In-Depth

The Industry: Now and Then

• By Charlie Simpson
• 02/01/2001

James G. Hayhoe: Within the context of my current position, as Business Operations Manager for a rapidly-growing Defense Industry division of a large aerospace company, it is to ensure that our users have the most productive IT tools available to them in a reasonable timeframe.

Bill Inmon: My number one responsibility, today, is running the many aspects of my Web site and consulting firm, billinmon.com.

Dan Greiner: Finding a new job. However, the prospect of having a relaxed atmosphere for holiday shopping is just too great - I plan on eluding this responsibility until mid-January.

William P. Crowell: Now that we've completed the development of a full set of e-business security products that include WAN, IPSec, VPN and PKI solutions, a smart card and smart-card reader, a USB security token and security management software, we need to deliver these products to market. As a result, my attention is turning to business development and pushing these products to market. I am spending time with customers to help IT managers understand there's no silver bullet in network security and that, instead, a combination of solutions are necessary for robust network protection. Also, I am working with partners and potential partners to develop solutions for particular vertical markets, like healthcare.

Jon "maddog" Hall: I am the Executive Director of Linux International, an association of companies that are promoting the use of Linux. My number one responsibility is to direct and expand that organization.

Peter Salus: I am a practicing enthusiast and propagandist for Matrix.Net and the Internet in general; for Open Source, in the UNIX/GNU/Linux "tradition."

Fred V. Provoncha: Lead Systems PGMR/MGR, The Estee Lauder Companies, 9672 IBM Mainframe, OS390/2.8.

Bill Hogarth: Assisting customers in the installation and use of automated testing software for client/server and Web applications.

Hayhoe: In 1991, we began the deployment of 8,000 PCs within our organization. IT was convinced that these PCs would simply be used as workstations to their central resources. When distributed tools became the demand overnight, it was too foreign to the old IBM-iron IT management.

Inmon: The Internet caught us off guard. It has turned out to be a tremendous media for building a consulting business.

Greiner: The demise of emitter-coupled-logic (ECL) processors and their replacement with complementary-metal-oxide semiconductor (CMOS). After years of trying to make the fastest processors, who would have thought that an effective marketing campaign could be won by telling customers that they needed smaller, coupled systems?

Crowell: What surprised me the most was the speed with which firewalls and other security products that provide only partial protection for networks entered the market and were accepted as complete, or almost complete, solutions. Another key surprise was the slow market acceptance of cryptographic and public-key based solutions. My personal view is that much of that has to do with market perceptions and how well public key is understood. In addition, there was a failure to launch public key infrastructures early on and, when PKI did arrive, a failure to enable applications to use it.

Hall: The abandonment of the desktop by the UNIX vendors, and the surrender of that desktop to Microsoft. I was so sure that no one would be so stupid as to think that Microsoft would be satisfied with only the desktop marketplace, leaving the server space to UNIX. I thought that even the dullest clod would realize that you need to keep your logo in front of the customer at all times, so they continue to think about your company instead of one in Redmond, Washington, but I was wrong. I guess this was not really a technology change, but a management fiasco. Linux is the last hope.

Salus: Wireless. I never imagined that we would actually efface geography and support mobility as we have.

Provoncha: Linux and UNIX on the mainframe.

Hogarth: Technological advances have made it so easy to forget the other advances we had made in defining business rules, functional requirements, system design and testing. The emphasis on new programming languages, and the rush to coding has wiped out the necessary disciplines it took so long to establish.

Hayhoe: A distributed architecture that allowed multiple servers to be clustered to serve a widely-distributed and diverse customer base.

Greiner: Without question, the Web. Initially, the ability to share information hinted at an exhilarating future, and I applaud many who have capitalized in expanding their "bricks-and-mortar" (God, I hate that cliché) enterprise into e-businesses - Land's End is one of my favorites. I, also, look favorably on the new businesses that sprang up - as long as they're in business, I'll shop at Amazon.

