With Oracle refusing to comment, theres no way to confirm that premise. But if it were in fact true that the deal proved too risky for Oracle, why would it be OK for Red Hat to buy JBoss?
Asked this question, lawyers raised a number of possibilities. Steven J. Henry, an attorney with Wolf Greenfield, in Boston, didnt care to speculate, but did raise the possibility that the two different teams—Red Hats and Oracles legal eagles—evaluated the facts to arrive at different levels of risk tolerance.
Given the possibility that code ownership was involved, it could be that the different parties have different stakes in the ownership, Henry said.
"[Oracle and Red Hat might have] a different bargaining position vis-a-vis the unnamed party that might own conflicting rights," he wrote in an e-mail exchange. "For example, if it appears that code of party A was copied, perhaps Red Hat has a lever to employ against A, should A make trouble, and Oracle did not. Or, perhaps Red Hat was willing to rewrite the [theoretical] problem code and Oracle was not."
Red Hat, the leading Linux distributor, announced on April 10 that it has entered into a definitive agreement to acquire JBoss, an open-source Java middleware company.
Another lawyer who formerly represented Oracle and therefore asked that his and his law firms names be withheld said his firm frequently sees intellectual property issues derail acquisitions.
"If [the potential] code issues are with Red Hat, and if Oracle was concerned that JBoss had been using some code that actually was intellectual property of Red Hat, then obviously theyd be acquiring a potential lawsuit for IP rights," he said.
In such cases, the lawyer said, one good way to solve an IP dispute is to merge the two companies. That can create problems, however, if youre talking about two major competitors, since it would eliminate competition.
While some would posit that such an acquisition would create a monopolistic situation, others might say that were there an overlap of IP between the two companies, the competition would be artificial anyway, he said.
At any rate, if there is a potential challenge associated with the code, and if that challenge came from Red Hats IP claims, Red Hat might well have priced that out of the risk column, the lawyer said.
Why Oracle would have wanted to purchase JBoss is another ripe area for speculation. The database giant has been a big supporter of open-source code, working with a diverse array of open-source projects. This staunch support of open source is what differentiates the company from Microsoft, and its also what allows potential Oracle database users to get off of Microsoft databases.
Microsofts hook in selling its database applications, after all, is that users are locked into Microsofts stack, the lawyer said. "If [Oracle] can break people out of Microsofts column, they have a better capability to sell databases to them," he said.
If JBoss remains open-source code, thus potentially making Red Hat a more effective competitor against Microsoft, and if achieving that doesnt take money out of Oracles pocket, Oracle has won the day. Even if the company doesnt have revenues associated with Red Hat, it can make it up on the database side, industry insiders said, as more Microsoft database users get freed up by a more viable open-source combination.
"Theres a difference between having an open-source company like Red Hat acquire JBoss and keep that openly available to open-source users as opposed to somebody who might make it more proprietary," said an industry insider who requested anonymity.
John S. Ferrell, a lawyer with Carr & Ferrell, in Palo Alto, Calif., said the acquisition decision boils down to five questions, the first two being "MBA" questions. First: Is there a good fit of products and resources?
Clearly, with Oracle and JBoss, there was.
Second, is there an acceptable cash flow and balance sheet? It appears that JBoss probably has all of those, Ferrell said, given that the company boasts of being profitable from Day One.
The other three questions are legal issues. These are where a lot of deals fall through, Ferrell said.
Question three is, What assets are owned by the target company, and how is that ownership secured?
"This is the SCO problem," Ferrell said, referring to the legal brouhaha that erupted from between Novell and SCO. SCO had claimed that it, and not Novell, owned Unixs copyrights and that Novells management knew this when the two negotiated a sale of Unix.
The question of who owns what was key in the SCO-Novell case. "Its not always too easy to determine, because copyright ownership can be somewhat complex," Ferrell said. "So at least with Novell and SCO there was an issue there."
Another question is, What are the outstanding contracts and license obligations of the target company?
This becomes a knotty issue with respect to open-source companies, Ferrell said, because theres the GPL license to contend with. "To the extent we incorporate code we have with GPL open-source code, we may have license obligations to the GPL community," he said.
One might suspect that was of concern to Oracle, Ferrell said.
"The last thing Oracle wants to do is to get tied up where their code starts to be owned by the open-source community," he said.
Thats not such a big problem for Red Hat, since its a pure open-source player. But Oracles lawyers "would have to be going nuts over this issue," Ferrell said. "And it may well be it became an insurmountable problem."
The final question when companies merge is, What are the legal liabilities faced by the target company? What are the pending lawsuits, threats from competitors, and patent issues from competitors sitting around and that have not yet matured, and what criminal investigations are floating around?
"We would speculate with any of these in respect to JBoss, because a lot of the time the public just doesnt know," Ferrell said.
Any one of those questions, whether legal or revenue-related, can derail a deal, he said.