search

rust-lang / rust-roadmap-2017

Tracking Rust's roadmap
215 stars 12 forks source link

Language ergonomic/learnability improvements #17

Open aturon opened 7 years ago

aturon commented 7 years ago

Point of contact

@nikomatsakis @aturon @withoutboats

Overview

As part of the effort to lower Rust's learning curve, we are interested in pursuing a number of language improvements that will make Rust both nicer to use and easier to learn. The goal is to enhance productivity without sacrificing Rust's other core values. See the README for more details.

A kind of manifesto for this work was written up on the main blog.

Status

There are a lot of ideas in flight! The lang team has been focused on writing RFCs for the biggest changes; there are lots of opportunities to help either there, or with more narrow RFCs. If you are interested in the ideas listed below, please contact @aturon or others on the team to hook you up with a mentor for RFC writing!

The potential improvements are organized by broad areas of Rust:

Ownership/borrowing

RFCs

Sketches

Traits

RFCs

Sketches

Module and privacy system

Landed

RFCs

Sketches

Error handling

RFCs

Primitive types/core traits

Sketches

Type annotations/verbosity

RFCs

Sketches

FFI

Landed

Other

RFCs

More ideas

Ideas that didn't make it

aturon commented 7 years ago

Update: added the trait aliases RFC.

aturon commented 7 years ago

Update: added Extending ? to work with more types.

aturon commented 7 years ago

Update: added Delegation of impls

aturon commented 7 years ago

Update: added Add else match

Ixrec commented 7 years ago

@RUSshy Could you be more specific about what "adding class" means to you? Most languages with "classes" have dozens of distinguishable features associated with their classes, and Rust already has analogues for many of them.

bestouff commented 7 years ago

@RUSshy why wouldn't this work:

pub struct Entity : Drawable, Updatable
{
...
}

?

(Otherwise to me, the two most useful items are the ones in "Early stage thoughts", i.e. non-lexical lifetimes and module system improvements)

Nemikolh commented 7 years ago

@RUSshy It would give wrong impression to beginners in my opinion. One could believe that all the trait being implemented by a type would appear next to the class definition itself. But this is obviously not true (see Sync, Send or Into for instance).

Besides how would you deal with where clauses? What about bounds on type that apply to only specific trait impl but not others?

Would the code really be more readable at this point?

It seems, that you would definitely end up splitting your class into multiple impls as your code evolves. Anyway, you can actually play with this class idea by writing a macro today. It would not look exactly as you have written it but would still save you a few braces. :)

Ixrec commented 7 years ago

i think having class to merge struct and impl can help something with readability, because i find myself repeating and write duplicate code

I don't see any duplicate/repeated code in your example, just three impl blocks merged into the struct definition. I do see the appeal of that sugar unifying "obviously related" blocks of code, but it is just sugar, and it doesn't appear to save very much typing or confusion compared to some of the other proposed sugars such as delegation, and with that particular keyword there's definitely a risk of misleading novices. I agree with Nemikolh that experimenting with a "class macro" would make more sense for now.

Though we should probably move this discussion to https://users.rust-lang.org/ or https://internals.rust-lang.org/ since it's at most a very tiny piece of the puzzle this issue is about.

withoutboats commented 7 years ago

My impression is that thread is for a high-level coordination of the work toward this roadmap goal. Brainstorming specific language feature ideas should be at either https://internals.rust-lang.org or an issue on the RFC repo https://github.com/rust-lang/rfcs/issues

aturon commented 7 years ago

Update: adding reference to this thread on internals which discusses some additional ideas.

aturon commented 7 years ago

Update: added match ownership ergonomics.

emberian commented 7 years ago

One idea that would make working with capn proto and similar systems so so much nicer would be something like attributes as exist in C# or Python, where one can use field access syntax but underneath call functions to get/set the field. The current state of the art in Rust is explicit getters and setters, which leads to terrible ergonomics for code interacting with data at the edges of the process boundary, to the point of Java-level verbosity. In the past (pre-1.0) I know there was some discussion and rejection of this, but I'd like to bring it up again.

aturon commented 7 years ago

Updated with a link to the new blog post and several in-progress RFCs. If there are pre-RFC internals threads that should be cited, please leave comments and I'll add them!

F001 commented 7 years ago

The "Delegation of impls" and "Delegation of implementation" both point to https://github.com/rust-lang/rfcs/pull/1806. They are duplicate.

crumblingstatue commented 7 years ago

I see default struct field values is listed here.

