[Standards] Response Stream Namespace Qualification
mwacker66 at gmail.com
Tue Feb 14 16:10:49 UTC 2012
On 2/13/2012 7:17 PM, Peter Saint-Andre wrote:
> It's not clear to me what you intend the behavior of such clients to
> be. When you say "expect that some clients would assume...", does that
> assumption have behavioral implications (e.g., return an error if the
> streams are not qualified consistently)?
I wasn't thinking all the way down to error behavior, but if the client
assumes the streams are qualified consistently, it may ignore the
response stream header and just assume it's qualified the same way as
the initial stream header. And then if they are qualified differently,
but the client assumes they are qualified the same, that mismatch would
likely cause the client to think it's received not-well-formed XML. I
wanted a second opinion on how much I should worry about that.
I don't have any strong opinions either way concerning whether the
client should explicitly check if they are qualified consistently and
whether it should return an error if they are not.
More information about the Standards