It _could_ be a performance penalty. However, compared to the
time and processing power consumed to perform the SIgnature
if is not that much. As far as I can remember about 60% to 70% of the
whole processing time was consumed by the Signature method,
all the rest by all other functions (did this with XML-SEC 1.0,
thus fairly old values) for a signed request.

Regards,
Werner

Granqvist, Hans wrote:
>>. . .. Just perform 
>>an additonla c14n as the last step after signing and/or 
>>encrypting a XML DOM. This is what the WSS4J handlers are doing.
> 
> 
> It could be a stop gap solution but scary:
> 
> 1. The performance hit of doing a courtesy c14n is considerable.
> 
> 2. There might be a security issue here somehow that we don't
>    immediately see, especially when signing and if exc-c14n is 
>    used, as exc-c14n is a destructive c14n algorithm (we thus 
>    might change what we think was signed in the first place)?
> 
> 
> I am not sure about a. below (been too long since I implemented
> that spec ;), but 1 and 2 seem quite serious in themselves.
> 
> a. IIRC, c14n only removes superfluous empty namespace decl, not
>    xmlns="x" where x is != empty.  Did you mean exc-c14n?
> 
> Thanks,
> Hans
> 
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
> 
> 


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to