[via Jivha] Technology Review (Simson Garfinkel) writes that "the blazingly fast Gmail service proves that Web-based applications can be just as good as ones running on your desktop computer":

Gmail validates a claim that Sun has been making for nearly a decade—that it’s possible to replace a network of PCs running Windows with world-class computers offering computing services to low-cost and easily-managed desktop machines—perhaps machines so inexpensive that they don’t even have a hard disk. Sun called such computers “thin clients.” While they are popular at some companies, they haven’t made real inroads against the Windows desktop because the applications just haven’t worked as well.

But Gmail does work just as well as a copy of Outlook Express running on the desktop. In some way, in fact, it works better. This is big news—bigger, in fact, then most people seem to realize.

Until Gmail, practically every Web-based application was a pale imitation of that same application running on a PC. Web-based applications had the advantage that they were accessible from any computer on the Internet on professionally managed servers, that the data was backed up, and that the applications themselves were constantly updated. But compared to applications running on your local machine the web versions had fewer features and performed more slowly. Most users—businesses and consumers alike—were unwilling to make that compromise.

Gmail is different. For starters, it’s blindingly fast—so fast that it feels like it is running on your local computer and not in some data center. Click on a message’s subject and it instantly appears. When you are done reading a message you click “Archive”—the message is instantly stored, and you’re looking back at your inbox. (As with other Web-based mail systems, you can report spam simply by clicking “report spam.”)

Gmail gets its speed from some of the cleverest JavaScript ever written. Lots of information is stored inside your browser and redisplayed from memory; this avoids the need to constantly download pages from Google’s servers. The JavaScript can "listen" directly to your keystrokes, allowing you to drive Gmail with single-letter commands: press “y” to archive a message, “c” to compose a new message, and so on. You don’t even have to depress the Ctrl key, making Gmail even faster to use.

Gmail shows that Web applications with thin clients can have advantages over software running on your desktop. The most obvious is reliability: Gmail runs on Google’s servers, not your hard drive, and Google almost certainly does a better job than you do with routine maintenance, backups, and the like. And because everything is kept on Google’s servers, you don’t have to wait for long downloads. Google’s computers are blazingly fast: searching through the few thousand messages stored in my Gmail account is essentially instantaneous. Searching through the same amount of mail on my local computer takes ten seconds or more.

This is about RIAs (rich internet apps), integrated web services and open standards being fused with productivity software, micro-content and social networking and offered as hosted experiences.

Does this sound like anything familiar?

Yahoo has defined what portals have been - since day one - but their UI just plain sucked! Even the valiant attempts at providing "customization" features in MyYahoo - were tolerable at best.

Yahoo supports RSS and has over 120M active end-users. Yahoo is showing how portals and ISPs can work together by providing software bundled with services - to the masses.

But ODDpost makes it a whole new ball game.

Now Yahoo can step up to the front on "end-user" experience. That holey grail that's been eluding them since day one. HTML was never desigend and will never fulfill the end-user quotient. The human factor.

The essence of compelling experiences.

HTML will also suck. But once you can truly integrate rich interactive experienecs in teh browser, and tie it into services and functionality - you got a winnign formula for digital lifestyle aggregation!

And once you have email, why stop there? Why not jukeboxes (like MySpace has) or photo blog objects (like Flickr) or Tribe Listings, Friends and Tribes appearing in blog gutters - as well?

Gen Kanai points to a Forbes article in TiVo and a survey conducted by the Forbes columnist Sam Whitmore which was released back to readers. From the Forbes article:

With only 304 employees, pipsqueak TiVo somehow continues to disrupt one institution after another. No other $141 million company has come even close to transforming government policy, audience measurement, direct response and TV advertising, content distribution and society itself.

Just look at what a new idea can do.

Barron's writes:

First, Veritas Software...then Siebel Systems, PeopleSoft, BMC Software and Computer Associates. They all missed the sales and earnings expected for the June quarter, with most blaming shortfalls on last-minute deals that should have closed but didn't.

Will the software industry wind up in the lost-and-found, or should you consider retrieving any of these shares? Bear Stearns analyst John DiFucci believes that investors -- and software executives -- are encountering a permanent slowdown in the industry's growth. At the same time, he sees reasonable opportunities for the likes of Oracle and Computer Associates.

