Re: Big-endian/little-endian (WAS: Re: can I burn the output of mpg123 -s?)
On Tue, Aug 18, 1998 at 01:22:53PM -0400, Michael Stone wrote:
> Quoting Stephen J. Carpenter (sjc@delphi.com):
> > On Tue, Aug 18, 1998 at 10:34:39AM -0400, Michael Stone wrote:
> > > The solution is to always run hton before putting stuff on the wire and
> > > running ntoh when pulling stuff back. That way you can be sure that the
> > > stuff on the wire is always network byte order without having to put in
> > > nasty test cases.
> >
> > thats all well and good BUT...
> > xfstt is a ont server for X11....So it has to work within the existing
> > protocols...it would be a "bad thing" if you had to modify the Xserver to
> > use a new font server.
>
> Are you saying that the xserver looks for host (e.g., random) byte order
> rather than network byte order on the network?!? (I never had reason to
> look, but I just might, now...)
Thetruth is I don't know.
xfstt , as part of its connection sequence uses some value to determine
the byte order of its peer and if it is not the same drops the connection.
This leads me to believe that the X Server expects its fonts (if not
everything else...the rest I don't know) to be rendered in host
order and sent in that order.
on the other hand it may just be parinoia? maybe X does fo ntoh's all
right? I could check by removing the check for byteoder and riunning the
server on my IPC (where host order == network order) and connecting my
i386 to it....
but the very existance of that check leads me to believe that it wont work
but...I was told it would not be hard to fix the swapping
-Steve
--
/* -- Stephen Carpenter <sjc@delphi.com> --- <sjc@debian.org>------------ */
E-mail "Bumper Stickers":
"A FREE America or a Drug-Free America: You can't have both!"
"honk if you Love Linux"
Reply to: