Chrome Apes IE8, Adds Clickjacking, XSS Defenses

CWmike writes "Google has announced that it added several new security features to Chrome 4, including two security measures first popularized (some later shot down as having 'zero impact') by rival Microsoft's IE8 last year. The newest 'stable' build of Chrome includes five security additions that target Web developers who want to build more secure sites, said Adam Barth, a software engineer on the Chrome team. The two aped from IE include 'X-Frame-Options'" a security feature that helps sites defend against 'clickjacking' attacks, and cross-site scripting protection.'"In Google Chrome 4, we've added an experimental feature to help mitigate one form of XSS [cross-site scripting], reflective XSS,' Barth said. 'The XSS filter checks whether a script that's about to run on a Web page is also present in the request that fetched that Web page. If the script is present in the request, that's a strong indication that the Web server might have been tricked into reflecting the script.'"

Cross-site scripting

[commlinx]

Recently I starting doing a bit of web development after being out of the loop for a while. I was working on a project and it was convenient to have the XHTML / JS running on my development machine while doing a few AJAX calls to my development server. After it failed at first I found I could add Access-Control-Allow-Origin: * to the HTTP header to allow cross-site access.

It made we wonder if you wanted to exploit cross-site vulnerabilities couldn't you setup a proxy in the middle that returned information from the original site but added that to the header? Anyway just got me wondering and maybe someone more knowledgeable could comment on it.

Re:Stay classy /.

[10101001 10101001]

Your house is seriously insecure, even if you have a steel door and have window panes are made of bullet-proof glass, you probably live in a stick frame building where a drill and a sawz-all can gain me access to the interior in an hour or two. Yet no one seems to get excited about the insecurity of our houses.

In large part because, as you point out, it's impossible to make a house physically secure (although security guards can hypothetically do a good job). Similarly, it's impossible to make a computer physically secure (after all, it's in a house or building and those security guards still aren't perfect). Meanwhile, software, being a virtual good, can actually provide absolute security within the confines of the computer that runs it being physically secure. Hence, there's a higher standard held on software.

When our houses get robbed, we recognize that the wrongdoing is being done by the criminal. Yet when our computers are hacked, we place the wrongdoing on the provider of the software.

No. In both situations, the wrongdoers are the criminals. The issue comes to the point, really, of whether any blame can be put upon the constructor of your house (or its parts) and the constructor of your computer (or its parts). For homes, if someone sold a lock that, as sold, should be reasonably able to stop being hacksawed through was in fact hacksawed through, you'd still have reason to blame the lock maker. Similarly, software that is clearly defective against what it reasonably should block would leave blame upon the software maker. The issue, then, is merely that Microsoft (and most software makers) regularly admit their software is faulty (the need for Windows Update). The only real thing left, then, is to point out that Microsoft has such a poor reputation, no person should reasonably expect their software to be secure; if that's your position, I agree that blame is being badly cast on Microsoft.

I have never really understood why software is held to such lofty standards, particularly on consumer desktops. It would be one thing if file sharing of your entire filesystem was enabled by default in typical software, but lets be real- hacks these days require really clever methods to exploit systems, and if it wasn't for very intelligent, very dedicated people constantly pounding and poking our software, we wouldn't have to worry at all. Yet an uneducated teenager can break into a house in a few minutes with little more than a stick to break a window, and we seem to all go about our day without any outrage at all.

Again, software can be actually made secure. Most the "easy" exploits have been fixed because they are actually fixable. There's nothing you can do to prevent a teenager from being able to break into a house (well, not legally, anyways); you can in many states/areas shoot the teenager after they enter. The comparison is rather apple and oranges.