|
 |
Articles: The History of The GUI
|
|
The History of The GUI
How did you get here today? This isn't a trick question and I'm not expecting a philosophical answer about God or the big bang. I'm not even interested in which bus or train you caught to get to work, how you got to your desk or any of that rubbish, but how did you navigate to this article? Unless you're using some bizarre, archaic piece of hardware, the chances are you navigated to it from your web browser on either your desktop or laptop computer (apologies if you're reading this in a book or on actual physical paper, but stick with me, you know that I like to ramble off topic a lot). For the purposes of this article though, we need to look lower than just firing up Safari, Firefox, IE or whatever your internet browser of choice is. Ok class, back to computing 101.
You turned on your computer, the operating system booted up, and, 99% of the time, you found your desktop displayed onscreen with the little mouse cursor all ready to do your bidding. Whether you're using OS X, Windows, Linux or something a little more obscure or defunct, the world of computing is now mostly accessed through mice, pointers, icons and windows. Did this happen overnight? Of course it didn't and this was part of the reason behind this article - to look at how we, as a planet, reached the stage where 'using a computer' involved moving a little arrow around a screen. There are other reasons (one of them being boredom) but we'll come back to that later.
To most people a modern computer relies heavily on a mouse and a graphical display in which you can move windows about, open files and documents by double-clicking them etc. If we lift the cover on the computer though, it's nothing more than a pile of silicon chips, wires and circuit boards (with maybe the odd hard disk, CD drive etc. thrown in). So how does a lump of plastic with a wire coming out of it manage to control this pile of components? The answer is, in simple terms, the operating system - a piece of software that makes your pile of components usable. The operating system acts as a buffer between us and the hardware. It translates what the hardware can do into tasks that we want it to do. It's the factory supervisor that tells the workers to do what we (the company directors) want them to do.
It's easy to think that an operating system must have a mouse, and it must have a desktop, and it must have... The truth of the matter is that the operating system doesn't have to have anything but, as time has passed, what has become accepted as the 'norm' has forced operating systems to incorporate features and approaches that have turned them into the ubiquitous mouse driven systems that we know and (sometimes) love. Sadly it wasn't always like this...
The Dark Ages...
At the dawn of computing history, machines were little more than intricate bundles of wire that could perform 'miraculous' tasks. They could compute missile trajectories, calculate payrolls, manage tea shop empires, and perform a myriad of other tasks. This was the public perception, but the fact was, early computers were behemoths. The likes of ENIAC and LEO could compute figures far faster than any human but telling them to do it was far from simple. Professor Big-Brain's new WUNDA-VAC computer may have been able to add 4,000,000 numbers together in a fraction of a second, but how could you tell it to subtract 4,000,000 numbers in a fraction of a second?
The big step forward in 1940's computing (and even in the 1930's if we look at Conrad Zuse's work) was to create machines that be told to perform generic tasks rather than a single specific task - the result was the programmable computer. You could build a machine to perform a task, switch it on and it would do it. A programmable computer allowed you to specify which task(s) to perform and then specify a whole new set of tasks to perform. But how to specify it? ENIAC required that operators physically re-wired it but the likes of UNIVAC were far more sophisticated and could 'read' their programs in while they were running. But how could you do this? Early programmers fed this information into their machines by toggling switches, and physically inputting every digit of the program. Hardly a doddle and very prone to mistakes but, in the fledgling computer industry, the only solution, and the only way that man could 'talk' to the machine.
The situation had improved as the industry had moved forwards into the 1950's and programs could now be fed in via magnetic tapes or punch cards, with results being output on printed pages or via a (seemingly) vast array of flashing lights. This may not have been an ideal solution but it opened computing up to a slightly wider user base. Now at least users didn't have to understand the complexities of binary sequences, the intricate language of 1's and 0's...but it was hardly Windows XP or OS X. What it also wasn't, was interactive. A user would typically prepare their stack of punch cards and take them down to the computer ("A computer in a desk? Why the very thought. This baby's the size of Norway!") where they'd be fed in, processed and any results output. The user never interacted with the machine itself and certainly didn't have any opportunity to intervene, feed in new data or affect how their programs and data were processed. Let's just say that it was very much hands off.
As the technology improved, so too did the need for new methods of interaction between man and machine. The cathode ray tube (CRT) and the keyboard proved to be the next big leap forward in terms of interface and users could now deal directly with the machine. No longer did 'computer work' involve a stack of punch cards and a hole punch. Instead users could type in commands on the keyboard (seeing the characters appearing on screen as they did so) and then see any results displayed before their very eyes. This was the age of the command line and, compared to the 'batch process' approach of before, was a God send. Programs could be developed which allowed the user to input data and values as the program ran, allow non-technical people to harness the power of computing, and even allow one user to instantly communicate with other users. This 'hands on' approach to computing was fantastic when compared to the previous methods and it opened up computing to even more people. To some though it was still too complex.
