Re: FW: found a way to mimic the digital identity

[Arshad Noor <arshad.noor@strongauth.com>]

Not to those who knew about the weakness back in 2004 when the flaw in
MD5 was first publicized, Marc.  Everyone knew it was only a matter of
time when this attack would show up - it happened yesterday.

Every PKI I have built since 2004 has used SHA-1 for the signing-digest.
Every PKI I have built since January 2008 uses SHA-256 (SHA-1 is deemed
OK only until the end of 2010 and NIST has already started the process
for moving past SHA-256).

All the ENML examples I've used for the Symmetric-Key and X509 Digital
Signature profiles use SHA-1 for the digest, so ENML is safe (for now).
As soon as the World Wide Web Consortium updates the XMLSignature
standard - which I expect them to do so in 2009 - we can incorporate
them into ENML by reference and issue an update to the standard.  It is
possible that by the time we vote on the standard, W3C may have updated
XMLSignature and ENML 1.0 could very well include support for SHA-256,
SHA-384 and SHA-512 hashes.

I anticipate that anyone building software using ENML in 2009 will be
smart enough to include SHA-256 support in it even if ENML doesn't
specify it - and as soon as ENML specifies it, it can be flipped-on in
the field by changing one option and restarting the software.  That's
how we built CSRTool back in 2007:

http://www.strongauth.com/index.php?option=com_content&task=view&id=32&Itemid=32

Happy new year, everyone.

Arshad

Marc L. Aronson wrote:
> Interesting. (Or not?)
> 
>  
> 
> http://www.washingtonpost.com/wp-dyn/content/article/2008/12/30/AR2008123001056.html?nav=printbox&sid=ST2008123001136&s_pos 
> <http://www.washingtonpost.com/wp-dyn/content/article/2008/12/30/AR2008123001056.html?nav=printbox&sid=ST2008123001136&s_pos>=
> 
>  
>