What Early Software Was Influential Enough To Deserve Acclaim? 704
theodp writes "That his 28-year-old whip-smart, well-educated CS grad friend could be unaware of MacWrite and MacPaint took Dave Winer by surprise. 'They don't, for some reason,' notes Winer, 'study these [types of seminal] products in computer science. They fall between the cracks of "serious" study of algorithms and data structures, and user interface and user experience (which still is not much-studied, but at least is starting). This is more the history of software. Much like the history of film, or the history of rock and roll.' So, Dave asks, what early software was influential and worthy of a Software Hall of Fame?"
The original Lotus 123 (Score:5, Interesting)
If you want groundbreaking early Mac software (Score:5, Interesting)
I'd say HyperCard [wikipedia.org] would be a better choice
For mechanical engineers/designers (Score:5, Interesting)
Autocad & PowerDraw (now PowerCADD) 2D CAD followed a decade later by SolidWorks 3D for turning concepts into executable designs that were within the realm of price and usability for individual designers.
Influential? (Score:5, Interesting)
dBase
Word Star
Turbo Pascal
Comment removed (Score:5, Interesting)
A few that have grown awful over the years... (Score:4, Interesting)
Here are a few that were great in the beginning but have become bloated and kind of overbearing since:
Word 4.0 for Mac (fast, stable, good UI, nearly perfect)
Photoshop 1.0 and then 3.0 (when they added layers)
Early versions of Excel (for Mac, then later Win95)
FreeHand (when it was Aldus)
PageMaker (when it was Aldus...see a pattern here?)
Aldus Persuasion (notice I didn't say PowerPoint?)
iMovie (compare to any version of movie editing software bundled with Windows ever...no contest)
Honorable Mention: Garage Band (too niche to be mainstream)
important bits (Score:4, Interesting)
Algol-60. RT-11. TECO. Hypercard (count this one twice!).
Article summary: "I am a Mac fanboi" (Score:5, Interesting)
OK, maybe that's a little harsh. But it's not completely apparent what value such a detailed review of early software programs would add to a computer science curriculum. It's probably sufficient to note the emergence of the GUI as the major defining element here, and let our poor undergrads get back to studying their bi-directional linked lists.
My opinion: it's not an accident that computer science is a more forward-looking than backward-looking discipline. Students will get more mileage out of downloading the latest version of OpenCV or playing with math in Python than sitting through a boring lecture about primitive computer software apps.
Software was written before 1980... (Score:4, Interesting)
How could they usefully study such software... (Score:2, Interesting)
...when the source is unavailable? I can see that these programs might be mentioned as examples of early efforts in a course on UI design, but what else is there to say about them?
POV-Ray (Score:4, Interesting)
This introduced a lot of people to 3-D rendering, and the free-enough license led to widespread adoption.
Freehand, Pagemaker and UltraPaint (Score:3, Interesting)
Re:VisiCalc (Score:4, Interesting)
VisiCalc was actually credited by a few business journalists in the 80s for starting the whole corporate raider business. They were now able to plug in all those numbers from SEC filings and other sources into the spreadsheet, run simulations of financing and figure a way to take the company over and make their billions.
They also used it to find out if the pension fund was over funded. See, back in the old days, companies would invest the pension in very low risk things like government bonds - at like 3%. The raiders said, "Hey wait a minute! If we put the money in the stock market, it could make 10% a year - because that's what it averaged for decades! They don't need all that cash in their and we can use it to finance the deal and pay our "consulting fees"!"
Flash forward to the '00s, and pensioners are getting their benefits cut left and right or they are just gone.
KKR, Icahn, T Boone, and Bain Capital (of Mitt Romney fame) were and are some of the players.
Now, many of those folks don't have the money that they counted on - their deferred compensation. Another way of putting it is those folks weren't fully paid for their work.
Under-appreciated (Score:5, Interesting)
Microsoft BASIC and later Visual Basic: Unjustly despised, but introduced many to programming (and the very first ones were marvels of micro-programming too). Also interestingly portable at a time where portability was on nobody's radar.
Spectre GCR, a Mac emulator on Atari ST. A precursor of virtualization in my opinion, and a very smartly done one at that.