But It's Still Too Hard
The military may have pushed the development of computing during WWII, but big business had quickly caught onto the possibilities that the new technology offered and companies across the globe (although mainly in the US) soon started to embrace the power of computing. In the 1960's though computing was still the domain of the intelligent and while typing a command in at a command line wasn't difficult, remembering what the commands were and what they did was still a daunting task. Many people believed that computers would be able to understand natural language and while typing 'open the reports for this financial quarter' was obvious, the computer required that you translate this into a language that it could understand: 'fopen finq0304.doc'.
This 'translation' from natural language into 'machine language' was logical to scientists and those of a technical persuasion, but it didn't make sense to the non-technical. This interface between man and machine had evolved a long way already but it had to evolve further and it was the machine that would adapt to man, rather than man adapt to machine.
The Graphical User Interface - GUI
The command line may have seemed revolutionary when it was first unleashed, and for many years and machines it was the ideal balance between usability and processing power (especially in the days of early home computers, which barely had enough power to perform even the most simple of tasks).
Machines sporting CRTs had brought computing to people and people now interacted directly with the machine. No longer was the computer a (very) big box in the basement which users had to physically go to, but instead it sat on user's desks and was an interactive tool for them. Business was making the computer a business tool and it therefore had to exist in a business world and work in a way that business people understood. The problem that faced researchers at the time was how to make computers work in the same way that people were used to working. The breakthrough was to look at how people actually worked and then adapt the machine to mimic this (as best they could).
On most desks (at the time, things have radically changed since then) a 'desk' would typically consist of an in-tray, an out-tray and filing system (of some kind - maybe a filing cabinet, maybe a small Rolodex). If the person wanted to look what was in his in-tray he/she simply leant over and looked. If he/she wanted someone's address, they simply looked in the filing cabinet. What the person didn't do was try to remember a cryptic command and an obscure filename, or have to stop doing writing a letter just so that they could look up a phone number.
The GUI transferred the physical elements of the office desk onto the screen so that the user could quickly identify an environment that he/she knew and understood. Just as in the 'real world', he/she could write a letter, look in the address book to get an address, and never have to worry about which file or program they were having to run. The computer and the operating system could worry about all of that and let the user concentrate on doing what they were supposed to be doing.
Over time the GUI would evolve, gaining new features and standards (and even imposing standards on users - totally in contrast to the original development of the GUI) but as hardware became cheaper and therefore allowed more complex and powerful operating systems to be run on them, the GUI was recognised as the (so far) easiest and most usable interface between man and machine. Now desktop machines all sport a mouse and a window based operating system, and users, both experienced and new, can happily use the power of the computer without having to delve into the arcane languages and commands of yesteryear, or worry about which pair of connections need wiring together.
Back To Year Zero
The inspiration for this article came from a discussion that started off about a certain computer manufacturer being accused of 'borrowing' ideas that had been supposedly developed by someone else several years earlier. This snowballed into how Microsoft had 'borrowed' ideas from Apple with ragards to Windows and Mac OS, and how Apple had 'borrowed' ideas from Xerox. Then there was the statement 'Xerox invented the GUI'. Oh how wrong you are... This brings us onto the second driving force behind this article - who did invent the GUI? Yes dear reader, all that rambling about command lines and punch cards was just a pre-cursor, it's here that the real fun begins...
Where Did It All Begin?
If you say 'GUI', most people will think of Microsoft Windows. Some might think of Apple's OS X or Mac OS. Probe back a bit further and you'll probably find the Apple Lisa and maybe Xerox basking in the limelight. The truth though goes back much, much further.
You could argue that the GUI was originally invented by two cavemen sharing an idea by drawing shapes in the sand. For our purposes this is probably a bit too far back but the history of what we recognise as the GUI does go back to before the development of the first digital electronic computers. In fact the far-back-o-scope takes us all the way to the 1930's and a gent by the name of Vannevar Bush.
Born in 1890, Bush found himself working in the field of analog computers by the 1930's (Note: Although analog computers were far in advance of their digital namesakes, analog computers (aka differential analysers) relied on gears, cogs and other mechanical parts to solve equations, and would quickly fall from favour in the coming years). This exposure to the power of computing (albeit in analog form) and various other work in the field of electronics led Bush to publish As We May Think, a revolutionary article that detailed a theoretical machine (which he named "Memex) that would enhance human memory by allowing the user to store and retrieve documents linked by association. In 1945 though (when the article was published) there wasn't a machine anywhere in the world that could come close to offering this type of power or functionality, and As We May Think ended up being little read or discussed...at the time.
