Six-bit BCD is the international standard encoding system for text, adopted by the US X3 Committee in 1964. It was originally used by IBM and a number of other large computer manufacturers and formed a sort of de facto standard. The standardisation on a six-bit byte imposed by space, cost and performance constraints which arose in the design of the IBM 7030 and were later adopted by the IBM 360 mainframe, combined with the the influence of IBM, meant that it was not practical to introduce an encoding system with a longer word-length. Instead, the six-bit BCD system was extended using "shift" codes analogous to pressing the control keys on a keyboard to produce different character codes. This led to a system where there were several character sets on different "pages". This was achieved by replacing the "!" character with a control character followed by another six-bit value indicating one of sixty-four possible pages. Although the first few pages are standard and widespread, if not always implemented, there are a number of approaches to other pages, including their use for pixel graphics characters, non-Latin scripts and symbols for APL.
Most significantly, the default character page has no lowercase letters.
The shorter word length bestows a few advantages on digital data absent in OTL. One is that the same baud rate is potentially capable of transmitting up to a third more characters in the same interval, though this also implies absence of parity. Another is that storage media have greater apparent capacity compared to OTL. For instance, a single-sided CD-ROM which would store 700 Mb of text in the Gordon Timeline in fact would have a storage capacity of 933 Mb.
Another way in which this is significant is in the design of computer display systems. Since each code page can only show a relatively small set of characters, the system has to switch pages to display other characters. The blanking interval provides an opportunity for this to take place at the start of each line of text. This means that each line is generally set up either to display the default page or the characters from a specific other page, which may be standard or non-standard depending on the implementation. Switching in the middle of a line is generally impractical. Therefore, most text on computer displays is upper case or lower case only per screen. The availability of pixel characters in one of the pages means that entire lines of pixel graphics can be displayed but they cannot be mixed with text because the pixels are 2 x 3, so there are sixty-four of them. Also, the time taken by the switching process means there will always either be borders or overscan.