|
|
By Dan Farber
March 29, 2004
SAN DIEGO--During an interview at the Gartner Symposium/ITxpo 2004, Bill Gates was asked to give his vision of computing in the year 2014. He cited the "magic of software" and the continuation of Moore's Law as the key components that would bring about new technology advances in the coming decade. Within 10 years, said Gates, hardware could be considered as almost free, with powerful server and desktop systems, high bandwidth networks and wireless technology bringing anytime, anywhere connections. The hardware manufacturers probably aren't happy to hear that the hardware riding on Moore's Law will get closer to zero cost for customers, but it appears the economic model is headed toward revenue for digital services used. Gates did throw a few jabs at the "free software" community, which is most likely a reference to the open source community. He said that "free software can do some good stuff, but not the really good stuff." He also said that the open source community would have to deal with compatibility problems, once it gets an installed base. Given how open source software is increasingly accepted in corporate environments, the comment about open source lacking an installed base is only accurate if you compare the desktop versions of Linux and Windows. In the case of Linux versus Microsoft's server operating system, the open source code is gaining an installed base and is already having to deal with the compatibility issues today. But he's right that Linux could suffer from incompatible distributions as vendors mix and match components and system tweaks. The crux of Gates' comments is that hardware could be almost free in 10 years, but free (open sourced) software won't be good for the best kind of applications during that timeframe. You could read into those two points that Microsoft believes that it will continue to provide a viable value proposition in the future, despite the significant growth of Java, open source, and the company's ongoing legal issues.
Regarding legal issues, Gate said that the EU situation would take time to resolve, and that it would follow what happened in the U.S., where the company was not forced to make drastic changes to its business and development practices. "There are some legal issues about how we package that up, how we license it, how we engineer it," Gates said. "Some models would slow [us] down or be a real problem for customers, and many models are very livable, even though they might create some complexity for us. In everything we've had along those lines, we've always come out with solutions that work well, and let us continue to innovate in Windows. We had that here in the United States. We solved that. We're moving full speed ahead. That's been a very good experience for everybody who was involved in that. And at some point the same thing will happen in Europe. It turns out that it's going to be several years of more process before we might get to that point." Gates also said that more model-driven approaches to software development would reduce the cost and complexity of code. He described the key breakthrough in software coding as writing less code. "The idea is to ask a business person what their business processes are--how they do forecasting, sales analysis, or an HR review--and capture those processes in a very precise manner," Gates said. "It should be easy to sit down at the PC and navigate those processes, and when something new comes along, such as a new partnership, you describe how that affects the various things that go on without writing a line of code. This is true for describing how applications ought to run in the data center and for typical business processes. There should just be a little visual model that lets you dive in and see detail." This is similar to Intentional Software's notion of pushing programming upstream as a PowerPoint-like design tool that allows stakeholders to describe an application in their own terms and then hand it off to the programmers to write a "generator" to produce the machine-readable code. The founder of Intentional Software, Charles Simonyi, is a former Microsoft top executive and the creator of Microsoft Word. Gates said that over the next decade companies could reduce amount the amount of code they need to write by at least a factor of five. In addition, he touted Web services and XML as key to software development trends. Software will also become more self-managing, leading to increased levels of automation. The underlying core of Gates' vision is not much different from today's world in which hardware prices fall, but software prices and licensing fees fall more slowly, or rise with market demand and associated fees for support and maintenance. However, as software becomes more self configuring and managing, will the cost of software--not just IT staff-- decline as a percent of the total cost of ownership? Gates also said that speech technology would become mainstream in every device in this ten year period. Microsoft launched Speech Server 2004 last week and touted it as the harbinger of mainstream speech. He also touted the company's Tablet PC as a mainstream platform. "The hardware transition has to be trivial," Gates said. He cited $100 as the maximum incremental cost for a tablet versus a notebook computer, and said that the pricing differential will reach that point this year. Nonetheless, tablet PCs today are mostly used for vertical applications, and the $100 differential will apply to higher-end notebooks rather than pervasively across product lines. However, combining speech and ink as standard human-computer interaction modes will give users and developers more interface options and flexibility in a single device. "This is like graphical user interface where as soon as you get the form factor right and the software right, it's just common sense," Gates said. "If the error rate is above a specific level, only certain people will want to use it. But once you pass the magic threshold, say for dictation, then everyone will want it." In addition, having standards for speech (SALT) and ink are necessary for mainstream deployment of these technologies. When Gates was asked to look into the past and identify Microsoft greatest and most innovative achievements, he pointed to the PC model (the compatible hardware reference model upon which the company was founded), the graphical user interface, speech, the NT kernel, and the Tablet PC. Not all of those are what one would call unique, Microsoft-pioneered technologies. But, Microsoft seems set on becoming more innovative today; the company is investing $6.8 billion in R&D this year. According to Gates, 75 percent of the $6.8 billion R&D budget is for projects in the development phase, with the majority going into creating more secure software. He predicted that within the next two years Microsoft would be able to remove security as one of its top five research and development priorities. Certainly, Microsoft will make more progress in securing its software with the launch of the Windows XP SP-2 update this year, but the holy grail is still creating software from scratch that is built with security in mind. That process will take more than a few years to resolve, and will require substantial education and training to bring about a culture of security within enterprises. Gates also gave his perspective on the consulting business. "You are not going to pay us $300 per hour and we will show you where your brain is," Gates told the audience of IT executives. "We want to help corporations become more self sufficient using low cost, high volume [computing] model." The low-cost, high-volume model plays to Microsoft's traditional strengths, compared to bigger iron systems. But, open source and new pricing models from Sun, for example, are challenging the the notion of low cost for enterprises. All of the platform vendors are chasing a similar holy grail of flexible pricing, high levels of automation, embedded security, low-cost hardware and standards-based software stacks. Microsoft will continue to innovate, as with the next major release of Windows called Longhorn. The company is spending billions to continue to dominate in the next decade and has a house full of high IQs who are contributing to this mission. However, the "magic" of software won't be measured by market share. Microsoft, for example, didn't invent the Web browser but managed to take over the market and innovate on top of the foundation that Netscape laid. That's more business acumen, blessed by the government, than software magic. The open source community, for instance, will need to create its own software magic and ecosystem that out innovates Microsoft and others and provides a value proposition that is attractive enterprises. So far, Linux has shown some magic in using the collaborative community model to create software. In reference to setting a date for the Longhorn, Gates said that the operating system release is not date driven. "People are speculating that we're out in 2006 sometime, and that's probably valid speculation, but this is not a date-driven release," Gates said. Longhorn will be in alpha testing this year. When asked about outsourcing development of Longhorn, Gates responded: "We are not about doing Longhorn 20 percent cheaper; we are more about doing it six months earlier or making sure that the fundamentals are exactly right." "If I had been born in India, maybe the one location would be in India or maybe it would be in China," Gates said. "It could be done there. Now, you wouldn't be closest to the world's biggest, most demanding market, but you could overcome that. You'd have to have a liaison group in to see those customers a lot. It turns out we are in the world's biggest, most demanding market, and I think that's some help to us. But the key point is, are you going to fragment it across a lot of different locations? Some companies in the technology industry do things across multiple locations, and we seen the inner lab fights and the lack of common architecture between labs. We have one common architecture, everybody working towards one way of doing modeling, one set of Web services protocols. That uniformity allows us to make the advances as fast as we can." On the other hand, Microsoft maintains research and development facilities in the United Kingdom, China, India and other countries.
You can write to me at dan.farber@cnet.com. If you're looking for my commentaries on other IT topics, check the archives. |
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|