Re: [MinGW-dvlpr] [Fwd: Re: #define Win32_Winsock fails in 1.1.8]

[Earnie B. <ear...@ya...>]

Danny Smith wrote:
> 
> --- Danny Smith <dan...@ya...> wrote: >
> > --- Earnie Boyd <ear...@ya...> wrote: > Danny,
> > >
> > > Can you take a look at Matt's problem please?
> > >
> > > Earnie.
> >
> >
> > I won't have time till Sunday (GMT+13).
> It's raining so cricket match delayed: I do have time.
> 

Good, my prayers were answered. ;^)

> > I don't think that recent
> > cahnges to winsock.h are involved.  As Matt indicated the problem is
> > probably in string.h and the inclusion of sys/types.h from there.
> 
> Yes that is problem.  There  is another problem if we include
> winsock[2].h before windows.h. I have fixes that work on my testcases
> with mingw and cygwin.
> 
> Will have patch against winsup cvs by Sunday.  Note that patch will not
> work against "latest" cygwin release w32api (20010131).  (200010131
> doesn't have my recent winsock patches)
> 

Good.  I know release won't match, the Cygwin CVS is what you should use
to produce the patch.  When I get some other patches in I'll make 0.5 an
official release and let Chris create a release.

> BTW: Where did the  "Win32_Winsock" macro name come from? Can we change
> that to something a bit more consistent with standard style?
> 

I've no idea.  Yes, change it; but, do so in such a way that if
Win32_Winsock is defined your new macro is defined with a #warning
message given that it will go away sometime in the future.

> Danny

Earnie

_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com