To achieve that goal, OpenDesign is stepping gingerly over the bodies of failed start-ups while eyeing a prize coveted by Microsoft, Sun Microsystems and Oracle.
The Bellevue, Wash.-based start-up is the brainchild of Intellectual Ventures, a Redmond, Wash.-based incubator headed by Microsoft's former chief software architect Edward Jung and former chief technology officer Nathan Myrhvold.
Web-based applications are seen as the next stage of evolution in computing, in which key functions of PCs will be transferred to more powerful machines, or servers, linked in a network. But Web-based delivery of core applications, such as word processing and spreadsheet programs, raises a host of problems that have limited widespread development and adoption of applications. As a result, most PC software today far outstrips Web-based rivals.
Like some of its predecessors, OpenDesign promises to do for Web-based applications what Microsoft (msft) and other operating system providers did for desktop-based applications: provide a platform to handle the low-level plumbing of application development and deployment.
OpenDesign proposes to let developers "publish" their applications through its system as a way to ease distribution problems, thereby taking advantage of scalability and load-balancing features, the lack of which company founders say hamstring today's Web applications.
"In a way, the Web was a huge step backward because we went from powerful PCs with huge applications and right back to a thin-client-server model, right back to the OS of the 1970s," said Alex Cohen, chief evangelist for OpenDesign. "OpenDesign and a lot of other companies are all trying to get to the next level."
Cohen and others say Web applications will be able to take off once they have resolved underlying infrastructure issues, such as easing heavy traffic demands on one server by distributing traffic and tasks evenly among multiple Web servers--a technique known as load-balancing.
"If we can solve those problems for the developer transparently, the potential for innovation increases exponentially," he said.
Rivals step in
That potential has drawn a hungry competitor in Microsoft, which has long regarded the Web as a potential replacement for the PC as a venue for application development, and therefore a threat to its dominant Windows operating system. In the Web-application sector, Chairman Bill Gates has wagered the future of the company on a high-stakes gamble known as Microsoft.Net.
Laboring in the shadow of Microsoft's Redmond empire, Cohen and the Microsoft veterans funding the start-up are entering an increasingly crowded field of companies providing infrastructure for Web applications.
Some of the pioneering members of the Web application infrastructure vanguard have already fallen by the wayside, felled by lack of developer and consumer interest in their offerings and by the recent doldrums in the venture capital markets for technology start-ups.
These include Desktop.com, a start-up launched in 1999 that provided an API (application programming interface) for Web-based applications. Desktop.com closed its doors late last year.
Another failed effort was My Internet Desktop, which abandoned its business-to-consumer model and was sold to Internet Appliance Network (IAN). Its technology is now being used in IAN's internal network.
"What we were trying to do was create an infrastructure simple enough for Mom or Dad to create something, but sophisticated enough so big corporations could use it too," said Marcelo Lewin, founder and chief executive of Level 2 Labs and former CEO of My Internet Desktop. "The way Web development works is comparable to building a house by first making the nails, then cutting the boards. We were trying to bring you all the pieces so you wouldn't have to do everything from scratch."
One of the troubles for the failed start-ups--and potentially for OpenDesign and the new crop of start-ups--is that the idea of simplifying Web application development and deployment is a principle focus for Microsoft, Oracle (orcl), and potentially Sun (sunw) as well.
Microsoft has placed great emphasis on its Microsoft.Net initiative; Oracle has countered with its own software release; and some expect Sun to throw its hat into the ring with a project code-named Brazil.
Another high-profile start-up tackling some of these issues is Loudcloud, the Web-based software infrastructure provider founded by Netscape co-founder Marc Andreessen.
"This kind of effort takes a lot of (research and development), a lot of money, and a lot of time to get it done," Lewin said. "Which is why companies like Microsoft and Sun are in a great position to do that. My suggestion to anyone trying to tackle that is to look at what Microsoft and Sun are doing and try to become part of it, rather than fighting it."
Start-ups face the giants
At least one company perceived by Microsoft to be stepping on its Web application toes has not fared well this year. Crossgain, a Redmond-based start-up briefly led by former Microsoft executives and staffed by former Microsoft engineers, was forced to fire its 20 former Microsoft employees under pressure from the software giant, which said they were violating the terms of their yearlong non-compete agreements, sources close to Crossgain have confirmed.
Although both Crossgain and OpenDesign have remained mum on the details, their plans appear to be in the same vein. Both companies compare their work on the Web with the function of the operating system on the PC. Both also promise to free time and resources for Web application developers by taking care of crucial underlying tasks such as load-balancing; fault tolerance, which lets one server take over when another goes down; and scalability, or the ability to increase output with demand.
OpenDesign executives have discussed their plans with Microsoft's Gates, and the start-up is not unduly concerned about the company's reaction to their plans, according to Cohen.
Jung, OpenDesign's acting chief executive, maintains that the start-up has plenty of room to maneuver around Microsoft's .Net initiative.
"From 30,000 feet, everyone seems to be working on the same thing," Jung wrote in a statement. "But if you get in there and really understand what each company is doing, you'll see people placing bets on different approaches and problems."
Analysts predict that the market for Web-application infrastructure services will follow a typical pattern of initial growth and ultimate consolidation, and that for companies to succeed, they must position themselves carefully to eke out a viable niche.
"Microsoft will not own Web services," said Daryl Plummer, an analyst with Gartner. "In the category of Web services framework or composition, companies are looking to address the underlying infrastructure of Web applications. There will be money to be made there."
Other analysts agree that the need for these services is great, demand is high, and competition is going to be stiff.
"The inadequacy of the present infrastructure is hobbling the ability for the whole software-as-services trend to take off," said Dwight Davis, vice president of Summit Strategies. "There's a lot of action, a lot of competition in that space."
Davis cited Citrix Systems as one example of a company with an early toehold in the market. Microsoft licensed Citrix's technology for hosting applications on a server while providing load-balancing, availability, monitoring and other services.