Slashdot Mirror
The Real History of the GUI
Big Nothing writes "Mike Tuck @ webmasterbase.com has written a piece on the development of GUIs. Like most other articles on webmasterbase.com it is fairly non-technical, but entertaining nonetheless." Update: 08/21 02:45 AM GMT by T : Note that the link above takes you to the print-friendly version of the story; for online reading, you might prefer this version instead.
fine, you click on the "Information" link and see where it takes you.
|---------------|
practically an AC
Once upon a time, way back in the Stone Age, lived two cavemen, Ugh and Glug.
Actually, the two cavemen were named "Ugh" and "Slug."
I hope this clears things up a bit for everyone.
I got my Linux laptop at System76.
The answer is you are using Netscape 4.7 under *nix and have not done anything to fix your fonts. Upgrade to Mozilla tweak your fonts and it should look much better.
Cypherpunks: Civil Liberty Through Complex Mathematics. Those who live by the sword die by the arrow.
Nice to see they mentioned the good ol' Atari ST! I managed to get one a couple years back and it's quite possibly the coolest antiquated piece of technology I own today, and the GUI is very, very impressive for the time. It works, (although finding discs for the 720k floppy drive is a pain in the ass) and has one of the best versions of Monopoly available today, not to mention that the OS is just as stable as Win98, and about 1/500th the size. Yeah, yeah, it doesn;t support everything, but then again, who wants a USB printer on a 16-year-old machine?
/* extreme sarcasm */. . . . .
Just need to figure out a hack to hook it into my network now
- Relativistic? That's barely Newtonian!
The world would be a better place if GUIs had never been invented.
Give me an Xterm, emacs and lynx over a point-and-slobber interface anyday.
* the first GUIs were written in smalltalk.
*smalltalk was the first cross platform portable bytecode language
* smalltalk is good pure OO unlike C++ (not pure OO) and Java (not good OO - the class hierarchy is a monster - ST's object heirarchy is much more clean)
I can't stand articles on technical subjects that include gibberish about "cave men" and the like. Who-needs-it? Obviously, the author is padding for his lack of knowledge about the subject.
I was going to rip this article a new one, but i'm glad they got it right. What I would consider to be the first GUI was Sutherland's "Sketchpad" system from the early 60's. The military had similar sorts of things predating Sutherland, but nothing quite flexible enough to really be called a full blown GUI.
Anybody with their brains in the right place can tell you that the GUI was not invented by Xerox PARC. They may have done a great deal to push the idea, or perhaps simply been at the right place at the right time, but the basic idea of using graphics as a means to interact with a machine predates PARC by about 20 years.
If you really wanna have some fun, check out Doug Englebart's 1968 presentation that introduced the world to the mouse, chordboard and other interesting stuff. There are plenty of links to it, but here's a good one incase you cant find any. A while back, there was a site that offered his entire presentation in RealVideo format, IIRC..I wish someone would post a link to it, or perhaps a better (re: DivX, or straight MPEG) link... It almost brings tears to my eyes when I watch it.
Bowie J. Poag
[random thoughts]
The GUI became popular because it really made most things that users need to do on a computer far easier than cryptic command lines. For years GUI's have been refined for ease of use. We're now coming to the limit of current icon-oriented design. There's just so many ways that an icon-based system can be presented to the user before usability starts going into the toilet.
We moved to GUI's because command line interfaces only got us so far, and some day someone will come up with a better-than-icon based system that is more logical. We'll all say "gosh, why didn't I think of that?" and everyone will jump into the "new way" of thinking.
I have visions of time-oriented interfaces that respond to "get me the spec sheet for the network I did last week sometime" and "set a new meeting for next Tuesday with Jim and Bob in the conference room". These new interfaces will be able to store and retrieve information based upon how we think, not in the traditional tree-like-structures we're currently used to. The concepts behind OO/RDBMS systems have some potential, such as nested tables and object oriented models, but don't present their information to a user very easily.
I don't see new interfaces becoming popular until they target the non-computer user market. I envision voice-activated systems, but they tend to be annoying to other people around. Mouse navigation doesn't seem to be viable because of it's limited 2D space, and thus the 2D GUI. The 3D systems (see spaceball on google) look neat, but aren't very intuitive to users. We may wind up with virtual filing cabinets, but hopefully we'll stay away from the Packard Bell Navigator!
Is there anyone (university or other) that is working on a new interface concept? I'd be interested in hearing what works and what doesn't. I know M$ and (Cr)Apple invest millions into GUI research, so I wouldn't be surprised if we saw something new out of those camps in the next few years.
And no, I don't count XP's "new and improved" GUI anything more than an over-hyped icon-based system.
[/random thoughts]
LOAD "SIG",8,1
LOADING...
