Paper Tape!

 "Electronic Computers Improve Management Control" UCLA 1957

These readers were mechanically very similar to the reel-to-reel decks that read and recorded to magnetic tape. But their history os so different. While punch cards and paper tape date back arguably to the work of Basile Bouchon 1725 the actual use of  paper tape is more recent. While Herman Hollerith was experimenting with the godfather of machine-read punch cards in 1889 he briefly experimented with paper tape before settling on punched cards. He described both media in his patent US395781, but there was prior art.. I'll get back to that in a moment.

It's hard to imagine now, but punched cards were used into the 1980 not because they were good, but because they were cheap. The UNITYPER, an input device for the UNIVAC I computer introduced magnetic tape for data entry way back in 1953.  During the 1960s, the punched card diminished in popularity but wasn't really extinguished until the advent of the floppy disc.

Paper tape had a  parallel history to punch cards. Joseph Marie Jacquard demonstrated in 1801 a chain of punched cards used to control a loom. While this was clearly is derived from the work of Bouchon, it was his assistant Jean Baptiste Falcon who (apocryphally) steered his work toward punched cards. Jacques Vaucanson had experimented with them as well, with working models as early as 1745, but for his trouble he was only pelted with stones.

Paper Tape Processing

For the most part paper tape is just one long continuous punched card.We are familiar with punched tape being fed to teletypes, and seeing it as the archaic output media of stock tickers and even telegrams. Binary transmitted data to paper tape dates to at least 1846 when it was used by Alexander Bain.  the first real commercial paper tape probably dates to 1937 with the The IBM Automatic Sequence Controlled Calculator (ASCC), also called the Harvard Mark I , which was used in WWII. The Mark I read its instructions from a 24-row punched paper tape. Later standardized tapes had 6, 7 and 8 rows. While most telegraph hardware of the day read Baudot, the ASCC did not.