RE: "Replace and encrypt will be Recommended (optional)."

Joseph,

>Ok, I tried to increase the symmetry. However, when we are doing a replace
>on encrypt are we:
>1. returning the octets representing the characters of EncryptedData XML
>available (I think so, but it's a shame to have to serialize!)
>2. returning the DOM nodes representing the EncryptedData element
available.
>3. executing a DOM function (replace).
>
>When doing a replace on decrypt, same thing, octets or nodes?

I think that when doing a replacement, something may be returned for
application's convenience, e.g., as a DOM node, but it does not have to be
required.

>When otherwise making the data available to the application, I presume
it's
>always as octets serialization of the characters of the XML.

The spec should be described as you presume, but I think that some
implementation may return a DOM node corresponding to the octets
serialization as far as applications obtain the same result.

>Now that I've considered it, I'd think it should be REQUIRED to implement.
I
>don't think it makes much sense to release this spec and implementation
>won't be able to be confident that a recipient can do a
>encrypt/decrypt&replace. Can you think of any reason to it shouldn't be
>REQUIRED?

Do you mean that both are required to implement and whether doing a
replacement or not depends on application's choice?  If so, I agree with
you.

Thanks,
Takeshi IMAMURA
Tokyo Research Laboratory
IBM Research
imamu@jp.ibm.com

Received on Tuesday, 31 July 2001 10:30:21 UTC