Overview

The printing process on Unix works roughly as follows:

1. An application generates a document to print.

2. You or the application sends the document to a print server by means of a print client program.

3. The print server puts the document on a queue.

4. When it is the document's turn to print, the print server sends the document through a print filter program.

5. The print filter changes the document to make it acceptable to a printer.

6. If the printer does not understand PostScript (the standard document format in Unix), another print filter must rasterize the document into a bitmap form that the printer understands.

7. The print server sends the result to the printer device.

Making your printer work on a Unix system can be a trying experience. There are several problems:

• There are few standards.

• Each stage of the process can present any number of errors.

• There are several different client, server, and filter packages for printing.

• It's sometimes difficult to determine the purpose of each printing system component.

• The existing software never seems quite satisfactory, so upgrades are frequent.

• Security holes run rampant in network print servers.

• There are two basic ways to configure network printing: one where all computers send their documents directly to a printer, and the other where all computers send their documents to a print server, which then talks to the printer. It can be confusing when you have a network full of machines that use both approaches.