The author made an interesting choice, to (essentially) lump the IBM
mainframe "greatest software" into one line item ("360 OS"). (OK, System/R
is another.) It's yet another area where the comparisons break down pretty
quickly. How do you compare CICS or IMS with other software, for example?
That's difficult -- they (and a lot of IBM-MAIN readers) keep civilization
running I guess you'd say. And I think he means more than "360 OS" when he
talks about the mainframe. (He kind of mixes up hardware and software,
although that's easy to do. Interesting how that works, though. So much of
the great software on his list depended on great hardware or at least was a
carefully constructed total package -- Macintosh, Apollo, etc.)
I disagree with Microsoft Excel, at least in terms of his dismissal of
VisiCalc and Lotus 1-2-3. Surely one aspect of great software is that the
world would be poorer for not having that software in existence. Is that
true of Excel? If 1-2-3 (or Quattro) was the world's most popular
spreadsheet today, would that somehow be a problem? No, I don't think so.
Visi Calc for Visi On preceded Excel for the Macintosh by about two years,
so Excel wasn't even the first GUI spreadsheet (which might be a possible
claim to greatness), and 1-2-3 became graphical in its life within a
reasonable time (and mouse-driven and single-application-graphical way
before that). Functionally Excel was pretty awful in its first several
releases, lagging behind 1-2-3 in the feature wars (such as number of
built-in mathematical and statistical functions), but the GUI look-and-feel
and bundling with Word were the primary advantages. No, I'd award this
slot to VisiCalc which was truly revolutionary and, thus, great. Mac OS
already has the slot for the GUI part of Excel.
I wouldn't have narrowed the "Apollo guidance system" to the software
aboard the lander. The Saturn V software (inboard and on the ground) was
remarkable, too.
Sabre is interesting since that begat what we now know as TPF (and z/TPF).
The IBM mainframe has three line items (exclusively), I guess.
I disagree with the Morris worm in a sense because it was an example of a
great bug. Morris didn't intend for the worm to spread as quickly and as
destructively as it did. Do note that it only affected Solaris systems,
back then called SunOS, and DEC VAX systems running BSD. In other words,
only UNIX. :-)
There are some missing ones that probably belong in (or at least near) the
top 12. Grace Hopper's A-0 or FORTRAN probably ought to figure somewhere.
Before compilers and programming languages getting a computer to do much
was horribly difficult, error-prone, and time consuming. It still is, but
programming languages helped a lot.
There's nothing on the list that directly addresses wordsmithing, such as
word processing or publishing. Electric Pencil? Wordstar? WordPerfect?
Aldus Pagemaker? Quark? Photoshop? TeX? There are some great ones in
there, I think. We generate more paper than ever, so that's got to count
for something.
How about Pong or MIT's Spacewar? Isn't entertainment and amusement
important, at least in many philosophies?
Apollo and Sabre aren't the only aerospace-related contenders. The U.S.
air traffic control system, which still owes its roots to the early 1960s,
is another "great software" nominee, I think. The Air Force's SAGE system
for strategic air defense probably deserves a nomination, too. The world
didn't get blown up, so that's pretty great.
I only partially agree with BSD UNIX. The BSD TCP/IP implementation has
long been the "gold standard," so I think I would have narrowed the award
to that -- a fine piece of software engineering. UNIX has otherwised
cursed humanity with such non-greatness as case sensitivity (rather than
solely case retentiveness) and a command line syntax with, shall we say, a
steep learning curve. The approach to file handling (hierarchical slash
paths) and I/O (everything-as-file /dev/whatever) is at least nice. The
kernel is a mixed bag. (Recompile for device drivers? Please! Yes, I
know this is sort of changing, and perhaps completely changed on
microkernel UNIX variants like Mac OS X.) In terms of influence and
spread, yes, excellent choice. Not sure about #1 because UNIX as a whole
is so mixed technically and in human terms, but it definitely should be on
the list.
Speaking of networking, how about the Apache HTTP Server (the analog to
NCSA Mosaic)? True, it's a pretty simple idea (file delivery, basically),
but simple ideas can be great ones.
I wonder if one of the timesharing systems should get mention, like
Dartmouth's, Michigan's (MTS), McGill's (MUSIC), or something else. The
author sort of positions UNIX in that role, but I don't think it really
was, at least in the first many years. Those systems did an awful lot to
bring computing to the masses in the 1960s and 1970s, to educate a whole
generation of computing professionals (and even non-computing
professionals).
We've really got to have z/VM on the list. Terrible oversight there.
Virtualization is hugely important (and great), and there's nothing like
z/VM and its predecessors. That one was (and is) just like Apollo: way,
way ahead of its time in innovation.
I would argue that compression software and algorithms -- and its natural
applications, like MPEG, JPEG, MP3, even the humble fax, etc. -- are hugely
important. We wouldn't have iPods without the greatness that is
compression software. Similarly encryption software. (iPods again, among
so much else.) I don't know exactly where you'd set a marker on those two
to choose a single work that represents the greatness, but I bet it's
possible.
Digital Research's CP/M probably deserves a greatness nomination as the
first general purpose, mass market microcomputer operating system. MS-DOS
is simply a bad copy of CP/M. CP/M had a profound and lasting impact well
beyond its own years.
If Microsoft belongs on the list I think it should get on the list for,
oddly enough, BASIC. Microsoft BASIC really was great software because it
brought a relatively simple high level programming language within a finite
hardware footprint to masses of people in a relatively standard way, and
the quality was at least good enough. Without BASIC I don't think small
computers would have been as approachable as they were in their formative
years. I think Microsoft BASIC was much, much greater than Excel.
Interesting article!
- - - - -
Timothy Sipples
IBM Consulting Enterprise Software Architect
Specializing in Software Architectures Related to System z
Based in Tokyo, Serving IBM Japan and IBM Asia-Pacific
E-Mail: [EMAIL PROTECTED]
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [EMAIL PROTECTED] with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