Are there any learnability improvements that it provides, or are the benefits purely ergonomic?

In the latter case, I find it odd that it would be tracked here. There are lots of ergonomic improvements (e.g. default/keyword args) that are currently postponed, and we couldn't possibly fit them all into this year's roadmap.

aturon commented 7 years ago

@crumblingstatue I went through the open lang RFCs looking for any that touched on ergonomics, just to have them tracked in a central place.

I'm open to keeping this tracking issue more focused, but I'm not sure what the best "cutoff" line should be.

LegNeato commented 7 years ago

What's the process for nominating an RFC for this? Sorry if I missed it.

As a Rust n00b I'd like to nominate https://github.com/rust-lang/rfcs/issues/1078 🍻

withoutboats commented 7 years ago

That issue has an RFC which will probably be merged soon :-) rust-lang/rfcs#1869

hgrecco commented 7 years ago

I find this example from the book quite annoying:

enum Message {
    Quit,
    ChangeColor(i32, i32, i32),
    Move { x: i32, y: i32 },
    Write(String),
}

fn process_message(msg: Message) {
    match msg {
        Message::Quit => quit(),
        Message::ChangeColor(r, g, b) => change_color(r, g, b),
        Message::Move { x: x, y: y } => move_cursor(x, y),
        Message::Write(s) => println!("{}", s),
    };
}

If you are matching an Enum variant, cannot Message:: (in this case) be implicit?

fn process_message(msg: Message) {
    match msg {
        Quit => quit(),
        ChangeColor(r, g, b) => change_color(r, g, b),
        Move { x: x, y: y } => move_cursor(x, y),
        Write(s) => println!("{}", s),
    };
}

I have seen ways to mitigate this (such as use Message::*) but I think some implicitness here will be nicer.

Manishearth commented 7 years ago

Please file bugs on the Rust or rfcs repo, this is for tracking efforts to improve readability and not discussing the specifics themselves.

(In that specific case, it used to be the opposite, actually, but then everyone namespaced enums as QuitMessage, etc.)

aturon commented 7 years ago

@Manishearth @hgrecco to clarify, I'd actually recommend opening threads on https://internals.rust-lang.org/ with ideas, and then posting a link here. I am using the comment thread/headers to track ideas floating around.

Manishearth commented 7 years ago

Yeah, that works too.

hgrecco commented 7 years ago

I apologize for the noise. Here is the post in rust internals: https://internals.rust-lang.org/t/elliding-type-in-matching-an-enum/4935

suhr commented 7 years ago

There's one old but still not fixed ergonomics problem: while you can do let (x, y) = ... you can't do (x, y) = ....

The open rust issue (from 2014!): https://github.com/rust-lang/rfcs/issues/372

Keats commented 7 years ago

There's this pre-RFC for named/default args: https://internals.rust-lang.org/t/pre-rfc-named-arguments/3831 which has stagnated a bit right now

aturon commented 7 years ago

I've extensively updated the issue with a lot more detail on the lang team's plans and various ideas that are currently in flight. I've tried to tag items with a lang team member where there's a clear person trying to push forward an idea. I've also called out opportunities for mentored RFCs (just search for "mentor").

Let me know what's missing!

aturon commented 7 years ago

Added coroutines; supporting async/await notation in some way could be a very important ergonomic/learnability improvement to the extent that we're leaning on futures-based abstractions for async IO.

mdinger commented 7 years ago

The Eliding type name in match arms topic has an RFC and a separate unrelated issue.

dpc commented 7 years ago

Would addressing https://github.com/rust-lang/rust/issues/40628 be considered to be added somewhere here?

MaikKlein commented 7 years ago

What about partial type inference? I know that there was an rfc for this but I can't find it anymore. playground

fn test<A, B>(b: B){}
fn main() {
    let i: u32 = 42;
    test::<i32, u32>(i); // u32 could be inferred
    // test::<i32>(i);
}
aturon commented 7 years ago

@dpc Oy, I hadn't seen that issue, but it's a doozy. Definitely something worth thinking about, though it's not obvious how to best address it.

@MaikKlein That definitely falls under the umbrella of ergonomics, but it doesn't seem particularly high impact. It does remind me, though, that we still need to sort out our story around defaulted type parameters and fallback etc.

suhr commented 7 years ago

@MaikKlein test::<i32, _>(i) seems to work just fine. While test::<i32>(i) looks like test has only one parameter.

aturon commented 7 years ago

I wanted to copy in a conversation from reddit:

... many on the libs + lang teams think that the distinction (between PartialOrd and Ord) is more trouble than it's worth, i.e. that the bugs being prevented here were pretty hypothetical, but it introduces friction for everyone, especially newcomers.

