#53.
If we’re selecting based on quantity, I withdraw my choice and go with Solitaire instead.
COBOL got used a lot, for sure, but had no influence on the development of computers and computer science except as a bad example.
Or, as Professor Dijkstra said
If you don’t know who he is, I rest my case.
Being used a bunch of places isn’t “influence”.
The prominence of COBOL is vastly overstated. There are orders of magnitude more lines of Java out there. Ditto C-based languages, Javascript, etc.
I remember decades ago telling people who thought that COBOL was the only language we CS professors should teach that they should walk down the aisle of CompUSA* and count how many of the software packages were written in COBOL.
If I said the same thing today but about app stores people would miss the point.
The question is: what influence did it have on the later aspects of the field? Very, very little.
- Like I said, decades ago.
The major influence of COBOL was the lesson in how not to design a computer language.
I second this. Most of the discussion here is about consumer products but without Unix there would a lot of things that would never happen. (I suppose there is an argument for Linux, but Linux is basically Unix.)
Lotus ripped off VisiCalc and called it 1-2-3.
I would go with Windows 3.1as well. Windows 3.0 made waves, but 3.1 was the killer. AFAIK, it intoduced TrueType fonts (can’t remember if 3.0 had them). It also made networking easy. It was simply much easier to work with than anything before. Windows 95 was the next step and showed a new direction, but from then on i was incremental improvements. including stability. Windows 3.0 and 95 were masters of the BSOD, Windows 2000 was way better.
If MS-DOS, then which version? 1.0 and 1.1 got the ball rolling, 2.0 brought in all the useful features such as directories.
Since some of the MS-DOS kernel was used to help boot Windows, then maybe DOS gets the nod?
Oh dear, the word processor wars. I would nominate Wordstar, simply because it was the first widely used and really kicked off the word processor n a PC revolution. I only used it a couple of times on borrowed PCs and never took to it, but it has its place in history. Word Perfect came in on its coattails.
In terms of influence, i would nominate VisiCalc. It spawned every single spreadsheet program ever since.
I am only looking at the PC world. I can’t even begin if you include mainframe of Unix programs, influential as they were, but in a much smaller computing world. I think it is fair to say that the most influential programs are applications used by Joe Public, hence WordStar and VisiCalc.
Next choice; the first browser.
ctrl+q !
The OP already mentioned Civ II, but for me, this was a really important game. It showed me that you might simulate just about anything using a computer. I played it soon after I got my MS in engineering, and it gave me a lot to think about in the years that followed.
Interestingly, I’m back in grad school now, after all these years, and my dissertation, arguably, is about writing every possible (space mission) planning problem as an optimization problem. This sort of dovetails with my early experiences with Civ II.
I completely disagree with the view that Windows 95 was somehow transformational in some fundamental sense, except in its popularity and $$$ for Microsoft. It continued the tradition of being architecturally just a GUI layer sitting on top of DOS – although with Windows 95 Microsoft tried harder to hide it – and as Windows grew in incremental releases it became more and more like an elephant balancing on top of a golf ball. And of course the elephant would frequently fall off. Under any kind of stress such as gaming, Windows 95 would crash at the drop of a hat.
It was Windows XP, as I said earlier, that was transformational because it brought the NT kernel into the mainstream consumer-OS environment, and ended Microsoft’s dual product line strategy of NT for business/scientific computing and Windows 9x for consumers. Finally and at long last, consumers had the sophistication and stability of the NT kernel along with all the consumer features they were used to, and more. Of course Windows 2000 had the NT kernel, too, but that’s because it was NT – it was just the name for NT 5.0. It wasn’t a consumer product, although that was Microsoft’s early design goal, which they didn’t finally achieve until Windows XP.
