Long-Range Broadband Wireless

A new wireless standard is in the works. This one is interesting because it promises to offer broadband speeds at distances up to 50 kms. That would make it an ideal connectivity solution for the villages of the world. WSJ has more:

Intel Corp., Proxim Corp., Fujitsu Ltd. and a handful of other technology companies said they joined an industry group called WiMax to help certify equipment based on a new wireless technical standard that could help greatly expand the availability of high-speed Internet access. Finland’s Nokia Corp. was already a member of the group.

The new standard is called 802.16. While Wi-Fi users generally cannot stray more than 300 feet from a special base station antenna set up in their homes or so-called hot spots in cafes and hotels, 802.16 technology has a range of as much as 31 miles. That means the newer technology could be used to quickly and cheaply extend high-speed Internet service to locations, such as rural areas, that aren’t currently served.

Products that use the 802.16 standard aren’t expected to be available until the second half of 2004, and carriers aren’t likely to introduce high-speed Internet service using it until 2005, companies said.

Continue reading

McNealy Interview

Information Week has a expansive interview with Sun CEO Scott McNealy. This is what he has to say about Linux and Mad Hatter, “which is our desktop strategy”, which is about thin clients and thick servers:

We’re using a whole bunch of open source–Gnome and Mozilla and Evolution and then the Linux operating system with the Java virtual machine StarOffice, Sun Ray Client, Java card readers, that sort of thing. And we’re going to build a wonderful Microsoft-free desktop. And if you want to know my longer-term opinion, I believe that Solaris is a fabulous answer on the server, and I think Linux is a fabulous answer on the desktop.

And what we’re doing also is we’re putting the same exact user interface that we’re putting on top of Linux on the desktop, so StarOffice, Mozilla, Gnome, Evolution, Sun Ray stuff, all the rest of it, all on Solaris for our workstations.

So your Sparc Solaris workstation sitting alongside an x86 Mad Hatter desktop will be indistinguishable because you can’t see the Solaris kernel or the Linux kernel. If you have the two screens side by side, you wouldn’t be able to tell. And then our thin-client strategy with Sun Ray will also have the identical user interface, so the only way you’ll know that it’s a Sun Ray is it won’t make any noise. The other two will make lots of noise. But visually you won’t be able to tell the difference.

Here is his vision of how it gets used:

So anywhere in North America I’ll be able to roam to my desktop from any Sun Ray from my smart card and roaming from home. And then by October I’ll have session transfer, and this is some new technology we’re working on where when I go to Europe or Asia I’ll stick my smart card in a Sun Ray in Europe or Asia and it will, unattended, in a matter of three or four minutes shut down my session in California on Sun Ray server and restart it on the local continent, and now I’ll be roaming on that continent with my smart card again.

So what are the barriers to this? Two. One is they’ve got to get off [Microsoft Office]. They don’t have to, but it certainly increases the mobility and the opportunity here by moving off of Microsoft Office to StarOffice because StarOffice runs on the Sun Ray, it runs on your Linux, it runs on your Windows. If you’re using StarOffice, you are inherently much more multiplatform and mobile, and by the way, you save a lot of money there, too. Every time I turn around there’s another way to save money.

The second thing is that if you can get your applications to be server-based and accessible from a travel browser as opposed to accessible from Windows, then you really can start to go wireless and you can start to use the 100 million Java phones out there to not only access any device that’s connected to the wireline network, but you can use these phones to access your mail or your CRM or your StarOffice documents from a wireless Java handset. Now you really do have mobility with security. So we think this is a big model change.

Continue reading

10 Qs for Tech Startup

From a ZDNet article:

1. Is there a ‘killer’ application for the first targeted niche? Or is there merely a solution looking for a problem?

2. Do you know exactly what your target niche is? Do you know why consumers in that niche will buy?

3. Do you understand the ‘adoption process’ among the customers in your target niche?

4. Are you aware of the specific needs of each phase of the adoption curve?

5. Are your budget estimates realistic? What percentage of the market must you gain for your business to work?

6. Passion and Trustworthiness: does the would-be entrepeneur have a passion to succeed?

7. What kind of track record does a would-be entrepreneur have?

8. Is the would-be entrepreneur coachable?

9. Leadership and Vision: Entrepreneurs who can motivate, persuade, influence, and communicate effectively are indispensable leaders.

10. Does the future entrepreneur have a realistic path to a liquidity event if needing investors?

Ellison’s Prescription

WSJ talks to Oracle’s Larry Ellison who gives his thoughts on how to survive the coming tech shakeout:

  • Recognize that simpler is better. He hopes to switch Oracle’s operations, which span 160 countries, from 1,500 servers to just 24 Dell Linux boxes.

  • Don’t reinvent the wheel. To fly to California, you shouldn’t have to design your own plane, build your own airport and learn to fly. But that’s how much of the tech industry operates.

