I would think that if the existing implementations support exc-c14n, then use of SHOULD has no impact (if they don't I'd be shocked).  Otherwise, RECOMMENDED appears to play nicely.

workable?

--
Matt Long
MV Squared Technologies
mlong@mvsquared.net
901-848-2640


--------- Original Message --------
From: jose.kahan@w3.org
To: "Rich Salz" <rsalz@datapower.com>
Cc: www-xkms@w3.org
Subject: Re: Namespace Inclusions
Date: 20/06/05 11:31

Hi folks,

A question for developers.

Following Rich's comment:

On Thu, May 19, 2005 at 10:42:21AM -0400, Rich Salz wrote:
>
> I think that since we no longer use QName's in XKMS, that this is not
> much of an issue any more. Also, since WS-Security and WS-I, et al.,
> are now all recommending exclusive-c14n, which doesn't have the problems
> caused by standard c14n and embedding content, we should strike this.
>
> It's not really an editorial change, although it can be treated as such,
> since it's removing a limitation. We can either remove the text, and
> let folks like ws-i, etc., advise what to do, or we can explicitly say
> XKMS messages that will be embedded in SOAP documents SHOULD be
> signed using exc-c14n.

Will either striking the text or changing it to request the use of exc-c14n
affect existing implementations? If the answer is yes, I prefer to defer
this modification to a subsequent edition of the spec.

I also think that mentioning exc-c14n is better than just striking out the
text.

Tommy, Vamsi, ... comments?

Thanks!

-jose



________________________________________________
Message sent using UebiMiau 2.7.2