Hi Boris,

I was using it more as a figure of speech (perhaps British only) rather than 
an exact ratio.  Perhaps "more often than not" would have been a better turn 
of phrase.

My may concern is that you said "Note that this change _will_ result in a 
different schema if there was a target namespace involved," which someone 
less experienced than yourself could have mis-interpreted.  If you had said 
"Note that this change _could_ result in a different schema if there was a 
target namespace involved" then I would totally agree with you.

Having said that, I actually think the case where Venetian blind schema 
design is used, and all local elements are unqualified looks quite pretty. 
Especially when types from multiple namespaces are used.  You get one 
namespace declaration at the start of your instance identifying your schema, 
and then you don't have to worry about namespaces anymore.  It looks a lot 
like Java packages and C++ namespaces etc.  Shame we're not allowed to use 
it :-)

Pete.
--
=============================================
Pete Cordell
Tech-Know-Ware Ltd
for XML Schema to C++ data binding visit
 http://www.tech-know-ware.com/lmx/
 http://www.codalogic.com/lmx/
=============================================

----- Original Message ----- 
From: "Boris Kolpackov" <boris@c...>
To: "Pete Cordell" <petexmldev@t...>
Cc: <xmlschema-dev@w...>
Sent: Friday, May 18, 2007 7:48 AM
Subject: Re: Defining recursive elements?

Pete Cordell <petexmldev@t...> writes:

> Although, just to clarify for those that are a bit fuzzy about namespaces
> in schema, commonly if people define a schema that specifies a target
> namespace, 9 times out of 10 they will also make
> elementFormDefault="qualified".

Any studies to support these numbers? I just did a quick check over
a bunch of real-world schemas in our repository. Out of 18 schemas,
7 are "unqualified" (~40%) and 11 are "qualified" (~60%).





From petexmldev@t... Fri May 18 09:54:34 2007
Received: from lisa.w3.org ([128.3