  • Take advantage of proven technology. “Every child’s unique. Every computer doesn’t have to be.”

  • Remember specialization of labor and economies of scale. “Companies will start doing less of their own computer operations and outsourcing more.”

  • Take cues from the customer. “We became the largest industry in the world by selling things that people didn’t want to buy.” That has to stop, he says.

  • Connection Money

    Households in the US are spending an increasing money on “information and entertainment” services. Writes NYTimes:

    An analysis of government data by academic researchers over the last 25 years shows that from the 1920’s through the 80’s, the average household spent about 6 percent of its disposable income on “information and entertainment,” a category embracing everything from newspapers and movies to telephone service, radios and television sets.

    In the 1990’s that figure jumped to 8 percent. Today, “it looks like it’s more than 10 percent,” said John Carey, who teaches courses in new media at Columbia University’s graduate school of business. Clearly, Mr. Carey said, the increase in consumer spending on information and entertainment coincided with widespread expansion of cable and satellite TV and with the rapid growth of the public Internet.

    Many market analysts predict that the surging popularity of TiVo, the digital television recorder, will inspire new home video servers that capture signals from multiple Internet, cable and satellite networks and feed them to computers, televisions and other devices around the house.

    What all these emerging services have in common is a business model based on subscriptions that are billed monthly or yearly.

    The clincher: “Research by Professor Katz suggests that the typical American household is nowhere near its limit – that the average family would be willing to spend as much as $500 a month for the right combination of subscription services.”

    Linux Winners and Losers

    The Economist separates the two. Winners: IBM, HP, Dell. Big Loser: Sun. Adds the Economist on the impact on Microsoft: “The most likely outcome is that customers will face a choice between Linux, which is cheap and cheerful, and Windows, which offers more bells and whistles, is tightly integrated with other Microsoft products and is easier for unskilled staff to use, but costs more. In short, Microsoft will be not so much a loser from Linux as less of a winner. In the server market at least, Linux is providing Microsoft with some much-needed competition.”

    TECH TALK: Transforming Rural India: TeleInfoCentre Differentiators

    What makes us think that TeleInfoCentres can work as a model to transform Rural India, where previous initiatives have met with only limited success? The TeleInfoCentre aggregates a number of innovative ideas and is different from previous approaches:

  • Presence in each village vs Sharing across villages: To succeed, access to the centres (or kiosks) must be such that villagers do not have travel significant distances at all. They should be able to use it multiple times a day, and on demand. This means the centres have to be present in each village.
  • Multi-functional vs Uni-functional: Most current approaches tend to work in silos, thus necessitating duplication of expensive infrastructure. As envisioned here, the TeleInfoCentre becomes a platform for various services which are applications on top of the basic computing grid.
  • Offline vs Internet-centric: The TeleInfoCentre can begin operations simply with periodic CD-based updates, and without the need for a telephone line or Internet connection.
  • Information-centric vs Transaction-centric: The initial focus is more on information, communications and community services which do not require real-time Internet connectivity, rather than transaction-oriented services which assume the presence of Internet access.
  • User-generated content vs Top-down broadcast: The TeleInfoCentre encourages users to generate their content and share it across vertical communities of interest and practice, to complement the information that is sent from the state and district headquarters.
  • Rs 5,000 computers vs Rs 50,000 computers: Most existing approaches tend to use new desktops with MS-Windows and MS-Office software resulting in a per PC cost of Rs 50,000 or more. (It is hard for the government and its agencies to pirate software so they have to buy legal copies!) The TeleInfoCentre approach uses thin clients for a fraction of the cost.
  • Multiple Computers vs Single Computer: The TeleInfoCentre has at least 3 computers available for use at any time, thus diminishing dramatically the chances that a villager will come in and not find a computer to use.
  • Linux/Open-source vs Proprietary software: The use of open-source software eliminates licence fees, while at the same time maintaining compatibility with file formats for document exchange with the external world.
  • Wireless vs Wired connectivity: The use of WiFi in the months to come can dramatically cut connectivity costs, and at the same time eliminate dependency on unreliable, wired telephone lines. (Cellular phones are still far away from making their presence felt in Indias villages, especially those away from urban and semi-urban areas.)
  • English and Local Language support vs English only: Support for at least one local language is important. People need to be able to interact with others in the language they feel most comfortable in.
  • Subscription model vs Pay-per-service: The TeleInfoCentre approach calls for each family to make a fixed monthly payment for a basic set of services, encouraging increased usage. The alternate pay-per-service model actually inhibits usage. That may be right when computing is a scarce resource but with cheap and multiple computers available at the TeleInfoCentre, one needs to take a user-friendly approach.

    Taking all these factors into consideration, the TeleInfoCentre is a more practical approach given the realities and constraints of Rural India.

    Next Week: Transforming Rural India (continued)

    Continue reading