Glen, Sam, Can you please initiate this as a huge concern through "appropriate" channels? A lot of TCK tests are coded this way and the behavior is not part of SAAJ spec.
Rick, Is the code for auto-magically adding "cid:"; prefix been removed yet? Any ETA? Thanks, dims --- Rick Rineholt <[EMAIL PROTECTED]> wrote: > Axis for attachments automatically generates "Content-Id" headers (CID) > when attachment parts are first created. In addition the logic tries to as > best as possible to make CIDs universally unique for each attachment as > required for MIME type attachements. This allows applications using Aixs > attachments to simply request the attachment's CID and immediately use it > as a reference for an href attribute without the need for first generating > this CID itself and having to set it in the attachment part. Further more > I don't believe that this behavior in any way contracts the JAX-M, JAX-RPC > or SAAJ specifications. However, the TCK as I understand it, has tests > which set headers and then validate that ONLY those headers are present. > Which fails for Axis. I believe this test should be changed to validate > that headers which it set are present but ignore additional headers. Or, > obtain the list of already present headers, add additional headers, and > then validate that the headers within the attachments have been updated > appropriately include the new headers added or modified. > > Rick Rineholt > > [EMAIL PROTECTED] > ===== Davanum Srinivas - http://xml.apache.org/~dims/ __________________________________________________ Do You Yahoo!? Sign up for SBC Yahoo! Dial - First Month Free http://sbc.yahoo.com