Let me add one more observation to the other comments about this silliness. The context here is innovation and influence. Nothing in your quote supports this for COBOL over more general and more important languages except the first statement, which is wrong. Computer manufacturers did not all have hardware-specific languages 50 years ago from which COBOL saved them; as already noted, FORTRAN was already an established language when COBOL was designed. The implication that COBOL led the way in this arena is just plain wrong.
Every university computer I have ever encountered, without exception, from the humble 12-bit PDP-8 to the DECsystem 10 mainframe, had some implementation of FORTRAN. None had COBOL*. And FORTRAN was not there for “scientific computing” per se, but for general-purpose research support, including all kinds of applications that had absolutely nothing to do with math. No doubt the administration computer had COBOL, though, the better to lay out the fields in the employee payroll stubs. 
From Technical Innovation in American History: An Encyclopedia of Science and Technology: Fortran allowed software to be produced for all types of hardware and was a pivotal programming language in the electronic revolution of the late 20th century. AFAICT, that huge encyclopedia doesn’t even mention COBOL, although there are pages missing in the Google Books version so it might mention it somewhere. From a computer science technical perspective, COBOL was not so much an innovation as an embarrassment. Yes, it was used a lot in early business applications, but as Voyager said, if “used a lot” is the major criterion, then I nominate Solitaire.
- In fact a COBOL compiler didn’t even exist for the PDP-8, though DEC had some proprietary abomination called DIBOL. Yet available for the PDP-8 was a FORTRAN II compiler, later a full-fledged ANSI standard FORTRAN IV, along with BASIC, FOCAL, a Pascal interpreter and a Pascal compiler, LISP, SNOBOL, and probably many others.
I’d object to this only because people who did design programming languages probably were never exposed to it. I can’t recall even one CS class at MIT 1969 - 1973 that taught COBOL, though maybe there was one in the business school. It never came up in any of my classes. It was never discussed in SIGPLAN notices except to be mocked - the Dijkstra quote I posted was from SIGPLAN notices. COBOL has no descendants as far as I know.
If I were going to nominate the most influential programming language, the choices for me would be Fortran, for being the first, Algol, as I mention before for introducing the structure found in most languages today, and Simula 67 for introducing objects. C++ was quite late in the game. I took a seminar on designing an object oriented language in 1976, and the language I designed and implemented for my dissertation in 1980 used object oriented techniques to model architectural features.
'Course, AOL got its start as Quantum Link for Commodore computers.
While I would agree that VisiCalc and later spreadsheet programs are big, let me throw out another piece of productivity software for consideration. PowerPoint. First out only on Macs back in 1987 and quickly acquired by Microsoft. Anyone who works in an office has had to sit through many bad presentations. It’s often badly used and abused (much like Excel, especially when a database would be better.) People like Mattis and McMaster think it’s a bad thing for the US military. It’s a piece of software that’s been debated for a quarter century.
Powerpoint was hardly a game changer. Long, long before Powerpoint, and long before digital projectors, I was creating presentation decks using whatever word processing software was handy. All you really needed in most cases was bulleted lists and a bit of text formatting. If illustrations were needed, you did them with whatever tools were available. Even if tools were really primitive, you could cut and paste illustrations with, literally, scissors and paste. Then all the slides got copied onto transparencies for use with overhead projectors. Realistically, in most cases when Powerpoint is used to do things well beyond this, like animation or ostentatious artwork, it sinks to about the same level of utility as Microsoft Bob and the Office dancing paperclip.
Yes, you could make slides. And it took time and effort and money for transparencies. But PowerPoint reduced or eventually removed some of the costs. My point is that it’s a piece of software often badly used and with potential costs that might not be recognized or defined and that it’s why it’s important. You can’t tell me you were making a deadly dull set of 150 transparencies for a meeting in the same way that I’ve had to sit through PowerPoint decks that large.
PowerPoint had nothing to do with moving away from transparencies. That was thanks to laptops and reasonably priced projectors.
At Bell Labs we used a macro packing sitting on top of nroff and troff, and if we had laptops running UNIX we could do presentations without transparencies just fine.