But, I take great exception with the phenomenal amount of investment that is being spent on expanding bandwidth, developing elaborate storage area networks, hyping higher-speed connectivity, et al - just to store and deliver one inch banner advertisements for junk that I'll never buy, and that get in the way of meaningful content.

Hayhoe: My own personal disappointment is in monitors. While, today, we have bigger, brighter and colored, it is just an extension of what sat on my desktop 10 years ago. Why don't I have a monitor hanging on my wall? Why don't I have a personal monitor that doesn't have to be cabled to a CPU? Why don't I have a continuous monitor affixed to my glasses?

Inmon: The list of failures is long - Object Orientation, EIS, AI, Knowledge Engineering, CASE, etc. Where are these people today?

Greiner: Aside from keeping up with new technology at a mind-boggling rate, aside from the constant pressure to do more with less, aside from absolutely impossible schedules that are needed to remain competitive, and aside from the insidious regulations resulting from politically-correct legislation, there's been little change at all.

Crowell: The biggest area where security failed to deliver was in virtual private networks, particularly for Internet protocol. This was true, primarily, because the standards for IPSec were slow to materialize once they had been agreed upon. Many of the early VPN products were focused only on meeting the standard. They delivered neither the needed performance because of the overhead the IPSec standards required, nor the manageability that made them easy to use. Cylink solved those problems with our NetHawk. The VPN appliance supports thousands of secure simultaneous connections and strong Triple-DES encryption at full wire speeds and is controlled by an IPSec-based network manager specially designed for demanding enterprise and global VPN environments.

Hall: Beta video tapes. For the same reason that Apple continues to fail. They both wanted to control 100 percent of their marketplace. If they had licensed out the technology, they would have found hundreds of companies trying to help them sell it. This is why VHS lives. Microsoft realized this in the PC industry.

Salus: GaAs was a promised technology as early as 20 years ago, but has never achieved either the success or the range that was anticipated. Projects at UMichigan, IBM Research and Oki Semiconductors have shown the potential, but I think that Intel, Motorola and AMD have too much invested to attempt new technologies.

Hogarth: CASE tools wasted more time and money than any other "silver bullet" that had ever captured management's fancy in the past. Non-technical management will continue to fall for the latest snake oil, which promises to get rid of IT "nerds" and "geeks."

How has the responsibility of the IT Manager changed during this past decade?

Hayhoe: Ten years ago, most IT Managers were still dealing with IBM-iron mentality. Central control was mandatory except in those "commercial companies" that could get away with new things. Today, the challenge is twofold: 1) keeping abreast of the office technology available today, and 2) selling upper management on continuing improvement and upgrade.

Inmon: The IT manager has become a manager of budget and purchasing. IT no longer builds systems, but is in a caretaker, passive world of watching outside firms build applications and tools. Selecting technology is the role of the IT department, today. Back in the old days, the IT department actually built things and occasionally got their hands dirty.

Greiner: Attracting and retaining high-quality, competent employees.

Crowell: The principal areas of change have been twofold: the absolute drive and demand to move organizations into the Internet space; and the huge demands those pressures put on IT managers to deal with the resulting security requirements. This has been very, very difficult for IT managers who in the beginning weren't really that familiar with the Web and all the consequences of connectivity. Also, most IT managers were not well trained or indoctrinated in security solutions. Finally, until recently, most of the security solutions that were available were not well suited to the demands of the marketplace in terms of transparency and performance.

Hall: You have to know so much more, at least on a 100,000 foot level. When I started out, all you needed to know was how to call IBM to fix your card punch and card sorter. RS232 was the only standard you needed to memorize. Fortran, COBOL and BAL were the only languages. There were no graphics to speak of, no networking specifications. Announcements of new products came relatively slowly.

Managers may fool themselves into thinking that they don't need to know even a little of the massive amount of standards and technical information thrown at them today, but they should at least know enough to tell when their engineers/operators/programmers are pulling the wool over their pointy hair.

