Mad Dog 21/21: iPhone Home

What did the extraterrestrial called ET in the eponymous 1982 Spielberg film use to save itself from a cadre of government agents? A phone. Twenty years later, people routinely phone home and everywhere else from anywhere. This hasn’t distracted the technologists caught up in arguments about thin versus thick clients, but now there’s a new Apple iPhone. It might ring loudly enough to be heard above the bickering advocates of various clients. A client as well as a phone, it could shift the focus of the user interface debate from technological means to budgetary goals.

In a sense, Apple helped spark the initial debate about clients. Its Apple II running Visicalc caught the attention of Wall Street types who used the spreadsheet in their work and also got their computing departments interested in the new gadget. By the time IBM delivered its PC, a year before ET hit the theaters, everyone in computing knew that green screen computer terminals were going to face some big challenges from displays that offered color and higher resolution, too.

The possibilities of client devices that were radically different from IBM’s 5250 and 3270 green screens were well illustrated in the 1970s. At that time, the examples were still too costly and exotic for deployment in business offices. Also, there were no business applications that took advantage of a graphical user interface. Nonetheless, devices running at Stanford University and its cousin SRI, along with inventions created at the incomparable Xerox PARC lab showed that computing could provide a rich visual experience and that this experience was improved if the end user had access to a pointing device such as a mouse.

By the time IBM and its competitors offered PCs based on Intel 80286 chips, the balance had tipped. Apple’s Macintosh came out the same year as the PC AT, 1984, and it won considerable acceptance in a number of business markets where graphics counted, such as advertising and publishing. The PC had to catch up, and to do that it needed better display cards and software that could do what character-based DOS could not. As it turned out, the PC wasn’t even in the running until Microsoft delivered Windows for Workgroups in 1992 and some users felt there was nothing about Windows to make it comparable to Apple’s offerings until three years later when Windows 95 and Windows NT 3.51 hit the market.

ET The Extraterrestrial: As endearing a spaced-out botanist as you can find outside Mendicino Count.

Even as the graphical user interface was becoming popular on PCs, all the suppliers of data processing computers (sometimes called systems, but certainly not called servers yet) for offices of every size still based their software on clients that had monochrome displays, keyboards, and no mice. Some displays had local functionality that helped when users had to fill in forms, but otherwise business software was written as if the PC didn’t exist. At the time, a visitor in a corporate office might see plenty of PCs, but when they were hooked up to central computers, the PCs used a combination of hardware and software to emulate what were at the time called dumb terminals. The term thin client would come much later, when advances in electronics (and, later, display technologies) actually made it possible for a terminal to provide functionality in a thin form factor.

Even so, there were plenty of disputes among computer professionals about what work should be done on the user’s desktop and what should be done centrally. In offices where PCs equipped with green screen emulation technology had replaced 5250 or 3270 screens (or DEC terminals), the PCs generally ran word processing and spreadsheet software, too. As the PCs got graphical operating systems and the applications took advantage of this technology, what had been called productivity software seemed to become more productive. There was even a case to be made that the PC, by giving the end user more power and freedom, provided motivation that increased the user’s pep. In short, with the help of improved presentation, productivity applications actually seemed to improve productivity.

In the meantime, in a parallel universe, the Internet was taking off. In 1994, Netscape delivered its first Navigator Web browser and related email software. All of a sudden, in home offices and small offices and, to a much smaller extent, in corporate offices, too, desktop machines were reaching out over external networks. That’s when all hell broke loose.

Corporate information processing departments might have been in denial at first, but before long the facts were inescapable. End users needed some kind of access to email and external Web sites. If corporate networks didn’t provide this, end users would hook up by other means. The upshot was that corporate computing departments ultimately provided adequate bandwidth, connectivity services, and along with these technologies they also supported the necessary security services.

The look and feel of Web sites improved very rapidly, initially as part of the competition between Netscape and Microsoft for dominance of the browser base. Later, the independent standards group W3C became stronger and eventually overpowered every company offering a Web browser. Today, every outfit offering a Web browser knows what is expected of its software and also what will be expected in the near future as evolved standards come into force.

Still, this evolution did not end the debate about how much computing should be done on a user’s desktop. If anything, it made the arguments more intense. And with every new generation of PC software, with every advance in central systems’ operating environments, and with every enhancement of the Internet, the situation has become even more contentious.

While giving an end user more freedom might enhance that user’s enthusiasm and productivity, it also provides an opportunity for mischief. There’s no question that business employees who can use email and surf the Web for personal use do so on company time. What’s not clear (but which may seem obvious on the surface) is whether these wandering users actually accomplish less in a day or a week than those who don’t (or can’t) digress. It might turn out that there’s no difference in output at all, or even that employees who take a quick personal Web break a few times a day actually get more done.

In any event, business managers seem to believe that their employees should not be fooling around on PCs during work hours. To curtail this behavior, companies have tried to restrict the way client systems can be used. The restrictions can be simply a matter of policy, but more often the policy is backed up by technology that constrains end users. Two common methods of limiting user activity are to restrict network access, which can be done with a firewall system or proxy server, and to limit the functionality of the client machine, which can be done by imposing software restrictions.

Sun Ray 270: This version of the Sun thin client has the screen and electronics in one unit; Sun also offers modular versions.

But neither of these techniques determines whether an end user should have a thin client or a rich client, because these days either type of client can do pretty much anything a company’s servers support. What has happened is that the thin client, which is in some ways a descendant of the green screen, may be driven by some a very powerful server. In effect, a thin client can be a rich client with most functionality moved to the other end of the wire. The thin client can look and act like a Windows PC even if its access to Windows is provided a virtual world residing entirely on a server (or virtual server) far from the desktop. But a thin client doesn’t necessarily provide a full operating environment; a thin client might be set up to run only one program, a Web browser, and the user may use this browser to go to corporate applications on an intranet or external sites in the Internet.