Bush had theorised this 'wonder' machine that was a radical departure from the conventional approach to computer use (not that there was much thought to computer use - the entire industry was still very much in its infancy with just a handful of machines worldwide) but it's influences wouldn't be felt for many years to come. For the next 15(ish) years the industry had other avenues to explore.
SAGE Council
The early computers were primitive in every area and user interface was no different, but an MIT team led by Jay Forrester set out to do something a little different. With the second world war having only finished two years earlier, the bulk of spending on the nascent computing industry was still coming from military channels and the US Navy approached MIT to develop a flight simulator that could help in the training of bomber crews - Project Whirlwind. Unlike similar systems of the time, this new project would be computer driven so that a variety of aircraft could be simulated and Forrester's team initially looked at using an analog computer but soon found that it was both inaccurate and inflexible.
Inspired by ENIAC, the team looked at the possibility of using a digital computer and this seemed to be far more possible and practical. There were hurdles to overcome though. ENIAC, EDSAC, UNIVAC, LEO and just about every computer of the time used the 'batch' approach to computing (in that data was fed in, the computer processed it and out popped the results). Whirlwind couldn't do this as it needed to constantly respond to user input - a concept that had never been considered or implemented before.
Although started in 1947, Whirlwind didn't come online until 1951 and the initially enthusiastic US Navy had lost interest (especially as it was costing them the best part of $1m per year). Whirlwind didn't face a great future but the Communist threat soon gave it a whole new lease of life, and the USSR's detonation of its first atomic bomb forced the US Air Force (USAF) to quickly become interested.
What made Whirlwind special (in the eyes of the USAF) was not only its ability to process data continuously (rather than in 'batch' mode) but also its use of a round cathode ray tube (CRT) to display the information to the user (far more useful than printouts or flashing lights) and a light-pen to select objects onscreen. It wasn't ideal though and Whirlwind was simply too slow (running at about 20,000 instructions per second) but Forrester's team replaced various components (notably the slow Williams memory tubes with 'core' memory) and Whirlwind II managed to double performance in just two short years.
Whirlwind may have worked but in 1954 it was still little more than a research project (a research project with some very interested parties, but still a research project). The race was on to create a military grade version and IBM eventually won the contract (RCA originally won but it was handed over to IBM later) to produce the AN/FSQ-7 which would be at the heart of the US's Semi Automatic Ground Environment (SAGE) defence system. SAGE was a massive project that pushed the limits of computing, not only in terms of user interface and speed but also communication as SAGE was not just a single site or a single machine - installations were spread across the country and every site had to be linked together to provide a complete defence system.
Sadly SAGE did not go onto greatness as, although it pushed technology, it hadn't allowed for it. Designed to identify incoming bomber attacks, the march of progress on both sides of the iron curtain had resulted in the perceived Russian bomber threat being replaced by intercontinental ballistic missles (ICBMs) - a delivery method that SAGE was hopelessly inadequate for.
Whirlwind (and later SAGE) had changed how users interacted with computers (in that they could interact) but, being a military project, very little of it was available to the outside world. It would fall to other, non-military, research to bring the graphical user interface back into the public domain.
Oh, Just Draw It
Just as Whirlwind became to symbolise the work of Jay Forrester, so did the Sketchpad program come to symbolise the work of Ivan Sutherland. Written in 1963 as part of his PhD thesis, Sketchpad changed the way that users interacted with the computer, and is the ancestor of every Computer Aided Design (CAD) package written since.
Like Whirlwind, Sketchpad used a CRT and a lightpen but, unlike Whirlwind, Sketchpad relied solely on the user's input (and not data being fed in from radar stations). Users could create lines on the screen using the lightpen and the software then stored them in such a way that they could then be manipulated again at a later time. No longer was a line just a shape on the screen, instead it was an 'object' that could be used with other objects to create more ever more complex objects. Sketchpad didn't just stop there though as it allowed users to create instances of objects. The user might, for example, create a triangle. He/she could then create several instances of the triangle object. If he/she then modified the triangle object, all of the other instances of the triangle also changed. He might not have realised it at the time (or maybe he did) but Sutherland had just taken the first step down the 'object orientated programming' path (trust me, in programming terms this was important).
The Vision Of The Future
Sutherland's Sketchpad had caused something of a quantum leap in terms of user the GUI (elsewhere command lines were becoming more common but 'batch' processing still ruled the roost) but it would 1968 that really brough the modern GUI to life - and all because one guy wanted to do something that would benefit all of humanity.
Having read a copy of Vannevar Bush's As We May Think (as published in July 1945 edition of Atlantic Monthly), a young engineer by the name of Doug Engelbart found that Bush's vision of the 'memex' machine sparked his imagination and, a mere 16 years later, published an article of his own: Augmenting Human Intellect - A Conceptual Framework.
