alt Hacker NewsI like this change. I was wondering if I would've preferred to have something on the function signature (eg `tcc_fn foo() ...` as in Tail Call Constrained fn) and when encountering that the rust compiler would make checks about whether the body of the function is tail-callable.
My fear is that adding yet another keyword it might get lost in the sea of keywords that a Rust developer needs to remember. And if recursion is not something you do often you might not reach for it when actually needed. Having this signal in the function signature means that people would be exposed to it just by reading the documentation and eventually will learn it exists and (hopefully) how to wield it.
What do you folks think?
Replies
The property we care about isn't a property of functions but of callers, so marking a function doesn't help.
`become blah(foo, bar)` is the same thing as `blah(foo, bar)` except that we, the caller are promising that we have nothing further to do and so when blah returns it can return to our caller.
If somebody else calls blah they don't want that behaviour, they might have lots more to do and if they were skipped over that's a disaster.
In some languages it's very obvious when you're going to get TCO anyway, but Rust has what C++ calls RAII, when a function ends all the local variables get destroyed and this may be non-trivial work. Presumably destroying a local i32 is trivial & so is a [u8; 32] but destroying a local String isn't, let alone a HashMap and who knows how complicated it is to destroy a File or a DBQuery or a Foo...
So in a sense "all" become does is try to bring that destruction sooner a little, so it happens before the call, leaving nothing to do afterwards. We weren't using that String any more anyway, lets just destroy it first, and the HashMap? Sure, and oh... no, actually if we destroy that Foo before calling blah which needs the Foo that messes things up... Rust's borrowck comes in clutch here to help us avoid a terrible mistake, our code was nonsense, it doesn't build.
Edited: Improve explanation
Given that it's not really that uncommon to see something like `pub(crate) async fn foo() ...`, the concern of function signatures starting to get unwieldy feels a lot more concrete than hypotheticals about a "sea of keywords". From looking at the list of keywords in the language currently (listed here: https://doc.rust-lang.org/std/#keywords), I don't really see a whole lot that I think the average Rust programmer would say is a burden to have to remember. At most, `union` and `unsafe` are probably ones that most Rust programmers are not going to need to use directly, and `SelfTy` might look a bit confusing at first due to the fact that the keyword itself is spelled `Self` (and presumably listed in that way to make it easier to differentiate from the `self` entry in the documentation), but even including those I'd argue that probably over half aren't specific to Rust.
As for being in the documentation, I'd argue that might even be an explicit non-goal; it's not clear to me why that should be something considered part of the API. If someone updates their tail recursive function to manually loop instead (or vice-versa), it doesn't seem like that should necessitate the documentation being updated.