11.11. Parallel Port Printers

When installing a parallel port printer, there are three elements of concern: the physical hardware connection (cable and connectors), the IO device and Operating System support, and finally the print spooler support for the device.

Originally most parallel port devices were strictly write only with a minimum amount of error information - only hardware signals for online, out of paper and fault were present. Due to the lack of any other way to interface devices to the Intel based PC platform, desperate hardware designers in search of a cheap (free?) way to interface their IO devices developed ways to manipulate the signals and do bidirectional communication. Needless to say, no two companies developed the same methods, and no two companies were compatible with any other companies method.

In 1994, the IEEE 1284 standard was first developed,. The following information is courtesy of Warp Nine Engineering, of San Diego, CA, from the http://www.fapo.com/1284int.htm web page.

When IBM introduced the PC, in 1981, the parallel printer port was included as an alternative to the slower serial port as a means for driving the latest high performance dot matrix printers. The parallel port had the capability to transfer 8 bits of data at time whereas the serial port transmitted one bit at a time. When the PC was introduced, dot matrix printers were the main peripheral that used the parallel port. As technology progressed and the need for greater external connectivity increased, the parallel port became the means by which you could connect higher performance peripherals. These peripherals now range from printer sharing devices, portable disk drives and tape backup to local area network adapters and CD ROM players.

The problems faced by developers and customers of these peripherals fall into three categories. First, although the performance of the PC has increased dramatically, there has been virtually no change in the parallel port performance or architecture. The maximum data transfer rate achievable with this architecture is around 150 kilobytes per second and is extremely software intensive. Second, there is no standard for the electrical interface. This causes many problems when attempting to guarantee operation across various platforms. Finally, the lack of design standards forced a distance limitation of only 6 feet for external cables.

In 1991 there was a meeting of printer manufacturers to start discussions on developing a new standard for the intelligent control of printers over a network. These manufacturers, which included Lexmark, IBM, Texas Instruments and others, formed the Network Printing Alliance. The NPA defined a set of parameters that, when implemented in the printer and host, will allow for the complete control of printer applications and jobs. While this work was in progress it became apparent that to fully implement this standard would require a high performance bi-directional connection to the PC. The usual means of connection, the ordinary PC parallel port, did not have the capabilities required to meet the full requirements or abilities of this standard.

The NPA submitted a proposal to the IEEE for the creation of a committee to develop a new standard for a high speed bi-directional parallel port for the PC. It was a requirement that this new standard would remain fully compatible with the original parallel port software and peripherals, but would increase the data rate capability to greater than 1M bytes per second, both in and out of the computer. This committee became the IEEE 1284 committee. The IEEE 1284 standard, "Standard Signaling Method for a Bi-directional Parallel Peripheral Interface for Personal Computers", was approved for final release in March of 1994.

Even if your hardware has support for the IEEE 1284 high speed bidirectional data transfers, your Operating System drivers must support it. Unfortunately, there is no universal agreement on the capabilities that should be provided by the low level printer port device drivers, and the method for supporting them. A good example of the problem of OS support and drivers is given by the Linux Parallel Port group, currently headed by Tim Waugh, and which is documented in the http://people.redhat.com/twaugh/parport/ and http://people.redhat.com/twaugh/parport/html/parportguide.html web pages.

Given this state of affairs, it should be no surprise that there is no support for bidirectional parallel port printers in LPRng. In fact, it turns out that there are severe problems with many Unix implementations that cause extreme headaches. These include:

• While a parallel port may be opened Read/Write, a read() will either block indefinitely or cause a major system failure (crash).

• There is no buffering of data in the low level driver; that is, once a write() starts, then the process will block until the data is written out. If the write() call is interrupted by a signal and not immediately restarted, then the status returned by the write() may be an error indication (-1) or the numbers of bytes that were actually written. If an error is returned, then an unknown number were written and the print job must be aborted.

• Many Operating systems do not implement a select() functionality for the parallel port, which means that it is difficult to do a multi-threaded implementation. Instead, a polling method must be used.

The good news is that on all known systems, if the parallel port device is opened exclusively for writing, and a blocking write() is used, and the write() is not interrupted, and there are no device errors, then data is delivered correctly to the device.

In most UNIX systems the printer port has the name /dev/lpt, /dev/prn, or something similar. On most systems the dmesg utility will print a list of IO devices found during system configuration. Use the following commands to get the information and scan for the device. You should also make sure that the printer device is available.

dmesg >/tmp/a
grep lp /tmp/a
ls /dev/lp*

Gordon Haverland <haverlan@agric.gov.ab.ca> supplied this little script, that will assist with this:

#!/bin/sh
#set -v -x              # uncomment for debugging
PATH=/bin:/usr/bin
printer=
for printer in /dev/lp* ;
do
    echo PRINTER TEST to $printer 1>&2
    for i in 1 2 3 4 5 6 7 8 9;
    do
        echo PRINTER $printer $i > $printer;
    done
    echo -e \\r\\f > $printer
done
exit 0;

If your printer is connected to the device name you provided, then you should get a page of something out. If the output suffers from the staircase effect, you will see the numbers marching across the page, otherwise the numbers will all be in a single column.