The problem of how to store electronic data goes hand-in-hand in with the evolution of computers.
In the beginning, there were so few computers in existence (none of them compatible with any others), digital pioneers had little need for portable software.
Soon, however, companies like IBM began selling multiple identical computers, and users sought a way to program the devices more efficiently, to move programs from computer to computer, and to reload programs later with a minimum of hassle.
Thus begins the history of removable computer storage, which in many ways is the story of software distribution: The first job of removable storage was to share software without requiring any reprogramming from scratch.
We've revisited dozens of ways that engineers have solved the problem over the past 60 years.
Punched paper tape
Most of the earliest computers used spools of paper tape, which stored information via punch holes.
Some early machines, like the Colossus Mark 1 (1944), operated entirely on the data fed in by the tape in real time. Later computers, like the Manchester Mark 1 (1949), read in programming from tape and stored it in a primitive type of electronic memory for later execution.
Various computers during the next three decades continued to use paper tape for both input and output, and the medium experienced a brief renaissance in the mid-1970s among personal computer hackers who were attracted to its low cost.
The punch card
Punch cards trace their origins to textile looms of the 1800s, where they conveyed instructions that defined and controlled machine weaving operations.
In 1890, Herman Hollerith applied the punch card idea to tabulating US Census data. He founded a company (which grew to become IBM) that used the cards in tabulating machines.
When IBM began building general-purpose computers in the 1950s, it used the cards for data storage and input, and soon many other computer manufacturers adopted punch card formats of various kinds.
Many used 80-column cards, which stored one character per column. As late as 2002, IBM was still researching punch card technology, in the form of a system capable of storing 25 million pages of data on a surface the size of a postage stamp.
NEXT PAGE: Data on tape