Bugs are hypothetical until they are real.

The problem with this perspective is that it doesn't acknowledge the tradeoffs involved. At the extreme, the only way to rule out all bugs is to work in a language where you can prove the correctness of programs (like Coq). But it's very impractical, and only worthwhile for the most critical software.

Rust strives, amongst other things, to be a practical programming language, and that requires us to avoid extremes and instead make judicious tradeoffs. In the design of the standard library, there were many such tradeoffs, and we learned over time that additional static distinctions are worthwhile only for cases where bugs are relatively common and otherwise hard to detect (or are extremely destructive). In a lot of cases where we could have used more type machinery to rule out potential bugs, those bugs would've been caught with even the most trivial testing.

For example, we could have used a typestate encoding to statically ensure that custom socket setup APIs are called only in valid states, but if you have test coverage you will catch this very quickly.

The motivation for PartialOrd and friends was to prevent people from doing things like using floats as a BTreeMap key. But that's an extremely obscure thing to try to do that would lead to much more obvious bugs in practice. The price for the distinction, however, is that every person learning Rust has to learn this unusual distinction, and people defining types often have to say #[derive(PartialOrd, Ord, PartialEq, Eq)] rather than #[derive(Ord, Eq)]. It's a paper cut, but paper cuts add up, and I really don't think this one is worth it.

burdges commented 7 years ago

Couldn't #[derive(Ord)], etc. imply #[derive(PartialOrd)], etc. unless PartialOrd, etc. already exists for this type?

frankmcsherry commented 7 years ago

A quick comment on PartialOrd. Whether this was its intended use or not, a partial order is actually a useful pre-existing thing. If you kill it off, some useful code is certainly going to break (all of timely and differential dataflow will have to pivot to a new custom trait). Yes, I understand "deprecate" doesn't mean "break", but it does mean "break and make others fix".

I suspect the bigger problem is "how can we have hierarchies of traits without forcing the user to acknowledge the complexity all the way down?". This problem doesn't seem to be specific to Ord and PartialOrd; you would have it with other hierarchies (e.g. Ring : Add+Sub+Mul) that are perhaps just less popular. What seems to have happened with the arithmetic traits (Add, Sub, Neg, Mul, etc) is that there is no attempt at hierarchy, and each is just "how you bind '+', '-', '*', etc). You could do that with Ord (it is just how you bind < etc; no guarantees).

Or, maybe I mis-understand and the real problem is "how few characters can a Rust beginner type before have a struct they can sort?". That a fine question to solve, but .. it would be a bit annoying to have to rewrite a lot of code because the solution was to damage the existing trait hierarchy.

Either way, it seems like there isn't much clarity on the Ord/PartialOrd distinction at the moment, in that there are some implicit laws that the standard library assumes about their implementations (that if you impl Ord, your PartialOrd impl is equivalent). It would be great to clarify things, and if that means dropping PartialOrd fair enough. I'd personally prefer that derive(Ord) get you an Ord impl (perhaps deriving a PartialOrd impl too, preventing the derivation of conflicting implementations), and PartialOrd lives on as how you hook < and friends, but without bothering people who don't care about the distinction.

Manishearth commented 7 years ago

how can we have hierarchies of traits without forcing the user to acknowledge the complexity all the way down?

This brushes on a very related issue with trait usability. We have a lot of "drill down" traits, like Eq, Serialize, Hash, etc. Most implementations are a #[derive()] away (with stable derive we may see an explosion of more such traits like this). Some are special. It's often a problem that a crate you use doesn't derive a particular trait (especially for non-std traits. this problem is assuaged by the automatic features RFC, but not fixed) and you need to go add it. Eq and Ord are the worst offenders here, because folks seem to often implement the partial ones but not the total ones.

I'm not really proposing a proper solution here, just mentioning that these traits have a class of problems associated with them. The class of problems comes from the fact that these traits are usually structurally-derived but sometimes need manual impls, and currently the structural-derive needs explicit opt in (which makes sense, really).

Throwing a half solution into the mix: in the past we have discussed things like deferred derive where derive(Trait) defers codegen until the crate that actually needs it is compiled. Similar mechanisms could be used to drill down to other objects, for example you could do something like #[derive(Hash)] pub struct Foo {x: u8, #[inline_derive(Hash)] y: somecrate::SomeType}, where SomeType doesn't impl Hash, but the compiler will try to structural derive Hash as deep as it needs to, if it can, and inline that code. Serde already has a (hacky) solution for this where you can specify a custom serializer function in an attribute which will be used instead. That feature has other, less hacky use cases, but when used to solve this problem you still have to do some annoying manual writing of derives.

