Canonicalization, much like life, is a bitch. Yet another way higher character encodings get downgraded into lower character encodings, bypassing IDS/IPS signatures.

Oh course, this is just another example of the fundamental problem: IDS aren't looking at the same bytes the destination service is looking at. Arian Evans does a good job scoping this:

Somewhere along the path from HTTP protocol --> to app untrusted entry point --> to parser, there are several possible layers of decoding. These could include:

-Web Sever itself
-Web Server plugin
-Canonicalization in framework (e.g.-some .NET modules)
-Canonicalization steps in web app code.
-Decoding and interpretation by shellscripts and the like.
-Decoding certain encoding types for normalization (see this a lot in PHP, or cookies base64 file-system encoded, etc.)
-etc.

This means that:

It is possible for an app to have one or more layers of canonicalization/conversion, allowing for even crazy things like double and triple-encoding, which IDS/IPS do not handle at all over HTTP

My homies in X-Force are going to have a shitty day tomorrow...

... but not as shitty as Bob Auger is going to have. I remember him starting to do this about 6 months ago, but he wasn't the one who broke the news. Bummer.

Web hackers 9999, IDS 0