[mod.computers.laser-printers] PostScript on LW steals control chars on the serial port

gnu@hoptoad.UUCP (John Gilmore) (09/29/86)

A "feature" of the serial port driver in the LaserWriter makes various
things hard to do.  I was wondering if there is a patch that can be
made to disable the feature.

When talking on the AppleTalk port, any of the 256 possible bytes can be
received and acted upon.  When talking on the serial port, the driver
intercepts ^S, ^Q, and ^T [and maybe others] and processes them specially.
This behaviour cannot be turned off, according to the PS Reference Manual.

This seems to me like a serious lose.  I would like to be able to write
PostScript programs that would make the printer emulate a variety of
other printers (eg Epson, HP LaserJet, etc).  This can't be done if the
PostScript emulator program can't read all the bytes from the serial line.
(One can certainly write a filter for a host computer which could
turn Epson output into PostScript, but that is not what I am interested in.)
Other than this serial port glitch, it seems like the LW could be made to
emulate just about any printer in existence with little trouble.

I would also like to print bitmaps without converting them to hex.
One might consider writing PostScript emulator that emulates a "rasterfile"
printer -- you dump a Sun rasterfile down the line at it, and it gets printed.
Of course, I could define my own protocol to quote the few byte values
that the LW eats, and write a protocol driver to run on the Unix end and
another to run on the LW end, but just turning off the damn serial driver
would be a lot easier.

Note that unless you are running an odd PostScript program that produces
voluminous output on the serial port, there is no need for ^S or ^Q to be sent
FROM THE SYSTEM TO THE PRINTER as flow control.  Virtually all the data
will be going from the system to the printer, so the flow control ^S's
will be sent the other way.  Thus turning off the interception of these
and other control characters in PostScript's standard input will not
injure the printer's ability to cope with high data rates.

Is there any hope for a patch for this, and a clean fix (e.g. like Unix
"raw mode") in future PS printers?