READY.
RUN
I prefer a decent command-line interface within an ergonomic GUI, i.e. best of both worlds. Windows definitely benefits from the addition of this . The shortcomings of the Windows CLI never cease to astound me. For instance, a command-line is not very functional without a decent egrep-like tool, IMHO.
It's amazing to see how so many beautiful and wonderful things happened as a result of two guys in the backseat of a car.
Wait a minute...
--SC
You read fiction? I write it! Lemme know what you th
OK, I've heard this quote a lot, and while it's hilarious, I have a sixth sense that says Bill Gates never actually said it.
Now, I've spoken to people who were there to hear, firsthand, Bill at a big computer show (SIGGRAPH?) in the early 90s, remark that "it's impossible to write a preemptive multitasking OS that runs in less than 4MB RAM" - while there were computers on the show floor doing precisely that. But that's not quite the same quote, and though I can imagine it evolving over the years to "640K oughta be enough" I just don't think that's what happened.
Any ideas? Anyone know where the quote supposedly appeared?
~ radiographite: art by john shepard
Out of curiousity, what is the site itself generally about? I thought it was a mainstream media site, but now I see there is a bit of complex coding discussed.
You know, though, in probably less than a hundred years that this will no longer be considered a joke. :)
Well, I beg to differ. You could say I've kind of been enlightened after listening to the epitome of computer cluelessness: my mother.
She was struggling with the Windows explorer GUI, trying to move a file. And then, she said, and I'm not kidding: "Oh, I prefered DOS, you know, you typed a command, and it worked!"
Maybe what simplicity is really about, is determinism in the way the computer behaves?
-- B.
This sig does in fact not have the property it claims not to have.
My mother made the same complaint. She was a touch-typist, after all, and hated having to move her hands away from the keyboard to use the fscking mouse.
Cut the ST some slack, man! It can't be that bad.
Oh, wait...you meant that as a compliment : )
The GUI is a tool, just like the CLI is. It's a tool that works if it is done right, and once a tool is done right, theres very little to no reason to reinvent it.
The "Classic" MacOS GUI works very well, yea maybe it's ripped off from Xerox, but whatever, it works. It's not themeable (well kinda...) but in the System 7 and OS 8 versions it was simple and it worked very well. So well that 3 year olds and 80 year olds could master the interface quickly and without help.
Aqua is just a step beyond the "Classic" GUI and with some refinement it could last another 20 years.
However...I do not agree that a new interface concept is needed. A screwdriver's interface has remained unchanged for centuries and it doesn't need a new concept. Same with the firearm, a Beretta flintlock from 1300 had the same interfact characteristics as a Beretta Gold Sable rifle made in 2001.
When a concept works...don't spend the effort on changing it. Spend the effort on making the OS run better behind the interface.
The irony of this article, and all the articles it links to is NO PICTURES.
Not even the history of computer graphics article has any pictures. You'd think that was a sure bet.
Since this Israeli company is trying to get rid of Windows, then we won't have a GUI anymore. We'd have a SUI (Speaking User Interface). Since Sooey sounds like a farm animal call or an oriental dish, it can't be used for geek-speak. Therefore we must abandon all research towards making computers talk to us since it will never be adopted by the geeky ones.
He refers to X as an OS, it is not an Operating Sytem - it's a Graphical Environment (and even that's putting it simple).
Also, Windows/386 - which was a full 32-bit version of 2.0 was the first Windows to take advantage of the 80386's features. He states that Windows 3.0 was, although it was actually an enhancement to W/386 that dropped support for the 80286 and relied exclusively on 32-bit mode.
He also skipped right over IBM LanManager, which was the precursor to OS/2.
OK, enough nitpicking... I guess the Ugh and Grub or what-have-you got to me more than I thought.
I AM, therefore I THINK!
Umm i dont understand
The fact that Altair Ran Basic (which at that time was considered and OS as there was no such thing as a Disk Operating System for the Altair as yet) is proven to be factually correct ? And regardless Gates wrote the First DOS for the Altair himself (on legal pads in fact and then coded it into a machine in machine language) so the statement is factually correct even if you read it as BASIC/OS
what are you trying to say then ? That history is lying (inclduing numerous non MS sources ?)
I refuse to argue with Anonymous Cowards - if you want a discussion get an account....
At last, somebody actually gets this right:
"Apple negotiated a deal with Xerox; in return for a block of Apple stock, Xerox allowed Jobs and his team to tour PARC, take notes, and implement some of the ideas and concepts being bounced around at PARC in their own creations."
Pirates seriously messed with history in this regard, having never touched on the deal Jobs made with Xerox, and the made-up commentary by the "Wozniak" character.