Written primarily for the Air Force (after having spent 7 years trying to drum up enough support for his 'wild ideas'), Engelbart's work built on that of Bush's and described a system whereby the user wasn't constrained by traditional work methods and approaches. Engelbart attempted to establish a system whereby the user connected objects, information and just about everything together in whatever order made sense to them. It was all high level stuff but he impressed enough people to gain a research project at the Stanford Research Institute (SRI) in 1962. The aim was simple: develop systems that would augment human capabilities.
Unlike many researchers at the time, Engelbart was keen to explore how the computer could be used to help enhance human intelligence rather than act as a replacement for it (as so many science fiction writers and doom-sayers would have had us believe). It would eventually take six years, a monumental amount of development and research, but the end result would be unlike anything anyone had ever seen before.
Let's Get Really Online
In 1968 Doug Engelbart's team from SRI was ready to show the fruits of their labours at the Fall Joint Computer Conference and the central product from all of his work was the oN-Line System (NLS). Elements of the project had leaked out before (the mouse, for example, had been demoed for use as a graphics input device in 1965) and there had been various research projects that had looked at certain aspects of what NLS could do (hypertext editing using a lighpen, and MIT's TECO screen text editor had both been demo-ed in 1967) but this was going to be the first complete demonstration of the NLS system to an unsuspecting audience. NLS was unlike anything that had been seen before and was far more than just a clever computer program. It was even more than just a clever computer program and some clever hardware - it was a new way of thinking.
Engelbart's work had focused on getting computers to augment human intelligence and the team had realised that getting people to work together and sharing information was an important factor. NLS allowed users to share information, work co-operatively and even video conference (radical stuff at the time). But it went further and NLS had networking at its core. Users could connect remotely and work together on the same screen of information even though they were physically miles apart.
Having developed a way to mix graphics and text onscreen, NLS had required a whole new interface method as the traditional keyboard was simply not up to the job. Even the lightpen, favourite of Jay Forrester and Ivan Sutherland, wasn't enough to augment the old QWERTY board and instead Engelbart's team had created a totally new input device - the mouse.
Although technically still the same as the mice now found on desktops around the world, Engelbart's original mouse was made of wood(!) and used two wheels directly connected to potentiometers (rather than the ball and, later, lasers employed nowadays). The technology may have changed but the concept remains the same to this day - by moving the mouse, the cursor moves on the screen. This is obvious to us but in 1968 it was revolutionary. The mouse was a great invention but it required some way for it to be represented onscreen...and so the mouse cursor was invented. Note: NLS also employed a 5 key 'chording' keyboard (in addition to the mouse and the regular QWERTY keyboard) which allowed users 'type' using just one hand and a set of combined key-presses - this input device didn't catch on though.
Using the mouse Engelbart dazzled his captivated audience with features such as emailing, instant messaging, hypertext links, context sensitive help and a swathe of other 'innovations'. It was radical but it wasn't without its problems and the limitations of the video hardware meant that the 'windows' (yep, they were new too) had no boundaries to them and it quickly became difficult to work out what was what. Similarly the lack of memory meant that the machine could only display upper case characters ('true' upper case characters were indicated by having a line drawn directly above them).
The crowd were enthralled though as, before their very eyes, Doug Engelbart had demonstrated not only a new set of technology but a whole new level of man-machine interaction. Some claimed it was all an elaborate trick, some were baffled and bemused, and some had their eyes well and truly opened, but all agreed, it was, without doubt, "the mother of all demos".
The Paperless Office
NLS had pushed the man-machine interface a long way and had introduced concepts and working practices that few could even begin to understand, never mind realise. NLS's 'virtual office' was a vision of the future, but this 'vision' wasn't the land of milk and honey for everyone. NLS had proved that the office could be moved from the real world into the virtual world and this was not a good thing if your business relied on the 'real world' office for its profits. Indeed the vision of the paperless office was pretty nightmarish if your business was office paper products.
The name Xerox had become synonymous with offices around the world and 'Xeroxing' had even become the generic term for document reproduction (certainly in the US). The office giant was certainly king of the hill and Engelbart's vision of the future had it scared...very scared. Xerox was faced with a problem: If the office was no longer going to rely on paper, who was going to buy Xerox products? Whether it liked it or not, Xerox had to make sure that, just as with the paper office, it was king of the paperless office.
Technology doesn't just happen overnight (it doesn't now and it didn't in 1969) and Xerox recognised that, while NLS pointed the way forward, it wasn't a commercial product and it would take years for the industry and technology to reach a point at which Engelbart's vision could be turned into something saleable. Xerox could have just sat back and then leapt on the bandwagon when the time was right, but it was far better to be ahead of the game when it came to technology, and the only way to do that was to have the technology before anyone else did. Of course the only way to that was to invent it.
