According to the docs:

class Foo
{
    int x;
}

void test(Foo foo)
{
    size_t o;

    o = Foo.x.offsetof;   // yields 8
    o = foo.x.offsetof;   // error, .offsetof an int type
}


Which is what I thought was correct, and I could have sworn I used before. But with DMD 0.143, the two cases are switched; Foo.x.offsetof is illegal, while foo.x.offsetof gives the proper result.

Is this a doc bug, or is DMD not adhering to the spec?  I think it should work as in the docs, as it's a class property, not an instance property.

"Jarrett Billingsley" <kb3ctd2@yahoo.com> wrote in message news:dr19a7$9s9$1@digitaldaemon.com...

I just found that this bug has been reported a few times over the past 8 months or so, and I reported this same bug last _April_.  Are the docs going to be fixed, or is this truly a bug?

Jarrett Billingsley wrote:
> "Jarrett Billingsley" <kb3ctd2@yahoo.com> wrote in message news:dr19a7$9s9$1@digitaldaemon.com...
> 
> I just found that this bug has been reported a few times over the past 8 months or so, and I reported this same bug last _April_.  Are the docs going to be fixed, or is this truly a bug?

I think this is a bug. Walter has been quite quiet about it, so it seems that it isn't that simple (or it's not top priority) to fix it. It would a lot better if we had a bugzilla instead of D.bugs-ng, I would certainly want to see a centralized todo-list for DMD.

-- 
Jari-Matti