But on the downside, the author doesn't spell Jef Raskin correctly.
Scott Stevenson
Tree House Ideas
The Alto itself didn't really have a GUI. What it had were graphical applications. One of them was Smalltalk, which had its own private windowed environment. Another was Bravo, the first WYSIWYG multi-font editor. You could write other applications yourself, in Mesa, Xerox's own language, or BCPL, in which the underlying tools were written. The underlying environment was a single-task command line environment comparable to early DOS.
Bravo was used as the programmer's editor. The internal representation of Bravo documents was ASCII text, followed by a control-Z, followed by the formatting information. The command-line tools, which understood control-Z as EOF, could thus happily process Bravo documents. Programs for the Alto were normally written in proportional fonts, with boldface and italics as needed.
The Alto hardware itself was built by Data General under contract to Xerox. It was basically a Data General Nova with custom microcode in a desk-height rackmount case containing the computer and a 14" removable disk cartridge drive (equivalent to a DEC RK05, if anybody cares.) The display, a portrait-mode b/w full-page display built at PARC, was the main hardware innovation, along with the 3MB Ethernet and the mouse.
The Alto project had several components. First was the concept of a number of single-user workstations connected by a network providing dedicated services. Each Alto had very limited disk space, but file servers were available. All serious storage was on the file server. There was also a print server, an Alto connected to a modified Xerox copier. PARC management was working on the assumption that, although all this was far too expensive to deploy, in time the hardware would get cheaper and it would be useful. They were right. The fact that they then blew the business aspect wasn't PARC's fault.
The other big push was Alan Kay's Dynabook. Kay was big on simulation and teaching kids to use computers. His real direction for Smalltalk was what we today disparaginly call "edutainment", games with educational intent. This seems strange now, but that's where he was going. He's continued with that direction, at the Media Lab and elsewhere. But it never took off.
PARC tried to commercialize the technology as the Xerox Star. This wasn't a general-purpose system; it was more like a really good dedicated word processor. Wang then ruled that market, with what was called "shared logic word processing", dumb terminals all running one common application on a time-shared host. This was cheap enough offices could afford it.
The Star, with a real computer at every desk, a big display, and a LAN, did roughly the same function, but at higher cost. It was cool, but didn't sell. It was a closed system; you ran the Star app, and that was it. PARC didn't trust the users to mess with the system, so you couldn't do anything they hadn't anticipated.
The computer scientists at PARC didn't see that the future was open systems running unreliable software. Really. That was the missed vision. Nobody dreamed that something like DOS, let alone non-protected mode multitasking, would end up in clerical offices. Obviously, it would break all the time, files would get lost, and the cost to the organization would be enormous. Remember, Xerox was a rental company back then; if the copier broke, it was Xerox's problem. So they took reliability very seriously. And, sadly, it cost them.
...than "move oldlocation newlocation" how?
The point wasn't that the guy couldn't explain how to do it, it was that it had to be learned, just like the old interface. On the whole, it's not more intuitive it's just cuter.
---
You'd be surprised at the broadband connection available to things crawling around in your hair.
Checked my copy of the dos 3.0 manual and i cannot find it - i suspect the quote is apocryphal for that and other reasons mainly related to the fact that by the time DOS 3.0 came out Apple was above 640k anyway - and this would not be consistent with the then Microsft business plan of sellling as much software as possible to build a business
I refuse to argue with Anonymous Cowards - if you want a discussion get an account....
There actually is a guy named David Gelernter who came up with something like this. Because of his computer science background, was a unabomber victim. While he was recovering, he came up with a system called "LifeStreams" that would record data throughout a person's life as if their entire life was some sort of filestream that is constantly added to.
t re ams_pr.html
http://www.wired.com/wired/archive/5.02/fflifes
While the icon metaphor is limited, part of the problems people are having with it are not so much related to the design itself, but the fact that so many programmers do so many cognitively unsound things that shouldn't be done in any interface design on any platform. And this is what is really causing many users to suffer through today's desktop interfaces. For example, some programmer might implement a button layout where it is not clear how one widget relates to another. One button on one side of the screen may have some relationship to a list that is in some obscure location somewhere else on the screen (as opposed placing the button right next to the list it acts on). Or one program might have both the menu selections "Customize" and "Options", which is ridiculously confusing for the user because both words refer to the same exact type of thing (configuring something in a program) but perform different actions. I'm not pulling that particular example out of my butt--I'm taking it directly from Microsoft. Before we eliminate the icons, we need to eliminate many programmers' lack of understanding about how to create usable interfaces. If we don't do this and simply go from icons to something else, they'll just end up making the next great interface as equally miserable as the current one.
"Talking about music is like dancing about architecture."
Can we get a "-1 Wrong" moderation option?