Rather than simply post links, please state your point. I don't see it.
For example, the Linux-watch article--first of all, the same problem as the article for these comments: no Microsoft response, just quotes from one side of the issue. Cox is right when he says that simply comparing the raw number of vulnerabilities isn't a good measure (although it is somewhat important nonetheless). He cites instead time-to-patch as a better measure, but this also has associated issues, such as full regression testing, localization concerns, etc. I guess I do lean towards getting the patch right away and then worry about testing it myself, but there is also something to be said about more thorough testing on the part of the vendor.
Now, the WMF flaw was indeed very serious. I applaud Microsoft for releasing it early and out of the patch cycle, but would have liked it even sooner. The Linux-watch article is misleading, though--it quotes a characterization of WMF as "one of those careless things Microsoft did years ago with little or no consideration for the security consequences." That fits their agenda, but WMF format dates back to 1990, when Microsoft was hardly the only one that was careless about security--it just wasn't that important of an issue at the time (Windows 3.0 didn't even have integrated networking). What was once a necessity is now a "design flaw;" it is hardly something "careless" considering the history, however. And, the "two new ways" to exploit the patched WMF flaw that the article mentions quickly (as to equate the seriousness of the new exploits with the first) do not have the same effect as the original flaw. In fact, they cause the program to crash... wow, serious stuff! I get a malformed WMF file, and it fails to display and crashes the viewer... not a big deal in my book. The WMF flaw is a notable and serious flaw, but singling it out as representative of Microsoft's security record is misleading at best.
In reality, this notion that security boils down to patching and technology issues is misguided. The security approach that needed is one of policy and procedures, built around both the people and technology. Threats don't need countermeasures, they need action plans. If that means cutting off web access to untrusted sites during a time of particularly malicious activity, then do it. That is not the mindset people have today, but it is the necessary mindset of the future.
|