VMware for making virtualization available to the masses and enabling the cloud.
AmigaDOS for being the first OS with built-in hardware-accelerated graphics and sound.
The RPL system in the HP28 and HP48 series of calculator. Reverse Polish Lisp and symbolic processing on a 4-bit calculator with 4K of RAM? Seriously?
The Minitel system in France, including nationwide phone directory and dubious innovations such as Minitel Rose (porn in text mode at 1200bps, basically).
Postscript and the whole desktop publishing revolution.
NeXTStep (or whatever the CorRect CapItalizATION is), so far ahead of its time that it took years for it to reach its full potential in the form of iOS.
GeOS (already mentioned by someone else)
Mathematica. Just wow. But also forgotten precursors such as TK! Solver.
Lisp, Fortran, Algol, Pascal, Ada, Eiffel, Smalltalk and a whole bunch of under-utilized languages.
Much lower on the name recognition scale, Alpha Waves [wordpress.com], arguably one of the earliest real 3D games, which also influenced the creation of Alone in the Dark.
Re:Lets create a list ... (Score:2, Interesting)
Crosstalk (the pricy RS-232 comms package)
Kermit (the open-source RS-232 and later network comms package)
Fastback (PC backup utility)
Norton disk explorer (disk drive maintainance)
Brief (another PC editor)
GED (another PC editor)
Fract386 (fractal explorer)
PHIGS (early 3D CAD library)
SRGP (Simple Raster Graphics Package)
Re:VisiCalc (Score:4, Interesting)
'nuff said
And as a hardware corollary, the 80 column video card that allowed visicalc to show a useful amount of screen real estate.
XtreePro (Score:5, Interesting)
That and Norton Utilities made DOS useable.
But XTP's superlative use of the screen area and hotkeys was stunningly competent.
Re:VisiCalc (Score:5, Interesting)
Before Wordstar, there was Electric Pencil. I also compared Apple Writer II vs. Wordstar for a technical presentation in some college class in 1982; I declared at the time that Apple Writer was far and away the most advanced and user-friendly WP on the market.
I find it amusing that some 30 years later, some of the old Wordstar keyboard shortcuts are still used in some programs today -- notably alt-X, ctrl-Y, and F1 still do essentially the same things they did in Wordstar.
I think someone else mentioned Colossal Cave, and yes indeedy -- CC begat Zork which begat the rest of Infocom's amazing library, which I still play from time to time today. My 20-something daughter just the other day complained about the difficulty of getting the babel fish in your ear! Tell me, Microsoft, what games of YOURS are still being played 20 to 30 years later?>
RUNOFF (Score:4, Interesting)
Re:Article summary: "I am a Mac fanboi" (Score:5, Interesting)
Napster...MS GWBasic...Windows 3.0 (Score:4, Interesting)
Napster - this is the software that kicked off the idea of music file sharing. Okay, the record companies hated this program but this is the first program that I can think of that really CONNECTED people as a group on the internet for exchanging data.
MS GW Basic - this was the basic that shipped with the IBM PC and was pretty much what much of its early software was written in because it was so simple to use and yet could be used to do quite a bit.
Windows 3.0 - This was the first version of Windows that people really used and really brought the GUI desktop with the mouse into the mainstream. Okay, the first Macintosh from Apple did that too and came before Windows 3.0 by a ways but it was not nearly as widely used, especially in the workplace.
Wrong premise (Score:2, Interesting)
As usual by self-centered people, rather than assuming your own knowledge should be known by everybody, why doesn't Dave Winer ask the teachers in charge of CS degrees why they are not teaching the software he presumes to be so valuable? Maybe then he would really learn why his comparison to shakespear sucks so much.
In fact, the whole logic is wrong, he first states knowledge of these programs should be mandatory, yet asks other people for knowledge about mandatory programs "he doesn't know". So they aren't mandatory if he doesn't know them? Otherwise why is he proclaiming himself the end all knowledge of what CS students should know?
The word for today is: confusion.
