🚀 go-pugleaf

In article <2127@shaman.wv.tek.com> andrew@frip.wv.tek.com writes:

>Ideally, PostScript jobs are transmitted through media which provide
>for out-of-band notification of control information such as
>end-of-file.  This is the case for AppleTalk and TCP/IP (and note that
>you'd better not send a control-D through these networks!), but not for
>serial and parallel, so these latter must use an in-band EOF indicator
>which is not part of the language.

Then wouldn't it have been a good idea to make ^D equivalent to white
space in postscript so the same data stream would work everywhere?
That way the ^D's needed to make serial/parallel connections work
(I'd guess the large majority) would be as innocuous as linefeeds
or carriage returns added by some systems for other reasons when they
are passed through the transmission media.  Instead we now have to make
the application aware of how the data is going to be transmitted to the
device in order to accomodate the most likely case where the application
must supply the ^D.  Programs that encapsulate postscript or add instructions
to the beginning would still need to strip off ^D's in case the actual
transmission is through a serial or parallel link, but that's trivial for
a special purpose program anyway.

Les Mikesell
  les@chinet.chi.il.us