"I don't think that the sky is falling in software," says DiFucci. "This is just the reality of the future." In the 1990s, he says, the software business grew at a compound annual rate of about 13%. With no novel applications on the horizon, DiFucci thinks software vendors will now grow sales just a bit faster than GDP growth -- say 7%-8% per year.

As software firms approach the limits of their original market, they've started to intrude on each others' turf. Oracle has tried to sell applications, in addition to its database. Microsoft's tried to sell databases and applications, on top of its operating system. Pricing pressure has increased. Mergers might relieve that pressure, says DiFucci, but antitrust objections are blocking the consolidation attempts of firms like Oracle.

News.com has more on the enterprise software market:

"The thing to realize is that the ERP market is a very small market," said Jim Shepherd, an analyst at AMR Research in Boston. "The reality is that the Fortune 1000 only has 1,000 companies."

Although much of the technology industry faces similar challenges, this change has been particularly harsh for enterprise software manufacturers. For three decades, the prime mover behind business software sales has been the promise of a "killer app" that can give customers new insight into their businesses, wring profits in the most efficient way and help them gain a competitive edge. That's what drove multimillion-dollar sales throughout the 1990s.

In recent years, however, many corporate customers have begun to rethink that notion, especially after the twin blows of the technology bust and the national recession forced them to live with less. Now, rather than buying more products to do more things, companies want the software they already own to more closely model how they do business.

Inquiring investors want to know: Is this a blip, or is the second-quarter slowdown the beginning of a longer-term malaise in tech?

Fred Hickey, editor of The High-Tech Strategist in Nashua, N.H., and a technology stock analyst who knows the industry down to its nittiest and grittiest, says the setbacks in the sector are just beginning.

Investors may have been lulled into thinking that the second-quarter results at tech companies would be sunny because reports of shortfalls had been relatively rare.

Companies typically alert investors to problems late in a quarter, but by June 30, that front was quiet. "Normally the third month in a quarter belongs to the confessors," Mr. Hickey said. But at the end of June, he added "there were more positive preannouncements than negative."

Nevertheless, two signs point to problems ahead for technology stocks, he said. First, semiconductor shares, which often lead the action in tech stocks, have gone into a nose dive. The semiconductor stock index, known as the SOX, is down 11.2 percent this year, and spot prices of computer chips are forecasting further declines.

But the biggest trouble spots on Mr. Hickey's horizon are the ballooning inventories on tech companies' balance sheets. Already rising in the first quarter, these inventories will probably show a surge for the second quarter, he said, because few tech companies appear to have cut production in recent months.

Business Week writes how wireless mesh networks can help in connecting rural communities:

Long-nelected rural towns and even recreational-vehicle parks are now surfing the Web on the Wi-Fi wave, thanks to new enhancements in the technology. Take the nearly 60,000 citizens of Rio Rancho, N.M., who on June 26 gained access to the Net from their homes, their cars, nearby parks -- virtually anywhere around town. No need for DSL or cable-modem service in Rio Rancho, a hamlet that was the 81st of 83 nationwide markets to get broadband from its local cable operator.

Rio Rancho didn't get just any Wi-Fi connection, mind you, which traditionally can reach out about 300 feet reliably. New "mesh" technology expands the reach of wireless networks, allowing Net signals to be beamed across miles. How? By using wireless routers installed every few hundred feet that communicate from point to point, passing along the broadband signal. Suddenly a Wi-Fi hot spot becomes a hot town like Rio Rancho.

"If an access point goes out, the system finds the nearest one to connect you to," says Ken Upcraft, executive vice-president of Usurf America, which built the Rio Rancho network. "That allows us to offer full coverage."

Usurf now bills Rio Rancho as the largest Wi-Fi hot spot in the world. It has installed about 500 routers on light poles throughout the city's 103 square miles. In addition, some are mounted on the roofs of tall buildings. "You can cover a large geography very quickly," says Craig Mathias, principal of researcher Farpoint Group. "You can literally be up and running in a few days."

What makes the technology most attractive to small locales is its affordability. Usurf spent about $2 million to construct the Rio Rancho network. It laid out $1.2 million deploying the access points and $830,000 for gear such as antennas and towers connecting to the Net backbone. The cost to residential customers is about $40 a month, and even less in other rural locales where it has already been rolled out. In Chaska, Minn., subscribers pay only $16 a month.

