Re: [pcp] Protecting AF_UNIX Code

[Nathan Scott <nathans@xxxxxxxxxx>]

Hi Dave,

----- Original Message -----
> On 06/10/2013 07:08 PM, Nathan Scott wrote:
> > Strictly speaking the macro being used to protect conditional code
> > compilation should match up with the thing it is protecting.   ...
> 
> I want to argue that the presence of the header implies the presence of
> the macro, but I don't know if that's actually true.

As discussed on IRC, MinGW headers define AF_UNIX for some inexplicable
reason which complicates things yet further.  I've pushed in a change to
add HAVE_STRUCT_SOCKADDR_UN which appears to be an unambiguous, portable
test we can use.

Also, as mentioned on IRC, even on the Unix variants AF_UNIX is defined
in sys/socket.h and not sys/un.h ... heh, what a twisty maze.

cheers.

--
Nathan