Salus: A decade ago, most places were single source: largely IBM, but also Sun, DEC or UNISYS with desktops featuring Compaq, Apple, etc. Today, I rarely walk into a site where SPARC, Intel and Motorola chips aren't all represented; where Solaris, Linux, BSD, some form of Windows and MacOS don't contend with one another; where three or four different mailers are in use.

Provoncha: Opened the decade with Big Iron, centralized, now multiplatform, diversafied. What do we have? How do we back it up? How do we recover from badness?

Hogarth: The manager now doesn't have to know anything about the details of how systems operate. He has changed to a recruiter trying to get staff, a buyer trying to select the latest and greatest development software, a salesman trying to sell the new gadgets and schemes to management, a student trying to keep up with the newest technology so he can at least talk to his technicians, and a skilled opportunist, always ready and willing to move on.

What's the biggest headache an IT manager faces in his or her job today?

Hayhoe: Again, the task of selling upper management is paramount. We used to think of IT in the same vernacular as capital brick and mortar - i.e., the investment could be depreciated across a period of time. Today, we must think and budget in the context of continuous improvement. This context is extremely difficult to sell to most COOs and CFOs today.

Inmon: The biggest headache is the misconception planted by the venture capital community that old technology is bad. As old technology is turned over, we throw out all the lessons we have learned and start all over again, making the same old mistakes with new technology. The VCs throw out the baby with the wash, and we all pay the price - over and over and over again.

Crowell: The biggest headache is trying to get everyone focused on delivering an interoperable set of solutions. There is no single killer application in security; there is no one thing that is going to solve all the security problems. And so, all of these pieces - PKI and VPN solutions, smart cards and readers, and other solutions - have to be integrated and interoperable. One of the biggest challenges any company in the security space faces is getting the industry to focus on producing solutions with this kind of interconnectivity and interdependency.

Hall: I don't have any headaches. I am 50 years old, and will (in all probability) retire at the age of 55 to start my last great project: Maddog's Manor of Microcomputing and Microbrewing, which will be where a bunch of energetic, smart CS students hang out, do distance studying for advanced degrees, make beer and learn about life.

Oh, I suppose that trying to explain market forces to Open Source developers causes a little grief from time to time. Or trying to explain Open Source forces to market drivers ... but, all in all, most of them are reasonable people.

Salus: Heterogeneity. I served on the ISO Working Group on Interoperability in 1989-90. We still ain't there.

Provoncha: Meeting deadlines with too little staff and finding trained staff.

Hogarth: The incredible increase in incompetence and inexperience, from managers to technicians to clerks that has resulted in more and more failed projects, unreadable documentation and unreliable software.

Hayhoe: Except for casual dress code now, I don't feel that my work environment has changed that much. I now have a Palm or laptop, a telephone that has WAP; I can read my work e-mail at home. But, I still have a desk, a secretary, a PC and way too much paper still crosses my desk and my physical filing space has not diminished much.

I can now do a lot of my work faster and away from the office, but I am not convinced that that is better. I can go for a whole day and only interface with people on my Blackberry, but I have lost the ability to sit across the table from someone and "read" their reaction to what I am saying. I believe I am a poorer manager than I was 10 years ago as a result of that.

[What would you change back?] Nothing! Keep the technology and tools and toys coming. It is our sociological task to figure out how to integrate them into our workplace and to change that workplace!

[What other changes would you make?] I would have us crusade for the paperless office environment. With the tools that we have today I don't need to see much paper and certainly not store much. Why not electronic signatures? Take away my filing cabinets! Take Data Warehousing out of the editorial commentary and make it an everyday event, not concept.

Inmon: Culturally, the IT organization changes have been that everyone is a caretaker, no one is a direct provider of information, and people can't make decisions without a committee. There is no vision, there is no leadership and all the good, young employees are working for more interesting shops, such a Internet startups. Furthermore, no one takes any risks. The changing paradigm of information is a mystery

Greiner: I kind of miss the three-martini lunch, although I can't remember having one for over 15 years (the wear and tear on my liver was too much).