While a number of vendors offer thin clients, I think that Sun Microsystems‘ Sun Ray is about as good a technical example as there is. The way it works, its real brain lives in a virtual server somewhere on network. Its local personality is defined for a particular end user. And the end user tells the Sun Ray to turn on a particular user profile by inserting a card (that looks like a credit card) into a slot in the terminal (or the processor box in the case of the more modular Sun Ray machines that have separate processors and displays). The Sun Ray and its support server take it all from there. It’s is also possible to built a thin client on a Linux or Windows foundation, and both of those approaches have their supporters.

Today, the main argument between advocates of thin clients and those who favor rich clients is about cost. Fans of thin clients say they are cheaper because they are managed centrally, and they also espouse their lower power consumption, too. Thin clients have fewer parts to fail, and in the rare event of a failure, field support is basically about swapping a component. Central client management can also simplify the use of clients that begin life as rich clients but are subsequently configured so unwanted functions become unavailable or significantly curtailed.

Advocates of rich clients disagree. They say that applications such as word processing can’t be run very well off a server unless that server has quite a bit of pep and this makes the server very expensive. Even then, they say, the user has to be lucky. If a lot of people with thin clients are doing word processing at the same time, networks will choke and the user will get poor response. The psychological impact on the user may compound the direct effects of slow response. In short, when an organization uses rich clients, servers and networks can support many more end users.

All this is old hat and computing folk who have listed to these arguments over the years often spot a pattern. The pendulum of opinion tends to swing from thin to rich and back again as technology evolves and business conditions change. Some of the loudest voices in the debate seem to be those of tech support teams and their executive managers. But server folk who are asked to figure out what it might take to add enough presentation server power to keep thin clients running nicely sometimes have a lot to say, too.

The debate makes a certain amount of sense as long as the underlying applications technology is stable or, if it’s evolving, changing in a way that makes it easy to get a glimpse of the future.

Right now, though, things may break loose, the way they did when Windows started offering a bearable GUI, Netscape started delivering high quality browsers, and Internet access became ubiquitous. Since then, the world has gone to broadband and the information highway, like any other highway before gas hit four bucks, managed to keep drawing traffic until it became jammed and maybe even after that point.

Now it’s starting to look like a new kind of Web site is going to become popular. At first it might look a lot like familiar sites, but underneath the surface there lurks another Web, one that works on a very small screen. It’s all the fault of the iPhone, which is creating challenges for Web developers and, ultimately, for companies that use the Internet and intranets to reach customers, employees, and shareholders.

The iPhone already has some imitators and while they’re not yet really up to snuff, it’s just a matter of time before there are a few good choices for Web surfers on the go. In business, the iPhone still has to beat the Blackberry and its rivals, because the iPhone doesn’t quite have the form, function, and gravitas Blackberry lovers demand. But that might more a matter of when than if. In the meantime, the iPhone has as a central feature something every high end mobile phone is going to need: a real Web browser.

The browser in the iPhone is a version of Apple’s Safari, itself based on the open source Konqueror software. Generally speaking, a Web site that works well in Safari on a Mac or Windows machine will work in the iPhone (if the site doesn’t depend on giant graphics or other big elements that can’t easily be shrunk). Opera has a browser for some mobile phones, too, and Opera is about as good as Safari. Mobile Firefox is in its infancy, but it’s a genuine effort that might yet become prominent. Internet Explorer Mobile is another contender; at the moment it looks like it need to be refreshed a bit to compete with Safari and Opera.

No mobile phone has a touch interface that’s as well thought out as the one on the iPhone. Some users might not like a touch screen at all and instead prefer keys, but even mobile gadgets with keys might have to offer touch screens, too, to stay in the race. For even though the iPhone doesn’t appeal to everyone, it does seem to be extraordinarily attractive to a lot of people. Now that there’s a second generation with a much lower price, the user base ought to leap head. Because iPhone is tied to AT&T Wireless, there’s a big opportunity for rivals that have really good technology and are network neutral.

Steve Jobs’s Home Phone: The man seems to be aging, but the inventions he promotes stay young.

There’s also going to be an opportunity for a touch screen phone on the Verizon network, but it’s not clear how Verizon will play this. It might be better off with a phone that runs on its network and also works on GSM networks outside the United States, or maybe with a few choices if no single vendor can make a phone that’s in the same league as the iPhone.

However the phone market plays out, it’s going to reshape the Internet because mobile phones are everywhere, the market for touch screen versions is still young and excited by upgrades and inventions, and because so much of what makes the Internet attractive is even more compelling when the opportunity to surf fits in a pocket.

So the Internet will get a new shape or a hidden second shape that becomes visible on phones. That will spur the development of technology that not only serves consumers using the Web, but also helps business users build and maintain Web sites and intranets more effectively. And it opens the door to a new generation of thin clients that offer not only the standard keyboard and mouse but also support touch features the way the iPhone does.

The iPhone concept also creates an opportunity for touch display features in Windows, but it’s not obvious Microsoft is up to the task right now. It might need a new generation of top executives who can fight the bloat that afflicts so much of Microsoft’s product line. I would be more optimistic about Microsoft if it had leaders who would press for streamlined software, demand simplification of every product the company offers, and got the company past its severe case of Google envy.

And Microsoft isn’t the only company that can suffer from inertia (to say nothing of embarrassing comparisons with Apple).

If Sun doesn’t have a Sun Ray Touch on the market soon, we advise its directors to make surprise visits to every key facility. They might spot their employees ignoring the Sun Rays on their desks and instead surfing and emailing on iPhones that look both thin and rich.

                     Post this story to del.icio.us
               Post this story to Digg
    Post this story to Slashdot