Playing In The PARC
Xerox knew that it had to get the best and the brightest in the emerging computing industry if it was to stand any chance of dominating the market in the future decades. It had to set the standards and become the trend-setter, and to do this it knew that it needed to do something different. The solution was somewhat risky but Xerox decided to create a dedicated team that would basically be given free reign to develop whatever technology they wanted without having to worry about the practicalities of turning their ideas and research projects into consumer items - that would come later.
Physicist Dr. George Pake had made his mark in the field of nuclear magnetic resonance but Xerox scientist Jack Goldman knew that he was just the man to create the hothouse of ideas that the company wanted. Pake was used to the research environment and wisely made sure that Xerox's research centre was situated far away from the prying corporate eyes (3000 miles away actually) and so it was that on 1st July 1970, 3180 Porter Drive (California) opened its doors and became the Xerox Palo Alto Research Center[sic], aka Xerox PARC. Its aim to create "the architecture of information".
There were other advantages to PARC's location though and Doug Engelbart's Augmentation Research Center was situated close by - a little too close it would end up. Being able to raid the NLS team was big bonus but when Engelbart's funding started to dry up in 1974, many of his team naturally gravitated towards PARC...and PARC was all too keen to welcome them onboard.
The influx of ex-NLS staff was a big boost to the PARC family but the original team had already had several successes in the intervening 4 years, having produced the first laser printer, investigated ethernet, and creating the Smalltalk programming language which would go on to underpin the majority of the work undertaken in the coming decade, as well as numerous other ideas and technologies that would push the computing industry further and faster than ever before.
PARC produced many innovations but key to its now legendary status was the Smalltalk language. Designed by Alan Kay (and implemented by Dan Ingalls) in 1971, Smalltalk was unlike any language before it and it employed an object orientated approach and philosophy, but it was far more than just a programming language and sported its own GUI and actually became the operating system. Eventually incorporating overlapping windows, documents, cutting and pasting and many of the graphical elements that NLS had demonstrated, Smalltalk was Doug Engelbart's demo system turned into a practical (although not yet saleable) system.
Getting Personal
PARC's success with the laser printer had highlighted a serious problem as, although the laser printer was fantastically fast and very powerful, there wasn't a machine that could prepare documents that really used it to its maximum potential. What the laser printer needed was a machine that users could use themselves (i.e. not have to share with someone else), prepare their documents on (i.e. in a highly visual manner as opposed to character based terminals) and then print. Here was a new need for a totally new type of computer - a 'personal' computer.
Originally computers had been the massive mainframe behemoths but, by the end of the 1960's, the smaller 'mini' computer had appeared. Mini-computers may not have been as powerful as their bigger brothers but they were the fraction of the cost and opened computing up to a much wider audience and had less of a physical impact on the workplace. While a mainframe might have taken up an entire floor, the mini took up the space of a filing cabinet. A filing cabinet was certainly smaller than an entire floor of a building but it was still far too big for the PARC team's needs and they were thinking of even smaller. Possibly something that would even fit on or under a desk?
The concept of a 'desktop' computer had been the stuff of science fiction for decades and the technology still wasn't up to being shrunk to such a degree, but this was the very reason that PARC had been setup...so that's what they did.
Originally outlined in 1972 and finally coming online in 1973, the Xerox Alto was a masterpiece of hardware. Contained in a single box that incorporated memory (128Kb to start with although expandable to a whopping 512Kb), hard drive storage (a removeable 2.5Mb cartridge), networking, and all of the input and output devices needed, the Alto was connected to a novel portrait display, a keyboard and a mouse (although Engelbart's 5-key 'chord-ing' keyboard was available as an option). What was astounding (at the time) was that all of this hardware was squeezed into a box that would fit under a desk.
The Alto brought real power to the desktop but it was far more than just a stripped down mainframe or mini computer. Essentially a full blown mini-computer, the Alto was capable of providing the kind of distributed computing power and information sharing that NLS had embraced so whole heartily. Its use of the new ethernet style of networking made it incredibly easy to network together and users could share documents, send messages and emails to each other, and behave in a way that computer users had never been able to before (outside of the NLS demo). It also provided the perfect level of power for Kay's Smalltalk system.
With the combination of the Alto's hardware and the Smalltalk graphical interface, the PARC people had managed to take the old paradigm of centralised computing and turn it into a true distributed model. Mainframes may have sprouted thousands of terminals but each of the terminals did nothing except act as a gateway into the computer. Now, with the Alto, users had real computing power at their fingertips but still had the ability to share information quickly and easily. This was great but it was Smalltalk that made it easy to use and access.
