On Sat, 20 Sep 2014 11:00:11 -0400, John Gilmore wrote:

>IBM has long had the habit of jacking up acronyms to put new
>interpretations under them.  The impetus has usually been to replace
>the notionally too particular and limiting with something more generic
>and innocuous for marketing reasons.  Thus,
>
>structured programming facility ==> system productivity facility.
>
>This practice is perhaps a sin against good language: it dilutes
>precision; but it does not strike me as reprehensible.  IBM's concern
>to protect its sometimes considerable investments in existing acronyms
>is entirely understandable.
> 
Primarily the investment in documentation.  In the earliest times of
the s360, "character" was used to mean an 8-bit datum; "byte"
was rare.  (Internet documentation uses "octet" because neither
"character" nor "byte" has a uniform meaning throughout the area
of supported platforms.)  This usage is obsolete since the advent
of DBCS.

What to do?  Declare that from now on, "byte" means an 8-bit datum
and "character" means an encoding of whatever length of a displayable
glyph, and change all documentation to conform?  Make the Humpty-
Dumptyesque statement somewhere in a glossary that "character"
means 8-bit datum except where clearly qualified as "DBCS character",
or "MBCS character", or "UTF-8" character, etc.?

Ouch!  Either way.

-- gil

----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to