On Tue, 20 Feb 2018 14:35:57 +0000, Seymour J Metz <[email protected]> wrote:
>The problem is not that the term "ASCII" is ambiguous; it isn't. The problem
>is that people don't understand what ASCII is and
> refer to things that aren't ASCII as "ASCII". Code pages like 437, 850 and
> ISO-8859-x aren't ASCII, and if IBM documentation
> refers to them as ACII then that is a disservice to your customers.
>
>There are, however, issues as to how ASCII should be converted to other code
>pages. For example, ASCII defines a broken
> vertical bar, nut it is common to display | as a solid vertical bar. Many
> code pages have both characters, and either choice
> can lead to problems.
At a point in history, "ASCII" and "EBCDIC" meant only one thing, but they have
since evolved. They retained, however, their core DNA. It is to that DNA we
refer when we say their names without qualification. The term "US-ASCII"
refers to the original 7-bit specification adopted by ANSI as X3.4. It is
preserved in IBM code page 367.
The nice thing about ASCII is that all descendants of US-ASCII simply added
characters 128 to 255. The first 128 code points remained untouched. How a
platform displayed the 32 control characters in a non-control context, or the
appearance of undefined code point 0x7f (i.e. in a character generator), was
beyond the scope of the ANSI standard.
EBCDIC, on the other hand, suffered under the strain of trying to maintain
compatibility with its BCD ancestor, the needs of the different programming
languages, a recalcitrant ANSI standards committee, and the limitations of the
hardware at the time ("Thank you, Mr. Hollerith!").
There was apparently an objective that EBCDIC have all the graphics of ASCII
and none that ASCII did not. "Men plan, gods laugh." There were too many
participants at the table and compromises were made that we deal with today.
The equivalence eventually made it into the literature as IBM code page 38.
Right. I don't use it, either.
The book "Coded Character Sets, History and Development" by Charles Mackenzie,
IBM (Addison-Wesley, System Programming Series, 1980) is as fascinating as it
is horrifyingly geeky.
Alan Altmark
IBM
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