The Alto had evolved (eventually sporting several different revisions and models) but so too had Smalltalk and the GUI itself. Icons, pop-up menus, scroll bars, radio buttons and even dialog boxes had all crept into GUI's vocabulary. Engelbart's NLS might have done the job, but it was, by comparrison, crude. Smalltalk now sported the most advanced GUI so far with windows that had proper lines around them, title bars and a whole swathe of features and functions that would become the 'norm' for GUIs forever more.
Practical Magic
The PARC people knew that they never had to produce 'consumer' goods (even the Alto, eventually running to several thousand units, was never considered as 'consumer' even though it spread outside of PARC and into other (non-Xerox) research departments) but Xerox expected to see some return for their vast investment and the pioneering work that had produced the Alto was soon seen to have some commercial value. What the Alto had done was make computing easy and practical and Xerox intended to exploit this power in the marketplace (Note: Several members of the Alto team wanted Xerox to produce the Alto itself as a commercial product but management figured that they knew best...). PARC was too busy creating yet more 'visions' of the future so Xerox created a new division for the job: the Systems Development Department (SDD).
SDD was the final link in Xerox's grand plan - the bridge between research and marketplace. Having seen what the Alto could do, SDD was given the task of taking the best of the Alto and creating a saleable office machine that was both easy to use and could perform every office task imaginable. The result was the Xerox Star Workstation (officially called the "8010 Star Information System") in 1981.
Although eventually sold as standalone machines, the Star workstations were originally intended to be purchased as part of a complete office solution which would typically consist of 2 or 3 workstations, a file server, and a print server. All well and good until buyers caught sight of the price tag: anywhere from $50,000 to $100,000 for a complete system, and even the Star workstations came in at $16,000 each! The Star may have provided easy networking, file sharing and the easiest interface ever seen on a consumer computer (despite having reverted back to 'tiled' windows rather than overlapping windows - a concept deemed far too confusing for the general public) but customers weren't buying.
Whether it was down to price tag, the deviation from the 'norm' (computers were still very much of the 'dumb terminal' kind in the late 1970s), the lack of speed from the machine's hardware (saving large files could take, literally, minutes and crashes could be followed by hours of 'scavenging' while the machine tried to recover) or poor marketing, the Star failed to ignite the marketplace and eventually a mere 25,000 were sold. The GUI might have been the darling of the IT industry's elite and may have made the jump from research project to commercial product, but it seemed that the world outside of the research lab wasn't at all interested...or maybe it just needed time, enthusiasm and...an apple?
Just Another Jobs
Small startup company Apple had shaken the computing industry with both the Apple 1 and the phenomenally successful Apple ][ but Apple knew that they couldn't rest on their laurels and, like every computing company in the late 1970's, were always striving to make the next 'must have' machine and make computing ever more user friendly and accessible. Or, in other words, sell more!
Started in 1978 Apple's Lisa project was intended to be little more than a natural evolution of the command line driven Apple ][ (from an interface perspective). Admittedly the Lisa (and the Apple ///) was aimed more towards the business end of the market but it was still stuck with the same command line and flashing cursor. And then came a meeting with Xerox.
PARC's reputation was known throughout the industry but getting access to the work and ideas being produced at Palo Alto wasn't always the easiest job as Xerox were, naturally, reluctant to let the competition know what they were up to. Apple employee Jeff Raskin had heard of the work being done at PARC and was especially intrigued by Smalltalk but failed to convince Apple's mercurial head Steve Jobs to pay a visit. Raskin wasn't alone though and when Apple's resident graphics genius Bill Atkinson also tried to convince Jobs that a visit might pay dividends, Jobs became curious about what exactly PARC had been up to.
Xerox might not have wanted to let people into PARC but Apple was, at the time, increasing in worth at an almost exponential rate. Business people understood cold hard cash if nothing else and Apple allowed Xerox to invest $1 million in return for two visits to PARC. It might have seemed like a very high price for two guided tours but, for Xerox, it would prove to be far more costly.
With the first visit scheduled for November 1979, Jobs and a small contingent of Apple employees were given PARC's vision of the computer in the shape of Smalltalk and the Alto. Totally blown away by what he saw, Jobs immediately decreed that the Lisa project would be the first to sport a GUI. Ok the hardware that PARC had developed was totally impractical from a sales perspective (estimated costs put the Alto at about $35,000) and the software was totally unlike anything that had been seen before but Jobs was adamant: Lisa would have a GUI and the Apple engineers would simply have to figure out how to do it.
In a sense Jobs was right: Someone else had done it so Apple could too, but Xerox, although letting Apple in the door, weren't about to let them walk away with everything they wanted. The Apple engineers might not have had the hardware or the software, but they had the inspiration and the knowledge that it could be done.
A further visit was still allowed (as part of their investment agreement) and Apple made sure that their best and brightest took this 'guided tour'. The actual PARC staff had happily shown off their work to anyone who'd visited and they expected to the do the same for this particular group of visitors but the Apple team had other ideas.
