Re: emacsclient horribly broken

[Tim Van Holder]

On 10/31/06, Juanma Barranquero <address@hidden> wrote:
> On 10/31/06, Tim Van Holder <address@hidden> wrote:
>
> OK. BTW, have you ever signed papers for changes in Emacs? The amount
> of patches till now is fine, I think, but more extensive changes would
> require signed papers.

I'm not quite sure - I know I have signed papers for autoconf,
automake, libtool and bison - but not sure about emacs.

> > -# include <sys/ioctl.h>
> > +/* FIONBIO can be in either sys/ioctl.h or sys/filio.h */
> > +# if HAVE_SYS_IOCTL_H
> > +#  include <sys/ioctl.h>
> > +# endif
> > +# if HAVE_SYS_FILIO_H
> > +#  include <sys/ioctl.h>
> > +# endif
>
> Shouldn't that last one be "# include <sys/filio.h>"?

Yes. The evil copy-as-kill/yank strikes again ^^

> > -#if !defined (HAVE_SOCKETS)
> > +#if !defined (HAVE_SOCKETS) || !defined(HAVE_INET_SOCKETS)
>
> So, both are required?

I don't know - I do know I had to add netinet/in.h, so
HAVE_INET_SOCKETS is definitely needed.  I figured it was safe enough
to just add the requirement.

> BTW: please send future patches as attached files. I don't like what
> Gmail does sometimes to patches included as text in the message.

Will do.

> I've installed the patch by Jan Djärv because people seems to agree
> that it is better to use O_NONBLOCK or O_NDELAY to FIONBIO.
>
> Could you rework your second patch and send it to me, please?

Unless I forget, I'll rework my patch tomorrow morning when I get in
to work and do my
daily build.