David:
The message from Chris earlier today (see attached)
states:
"The thing I have a problem with in using URN's for
the namespace is that many parsers want to find the
schema at the URI of the namespace. If we were to
use a URN there, it would necessitate that either
xsi:schemaLocation be specified explicitly in ALL
messages or else a mapping would need to be made
between the URN and the schema location."
I am not quite clear on what Chris means by
"or else a mapping would need to be made
between the URN and the schema location".
If we use an unresolvable namespace like:
http://oasis-open.org/messageService
then it will be necessary to use xsi:schemaLocation to
identify the schema location.
Likewise, if we use the same namespace for different
versions of the schema (for example, when we have a
2.0 version), we will have to use xsi:schemaLocation to
specify the schema location.
Ian:
Can we schedule some time next week to discuss the
namespace issue at the F2F meeting?
In fact, Suresh suggests that the namespace issue
should be brought up to the OASIS ebXML JC.
Regards,
-Arvola