Despite my lampoon of the Web, one good thing about contemporary communications is that it makes telecommuting far more feasible. I tried it 10 years ago with a 9600-baud modem (hot stuff in 1990), and it was a difficult sell to my company. Now, even blue-chip companies describe their telecommuting policies in their recruitment literature.

Working in California, it's nice not to have to dress up for work anymore. I have a closet full of suits that I used to wear when I was in Customer Services, and it's about time to give them to Goodwill.

Crowell: There are no more three-martini lunches, smoking at work has disappeared and the dress code has relaxed - these are all for the better. I wouldn't change any of these and I think the cultural workplace changes will continue to focus more on the business of delivering solutions to customers' real problems. The relaxed dress codes and the prohibition against smoking in the workplace are very conducive to a productive and creative work atmosphere.

Telecommuting hasn't lived up to its promise yet. Cylink has telecommuters, but only a few. They will remain small in number as long as the tools for telecommuting do not permit a high-quality experience as part of a team, particularly the availability of high-quality videoconference equipment at low operating cost. Most high-end technology projects require a large team, working very closely together to get the job done. The demand for face-to-face team interaction is very high.

Hall: Well, in the 30 years I have been in CS, I have never had a three-martini lunch, and my dress code has been more or less static over the past 30 years (T-shirts and shorts in summer, sweat-shirts and long pants in winter). I did discover "cargo shorts" last year, and they get "the maddog sign of approval." I am glad that smoking in work has disappeared, not that it was ever encouraged in the large machine rooms anyway.

I think the biggest thing that has changed is the concept of the 70-hour week, and the blending of "home work" and "office work." If I had one message that I would like to get out to some of the younger programmers these days, it would be, "Make sure you take time for your family ... time for yourself." I am a firm believer in the Scott Adams' OA5 (Out at Five) club, and, if managers cannot schedule load to make sure that hard-working people are "out at five" (on the average), then they are not managing properly.

Airlines are now giving back the leg room we had in 1973. I think it is time for employers to give us the "joy time" we had then, too. [Just this morning, I got word that a fine gentleman who was the first manager of the Ultrix group, Bill Munson, has passed away at the age of 55. Even more reason for "OA5."]

Salus: I never went to a three-martini lunch. They may be mythic. I'd love to prohibit all use of idiot Word attachments to documents and to mail. They just bulk things up, and I can't read them anyway.

Provoncha: [I would change back] vendor attention[and encourage]more telecommuting.

Hogarth: Having to wear a suit at work was always considered pretentious; a non-smoking IT professional is not to be trusted; three martinis at lunch (with one's boss) promotes camaraderie; telecommuting is the best thing since the TV remote control; cubicles are demeaning: professionals deserve offices with walls and doors.

Hayhoe: I have always emphasized interpersonal communication skills to those who work for me. I think it will be a special challenge in the future as face-to-face time continues to diminish.

Inmon: For new employees, there is a demand for practically every skill you can think of. But, the bright, new, young graduates are not going to work in the IT organizations. They are going to work for the vendors and startups. Wouldn't you?

Greiner: Adaptability! The marketplace is changing too fast to expect to be doing the same thing next year, next month ... maybe even next week. Plan on continuing your education and find a company that will support you in the effort.

Crowell: IT is now a team sport. There is no one I know who knows enough about information technology to be able to single-handedly solve an IT problem.

The biggest skills IT professionals need are the ability to work as partners and integrators, to suggest ideas and adapt to other ideas, and to focus on a common solution. A secondary skill is a thorough understanding of how software and network systems work together and how security can help overcome their shortcomings, since networks are inherently insecure and you don't always know who you're talking to.

Hall: The ability to learn on your own, from books, magazines and the WWW. Today, it is like the Olympics: You need to start your interest in computers when you are eight years old, just so you can participate in the workforce when you are old enough to work. I tell kids that I feel sorry for them, since they have to have the equivalent of my Master's Degree (1977) just to start coding. Or, at least, coding correctly. Internationalization and localization issues, graphics, networking, high availability ... and all for the only engineering degree which does not have "professional" status.

