SIGPC only scratched the surface of news-worthy items in 1997. Some stories were outside our scope. Other stories were sufficiently well covered by other publications. And there are only so many hours in the day. But we did cover enough topics in 22 issues for a few themes to emerge:

1. Net-centric computing
2. Lawsuits
3. Consolidations
4. Apple
5. Java

For the year's final musing, let's review the SIGPC top five themes of 1997.

Net-centric computing (NCC) was a primary theme for the first year of SIGPC. This might just reflect my own personal interest in the topic, but I think it reflects an increasing interest in the topic on a wider scale. NCC encompasses a large number of issues, including thin clients, extranets, virtual private networks, total cost of ownership, Java, and network computers (NC), just to name a few.

Issue V1N11, Dilbert's view of the Network Computer (NC), was the most accessed of all the SIGPC musings in 1997. I guess many people could relate to Scott Adam's wonderful description of the pro- and counter-NC arguments. I still recommend people read his book "The Dilbert Future: Thriving on Stupidity in the 21st Century" (NY, NY: HarperBusiness, 1997).

I'm sure Apple owners felt somewhat gratified to read about the "joys" of using a PC in issue V1N19, TCO: A Personal Experience. This musing certainly struck a cord with many users of personal computers. TCO (total cost of ownership) is an issue that will continue to play an important role in 1998. Even users of thin clients won't be able to completely escape TCO issues in an NCC world: instead of disk crashes, they may have to endure network brownouts.

V1N20, 56K Modems was the issue I received the most feedback about. Operating in a NCC environment is not very feasible without a high-speed network connection. For now, 56K modems are about the fastest possible connection---at a reasonable price---for most users. I think that may change in 1998. For example, residents of  Phoenix can choose between cable modems and ADSL, both offering high-speed access at low cost. For many Canadians, this type of bandwidth is already available.

In V1N22, Networked PCs Crack 56-bit Crypto I described how networked PCs were being used as virtual, parallel supercomputers. Several Internet-based collaborative computing efforts took place in 1997, attacking problems such as cracking 56-bit ciphers and finding large prime numbers. I have a feeling this type of practical application of NCC will have profound implications for future research in computer science and software engineering.

Lawsuits were a major theme of 1997, an unfortunate occurrence for everyone except litigation attorneys. SIGPC didn't even cover the two biggest legal battles, the first between the US Department of Justice and Microsoft, and the second between Sun Microsystems and Microsoft. But we did cover several other lawsuits. The legal dustup between Intel and Digital was covered in V1N10, Digital Sues Intel Over Pentium Patent Infringements and V1N16, Lawsuit Inside: Intel Sues Digital. Digital struck first in May, claiming Intel had infringed on 10 patents in their Pentium product line. Intel struck back in August with a counter suit, alleging Digital had infringed on 14 of their patents. For now, the lawyers seem to have worked things out. V1N21, Digital to License Alpha to Intel described how a $1.5B settlement will see Intel producing the Alpha microprocessor for Digital. Many industry pundits saw this as a way for Digital to gracefully exit from the chip design and fabrication business and hitch its wagon to the Intel/H-P IA-64 (Merced) architecture, due in 1999. This development also was part of the next SIGPC theme, consolidation between some of the biggest players in the business.

Consolidations in the market continues apace. Witness the growth of WorldCom in the telephone and data business through its acquisitions in 1997. In my opinion, the settlement of the Digital/Intel lawsuit effectively removed Alpha from the microprocessor landscape, leaving only Sun's SPARC and perhaps the IBM/Motorola PowerPC chip to compete with Merced. No matter what happens, Intel's position as market leader has only been enhanced.

A shakeout in the other chip manufacturers is also taking place. V1N13, National Semiconductor Buys Cyrix described how National Semiconductor has entered the Intel-compatible market by leveraging Cyrix's 6x86 chips. Actually I think this bodes well for the consumer, given the huge increase in sales in sub-$1000 PCs this year.

On the software side, consolidations are also the order of the day. Pure Software and Atria had merged in June of 1996. In April of this year, issue V1N8, Rational Buys Pure Atria described the acquisition of this new company by Rational for $515M in stock. Similar to Intel, Microsoft just seems to get bigger and bigger (lawsuits not withstanding). Much of their success is due to very shrewd marketing. V1N15, Internet Explorer's Web Buddies described how Microsoft had out-maneuvered Netscape to partner with Apple, Lotus, and others. This was quite a coup, making IE4 is the default browser for Lotus Notes and for Apple's MacOS.

Apple Computer was in the news all year, but for all the wrong reasons. It seems that Apple has been in death throes for the last few years, but somehow it still keeps going. V1N3, OpenDoc Book Closes described the demise of OpenDoc, the compound document architecture that was also advocated by IBM. OpenDoc just couldn't get developers to commit to it, and tools were continually late in arriving.

Apple bled red ink all year, with its market share dropping to about 3%. Nowhere was this better reflected than in April, when the first 1997 financial statements were release. As discussed in V1N6, Microsoft and Apple 1997 First Quarter Earnings Wildly Divergent, it seemed that for every dollar per share that Microsoft made, Apple lost seven.

There is still a lack of leadership at Apple. It remains without a CEO, as reported in V1N14, Apple CEO not a Job for Jobs. Perhaps 1998 will show if Apple can survive, either with Jobs at the helm (officially) or with someone else. Several proponents of NCC, such as Larry Ellison of Oracle, have said Apple should refocus and manufacture NCs. Powered by Java of course.

Java was constantly in the news in 1997, but I think only in 1998 will it really show if it can truly be used in the enterprise. Although there have been success stories, there have also been failures. In V1N7, Corel Office for Java (Beta) and V1N17, Corel Cancels "Office for Java", SIGPC readers learned about Corel's experiences with Java. In April I tried an early beta of the all-Java office suite and found it to be very promising. But Corel canned the effort in August. Time will tell if Corel made an expensive mistake with this experiment (their plummeting stock price would seem to say yes: it has lost over 70% of its value in 1997), or if they merely made a misstep on the road to NCC.

Have a Happy New Year and a very productive 1998. SIGPC will return next month with more musings on PCs and their impact on the wider software engineering community.

Sources: SIGPC

Copyright © 1997 Scott Tilley

www.srtilley.com