Re: PATCH: autoload with explicit path

[Peter Stephenson <p.stephenson@xxxxxxxxxxx>]

On Thu, 12 Jan 2017 16:23:38 +0000
Daniel Shahaf <d.s@xxxxxxxxxxxxxxxxxx> wrote:
> The point is that the next time writes:
> .
>     foo(shf->filename);
> .
> that won't compile, so he will be forced to take into account the two
> distinct overloaded meanings.

OK, I can see that might be useful.

> This might have prevented the bug you fixed in 40335 from being
> written.

No, it would still have been possible to resolve all uses into one form
or the other.  The problem was it was done implicitly by location in the
code and needed discriminating by something else such as the new bit
(or the implied NULL / non-NULL test of a separate field).

> I'll see if adding it makes things clearer, will post if it does.

Fine, you might want to see what happens with the change for caching the
autoload directory first as it hits the same lines.

pws