See also: IRC log
No chair, no quorum, so AM and HST reviewed the state-of-play wrt issues around p:http-request, particularly encoding and content type.
AM: Clarified misunderstanding
last week, wrt body encoding issues
... There is only one set of serialisation parameters
... regardless of how many parts there are
... No change planned there
... We did clear up how things work on the issue HST was
concerned about
HST: Good, I'm happy if you have concensus
AM: Looking at http://www.w3.org/XML/XProc/2009/02/26-minutes.html
... If you ship base64, all is well
... On the way out, the content-type attribute just sets the
header, it doesn't determine serialisation, but if it has a
charset parameter, that must match the serialisation
controls
... All the pieces have to match
... If they don't, that should provoke an error
HST: If the WG and implementors/active commentators are in agreement, that's good enough for me
AM: Wrt the details of setting
the content type, media type parameters are allowed
... There are some things that aren't covered, notably XML into
multipart with different serialisations, but a) there's a
workaround (serialise in separate steps and base64 encode) and
b) it's a corner case anywhere
... Then there's the XQuery example I sent, with an XQuery
content type but element content: that's an error according to
the spec as it stands
... We agreed we can live with that
Meeting adjourned, no quorum