Dan Pop <
[email protected]> wrote:
And consecutive, which is actually a stronger requirement.
That is a restriction on the character set. It also happens to be a
very desirable characteristic of a coded character set; so desirable
that no one has ever reported meeting one that doesn't have it.
It's desirable, as are consecutive or at least ascending letters, only
for machine processing of the data represented therein, which was not
the purpose (or application) of many character codes. In fact, until
the rise of electronic digital computers, I believe pretty much the
only code for even limited processing (rather than transmission or
storage) of data was Hollerith card (which survives almost unchanged
as a subset of EBCDIC).
Important examples of nonconsecutive digit codes:
International Alphabet 2 aka "Baudot" code, 5-bits with 2 shift states
(letters and figures); the digits were the FIGS-shift of the top
letter row of the (US standard) keyboard QWERTYUIOP, which wouldn't
have been consecutive even if letter codes had been, which they
weren't: much like the telegraph and later radio "Morse" code, they
were originally designed to use fewest "mark" bits for the commonest
letters to reduce power usage over long wires. Used AIUI by Teletype
models prior to 33 (the first IA5/ASCII model), and in the Telex
public switched network into the '80s at least, even though probably
few if any of the terminals were still Teletypes, and many were
computers and not really terminals at all.
I think I still have some software containing Baudot/ASCII tables --
stored somewhere among a bunch of files in a now-unsupported format
written on 9-track magtape, if you can read that
TeleTypeSetter or TTS, 6-bits, 2-shift but digits had their own codes
i.e. corresponding approximately to the 2 shifts of a 4-row typewriter
keyboard including both upper and lower case letters. Originally used
to operate Linotype machines in duplicate or remotely, used at least
into the '70s in a variety of pre-press equipment. Also chose fewest
'1' bits for commonest codes, but I think by this time the motivation
was more to reduce average punch wear.
I'm pretty sure Frieden Flexowriter, a classic typewriter mechanism
(i.e. typebar basket and platen on moving carriage) modified into a
computer terminal, used a code that did not have consecutive digits,
but I didn't spend much time looking at its tapes and don't remember.
Although none of these was ever used, and I don't think would even
have been considered, as the internal code for a computer, and thus
irrelevant to the design of C or any other programming language.
- David.Thompson1 at worldnet.att.net
