and no other sizes. If you tried to pass a list of length 4, the program wouldn't compile, preventing this kind of mistake early on.

It's a bit like having an extra layer of safety checks, where the types are more expressive and can encode more intricate relationships between variables.

[howenterprisey]

I never care much for chatgpt answers. I don't know why people post them on here.

In the first sentence, "another" is wrong because you don't need two variables, you just need one. Final paragraph's wrong for the same reason.

The example given is poor given that I can write [i8; 3] or int[3] in Rust or C and those very much do not have "dependent types" in the popular sense (Rust's const generics excepted). To be fair, those examples are technically dependent types, but it would be better to give an example that's impossible in those languages, such as "array with length at most 3" or "even integer".

Finally, to nitpick, "a bit like" is unneeded hedging.

Stack Overflow did a much better job: https://stackoverflow.com/questions/9338709/what-is-dependen.... Wikipedia's article is pretty bad and maybe I'll get around to fixing it at some point.

[ChadNauseam]

I like ChatGPT’s answer better than yours. Rust’s cost generics and C++ aren’t dependent types in any sense. And saying “the type of one variable can depend on the value of another” is vague but a good intuition pump.

Since I guess we’re doing dependent types explanations I’ll give it a go. Dependent types extend a type system with two new type thingies: 2-tuples where the type of the second element depends on the value of the first, and functions where the type of the function’s return value depends on the value that was passed to it.

[howenterprisey]

Ha! Looks like I was wrong and confusing refinement types with dependent types: https://cs.stackexchange.com/a/21733 (unless of course this person is also confidently wrong, heh)

> Dependent types are types which depend on values in any way. A classic example is "the type of vectors of length n", where n is a value. Refinement types, as you say in the question, consist of all values of a given type which satisfy a given predicate. E.g. the type of positive numbers. These concepts aren't particularly related (that I know of). Of course, you can also reasonably have dependent refinement types, like "type of all numbers greater than n".

[mettamage]

Because I had the question and I figured this was quicker. I didn’t know what dependent types were.

So now you know why I do it. Also, I believe this is my first time doing it. I might be wrong.

Is it better to ask and wait for an answer instead?

There is nothing in the guidelines on HN about it. I don’t know what’s reasonable and I haven’t seen strong cultural norms from HN yet. I at least labeled that the text was from chatgpt as to not confuse it was my own text.

[howenterprisey]

It's of course fine to ask chatgpt. I also appreciate that you wrote that it was from chatgpt.

However, I wouldn't recommend posting the result here if you don't know if it's correct. Moreover, anyone can ask chatgpt themselves. It's better to wait for someone here to post an answer.

Yes, there's nothing in the guidelines, but they're (deliberately) not all-encompassing. Besides, I would hope it's part of basic Internet etiquette; just because we now have access to tools that generate plausible-sounding but not necessarily correct answers to questions doesn't mean we need to post what they create in the place of genuine answers.

[mettamage]

> Besides, I would hope it's part of basic Internet etiquette

Yea that got lost on me. I think I view these things a bit differently than most on HN. I've noticed in general that the more I'm not at uni, the more my thinking has become heuristic and quick-ish. It used to be more thorough and in-depth. The trade-off of it being that such type of thinking is more time consuming but the answer is more comprehensive and/or accurate.

[gosub100]

> Is it better to ask and wait for an answer instead?

No, around here it's better to say "So dependent types are pretty much $SOMETHING_COMPLETLY_WRONG ?" and wait for all the "corrections" (aka Cunningham's law someone linked to nearby).

[kccqzy]

Did you forget about search engines that enable you to do your own research rather than asking a human or an LLM?

[mettamage]

I could use google but chatgpt wins on speed and asking on HN wins on vine. When someone asks me a question I am mostly quite enthusiastic in explaining it. You can’t get that as easily from a search engine.

It all has trade offs.

[kccqzy]

I don't fault you if you explicitly emphasize speed rather than accuracy. Just don't post these on HN.

[xeonmc]

ChatGPT = Cunningham bait

[gosub100]

I don't see anything wrong with GPT's answer other than pedantic quibbling. I just don't like it because I don't want to share this space with non-humans.

[mettamage]

What does that mean?

[Jtsummers]

https://meta.wikimedia.org/wiki/Cunningham%27s_Law

Cunningham's Law: Post something wrong online so you can get a correct answer.

Half the ChatGPT answers on here seem to be wrong in obvious ways or, worse, subtle but critical ways. When people post them they get downvoted, and other people chime in with "Why are you trusting a fabulist like ChatGPT instead of going to actual resources with definitions and explanations that aren't garbage? Here's what it actually is..."

[gosub100]

I wonder if it would create a singularity in spacetime by posting the GPT answer to "what is cunningham's law" here?

[howenterprisey]

Worked for me :<

[bjornasm]

That is really interesting. Whish there was some better name for that, as I feel like it isn't that descriptive. However their benefit seem really obvious, saying that something is an areay, or even an array of integers is barely half the picture if in reality it is a array of length 5 of even integers to loan from your example. I guess you would try to implement it in other languages creating objects?

[icen]

Another classic use case is that you can have expressions in the types of your function, for example this won't compile:

    stringOrInt : (x : boolean) -> int -> (if x then String else int)
    stringOrInt true x = toString x
    stringOrInt false x = x + 1

    1 + stringOrInt true 37 # this will error, because it knows you've not returned an int

The other example that you can do in depedently typed languages, but is too involved to write out here, is make a type-safe printf, where the format string produces the types for the other arguments.

[turndown]

It’s simplest explanation is that dependent types are types whose exact type use some kind of parameterized value in the definition. So a dependently typed array would use a parameterized length value as part of it’s definition(this is usually the easiest example to understand). So an array of length 4(of whatever type) could be identified as different from an array of length 5, etc.