Security Watch

Dr. Watson Gets an Extreme Makeover

Changes to Microsoft's error reporting tool are more than just cosmetic, but are they for the better?

Microsoft Chairman and Chief Software Architect Bill Gates said at a recent conference that Microsoft is enhancing the features of the "Dr. Watson" error reporting tool. Dr. Watson has been part of the Windows operating system since the inception of Windows NT more than 12 years ago. As with other features in the operating system, Microsoft plans to extend the granularity of Dr. Watson's controls to allow consumers to send data in context, rather than just as a memory image. Further, consumers will be able to decide what data will be sent to Microsoft and selected third-party software vendors.

When an application crashes today, Dr. Watson typically stores a dump of the entire memory image, together with specific details about the application that crashed. If, for example, Outlook crashes while an e-mail's being created, the data contained in the e-mail up to that point is part of the memory image. The consumer has no way of removing it, although he can decide not to send the entire image. In the future version, you'll be able to clear out the data in that e-mail and still send the balance of the memory image to Microsoft for inspection.

Microsoft has realized that in order to make better use out of the data it receives when a crash occurs, it needs additional information, such as what other programs were running, what other data was in memory, the status of Registry keys and so on. Consumers will have the ability to choose what data to share—and what not to share—with Microsoft. Corporations will be able to control these details through a Group Policy Object.

Privacy advocates aren't going to like this new change in the functionality of Dr. Watson, since the vast majority of consumers won't be able to navigate through the volumes of data to make informed decisions as to what they don't want to send. Microsoft has said the data will be submitted anonymously, but it's hard to see how a submission will be useful to the person who submits it if it's done completely anonymously.

Want More Security?

This column was originally published in our weekly Security Watch newsletter. To subscribe, click here.

Further, in a corporate environment, the fact that significant and potentially confidential data would be transmitted automatically across the network due to an application crash may lead to an entirely new type of Denial of Service attack. If it's possible to crash a machine at precisely the right time, then intercept or eavesdrop on the Dr. Watson dump transfer, the attacker could obtain whatever was in memory at the time of the crash.

Currently, Automatic Error Reporting yields little useful information for the consumer whose application has just crashed. This new information and greater detail may help Microsoft understand the failures better, but in reality it makes everyone a beta tester. This will likely become a popular feature to turn off, except possibly on developers' systems. Also, imagine the overhead requirements to keep such data available for dumping during a crash—it will likely be significant.

U.S. Federal Communications Commission (FCC) Chairman Kevin Martin is moving to impose some regulation on Voice over IP (VoIP) phone providers to ensure that 9-1-1 calls get through. In some cities, VoIP customers may not reach 911 at all. Even if calls get through, some necessary information, such as the address, may not appear. The problem stems from a lack of data access between the VoIP provider and traditional landline providers. Vonage, the largest U.S. VoIP provider, continues to make agreements with landline providers to give 911 centers all the relevant information on the Vonage customer, but progress is slower than the FCC believes it should be.

Although such regulation was inevitable, it may do more harm than good. There is a tangible correlation between a physical piece of copper and an address; the same can't be said of a VoIP terminal. It's inevitable that we'll hear about some fire or police department responding to an emergency call and ending up at the wrong place. The burden of accuracy will end up falling on the consumer, who will likely have to ensure the information they have with their VoIP provider is up-to-date and accurate enough for 911 call centers.

Retailers have until June 30 to comply with the Payment Card Industry Data Security Standard. In order to receive certification, retail merchants are required to meet numerous security requirements, including use of a firewall, non-default passwords and restricted data access amongst other controls. In addition, businesses must not store card verification codes or data from magnetic strips. That may be more difficult than either party realize, as recent information suggests that some credit-card processing software temporarily stores PIN numbers, which can be vulnerable to data theft. Any merchant who fails to comply will face fines, or may be banned from taking credit transactions.

Banning merchants may prove more difficult than the payment card industry believes, as so many companies are offering services to process payments on behalf of retailers who want or need the service. On the other hand, the new standard may well reign in resellers of processing services, making them more careful to whom they offer their services.

About the Author

Russ Cooper is a senior information security analyst with Verizon Business, Inc. He's also founder and editor of NTBugtraq,, one of the industry's most influential mailing lists dedicated to Microsoft security. One of the world's most-recognized security experts, he's often quoted by major media outlets on security issues.

comments powered by Disqus
Most   Popular

SharePoint Watch

Sign up for our newsletter.

Terms and Privacy Policy consent

I agree to this site's Privacy Policy.