Fully briefed by Jobs and Atkinson beforehand, Apple's team knew precisely what they wanted to ask and what information they needed to gather. Technical details may have passed them by but the Apple team quickly learned what they needed to learn: how it worked. Yes they'd still have to replicate the functionality and programming, but they understood what it was supposed to do, how the mouse interacted with the elements of the desktop and how the user interacted with the machine.
PARC had certainly developed a fantastic system but Apple effectively had to re-develop everything from scratch for use with their, by comparrison, limited hardware. Of course it didn't hurt when 15 of the original PARC engineers later jumped boat to the good ship Apple.
The Lisa project eventually appeared to the world in 1983, two years after Xerox's Star, but it was not the all-conquering machine that Apple had hoped for. Lisa's interface was excellent and the Apple team had not only captured the spirit of the Alto but had also expanded and built on it, adding features such as drop down menus (and overlapping windows which the Star had removed so as to avoid 'confusing' users). If this wasn't what stopped sales though, what was?
Lisa had started life as a project promising a $2,000 business machine but, by January 1983, this had crept up to a far more worrying $9,999 and that was just a bit too hard on the wallet for most buyers. Apple had aimed for the low end business market but, by reaching the near mythical $10,000 mark, Apple had inadvertantly strayed into a whole new territory, a territory inhabited by the established 'big' computer firms.
The Lisa was expensive but this wasn't its only problem. The Alto had been able to rely on dedicated hardware to accelerate its performance but Lisa had to work in a far more cost effective framework and the end result was a massive hit in performance. Yes Lisa could do everything that users asked of it but it really pushed the limits of desktop hardware to the edge. In short, Lisa was s-l-o-w.
Despite its flaws, the Lisa had brought the GUI to a far wider audience than Xerox had managed in 1981. Lisa didn't sell especially well (and was one of the companies biggest losses after the disasterous Apple /// in 1980) but it did capture the public's imagination and was, to Apple, the only way forward. Xerox might have had massive presence in the office but Apple had a massive foot hold in the home user market and, while users didn't buy the Lisa, they were far more aware of it. Now if only Apple could get the price down (a lot)...
Another Bite Of The...Apple
The Lisa project had been aimed primarily at the business market but Apple knew not to put all of its eggs in one basket. Lisa was the high end machine which, obviously, left a gap for a low end machine: Macintosh.
Originally proposed as a $500 consumer machine, Macintosh was the brainchild of Jeff Raskin who hoped to create an affordable machine that users (both experienced and novice) could quickly get to grips with. Like Lisa though Macintosh underwent massive changes in its development lifetime (both in terms of technology and staff - Raskin eventually left and Jobs took charge after having been ousted from the Lisa team) and by the time it was eventually released in January 1984, Raskin's $500 machine had skyrocketed to a compartively huge $2,495. It might have cost a lot more than it was originally intended to be but Macintosh was still a fraction of the price of the Lisa.
It wasn't just the price that had changed though as the Mac had adopted various different hardware changes and, encouraged by the Lisa, a GUI. Macintosh and Lisa both look similar when compared side by side (they both had the same CPU (although the Mac ran at 8Mhz compared to Lisa's 5MHz), they both sported black and white screens (although Lisa's was larger), they both had GUIs etc.) but Macintosh and Lisa were two entirely different beasts. Physically Lisa was bigger but Macintosh was arguably faster and was able to offer just about everything that its bigger 'sister' could. Both did have a GUI but, although the Mac team had 'begged, borrowed, and stolen' bits and pieces from the Lisa team, the two systems were incompatible with each other and actually looked and behaved differently.
One other key area in which Macintosh and Lisa differed was in their approach to software. For $10,000 Lisa buyers were getting a complete office system (in fact Lisa OS stands for 'Office System' rather than 'Operating System') and a massive amount of software was bundled with the machine. This was great for the buyer initially but meant that third party software development was practically non-existant. Macintosh, by comparrison, actively encouraged external software developers to write apps for it. This certainly made Macintosh a far more attractive proposition than the Lisa but it also created Apple's biggest corporate foe: Microsoft.
Write Me A Program
Apple's decision to allow third party developers to create software for the Macintosh had brought onboard fellow startup company Microsoft (although by the eary 1980's both companies had been around for several years and Apple was certainly an established form). Although the Mac shipped with MacDraw and MacWrite, many still perceived the machine as a 'toy' and it's miniscule 128K of main memory was soo small that it was difficult to do anything with it and creating software that took advantage of the GUI was a challenge that many decided to decline. As time passed though, the Mac was revised and more memory became available. And with this extra memory came more opportunities.
