[ConceptGCC] |
ConceptGCC :Re: missing detail in error for a failed where clause |
From: Doug Gregor (dgregor_at_[hidden])
Date: 2007-02-28 13:21:42
On Feb 27, 2007, at 6:32 PM, Martin Sebor wrote:
> Consider the program below (taken from 3.6.1 of N2042). I find
> the error message I get from Concept gcc to be worse than the
> error issued by stock g++ in that it doesn't tell me why the
> function doesn't match. Is this a known limitation or should
> I open an issue for it?
Please go ahead and open an issue. I know what's going on here, but
I'm not sure if it's just a ConceptGCC bug or if there's something we
need to address in the specification. Essentially, when we
instantiate A<B>, we check the requirement C<B>. Since it fails, the
default constructor is completely omitted from the class. So, the
compiler isn't just skipping the default constructor in that call...
there really isn't any default constructor.
Cheers,
Doug