Early versions of Unix included an elegant utility called graph, which accepted simple text files such as:
1 1 2 4 3 9 4 16producing from them neatly scaled and labelled 2D plots in a device-independent format. graph was accompanied by plot, which consisted of a family of interpreters (drivers) that translated graph's output for a wide variety of mostly obsolete graphics terminals. I wrote a number of plot drivers for slightly-less obsolete devices around 1982, and Paul Albrecht wrote the first version of plt as a replacement for graph when it disappeared from the standard Unix toolkit, sometime after Version 6. (Try it out: type the numbers above into a file named squares, putting two numbers on each line, then run the command ``plt squares''. It's that easy!)
Early versions of plt used drivers I had written for the Tektronix 4010 (a vector graphics terminal with a storage display; we had a few of them around the lab), the Tektronix 4662 XY plotter that was capable of changing pens (the first color output device we had), the NEC Spinwriter (an impact printer with fine-resolution paper control, for which a special type wheel with a brass ``.'' was available, for plotting applications such as ours), and even dumb terminals (with 80x24 resolution). During the mid-1980s, MIT's Project Athena produced the first versions of the X Window System, and Apple produced the first PostScript printer (the LaserWriter). Paul wrote X11 and PostScript drivers and integrated them with plt, in order to make better use of the enhanced capabilities of these environments than was possible using only the Unix plot API. It is these two drivers, with updates I have made since the early 1990s, that are the nucleus of plt today.