Microsoft had already created the operating system for the IBM PC (MS-DOS) and they soon started to creating applications for Macintosh that were ideally suited to the GUI and really showed how it was far a more user friendly and productive environment than the old command line. Indeed applications that would have been impossible on a command line system soon came about simply because of the GUI that Apple had put in user's hands. The desktop publishing boom was born and Apple had brought computing to the masses all over again. It had also brought the GUI to the attention of others.
IBM's PC had, although technically inferior to the Apple machines, become a big seller and Microsoft had earned a healthy profit by licensing it's MS-DOS operating system to the industry giant. The problem with the PC though was that it used a command line interface and, as Microsoft had seen through their work with Apple, this was decidedly outdated and awkward to use when compared to the Macintosh. Perhaps the same technology could be applied to the PC...
Announced in 1983, Microsoft's Windows would take another 4 years of development before it was released but Microsoft still knew that it was in potentially troubled waters. Fearing a lawsuit from Apple, Microsoft approached Apple's (then) CEO John Sculley and an agreement was reached whereby Microsoft would continue to develop Macintosh applications in return for the right to use Macintosh technology. It was one of the biggest mistakes in computing history.
Released in 1987, Windows v1.01 brought the GUI to the PC but is was instantly derided mainly due to its crude interface (no overlapping windows or even icons) and most users found it unusable. It was enough to upset Apple though and Steve Jobs started to voice complaints about how Microsoft had stolen the Macintosh's interface design and this led to the infamous response by Bill Gates: "Hey, Steve, just because you broke into Xerox's house before I did and took the TV doesn't mean I can't go in later and take the stereo".
Windows v1.01 might not have been brilliant (by any stretch of the imagination) but the Microsoft engineers kept plugging away at it and Windows v2.03 was released in January 1988. Although still not perfect, this new version of Windows featured overlapping windows and Mac-like icons...and was enough for Apple to start legal proceedings.
The courts battled for 5 years on the issue but in the mean time Windows v3.1 was released and here, at last, was a usable version of Windows. Now PC users could realistically use a GUI on their machines and fully explore the power and simplicity that the GUI brought (even if the underlying PC hardware was more questionable). Windows v3.1 went on to sell millions of copies and soon became the de facto standard for PC users. This, coupled with the ever falling price of PC hardware, soon allowed the PC and Microsoft to gain market dominance over Apple.
Monkey See, Monkey Do
While Apple and Microsoft battled in the courts, other manufacturers (especially in the home computing industry) realised that the GUI was the way forward. Add to this the fact that hardware had advanced enormously (while also dropping in price) and every machine now had to ship with a GUI as potential customers simply wouldn't entertain anything else. Whether it was Commodore's Amiga, Atari's ST or Acorn's Archimedes (amongst others), the GUI was the way forward, and while the most 'common' elements remained (overlapping windows, icons etc.) every implementation brought its own unique flavour and nuances. The Archimedes brought us the fixed menu bar at the bottom of the screen, SGI's IRIX allowed users to resize icons in windows...
To Infinity And...Beyond
Apple's case against Microsoft eventually ended in Microsoft emerging victorious due to John Sculley's 1983 agreement, and a similar case of Xerox against Apple resulted in the same result (Xerox claimed theft on Apple's part and...lost). The situation was clear: the GUI didn't exist to any one company or organisation anymore.
Venevar Bush's 'Memex' machine might never have seen the light of day but his work undoubtedly lives on to this very day. Similarly Doug Engelbart's vision of the future might not have come (totally) to fruition, but his influence is still all too apparent on desktops across he face of the planet. Xerox never did manage to turn their millions of research dollars into sales but they did provide a vital bridge between the research world and the consumer world, providing Apple with the push and inspiration to make it work. Even the popular 'bad boys' at Microsoft have played a part in progressing the GUI.
Admittedly the GUI has now reached a state whereby it's an almost static piece of technology (rather than just a grand concept to be interpreted as and how) with so many elements now considered 'standard' that to move away from them would be tantamount to commercial sucide. As Microsoft and Apple now strive to innovate and invent ever cleverer ways to improve the man-machine interface, it's clear that the GUI has brought computing to the masses. Engelbart's dream of 'augmenting' human intelligence (rather than replacing it) has been very much realised (although stick an idiot in front of a modern computer and, no matter how advanced the interface, they're still an idiot) and it's now possible to share information, work with colleagues on the other side of the planet, video conference, multi-task, and perform any one of a million tasks with just a few clicks of the mouse or taps on the keyboard.
Where will the industry go next? Who can tell, but one thing's certain, it'll be fun finding out.
Want To Know More?
http://applemuseum.bott.org/sections/gui.html
http://en.wikipedia.org/wiki/History_of_the_GUI
http://www.livinginternet.com/w/wi_engelbart.htm
http://arstechnica.com/articles/paedia/gui.ars
|
|