Salus: Adaptability. My crystal ball is opaque, but I know that things will change a lot. Be smart and be adaptable.

Provoncha: Less than were needed 10 years ago. I see less commitment and generally lower expectations.

Hogarth: A birthdate later than 1975, two years Java experience in e-commerce, and a rudimentary understanding of English as a second language.

Greiner: Using "wishful thinking" as an alternative to data security, system reliability and development. Security and reliability that was built-in to many mainframe applications a decade ago is considered to be new technology to Web application designers. Some enterprises' approach to back-up is, "We'll get it back up as soon as possible." Aside from an unamusing joke, this is little more than gross negligence.

Crowell: It's the fact that there are 300 million users on the Internet and 300 million potential hackers. Without security, the Internet will never fulfill its promise. If I were to rank industries currently leading the way in security expertise, I'd put financial institutions on top because they have such a compelling need for security, followed by large aerospace and large high-tech multinational companies. They understand the threats to international communications.

Hall: Microsoft.

Salus: Microsoft and stupidity are the greatest threats to IT today.

Provoncha: Forgetting the lessons we learned 25 years ago. Back it up! Store offsite. Test recovery. Keep an audit trail. Believe in security.

Hogarth: The declining respect for computer "geeks" and "nerds" shown, not only by corporate management, but by society, in general; when all of the experienced and talented professionals have been replaced by amateurs and H1b immigrants, who will keep the business running?

Inmon: The biggest threat to IT, today, is the continuing complacency of trying to keep a status quo of a world of IT technology that is 20 or more years old. The IT function has slipped into the hands of the end users - finance, marketing, sales, accounting, engineering - and the vendors. IT has become a tired, visionless place where people are counting the days to retirement. I hope this helps. (Sorry about the gloomy outlook, but at least it is honest.)

Any final thoughts ... ?

Hayhoe: It is important for individuals like yourself to create an environment where change is encouraged. The technical press has a responsibility to, not just report on the black boxes, but to address the environment in which those black boxes are used. And, to challenge the workplace to deal with those cultural changes that must take place.

Greiner: In a recent speech, Dr. Gene Amdahl indicated that Moore's Law (that computers will double in speed every 18 months) may soon collide with the laws of physics. In order to sustain the rate of growth that has been seen in the past 10 years into the next 10 years, major, fundamental changes will be needed in the design and implementation of computer systems. I expect that the next decade will see an increase in parallel processing and software optimization.

Crowell: The greatest challenge today to harnessing the full power of the Internet is security. Whether we realize it or not, without security, our businesses are vulnerable. If we build our business systems on top of the Internet without adequate security, we're creating an electronic Tower of Babel. The industry really needs to focus on a set of requirements and standards for liability and recognize that everyone is responsible for security. If there are weak points in the chain of security within my business or yours, they will make all of the rest of us vulnerable.

Hall: I think that most of this interview could be described as pretty "negative," which is very unlike me. So, now I will put on my Pollyanna hat and tell you what gets me out of bed in the morning:

In the past 30 years, I have seen computers turn from large, clunky, fantastically expensive machines to things that I can even wear on my wrist. Today, I can buy more main memory for my computer than I could disk space 10 years ago, and for less money. Data communication speeds are now a hundred-fold over what they were 10 years ago (assuming the telephone company every gets around to installing DSL at my house). But, more amazing than any of this is the teenager who can hold his own in a discussion on building a compiler or designing an operating system.

It is people that keep me going, not machinery. IT is only a means to an end, and that end is helping people reach their goals.

Salus: The computer is the best screwdriver ever invented. It is a wonderful, versatile tool. Together with the Internet, it has made the world smaller and promulgated both knowledge and rumor.

Provoncha: It's a fun time to be involved, lots of possibilities, the Web has changed everything, but let's not be lemmings, look before you leap.

Hogarth: According to The GartnerGroup (November 1996), "Approximately 40 percent of mission-critical mainframe projects fail, as do approximately 70 percent of mission-critical client/server projects." Gee, I wonder why?