The Ownership guide does not clearly explain what happens when there are two or more elided lifetimes

[snowe2010]

Reading over the Ownership section in TRPL and came across the example

fn frob(s: &str, t: &str) -> &str;  // ILLEGAL, two inputs

As a newbie, this was very confusing. I believe the situation could be explained by simply adding another line.

fn frob(s: &str, t: &str) -> &str;  // ILLEGAL, two inputs
fn frob(s: &'a str, t: &'b str) -> &'? str; // Expanded: Output lifetime is unclear

I'm not sure if this is the best way to do this.

[Gankra]

This sounds like a great fix to me, (although I would add the <'a, 'b> to the signature for clarity). PR welcome! 😄

[snowe2010]

I've been waiting for input on my PR for several days. Still have no response. The ? didn't compile, and so I'm not sure what to change it to. I thought this section was ignored by the compile, but it doesn't seem that way.

[bombless]

How about:

fn frob(s: &str, t: &str) -> &str; // two inputs, failed to expand

[Gankra]

Oh, sorry. I didn't even parse that there was a PR referring back to this issue 😅

[snowe2010]

@bombless That's what the original is, I was very confused by that, and I was trying to update the docs to sufficiently explain the problem with just a short bit of code. The second bit of code was much better at explaining the problem to me, as the question mark implies that the return lifetime is ambiguous.

[steveklabnik]

It's also the only place that this notation is used, which makes it
confusing.

[snowe2010]

So what should I change it to? @gankro suggested fn frob<'a, 'b>(s: &'a str, t: &'b str) -> &str;

[steveklabnik]

I think that's fine