The Times of India has an (audio) interview with me, done by Sudhir Chowdhary. (The full text interview seems to have vanished from the site!)
The consolidation o the IT industry has begun. In the recent past, PeopleSoft has announced plans to buy JD Edwards, and then itself is now the target of a takeover bid by Oracle. Palm has bought Handspring. Barron’s writes:
There are far too many technology companies, and they will not all survive. Seventy percent of all publicly held technology companies have market caps of $250 million or less. “And here are J.D. Edwards and PeopleSoft, who clearly are not in this category, thinking that they all need to bulk up,” Deninger says. “So the question is: What are those in the seventy percent category thinking? It’s time to wake up and smell the coffee.”
The number of publicly held technology companies has been winnowing since the bubble burst, but naturally most of them have been in Deninger’s 70% Club. In 2002, the number of publicly traded tech concerns declined for the second year in a row, for the first time since 1989, and Deninger predicts the trend will continue in 2003. “It’s going to be the first time ever for three years in a row,” he says. “There simply are too many of them.”
Some 200 to 300 tech companies will disappear this year and probably next due to mergers, buyouts, and bankruptcies. At least, that has been the case during the previous two years. There are 30% fewer companies in the Nasdaq from its peak.
Adds Kevin Werbach: “Everyone in high-tech was hoping that the economic downturn would be over by now, but instead we’re still bumping along the bottom of the IT spending trough. That, combined with ruthless commoditization pressures, means no one can grow their way out of the current situation. Everyone must either find a small but lucrative niche, or get big enough to compete with gorillas like Microsoft, IBM, Cisco, Nokia, and Sony.”
Technological innovation, as well as deregulation and trade liberalization, ”fostered a pronounced expansion of competition and creative destruction,” observed Alan Greenspan. ”The result through the 1990’s of all this seeming-heightened instability for individual businesses, somewhat surprisingly, was an apparent reduction in the volatility of output and in the frequency and amplitude of business cycles for the macroeconomy.” Translated from Fed-speak, that means that for the past 20 years, individual companies have prospered or failed, entire industries have grown up while others have vanished and, when all that frantic sinking and swimming in the economic waters is plotted as a graph of overall output, it looks like a gently rising curve. Meanwhile, the periods of economic expansion keep getting longer, while the recessions get shorter and less severe.
That’s the sign of a robust system, says Michael Mauboussin, who is paid by Credit Suisse First Boston to think big thoughts about markets and financial behavior. ”It takes two essential features to make a system robust,” he says. ”You need diversity, and you need interaction. That describes the American economy. You have diversity, a lot of local agents doing their own thing based on local information. And these agents interact in the marketplace; at some point, two agents will meet at a price. Then you have a big diversity in outcomes — some buy, some get bought, some win, some lose — and that makes for a robust, stable system.”
Wired has a story by Paul Boutin on how the Slammer worm worked to”crash the Internet in 15 minutes”. The worm source code is also listed.
Slammer owes its speed to UDP, an Internet protocol that’s lighter and quicker than the TCP used for Web sites, email, and file downloads. TCP requires sender and receiver to acknowledge each other in a handshake before exchanging information; UDP can carry a message in a single, one-way packet. Microsoft’s SQL Server 2000 software has a UDP-powered directory service that lets applications automatically find the right database. Moreover, SQL code comes built into other programs the company sells. Many Slammer victims didn’t even realize they were running SQL.
The worm takes advantage of a common software bug called a buffer overflow. Buffers overflow when a data string is written into memory without its length being checked by the program. If the string is too long, the tail end of the data overwrites the program’s own code.
The genius of Slammer is how it uses an attack on just one type of software as leverage for a general attack on the Web itself. Machines infected by the worm swiftly spam the Net with randomly addressed traffic, hitting other vulnerable servers. As the number of computers spewing Slammer packets rises, the situation reaches critical mass, potentially creating a denial of service attack on all 4 billion IP addresses on the Net. Sounds crazy, but Slammer is fast enough to pull it off.
Mark Hurst has a set of good ideas on dealing with spam. I have been doing much of what he mentions. Have set up filters and multiple inboxes. My spam management time is under a couple minutes a day. recently, we have also added SpamAssassin on our mail server, and that too does a pretty good job in identifying spam.
Dave Winer makes an excellent suggestion: “If I were starting a new company in 2003, I would put in the charter that, in addition to whatever else my company did, the new company would be a publication. I’d hire an editor in chief, parallel to the CFO and CTO. This person’s charter would be to cover the company, much the same way the editor of the San Jose Mercury News covers Silicon Valley.”
I think a compnay blog is going to be increasingly important. How to get the views of a new company across to the interested set of people is a challenge that can be addressed by the emerging micromedia like blogs. It is one way to get out of the marketing trap that small / new businesses face.
Think of the blog we create as a microcosm of our interests, reflected in the links, analysis, comments, blogroll and categories. We think that this blog reflects who we are. But more accurately, the blog reflects who we want to be seen as how we want others to see us as.
Now, imagine another blog. A Mirror Blog. This has links to what we are reading and would probably like to read, based on our interests. It has a blogroll which reflects our friends, and friends of friends. This blog is created from the blog we create and its interactions with the rest of the blogging ecosystem. The Mirror Blog is a more accurate reflection of us who we really are. It is constructed automatically by what we do, and what others do. It is a product of emergence shows us a top-level view of the world around us.
The best part about the Mirror Blog is that we do not have to do anything special other than blogging to create it. The more we tell it about what we like, the more we link into the world around us, the better it gets. It is a content recommendation engine, distilling out what our neighbourhood is reading, writing and discussing. It provides us with new inputs and insights. In fact, even if the personal blog is not updated on a daily basis, the Mirror Blog continues to be because the world around us is constantly changing.
The Mirror Blog also becomes an extension of our memory, in conjunction with our blog, ensuring that it is always possible to search the network of friends and related content, a kind-of middle between the compactness of our own brain and the breadth of Google and the other search engines.
The Mirror Blog is indeed the heart of MyMemex.
Gelernter envisioned all kinds of Mirror Worlds for hospitals, cities, universities, communities. This is how the blogosphere is emerging. Today, we see blogs created by people. Already, blogs are being embellished by photos from digital cameras. Soon, we can expect sensors to send out information to specific log pages. Each entity whether human or machine or an aggregation can be expected to have a weblog, which is a representation of itself. For all practical purposes, the weblog is akin to a Whats New page of a website. It mirrors the latest state of the entity, with access to the history also. The weblogs become the information sources, publishing their data in RSS, which we can subscribe to as part of our Memex. The interaction of the weblogs with the rest of the blogosphere creates the Mirror Blog.