restriction

Hi!

I have some questions regarding derivation by restriction. While
browsing through the mail archive I noticed that this is a
controversially discussed topic.

OK, a restriction should contain a list of particles with which a
one-to-one relation to the supertype particles can be built. Renaming is
not allowed. 

- Is it right that the particles of the derived type are not obliged to
have a "name"-attribute?!
- Is it possible to change the type of a particle as part of the
derivation by restriction process?
- If there are model groups in a complexType how can they be restricted?
Is it possible by recursive restriction? Do all the rules for
restriction (on the top level of a restriction) apply inside model
group, too?

If all those questions are answered by a current, unofficial
specification, is it possible to post those parts? Or: when can the next
specification be expected.

Volker