The Web is like an 8-track tape
I found this in my archive and just had to post it here. It’s a review of Internet World ‘97 for the February 1998 issue of IEEE Spectrum Magazine, where I briefly contributed to a column called Websights. As I post it now almost 10 years later, I think oh my, how young and naive we were.
Grumblings from the trade show floor
Internet World ‘97, held in December at New York City’s cavernous Jacob Javits Center, proved to be more style than substance, as 50 000 attendees discovered few new developments buried beneath mountains of hype. Many of the 600 exhibitors were start-ups offering their own versions of existing products, but industry giants like Microsoft, IBM, Netscape, and Sun Microsystems were equally lacking in cutting-edge developments. For them it was more like damage control, affirming that this year their existing products would really perform as advertised. The big surprise was Apple Computer’s withdrawal from the event, its second major trade show absence in less than a month. (Apple missed Comdex in November.)
But despite the absence of notable developments and developers, curiosity seekers were out in force. The percentage of nontechnical attendees was so great that it prompted one Javits Center employee to inquire whether this was a trade show or a consumer show. One recurring theme among first-time visitors was the anxiety over corporate America’s dictum to “Get us on the Web.”
The people responsible for doing just that were among those who attended the show’s Webmaster’s Symposium. Some informal polling revealed frustration with the standards bearers for making an already complex, ever-changing medium even harder to produce and maintain.
Overheard at the coffee cart
Almost all the Webmasters queried at the meeting believed a lack of universally adopted standards was the issue of greatest concern. With standards in different phases of acceptance or refusal by different developers, many Webmasters expressed frustration at having to design entire domains to the lowest common denominator–that is, reducing the level of eye-catching content and services to allow access to the greatest number of users and their plain vanilla gear.
The alternatives are either exclusive (barring access to many users) or expensive (replicating the entire site content to provide optimized services based on the user’s equipment).
Surprisingly, only a small minority of those surveyed were concerned with connection speeds, predominantly those involved in commercial ventures (e-commerce was all the rage).
A Webmaster for one retailer said, “Sites with flashy, interesting, and functional content are fine for users with a T-1 connection at work, provided they have the time and permission to use it. Most home users prefer simple, graphically challenged pages that load quickly on their 386 with a 14.4K modem. If our pages take too long to download, we risk losing a sale. On the other hand, if we look primitive, users may interpret that as inexperience and shop elsewhere.”
Most feel the problem will taper off when cable modems–and their speeds of upwards of 200 kb/s–become more prevalent, but others argue that in computer and communications technology, there is no such thing as “fast enough.”
For many, these difficult decisions are compounded by a lack of technical experience. Over three-quarters considered themselves “accidental” Webmasters–mid-level administrators and small business employees with limited technical knowledge who have been told to sort this whole Internet thing out, report back, and be ready to set it up.
A corporate communications manager for a mid-sized engineering firm summed it up best: “We know it’s big, and we know we have to be on it. But we’re not sure why, and we’re not sure how.”
One feeling shared by Webmasters throughout the show was directed vociferously to software developers: please work together. For those in the trenches, the daily struggle to appeal to the greatest number of users, while still serving up the latest services, is stunting the growth of the Web’s capability. All the Webmaster can do is run in place, while industry suppliers try to eliminate each other, the standards consortiums try to buy time, the Government tries to control what it can, and the marketers try to capitalize on the hype.
The year 1997 was when the Internet came to the common man. The growth was staggering, but many Webmasters are beginning to stagger, too.
Removing bugs from the Web
Now that computers’ proximity to each other is growing, so, too, are the risks to system health. Whereas computer viruses often took years to spread worldwide, now, thanks to the Internet, they can migrate between hosts in a matter of hours. Fortunately, the cause of this increased rate of infection is also the means to its cure.
Using human biological parallels, a unit at IBM Corp.’s Thomas J. Watson Research Facility, in Hawthorne, N.Y., has developed and successfully tested what it calls an Immune System for Cyberspace. Invisible to the user, the system automatically detects, removes, and vaccinates against new computer viruses by using the Internet. It works like a mirror of human immunology, only better, since it doesn’t rely on humans having the good sense to protect themselves. (When was the last time you had a tetanus booster?)
Rather than expecting users to retrieve downloadable updates from an anti-virus supplier’s Website, Jeffrey Kephart, a manager in IBM Corp.’s Massively Distributed Systems Group, believes that the most effective way is to remove the human element from the delivery system altogether. “A computer’s immune system should get the updates itself, because people aren’t going to want to do it,” he said. “People shouldn’t have to. Even if it’s just pressing a button, that’s too much.”
When a protected PC’s virus-checking software detects an unknown virus, the system samples it by removing the infected data, attaching it to an innocuous host, and sending it either to the network’s administrator or directly to IBM over the Internet.
If the sample is not already known to IBM, the virus is automatically replicated and analyzed to determine its workings. The cure is then sent back to the administrator, where it is disseminated to the original host and other computers on the same network. In addition, the virus’s signature and vaccine are both added to the master database at IBM, which will ultimately begin updating the immune systems of all its users via the Internet.
At a San Francisco demonstration last October, IBM researchers infected a laptop computer with viruses unknown to its anti-virus software and watched the computer automatically cure itself. Roundtrip time from point of infection to point of cure was an impressive 3-5 minutes, including the file transfer time between San Francisco and Hawthorne, N.Y., the location of IBM’s virus detection unit.
Even with full implementation of an immune system, there will always be new computer viruses, just as there are always new human ones. At the current rate of Internet growth, Kephart believes there exists “a recipe for widespread infiltration of the Internet if you don’t have an immune system. It’s not a panacea, but just a necessary step to keep pace with the virus problem, to just keep it at the nuisance level.”
While no specific product information has been announced, expect the Immune System for Cyberspace to be incorporated into IBM AntiVirus sometime this year.
Leave a Reply
You must be logged in to post a comment.