FORTH (Score:4, Interesting)
C compiler (Score:5, Interesting)
Most valuable program(s) ever. From day one, and still today. Hands down. Best positioned language in terms of "to-the-metal", changes from tool to uber-tool in the hands of anyone who masters assembler and arrives at learning C with that under their belt, can create extremely fast executables if the CPU is really taken into account, or can be extremely simple to implement if a CPU is treated simplistically -- yet your code will still work fine, if a bit more slowly. Made portability something achievable instead of just desired. C is so well positioned that implementing the language's constructs on top of [some random] CPU is a relatively simple exercise, and then you have immediate access to oodles of goodness.
Also the source of a lot of whining and bad programming from poor programmers. But hey, a fine carpentry set doesn't make you a great carpenter, either.
Also a nod out to standard libraries -- also a boon to portability and more.
C++, oC, C#... also worthy of nods, but C is the king.
VSAS (Score:5, Interesting)
Variation Simulation Analysis Software.
It's a technique for simulating variations in product assemblies. Usually mechanical, but could be of other natures, as well. You model the assembly and it's manufacturing variations, and then "build" some quantity of parts. One can determine how many assemblies will likely meet specifications, the major contributors to out-of-spec assemblies, etc. etc.
The technique was developed during WWII at Willow Run Labs, where it was implemented by the classic "banks of women operating calculators", and is one of the reasons we were able to crank-out all those airplanes that actually worked.
By the 70's it was implemented in an academic setting on mainframes.
A company I worked for obtained rights to VSAS and we ported it to the IBM PC. I did the initial port to Watcom Fortran (there's another one for you!), and then designed a domain-specific language (VSL) and implemented a compiler in C and interpreter in Fortran, so that mechanical engineers didn't have to write their models in Fortran any more. The Fortran models were bulky - with line after line of function calls with zillions of parameters, passing separate X,Y,Z values in the calls. I'd imagine the engineers wore-out the parenthesis keys on their keyboard pretty fast. VSL, on the other hand, had data types for points, lines, vectors, planes, etc. Using an interpreter didn't slow things down, because most of the time was spent in geometric library routines, which were in carefully-optimized Fortran.
I insisted on their hiring a mathematician, and between the two of us, we tweaked it to run faster on the PC than it did on the mainframe. (Engineering professors don't write code that is either fast or mathematically-correct, it turned out...)
And that's when it's use took off. The company founder started as a manufacturer's rep for some Finite Element Modelliing software, so had lots of contacts in the auto industry. (And the company was located near Detroit.) They both sold the software and did also did in-house projects for the auto companies until they ramped-up their own engineers. This allowed the auto makers, for example, to start treating windshields as structural elements (because the hole for the windshield could be manufacturered to precise tolerances), and allowed them to eliminate costly alignment operations, such as when fitting hoods.
It's used by every auto and aircraft manufacturer, every hard disk manufacturer, etc. etc. etc. Basically just about any complex mechanical product you touch was touched by VSAS during design.
I'd imagine you couldn't build an iPhone at an affordable cost or with the quality level of an iPhone without VSAS (or it's equivalent). You wouldn't be able to buy a terabyte hard drive for less than $100.
There's more info on it here:
http://www.plm.automation.siemens.com/en_us/products/tecnomatix/quality_mgmt/variation_analyst/ [siemens.com]
(The company was acquired by Siemens many years ago.)
Maybe not quite what this post was looking for, which I think was more consumer PC software. But it runs on a PC and has from the beginning of PCs, and has had a large but mostly-invisible influence on just about every tech product we use every day.
A 30-year run is nothing to sniff at, either.
I met Michael Shrayer, author of Electric Pencil (Score:3, Interesting)
I quietly asked my friend if he would ask the homeless person to leave; maybe there would be a concern about theft. My friend laughed, "That's Michael Shrayer [wikipedia.org], he wrote Electric Pencil, he's a multi-millionaire".
Re:VisiCalc (Score:4, Interesting)
"CS is not about software development, it is a branch of mathematics."
That depends entirely on what college or university you are attending. The definition is still pretty much dependent on the school. Although it has been getting somewhat more consistent.
However: at least in the U.S., computer engineering is definitely NOT a software discipline. It is engineering of the computers themselves, that is to say, hardware (though firmware is involved, naturally).