My killer robot AI but being held at plasma rif…
The AI is actually benevolent, and wants to give humanity all that it deserves. All hail the AI. beep
Signed
Little Girl
Sent from my SM-G930W8 using Tapatalk
Readers may be interested to learn that the original BASIC was a compiled (Line Compiled) language. And not “compiled to bytecode” like JAVA or .NET. Line oriented languages like BASIC and FORTRAN were replaced by block oriented languages like C and Pascal and Fortran :). And, in an interesting twist, block comments turned out to be not such an exciting advance as first supposed 
In that line, I would suggest WordStar. 20 years later, my programming environment still supported the WordStar shortcut keys – although by then the interface was undocumented. For most of the world, WordStar was the first WYSIWYG word processor. It was the word processor replaced by Wordperfiect. WordStar had show what people wanted, and that they wanted it.
That was Windows For Workgroups, 3.11.
In the same way that Linux was just a GUI on top of GRUB? Actually, no, that’s ridiculous, that would be like saying that Win95 was just a GUI on top of DOS, like Win 1.0. :). MS had made the first steps in their development of their 32-bit operating system with Win32x in Win 3.1 / Win 3.11. Even Win 3 virtualised the hardware to a certain extent, so ‘just a GUI’ wasn’t true even then. *You don’t have access to the hardware under Win95 * – it was transformational in a fundamental sense. *Win95 virtualized the entire machine. *
FORTRAN was also used to write things like your general-purpose accounting package, your stock control system, whatever, etc. It was just the go - to general purpose programming language before BASIC, Pascal and C.
And yes, on Mini-Computer systems BASIC was used to write your general-purpose accounting package, your stock control system, your GIS / Surveying package or your FX (Currency Trading) system.
“Ridiculous”?
Caldera claims that, by monitoring operations on a second PC, it found there was far more than the odd handshake going on between Microsoft’s own MS-Dos and the Windows 95 user interface. It detected thousands of calls, and claims this shows that Windows 9x is still basically a “graphical interface sitting on a Dos operating system”.
Now granted, there are nuanced aspects to this and there was controversy about it. IBM, pushing OS/2 at the time, loved to refer to Windows as a “DOS enhancer” or something to that effect. A bit of hyperbole, but note the above quote.
But here is the point. Notwithstanding the protected mode offered by the 386 and newer class processors, Windows 9x fell far short of supporting their theoretical capabilities:
Both NT Workstation and Windows 95 are 32-bit operating systems for the desktop. This means, in theory at least, they should be free from the constraints that have plagued MS-DOS and Windows systems for over a decade. In theory, the 640KB memory limit should vanish; the problems with handling protected-mode memory should disappear; and device drivers shouldn’t have to compete for scarce resources. In theory!
In reality, only NT Workstation actually delivers on the promise of true 32-bit operations. Windows 95 delivers part of the promise, but deep down inside, MS-DOS still lurks. Those memories of days past are well hidden, which is why Windows 95 can give you long file names and why it can handle 32-bit device drivers. What it doesn’t give you is the robust protection against problem software that NT Workstation can deliver, and it doesn’t give you the rich networking environment.
https://www.itprotoday.com/compute-engines/windows-95-vs-windows-nt-workstation
See also this comparison: “System completely protected from errant Win16 and Win32 applications?” – Windows NT = YES, Windows 95 = NO.
I know this very well since my son was and remains an avid gamer, and back in the Windows 95 days, for many or maybe even most games, a total system crash once in a while was pretty much routine. It sort of became a question of “when”, not “if”.
The problem went away, totally and forever, with the advent of Windows XP. For the first time, no user-level program could ever affect the OS or other programs. For other reasons, like the standardized use of DirectX, games inherently became more stable as well. For gamers, like for everybody else, Windows XP with its NT kernel was Da Bomb. Its introduction into the consumer market was revolutionary!