On 8/26/25 19:07, Jon Perryman wrote:
On Tue, 26 Aug 2025 10:10:11 -0600, Paul Gilmartin <[email protected]> wrote:
Ask yourself, why are unix commands so strange (e.g. testcmd -x
Most UNIX facilities support search for 2 (or more) different
characters, often by [list].
Really? The "-" simplifies parsing. Single characters simplify parsing. IBM
never had this problem because TRT and TR existed in the original design.
I don't understand " The "-" simplifies parsing." Do you mean a
range of characters such as A-Z? TRT and TR work equally with
a sparse list such as [aeiou]. Generally, TRT and TR are
overused; to IBM developers they're the shiny new hammer,
making everything look like a nail.
Again, IBM had the TR command from the beginning and uppercase is readable &
consistent. Anyone remember back people started using mixed case edit that caused
lowercase DSN on disk but upper in catalog needing SPZAP to rename the dataset?
The right reaction would have been to make DISABLE(DSNCHECK) the
default.
IBM had typewriters before keypunches. The IBM 026 was a choice (not a
limitation).
It was a limitation by choice. The motivation may have
been the area of a mechanical character generator used
by INTERPRET.
aggravated by the 3277 which, unforgivably, transmitted
minuscule while it displayed majuscule.
3277 keyboard was dual case keyboard. I don't remember using caps lock so I'm
guessing it defaulted to uppercase.
It didn't have CAPS LOCK. I believe it had SHIFT LOCK which
mechanically latched the SHIFT key making numbers om the top
row inaccessible.
Outgrow your 026 mindset. Think of 52 different characters,
not two typographic variants of 26 characters.
It was not a mindset. It was a choice. Are you saying IBM built typewriters but
couldn't design a fully functional keypunch?
The evidence is that they didn't. Lower case could be
entered only painstakingly with Multi Punch.
IBM made a choice that was in line with a world to uppercase. At least they
didn't screw the pooch with commands.
...
A woman I knew, ca. 1985, was using MS Word to write her
dissertation; English but containing many French words.
Her professor marked down for incorrect capitalization
or accents. Case and accents mattered to her. We could
add the French words, correct, to the spelling dictionary
but we could find no way to remove English words
different only by lacking the accents.
About that time I adopted the habit of using mixed case in
comments so I could search Assembler source for identifiers
without being distracted by my comments. ISPF delimited
strings were case-sensitive. But IBM, uncharacteristically,
changed a default by making then insensitive.
(Before you comment, become familiar with Hyrum's law.)
--
gil
