21st: License to code

Experimenting with ways to cash in on software research, universities try a dash of habanero.

Published December 12, 1997 8:00PM (EST)

Many of the technologies that make the Net go round -- from the arcane details of routing protocols to the Web browsers we use every day -- were hatched at universities and bestowed on the world largely for free. But as universities struggle to find their place in the anarchic digital marketplace, they're getting serious about trying to cash in.

For a look at how, pay a visit to Team Habanero. Ensconced at the University of Illinois in a linked series of offices packed with a cornucopia of Windows NT, Sun and SGI workstations, these front-line researchers of the National Center for Supercomputing Applications (<a NCSA) at the University of Illinois have just unveiled their software baby -- a Java-based product called Habanero.

Habanero promises its users the ability to create and view documents, chat, throw in some graphics and change the documents, all in real time and -- thanks to the "write once, run anywhere" principle -- from any machine. Habanero is intended to be the cornerstone of the next generation of collaborative groupware (think Lotus Notes). And yes, it's for sale.

Or, to be more precise, it's up for licensing. If you only want to use it -- say you and your boss want to haggle over a memo together -- then just head to Habanero's home page and download a free copy, which is currently available for UNIX, Windows and the Mac. If you're more interested in borrowing its source code to use as the foundation for your own new product -- a kind of "Habanero Plus" -- the NCSA is still more than willing to accommodate you. Just drop by the Web site, sign a few forms and voil`! You're only a few phone calls and lawyers' fees away from owning some advanced technology of your very own.

University-funded software projects like Habanero have become the middle ground between free software movement programs like Apache and Linux, with their traditions of full source-code disclosure, and commercial software, where the code is usually held under lock and key. You can use programs like Habanero for free, but their ivory-tower owners will only hand over the key if you're willing to promise royalty fees. Most of these projects are the result of collaborations between senior researchers, who often have real-world projects under their belts, and talented students, who hope to go professional themselves someday.

Since researchers don't have to worry about end users, they often create software that's buggy, weak on interface design and intended primarily for researchers -- the people who wrote it in the first place. But at the same time, the increasing complexity of the software means the old strategy of just dumping half-finished programs on a server (as the NCSA did with Mosaic, the original Web browser) just doesn't cut it anymore.

"Universities are at a crossroads," says Karen Hersey, the president-elect of the Association of University Technology Managers (AUTM) . "They have to decide what is the best way to distribute their software. Should it be put on a server and be distributed via FTP for free, or should the university look at the commercial potential?"

"Most often," added Hersey, the MIT intellectual property counsel, "it just can't be distributed at will. [The software] is too complex. It needs maintenance. It needs documentation. It needs someone who understands the software. What you're talking about here is a totally different type of software than, say, shareware; a type with commercial value, but also a type that demands this kind of maintenance. If you just distribute that from an FTP server, then no one will know how to use it, it will lose credibility and will eventually be thrown away."

- - - - - - - - - - - - - - - - -

Hersey and AUTM are representative of the new managers emerging in universities across the country, who not only oversee the distribution and safeguarding of intellectual property, but also are charged with making some money off the product. Although AUTM has not yet polled its members for their annual software revenues alone (most universities count them as part of their larger technology licensing efforts), Hersey says such a poll will be conducted early next year.

In the meantime, an utterly subjective sampling of just two research universities revealed numbers that could keep a start-up happy for months: MIT's <a Technology Licensing Office grossed $21.2 million in its last fiscal year, and the University of Illinois' <a Research and Technology Management Office pulled in $4 million. Not too shabby in either case -- but of course the real money is being made by the companies who are pulling this software off the shelf and plugging it into their existing technology to get the crucial edge against a competitor. Just ask Microsoft.

Those of you reading this in Microsoft's Internet Explorer should take a moment to take a look at the application's "About" section for an illuminating example. Back in the early days of the browser wars, Microsoft decided not to fool around with coding a browser from scratch and instead sub-licensed the original Mosaic from U of I-approved <a Spyglass Inc. Although the royalties have dried up (Spyglass accused Microsoft of falling behind on payments; the two sides settled on a one-time payment of $8 million), the speedy progress Microsoft's IE team was able to make illustrates how well a licensing strategy can work for both sides.

That kind of fame and success is precisely what the Habanero team is counting on to round up potential licensees for their work. Although they have no financial stake in the software themselves, the project's leaders, Ed Grossman and Larry Jackson, have been working on Habanero since the summer of 1995, virtually since the birth of Java itself. And the project's roots go back further, to an eventually abandoned groupware project named Collage, which future Netscape bad boys Marc Andreessen and Eric Bina worked on before their fateful reassignment.

So far, Jackson says, only a few small companies working in Java development have taken a serious look at the software. The U of I Research and Technology Management Office offers Habanero for licensing but has no plans to give the software an extra push into the marketplace. "We usually don't go looking for licensees," says Sharon Tipsword, technology manager in charge of multimedia and software "We wait for them to come to us. Our goal isn't to make money, it's to make sure that people use the software. That's all."

That's why it was Jackson who went to Tipsword with the idea of licensing Habanero. At all of the universities I spoke to, it's always up to the researcher to step forward with his or her invention and help map out a strategy. Don Walker, a technology transfer officer at Harvard, says, "We only hear about the software that is revealed to us by the inventor. I think the stuff that I see is just a fraction of what is actually going on in the university."

In the case of Habanero, Jackson and Tipsword met and collaborated on the licensing agreement that now adorns Habanero's home page. Should a company come along and decide it's interested enough to invest in the technology, the U of I has several options, ranging from licensing the software itself, which is the most common, to appointing one company as the official commercial licensee (à la Spyglass and Mosaic) to accepting equity in the company itself.

Although this last practice is not uncommon, according to AUTM's Hersey, it immediately raises the specter of universities as the next generation of venture capitalists, possibly financing their own graduates' start-ups in a gamble to recoup their research budgets in one fell swoop.

That's not likely to happen, says Hersey: "We really can't expand into the for-profit marketplace because of tax exemptions. We can't sell products and we shouldn't be competing with commercial software manufacturers, anyway. They are devoted to building a product and updating it, adding features and fixing bugs. University research is about completing one challenge and starting another. I don't see that we will get into the stream of commerce any more than we already are."

And neither, it seems, does Team Habanero. Both Grossman and Jackson came to the NCSA from the private sector, and neither shows signs of storming off to form his own Habanero-clone company. As Grossman, who has already lived the start-up life once, put it, "I came here because I liked the laid-back and relaxed atmosphere. I don't see myself going through the start-up crunch all over again." But, he added, referring to the proud papa of Netscape who gave Marc Andreessen all the money he needed, "If Jim Clark came to me and wanted to start a company around Habanero, that would be a different story."

By Greg Lindsay

Greg Lindsay is a frequent contributor to Salon.

MORE FROM Greg Lindsay

Related Topics ------------------------------------------