Error when forgetting a :: when specializing a generic is confusing

[jdm]

If there's no way to fix this, so be it, but it would be nice to have a heuristic that could help the user figure out what's going on.

../src/comp/middle/debuginfo.rs:61:44: 61:52 error: unresolved name: ValueRef
../src/comp/middle/debuginfo.rs:61     unsafe::reinterpret_cast(std::ptr::null<ValueRef>())
                                                                               ^~~~~~~~

[jdm]

N.B. the confusion here is that std::ptr::null<ValueRef>() is parsed as (std::ptr::null) < (ValueRef) > (), and there's no value named ValueRef.

[catamorphism]

Confirmed that HEAD still exhibits the same (confusing) behavior.

[catamorphism]

Still just as confusing, as of 86b8bf3

[catamorphism]

This is still just as confusing as ever. Nominating for milestone 5, production-ready -- I think it's really important to fix this, as it confuses new users.

[graydon]

yeah, this is a special-case warning, should look at unresolved name when occurring within the RHS of a binary less-than expr and check for a type name, if matching, issue special-case warning.

[graydon]

accepted for production-ready milestone

[huonw]

Triage: still just as confusing as of 4ecb0a3.

[catamorphism]

Not 1.0

[flaper87]

Visiting for triage. Nothing to add

[steveklabnik]

Today, with this example:

fn main() {
    std::mem::size_of<i32>();
}

I get this error:

hello.rs:2:26: 2:27 error: Chained comparison operators require parentheses
hello.rs:2     std::mem::size_of<i32>();
                                    ^
hello.rs:2:26: 2:27 help: Use ::< instead of < if you meant to specify type arguments.
hello.rs:2     std::mem::size_of<i32>();
                                    ^

Seems good! I don't know when this changed, but it has.