On Nov 19, 2018, at 2:52 PM, Tony Harminc <[email protected]> wrote:
> 
> I'd say it's an "EBCDIC in an ASCII world" problem; not anything
> fundamentally wrong with EBCDIC. Imagine if the original IBM PC had
> been an EBCDIC machine. OS/2 and Windows would surely have followed,
> Unicode would've been EBCDIC-based, and we'd live in a different but
> not necessarily worse world.
> 

What’s “fundamentally wrong” with EBCDIC is it’s non-standard. Sure, as far as 
just one set of mappings of codes to characters there’s nothing inherently 
superior about ASCII, but being a standard across different vendors and 
operating systems makes ASCII superior.

By the time the IBM PC came along, everybody outside the mainframe world had 
standardized on ASCII; if it had used EBCDIC it probably wouldn’t have had as 
much uptake.

The truly sad thing is that Bob Bemer was working for IBM and had convinced 
upper management of the importance of ASCII during the development of 
System/360, but this understanding wasn’t communicated to the engineers, 
programmers, and (notably) customers who needed to implement it. As a result, 
those of us who work with z/OS, z/VM, and z/VSE are burdened by character 
encoding issues that don’t face people on other systems/OS’s.

I just spent an hour this very morning explaining to a developer why a 
EBCDIC->ASCII process wasn’t translating square brackets and a few other 
special characters the way he expected. If the System/360 community had 
embraced ASCII, as IBM’s management had intended, I could have spent that time 
doing something useful.

-- 
Pew, Curtis G
[email protected]
ITS Systems/Core/Administrative Services


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

Reply via email to