In 1995, I stood before a class of journalism students who were eager to explore topics relating to information technology and told them that soon the only software they would need would be Internet Explorer.
Even though "WWW" meant little or nothing to anyone at that time, the Web appeared to be an unstoppable juggernaut bent on sacking the concept of the task-specific IT department. At least, that's what I had read.
You may recall that Internet Explorer 1.0 and 2.0 were both released in 1995. IE wowed Windows users with built-in dial-up network support over TCP/IP. In 1995, the total number of "information super highway" users was pegged by IDC at 16 million worldwide. This was roughly the population of greater Los Angeles at the time of the last census. Notably, most of these folks were scientists and professors. By comparison, the number of people wired to the Web today is 1,133 million, or 17.2 percent of the earth's inhabitants, according to Internet World stats. The Web's magnetism is so strong that behaviorists are counting it among other notoriously addictive substances, like opiates.
Anyway, I remember that as I spoke these words with the confidence that you'd expect from a guest lecturer, I wasn't sure that it could ever really happen. The Internet was still embryonic, and I didn't know exactly why, but it seemed like a stretch. Security, network bandwidth, customization requirements, and issues of speed had not yet been addressed. Former Secretary of Defense Rumsfeld said it best: "There are known unknowns.... There are also unknown unknowns."
Now, in 2007, I still wonder whether, in the not-too-distant future, all of the tasks performed by typical business computers with desktop IO devices will be replaced by huge central servers, a Web browser, and an elaborate network. How many companies will switch from a conventional information-management infrastructure to one that's substantially Web-based in the next five years?
Certainly, comfort is important, but can we ever break the emotional bond that we have with green-screens? When it comes to heads-down data entry, are they really superior? I've seen Web apps that practically fill in an entire screen once you select a single value in a drop-down box.
And what about System i servers coupled with PC-based clients? This configuration works well in business settings because users are presented with a GUI and can customize their desktops, but maintaining a client/server environment isn't as cheap or as simple as you-know-who would like us to think. New releases of application software have to be loaded onto each desktop when they come out, viruses threaten productivity, and disk space has to be policed.
How nice would it be to have employees gain access to your system from anywhere that the Web is available? I once read that you could make a powerful WiFi antenna by winding copper wire around a Pringles can. Besides the simplicity and abundance of access points, dazzling new wireless devices are launched every hour. And wouldn't it be easier to bring new employees into your organization if they already knew how to navigate their way through your business system?
Web technologies have come a long way in 10 years. Only a maharishi could have looked at a static, brochure-style Web page 12 years ago and imagined that someday the Web would entirely displace traditional business computing modalities. Heck, back then you couldn't even keep track of visitors to your site. In '95, cookies were edible and came in decorative metal tins.
Now, Web application development techniques like AJAX support the creation of dynamic Web applications that feel like the best apps you ever imagined. AJAX-derived apps are impressive because they exchange only a small amount of data with the server. Entire Web pages do not have to be reloaded each time the user enters data into a field. Furthermore, AJAX is based on open standards, so its progeny transcend the barriers imposed by disparate platforms.
And new cryptography protocols like Transport Layer Security (TLS), the successor to SSL, facilitate secure, encrypted communications across any network and eliminate tampering or eavesdropping from the man in the middle.
Many Web technology–centric discussions now seem to focus on Software as a Service (SaaS). The acronym SaaS has eclipsed the concepts of application service provider, on-demand, and utility computing. SaaS is essentially a software application delivery schema that works like cable TV but better. Once you determine what you need in terms of information-processing capabilities, a vendor either configures existing Web-native software to suit your order (or develops new stuff) and then hosts and manages it for you. All you need is a fast connection to the Internet, a thin desktop, and a checkbook.
In the pioneering days when every business application was hosted by a big, centralized computer, IT people with specialized skills spent their sleepless youth making sure that every valid user had a blinking cursor. Then, the client/server model turned centralized computing on its ear, more or less, but the decentralization of computing resources diminished the amount of control that IT has over the availability of information. Now, centralized computing is coming back to town, and SaaS is the bus that it's riding on. The question is, do companies want to hand over control of their IT resources, knowing what we now know about the loss of control?
The economies of scale posed by SaaS are making this computing model attractive to companies that are poised to offer these services. These vendors are busy tweaking their services packages and price points (which will be a challenge, given the low cost of maintaining a System i environment). Small companies that have no IT department and a closet full of outdated IT gear could easily get caught in the gravitational pull of SaaS, but what about SMBs and large companies that have grown quite accustomed to having in-house tech departments that cater to their corporate computing whims?
And, sure, the Web has built-in redundancy, and insomniacs I know tell me that it's up 24x7, but thousands of System i shops have already figured out how to attain that kind of accessibility on their own.
To be sure, the efficacy of the Web for business has been proven, but in my humble opinion, we are not yet ready to hand over the keys to the data center. The needs of users who are currently served by IT topographies that involve System i technology should continue to be served with in-place systems and procedures that are embellished with browser-based desktops, portals, and Web services. Accessing the system through a desktop portal adequately addresses issues relating to ease of use in multiple-server environments. Since the applications that serve these desktops are consolidated and reasonably centralized, maintaining them is much easier.
Admittedly, I wasn't sure where things were going back in 1995. Now I am absolutely certain that the only things that you can be absolutely certain about are the things that have already happened. I am also fairly certain that hysteria, in and of itself, will not drive a System i shop to do anything that cannot be justified by tangible user and business requirements.