Discussed at 2005-10-21 telecon [1]. A few amendments were proposed and 
accepted by the WG.

Resolution: Proposal accepted unanimously as amended.

[1] http://lists.w3.org/Archives/Member/w3c-xml-schema-ig/2005Oct/0019.html

There appears to me to be an error in the text approved by the WG at
its meeting of 21 October.  The proposal acted on there reads in part

    The {context} property is only relevant for anonymous type
    definitions, for which its value is the component in which this
    type definition appears as the value of a property, e.g. {item
    type definition} or {base type definition}.

But there are some anonymous type definitions which will not have a
{context} property, namely definitions shadowed by a redefinition.
Such definitions have no name, and thus would seem to qualify as
anonymous, but we agreed in Edinburgh that they would not acquire a
context property, indeed that no global / top-level components would
have a context property.

I believe the paragraph would be correct if it read

    The {context} property is only relevant for type definitions local
    to some other definition or declaration, for which its value is
    the component in which this type definition appears as the value
    of a property, e.g. {item type definition} or {base type
    definition}.  

This has the disadvantage of being clunkier: we are punished for
our failure to define simple terminology for global vs. local
components.
(I would change the status of the issue to reopened, except that
it is already in that status, for reasons which don't seem to have
been recorded.)

Closer examination of the Edinburgh minutes shows my memory was at fault:
they say clearly that redefined components will get a {context}
property pointing to their redefiner.  The state of affairs I was
remembering was an earlier stage of the proposal finally adopted.

The text approved by the WG does not appear to me to be 
particularly clear, and it assumes silently that only one other 
component will ever have an anonymous type definition as the value 
of one of its properties, which seems error prone.  But it is not, 
as far as I can tell, in error.

Integrated into status quo documents, 8 December 2005.

Although no formal request for closure was made, since the reporter as a member
of the WG was aware of the resolution of this bug over two years ago, I'm
marking it closed.