I wonder if anyone is looking at wireless mesh networks in India.

The New York Times writes:

It is known variously as B.P.L, for broadband over power lines, or as HomePlug. As a concept, it has been around for a long time. What is new in the last two years is a series of technical breakthroughs, mainly in chips designed by Intellon, a tiny company in Ocala, Fla. These chips have made power-line transmission fast enough, cheap enough and reliable enough to merit serious attention. A standards-setting group called the HomePlug alliance has also played an important role.

The idea behind this approach is that plain old electric wires can do double duty in carrying high-speed digital data, much the same way that cable, fiber-optic and D.S.L. networks do. The advantage is that the needed electric wires are already there, bringing power to nearly every house in the nation and almost every room in each house. So for a tiny fraction of the cost of building new connections, this approach could help solve the familiar "last mile" problem: how to bring Internet service from trunk lines to each school and household. It can immediately deal with the increasingly vexing "last hundred feet" problem: how to bring broadband service to every nook and cranny of a building.

Bill Berkman, the chairman of Current, said the power-line system had advantages for utilities like Cinergy because it made their electric grids "intelligent." The systems can automatically sense service interruptions, problems and performance levels more quickly and precisely than they otherwise could. But no one, including Mr. Berkman, contends that power-line transmission is the exclusive or final answer to broadband problems.

"It's another tool in the toolbox," he said, to be used with cable, D.S.L., a promising wireless technology called WiMax and other systems, employing each one where it is most efficient. John H. F. Miner, the president of Intel Capital, said that in the long run, all of today's data networks might have to give way to fiber-optic systems with even higher capacity. But that could take a long time - and in the short term, power-line transmission could be valuable not only in the United States, but also "in places with inferior telecom infrastructure but great power grids, like Russia."

Indian organisations need to take security far more seriously than they currently do. There are many elements which need to be understood and acted upon. While security services can be expensive to deploy, the solution of not using them can be even more costly. Security needs to be considered at four levels:

Internet: An Internet-based can help filter email for spam and viruses. Considering that the majority of email received is spam, this can help kill spam earlier in the value chain. In India, this also helps save on bandwidth. The same applies for viruses. In addition, ensuring that most spam does not reach the user also helps reduce the frustration levels in having to deal with spam.

Local Network: The perimeter of the enterprise needs to be guarded. A security appliance which provides the essential functionally of a firewall, virtual private network and intrusion detection is what enterprises needed. This could also be integrated with the messaging server, and complemented with local anti-virus and HTTP virus scanning. Care must be also taken to put in place the right security policies.

Desktop: User actions at the desktop are a source of vulnerability. By clicking on innocuous-looking attachments in email, viruses can be unleashed. In today’s world of connected networks, all it takes is one unprotected computer to create havoc. Server-centric computing with open-source applications on the desktop can help almost eliminate the threat at the local desktop level.

Application-level: An oft-ignored area is application security. Since more and more companies are putting their business online, it is important to also ensure that web applications are protected. A presentation by Jerry Berkman captures the concepts involved.

Looking at from another side, there is an opportunity for Indian software companies to create security solutions and services and offer them to organisations globally. Walter Mossberg provides a sense of the opportunity:

[Today,] you have to buy each of [the security components] separately, because each takes care of only a narrow slice of the growing problem of criminals and slimeballs who want to invade your computer. Antivirus programs can't stop hacker intrusions or recognize spyware. Firewalls and spyware programs can't detect viruses. Windows updates close vulnerabilities the criminals use, but don't clean up any damage done.

Yet, few consumers really care whether an invasion is classified by the experts as a virus, a worm, a Trojan horse, a browser hijacker, spyware, adware or just spam. Focusing on the difference between a virus and a spyware program is like focusing on what kind of lock-picking equipment was used by the burglar who just broke into your home. The experts may care, but all you know is that you feel invaded.

What we consumers need is a simple, unified protection plan to counter all of these threats. And the computer, software and Internet industries have badly failed us in this regard. They would rather dump the security mess in the laps of users than solve it at the level where a solution really belongs: in the operating system, or the hardware, or the online provider's servers.

Tomorrow: e-Business