The long-standing tension between software vendors and independent researchers who try to find security holes in products came into public view late last month, when Oracle criticised bug hunters after it came under fire for its security practices.
In a message posted on November 27 in a blog on Oracle's website, Eric Maurice, manager of security in the company's global technology business unit, said Oracle wouldn't let external perceptions drive its software security policies. Maurice reiterated Oracle's commitment to strong security practices but said it would continue to prioritise vulnerabilities based on their criticality and not on who had discovered them.
He also blasted security researchers who disclose so-called zero-day flaws before vendors make fixes available for them. "We consider such practices to be irresponsible, as they can result in needlessly exposing customers to risk of attack," Maurice wrote.
The blog post was an apparent response to what Maurice described as "a flurry of articles and blog entries" about Oracle security issues.
For example, Next Generation Security Software, a Surrey, England-based security research firm that has consulted with Microsoft on security issues in the past, released a study showing that Oracle's databases have had far more vulnerabilities than Microsoft's SQL Server has had over the past six years.
Meanwhile, a security researcher in Argentina announced -- then abruptly cancelled -- plans to release information about an Oracle zero-day flaw every day for one week in December.
Cesar Cerrudo, founder of Argeniss, an IT security firm in Buenos Aires, wouldn't explain why he dropped the bug-disclosure plans. But via email, Cerrudo defends the work done by security researchers and says vendors should be more concerned about "responsible software development" than about proper vulnerability disclosure practices. "Vendors are used to researchers playing nice," he writes. "The situation should change. Research costs thousands of dollars, and right now vendors are getting [it for] free."
H.D. Moore, founder of the controversial Metasploit Project, which releases vulnerability information and tool kits for writing attack code, rebutted the notion that such initiatives only benefit malicious hackers. The information made available by Metasploit "puts the 'good guys' on equal footing with the folks who already have the skill to launch these types of attacks," Moore writes as part of an email interview.
Security flaws are unlikely to remain undiscovered for long, whether bug hunters go looking for them or not, says Robert Palmer, vice president of IT at Lenox., a Lawrenceville, N.J.-based maker of tableware and giftware.
Independent researchers provide "a valuable service," not just to users but to software vendors as well, Palmer says. He adds that he wants to see vendors bring bug hunters into the software development cycle. One way to do so would be to give researchers access to alpha or beta code "with the express intent" of letting them try to crack it before the software is commercially released, Palmer says.
But Andrew Plato, president of Anitian Enterprise Security, a consulting and systems integration firm in Beaverton, Ore., says researchers should give vendors at least 30 days to address vulnerabilities before reporting them publicly. "One of the largest problems with independent vulnerability research is blackmailing and grandstanding," says Plato.
He adds that as long as bug hunters follow generally accepted flaw-reporting practices, they serve an important role. "Obscurity is not security," says Plato. "It's better to know about a bug and get it fixed than to have it hidden."