Re: [docbook-tc] DocBook 5.0 namespace

[Norman Walsh <ndw@nwalsh.com>]

/ Jirka Kosek <jirka@kosek.cz> was heard to say:
[...]
| to kick-off discussion about namespace for DocBook 5.0 elements I did
| small survey of namespaces used in other OASIS based XML vocabularies:

Thanks, Jirka.

[...]
| So nothing very uniform. Some languages use URNs, some URLs often
| borrowed from time when specification was developed outside of OASIS.
|
|  From this I suppose we could use namaspace name like
| http://docbook.org/docbook-ng/ which Norm is currently using in his
| prototype implementation.
|
| However I personally don't like URLs as namespaces.

I used to believe in the distinction between URLs and URNs. I don't
anymore[1].

The Architecture of the World Wide Web, Volume One says:

  Good practice: Identify with URIs[2]

    To benefit from and increase the value of the World Wide Web,
    agents should provide URIs as identifiers for resources.

and

  Good practice: Available representation[3]

    A URI owner SHOULD provide representations of the resource it
    identifies

Taken together, these two good practice notes argue for the use of an
http: URI for our namespace.

| It is because I do
| a lot of XML training and for many people it is very hard to get
| familiar with fact that http://something is not referenced and fetched
| when XML document in http://something namespace is processed.

Yes, but the sooner they learn that, the happier they will be with
XML namespaces. WebArch goes on to say:

  Principle: Reference does not imply dereference[4]

    An application developer or specification author SHOULD NOT
    require networked retrieval of representations each time they are
    referenced.

What's more, a niave user who *does* try to retrieve a document
located at the namespace name *should* get something useful back. I
propose that we publish a RDDL[5] document there.

| There is RFC that defines structure of URNs for OASIS
| (ftp://ftp.isi.edu/in-notes/rfc3121.txt). If we will decide to use URN
| for namespace it should conform to one of the following templates:

I regret my part in that RFC :-)

| I hope that this summary will help us to sort out namespace quicker
| during the following TC.

I wasn't aware of the precedent for using namespace names outside of
the OASIS DNS. On that basis, I have a strong preference for one on
docbook.org. Perhaps:

  http://docbook.org/schemas/docbook/

                                        Be seeing you,
                                          norm

[1] http://norman.walsh.name/2004/03/03/266NorthPleasant
[2] http://www.w3.org/TR/webarch/#pr-use-uris
[3] http://www.w3.org/TR/webarch/#pr-describe-resource
[4] http://www.w3.org/TR/webarch/#implied-dereference
[5] http://www.rddl.org/

-- 
Norman Walsh <ndw@nwalsh.com>      | The most universal quality is
http://www.oasis-open.org/docbook/ | diversity.-- Montaigne
Chair, DocBook Technical Committee |

PGP signature