This "solution" if used too much would have codesize bloat problems, and there's also the whole interaction with safety (you don't want someone to force-derive on an unsafe abstraction), but it's something that can be built upon.

djc commented 7 years ago

It seems like @Manishearth's comment is really concerned with what I call deriving-at-a-distance. However, if derive(X) today is really syntax sugar for writing out an impl X for T, should Rust not provide you with the option to derive X for T for use in the local crate even if T's origin crate doesn't derive X? I was thinking about this recently in the context of Askama templates and wanting to implement a template for some context in another crate (mostly to allow plugin templates as dylibs).

Manishearth commented 7 years ago

should Rust not provide you with the option to derive X for T for use in the local crate even if T's origin crate doesn't derive X?

Well, you have coherence problems there.

Unless you're the one defining the deriving stuff. Crates like serde/heapsize may have to do manual tinkering to implement these traits since they themselves can't apply the deriving syntax to things from std.

My solution to the drill down issue was to ignore the part of generating impls at a distance, and just make the codegen accessible. Impls come with a whole host of other problems (coherence) 😄

Anyway, this is kind of off topic I guess.

eddyb commented 7 years ago

@Manishearth Crate introspection is possible after loading their metadata, so perhaps a feature along the lines of #[derive(Serialize, Deserialize)] use std::ops::Range; could be added.

Manishearth commented 7 years ago

Yep. We have discussed this before, I think 😄

I understand that it's feasible, it's just designing the exact API (and figuring out if we actually need it) that needs doing.

burdges commented 7 years ago

Is there any problem with using specialization to reduce the boilerplate around PartialEq and PartialOrd?

impl<T> PartialEq<Self> for T where T: Ord {
    default fn eq(&self, other: &Self) -> bool { Ord::cmp(self, other) == Ordering::Equal }
}
impl<T> PartialOrd<Self> for T where T: Ord {
    default fn partial_cmp(&self, other: &Self) -> Option<Ordering> { Ok( Ord::cmp(self,other) ) }
}

I suppose impl<T> Eq for T where T: Ord { } might work too, but maybe you want Eq to serve as a reminder.

nikomatsakis commented 7 years ago

@eddyb

Crate introspection is possible after loading their metadata, so perhaps a feature along the lines of #[derive(Serialize, Deserialize)] use std::ops::Range; could be added.

Can you clarify what you envision this syntax as doing? It seems to me that the interesting (and hard) problem here is deciding what to do about coherence. Ah, would this be something we would do in the crate that defines Serialize?

eddyb commented 7 years ago

@nikomatsakis Yupp, giving the derives an approximation of the original AST, from crate metadata.

shepmaster commented 7 years ago

Is there a more detailed location to add thoughts about "Allow owned values where references are expected" ?

aturon commented 7 years ago

@shepmaster There's not currently, but you're welcome to make an internals thread on the topic (and I'll link it here)!

FWIW, the strawman proposal is to automatically coerce via AsRef or AsMut, which will have the effect of allowing owned where borrowed is expected. This automatic coercion is not like full-blown implicit coercions, because the types involved (going from a reference to a reference) strongly limit what you can do, much like with Deref and deref coercions.

We already use this pattern in std, but it'd be nice to have it apply uniformly and not have to muck up signatures with it.

aturon commented 7 years ago

Added:

shepmaster commented 7 years ago

you're welcome to make an internals thread on the topic

@aturon I've created a thread as requested. We'll see where it goes ;-)

aturon commented 7 years ago

Updated to add some ideas around coercions, e.g. https://internals.rust-lang.org/t/ergonomics-initiative-discussion-allowing-owned-values-where-references-are-expected/5161/22?u=aturon

aturon commented 7 years ago

Updated with the latest status info.

le-jzr commented 7 years ago

Any news regarding PartialEq/PartialOrd? I find @burdges's solution elegant and nobody suggested any problems.

It would also solve the weirdness where current bound Ord: PartialOrd<Self> means "for a type to be Ord, it must be PartialOrd first", whereas the true mathematical relation is that Ord is automatically PartialOrd<Self>, Eq and PartialEq<Self> as well, by virtue of Ord subsuming all those properties, so encoding the relationship in a trait bound actually makes no sense.